Documentos de Académico
Documentos de Profesional
Documentos de Cultura
Brills Companions in
Classical Studies
History
Disciplinary Profiles
Edited by
LEIDEN | BOSTON
Cover illustration: Venetus A: Marcianus Graecus Z. 454 (= 822), folio 12, recto. Center for Hellenic Studies,
Harvard University.
Brills companion to ancient greek scholarship / Edited by Franco Montanari, Stephanos Matthaios, and
Antonios Rengakos.
pages cm. (Brills companions in classical studies)
Includes bibliographical references and index.
ISBN 978-90-04-24594-5 (hardback set : alk. paper) ISBN 978-90-04-28190-5 (hardback, v. 1 :
alk. paper) ISBN 978-90-04-28191-2 (hardback, v. 2 : alk. paper) ISBN 978-90-04-28192-9 (e-book)
1. Classical philologyHistoryTo 1500. 2. Classical languagesGrammar, Historical. I. Montanari,
Franco, editor. II. Matthaios, Stephanos, editor. III. Rengakos, Antonios, editor.
PA53.B74 2015
480.09dc23
2014032536
This publication has been typeset in the multilingual Brill typeface. With over 5,100 characters covering
Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities.
For more information, please see www.brill.com/brill-typeface.
issn 1872-3357
isbn 978-90-04-24594-5 (hardback, set)
isbn 978-90-04-28190-5 (hardback, vol. 1)
isbn 978-90-04-28191-2 (hardback, vol. 2)
isbn 978-90-04-28192-9 (e-book)
Volume 1
Prefaceix
List of Contributorsxvi
Part 1
History
2 Hellenistic Scholarship60
Fausto Montana
Part 2
Disciplinary Profiles
2 Definitions of Grammar515
Alfons Wouters and Pierre Swiggers
Volume 2
Part 3
Between Theory and Practice
Section 3.1
Scholarship
Section 3.2
Grammar
3 Syntax850
Jean Lallot
contents vii
6 Orthography949
Stefano Valente
Section 3.3
Philological and Linguistic Observations and Theories in
Interdisciplinary Context
3 Mythography1057
Claudio Meliad
Bibliography1267
General Index1427
Passages Index1463
Preface
During the second half of the Twentieth century, studies on Alexandrian phi-
lology and more generally on the history of erudition, exegesis and grammar
in the cultural panorama of the ancient world experienced a renewed period
of great flowering, which continues unabated in these opening decades of
the Twenty-first century. In comparison to the state and tendencies of studies
and research on this field in the first half of the last century, today the picture
appears radically changed. Undeniably, the breakthrough was achieved by the
celebrated work of Rudolf Pfeiffer, whose book was published in Oxford in 1968
and exerted great influence on studies of the ancient world, gradually lead-
ing to a general change in the approach to this very broad field. The research
themes springing from Pfeiffers work have progressively grown in importance
and presence in the current panorama of classical studies, and now rest on
different cultural foundations and orientations compared to the manner in
which they were considered and treated, albeit with abundance and atten-
tion, in classical philology during the Nineteenth and the first two-thirds of
the Twentieth century.
As new and adequate working tools are devised and editions of texts become
available, the effort to construct solid bases for research in this sector is acquir-
ing more concrete form, while studies and essays continue, at the same time,
to shed light in greater depth on a number of themes that are relevant for the
study of ancient literary civilization. One significant element of this evolu-
tionary process has involved a reappraisal of the role of ancient scholarship,
which is now considered as one of the most important features characteris-
ing the cultural horizon of antiquity. No longer is ancient scholarship regarded
merely as a question of erudition, defined in terms of the different trends and
positions and constituting a potentially useful but essentially ancillary sci-
ence, knowledge of which was held to serve mainly as a source of fragments
of lost works and information of a historical and antiquarian nature. Rather,
the new approach can on the one hand be seen as forming part of movement
towards a positive reassessment, which by now seems well consolidated, of the
postclassical historical phases of ancient Greek culture, from the Imperial to
the Byzantine age; on the other hand, it can rightly be described as one of the
important components of a definitive transition away from the aestheticizing
and intuitionist tendencies of the misguided and often a-historical classicism
has the cultural and intellectual tools to perform the interpretation; further-
more, exegetes claim for their own workand indeed for themselvesthe
right to extract all kinds of meaning from a text and to construct their own
line of argument, focusing on what they regards as useful and important to
develop. An exegete not only enjoys great freedom (which can go as far as arbi-
trary discretion), but also wields potentially enormous cultural influence. So
powerful can this influence be that an exegete can sway widely held opinions
and shape general attitudes, all the more so if the text under consideration
is recognized as a highly authoritative work and the interpreter enjoys great
authority. The critical currents linked to philosophical, political and religious
ideologies belong to this general and generic framework, in which even the
most blatant anachronisms may be present and are in effect admitted. History
right up to the modern times is studded with manifestations of this attitude
and these cultural operations, and each of us must decide individually whether
to consider them acceptable or not, useful or not, and to what extent, in the
history of the reception of an author.
One fundamental element in this perspective, as mentioned at the outset, is
that the investigations on the philology and erudition of the ancients no longer
have an exclusively or predominantly ancillary value, and are no longer con-
sidered essentially or only as a repository of fragments of lost works, antiquar-
ian curiosities, historical information or potential aids to modern philology.
Ancient erudite and philological-grammatical production, in a word ancient
scholarship, has acquired an independent meaning of its own, inasmuch as it
is now seen as an expression and manifestation of a precise intellectual sphere
and as an important aspect of ancient civilization. The exegetic observations
and the erudite knowledge of the ancients are no longer considered as of ben-
efit only for the information that can be gleaned about a work the ancient
scholars are interpreting or a phenomenon they aim to explain. Rather, today
ancient scholarship can and must be perceived as useful and interesting for
what it reveals about the ancient scholars themselves, i.e. about the ideas,
intellectual attitudes and culture of which they are an expression in the many
different historical macrocontexts in which ancient scholarship developed,
over the centuries ranging from the Hellenistic kingdoms to the Roman empire
and right up to the Byzantine millennium. Studies on ancient philology are
and must be rooted in a historical-cultural perspective, capable of highlight-
ing the encounters among diversified historical situations and their reciprocal
influence.
Yet even today one still too often notes the tendency to discuss the data of
ancient philology and grammar on the basis of the principle of what is right
or wrong from the point of view of modern science; in other words, the
xiv preface
tendency to try to gauge how far the ancients had drawn close to the correct
interpretation and to what extent they missed the point, whether they were
good or bad philologists, with regard to their textual choices as well. These
are evaluations that distort the historical perspective and lead to mistaken
use and inappropriate evaluation of the available testimonies and evidence.
Moreover, too often the criterion for selection of materials considered worthy
of interest and study remains based essentially on what appears to be useful or
useless for the specific purpose of interpreting today the ancient author who
is the focus of attention, according to our criteria and for our own ends. In
other words, too often the body of knowledge represented by ancient scholar-
ship is viewed as potentially interesting and significant only when it is of aid
in solving a problem of modern scholarship. But this is a drastically limited
and reductive viewpoint. Instead, everything that is of no aid in specifically
interpreting Homer or Pindar or Aristophanes from our own point of view,
is of the greatest aid in interpreting Zenodotus, Aristophanes of Byzantium,
Aristarchus, and in understanding their cultural context and their intellectual
milieu. This revised perspective is making an important contribution to the
undeniable progress in the general historical vision of the ancient world:
the products of scholarship have begun to be subjected to investigation for the
purpose of discerning the critical principles, the ideas on literature and lan-
guage, the interests, the thought of the scholars themselves in their own cul-
tural context. This approach must be further consolidated and must become
the norm, resulting in heightened awareness that knowledge of the intellectual
history of ones own discipline has an essential value. This is the best and most
elevated form of utilitarianism to be attributed to the history of philology:
namely, the mission of prompting militant philology to enter into a dialogue
with itself and with the history of its own objectives and methods, its successes
and failures: for viewing oneself through the mirror of epistemological self-
observation is the high road that will build a solid foundation to guarantee the
scientific legitimacy that philology, like all the sciences, constantly pursues.
The first idea of composing this book dates back to 2007, when the three
Editors began to reflect on the possibility of putting together a Companion
to Ancient Scholarship and to consider what might be the most appropriate
structure for a production of this kind. In-depth discussion developed around
the ideas and orientations outlined in this preface, although the focus of the
original proposal naturally underwent adjustment, adaptation and corrections
as work progressed. Firstly, as has already been made clear, a historical outline
was an inescapable requisite (Part 1. History: this section constitutes the only
up-to-date systematic compendium of history of ancient scholarship from the
origins up to the Byzantine age). Secondly, it was crucial that the historical
preface xv
Franco Montanari
Stephanos Matthaios
Antonios Rengakos
List of Contributors
Eleanor Dickey
is Professor of Classics at the University of Reading in England and a Fellow of
the British Academy. She was educated at Bryn Mawr and Oxford and has pub-
lished more than eighty scholarly works, including books on Greek forms of
address, Latin forms of address, ancient Greek scholarship, and the Colloquia
of the Hermeneumata Pseudodositheana.
Markus Dubischar
is Associate Professor of Classics at Lafayette College. He is the author of Die
Agonszenen bei Euripides: Untersuchungen zu ausgewhlten Dramen (2001) as
well as of Auxiliartexte: Studien zur Praxis und Theorie einer Textfunktion im
antiken literarischen Feld (Habilitation, LMU Munich 2007). He has also pub-
lished articles and book chapters both on Greek tragedy and on knowledge
transmission in Greco-Roman antiquity.
Oliver Hellmann
is auerplanmiger Professor at the Department of Classical Philology at Trier
University. His main fields of interest are Greek Epic and ancient natural sci-
ence and its tradition, especially biology. He is the author of Die Schlachtszenen
der Ilias (2000).
Richard Hunter
is Regius Professor of Greek at the University of Cambridge and a Fellow of
Trinity College. His most recent books include Critical Moments in Ancient
Literature (2009), (with D. Russell) Plutarch, How to Study Poetry (De audiendis
poetis) (2011), Plato and the Traditions of Ancient Literature (2012) and Hesiodic
Voices: Studies in the Ancient Reception of Hesiods Works and Days (2014). Many
of his essays have been collected in On Coming After: Studies in Post-Classical
Greek Literature and its Reception (2008).
Casper C. de Jonge
is Assistant Professor of Ancient Greek Language and Literature at Leiden
University. His research focuses on Greek and Roman rhetoric and literary
criticism. His publications include Between Grammar and Rhetoric. Dionysius
of Halicarnassus on Language, Linguistics and Literature (Leiden/Boston: Brill
2008). The Netherlands Organisation for Scientific Research (NWO) awarded
list of contributors xvii
him research grants for two projects: The Sublime in Context (20102013) and
Greek Criticism and Latin Literature (20142018).
Jean Lallot
formerly matre de confrences in Greek linguistics at the cole Normale
Suprieure, Paris, is now emeritus. He has translated and commented upon
Aristotles Poetics and Categories (in collaboration, 1980 and 2002). In the
domain of Alexandrian grammar, he has published translations of Dionysius
Thraxs Techne Grammatike and of Apollonius Dyscolus Syntax (Paris 1989 and
1997, both with commentary). In 2014, the French Academy awarded him the
Georges Dumzil prize for a collection of 22 papers published under the title
tudes de grammaire alexandrine (2012).
Walter Lapini
is Professor of Ancient Greek at the University of Genova. His researches
concern classical historiography, 5th century BC theatre, and philosophical
authors, especially the Presocratics.
Raffaele Luiselli
M.A. and Ph.D. in Classics, University of London, is Lecturer in Papyrology
at the University of Rome La Sapienza. His publications include editions of
papyri from several European collections and studies on Greek letters on papy-
rus. In 2011 he has reedited the papyrus evidence for the practice of commen-
tary and annotation as exercised on the Phaenomena of Aratus.
Daniela Manetti
is Professor of Classical Philology at the University of Florence. She has pub-
lished editions of ancient Greek medical texts (Hippocrates Epidemics VI,
Anonymus Londiniensis De medicina, Hippocratic papyri) and many essays
concerning the history of ancient Greek medicine.
Stephanos Matthaios
is Assistant Professor of Ancient Greek at the Aristotle University of
Thessaloniki. He has published a monograph and a series of articles on ancient
scholarship and linguistics. He is co-editor of the volume Ancient Scholarship
and Grammar (De Gruyter, 2011; together with F. Montanari and A. Rengakos)
and associate editor of the Encyclopedia of Ancient Greek Language and
Linguistics (EAGLL; Brill, 2014).
xviii list of contributors
Claudio Meliad
is Researcher of Ancient Greek at the University of Messina. His interests
mainly concern Greek epic, Hellenistic poetry and ancient scholarship. He has
published E cantando danzer (2008), in which he reedited P.Lit.Goodspeed
2 with translation and commentary, and is preparing the first critical edition of
the Scholia to Aelians Historia animalium.
Fausto Montana
is Professor of Ancient Greek at the University of Pavia, Department of
Musicologia e Beni Culturali in Cremona. He is a member of the Italian staff of
the Anne Philologique and coeditor of the Lessico dei Grammatici Greci Antichi
on line. His research interests include Attic comedy, the history of ancient
Greek scholarship, and Byzantine scholiography, with special focus on exegesis
to drama and historiography. He re-edited and commented on all known papy-
rus exegesis on Aristophanes comedies for the Commentaria et lexica Graeca
in papyris (2006; 20122), and is the author of The Making of Greek Scholiastic
Corpora (in From Scholars to Scholia, eds. F. Montanari and L. Pagani, 2011) and
coeditor of the collection of papers The Birth of Scholiography. From Types to
Texts (2014).
Franco Montanari
is Professor of Ancient Greek at the University of Genova. He is the author
of the Greek-Italian dictionary, to be published by Brill as Brills Dictionary of
Ancient Greek. He has published monographs and articles and co-edited vol-
umes on Homer, Ancient Greek scholarship and Greek papyri, including Brills
Companion to Hesiod (2009; with A. Rengakos and Chr. Tsagalis).
Roberto Nicolai
is Professor of Ancient Greek at the University of Rome La Sapienza and the
editor of the journal Seminari Romani di Cultura Greca. He has published
several articles on historiography, geography, epic and tragic poetry, and two
monographs: La storiografia nelleducazione antica (Pisa 1992) and Studi su
Isocrate (Roma 2004).
Anna Novokhatko
is Assistant Professor of Classics at the Albert-Ludwigs-Universitt of Freiburg
(Germany). Her research interests include Sicilian and Old Attic comedy, the
symbiosis of Classics and the digital humanities, the history of Greek and
Roman scholarship with special attention to the development of linguistic
list of contributors xix
Ren Nnlist
is Professor of Classics at the University of Cologne and a co-founder of the
Basel commentary on the Iliad (2000). His most recent book is The Ancient
Critic at Work: Terms and Concepts of Literary Criticism in Greek Scholia (2009,
pbk. 2011).
Lara Pagani
is Research Fellow of Ancient Greek at the University of Genova. Her main
interests concern ancient Greek scholarship and grammar, Homeric studies,
Greek lexicography and the reflection on language in ancient Greece, literary
papyrology. She is co-editor of the encyclopedia Lessico dei Grammatici Greci
Antichi (LGGA), rdacteur of LAnne Philologique, member of the lexicographi-
cal project Poorly Attested Words in Ancient Greek (PAWAG) and of the quipe
which accomplished the third edition of GIVocabolario della lingua greca
(2013). Her publications include a critical edition with commentary of the frag-
ments of Asclepiades of Myrlea (2007).
Filippomaria Pontani
is Associate Professor of Classical Philology at the University of Venice Ca
Foscari. While primarily concerned with issues of manuscript transmission,
he is currently editing the ancient and medieval scholia to Homers Odyssey
(Rome 20072010). He has published extensively on Greek and Latin texts
(Sappho, Simonides, Pindar, Callimachus, Lucilius, Catullus, Virgil, Petronius
etc.), and on Byzantine, humanist and modern Greek literature (Politians
Greek epigrams, Roma 2002; Poeti greci del Novecento, with N. Crocetti, Milano
2010), focusing inter alia on the rise of ancient grammar and scholarship, on
rhetoric, geography, on Homeric allegories, and the literary facies of ancient
myths.
Philomen Probert
is University Lecturer in Classical Philology and Linguistics at the University
of Oxford, and a Fellow of Wolfson College. She has written A New Short
Guide to the Accentuation of Ancient Greek (Duckworth 2003), Ancient Greek
Accentuation: Synchronic Patterns, Frequency Effects, and Prehistory (OUP
2006), and Early Greek Relative Clauses (OUP 2015).
xx list of contributors
Ineke Sluiter
is Professor of Greek at Leiden University. She has published widely on the
history of ancient and medieval thought. Her most recent book (with Rita
Copeland) on this topic is Medieval Grammar and Rhetoric. Language Arts and
Literary Theory, AD 3001475, Oxford 2009.
Pierre Swiggers
is Professor in the Faculty of Arts at the University of Leuven (KU Leuven) and
Research Director of the Flemish Fund for Scientific Research (FWO). His his-
toriographical research focuses on the evolution of grammatical description
and grammar teaching, of logic and philosophy of language, and sign theory,
in Classical Antiquity as well as in later periods.
Renzo Tosi
is Professor of Ancient Greek at the University of Bologna. His main interests
concern ancient lexicography, ancient and modern proverbs, Greek theatre,
Thucydides and the history of classical philology. He wrote Studi sulla tra-
dizione indiretta dei classici greci (1988), Dictionnaire des sentences latines et
grecques (20102), La donna mobile e altri studi di intertestualit proverbiale
(2011), I carmi greci di Clotilde Tambroni (2011).
Stefano Valente
is Researcher at the Institut fr Griechische und Lateinische Philologie at the
University of Hamburg (DFG-project: Wissenschaft und Naturphilosophie in
der byzantinischen Welt: Das Physiklehrbuch des Nikephoros Blemmydes).
His research interests include Greek and Byzantine scholarship and lexicog-
raphy as well as Greek palaeography. He is the author of I lessici a Platone di
Timeo Sofista e Pseudo-Didimo. Introduzione ed edizione critica (2012) and of
The Antiatticist. Introduction and Critical Edition (2015, forthcoming).
Alfons Wouters
is Professor Emeritus of the Faculty of Arts of the University of Leuven (KU
Leuven). His research focuses on grammatical papyri and on the evolution of
grammatical theory and grammar teaching in Classical Antiquity. He is co-
founder, with P. Swiggers, of KU Leuvens Centre for the Historiography of
Linguistics.
part 1
History
chapter 1
Anna Novokhatko
1 On the tasks of scholarship as understood in Ancient and modern thought, see the appendix
on the art of philology in Gentili [1990] 223233.
2 See Montana and Montanari, this volume.
greek scholarship from its beginnings to alexandria 5
poets were also quoted and discussed) and thus linguistic and stylistic stud-
ies dealt with Homeric examples. Nonetheless, distinguishing the three main
branches remained important. One branch led to the criticism, emendation,
interpolation and editing of the Homeric text by the classical philologists of
Alexandria, a second led to the development of Greek (and later Roman) lin-
guistics, mainly grammar and semantics, a third led to literary and aesthetic
criticism.
In the cultural milieu of Archaic Greece, poetry was a public medium. Vases,
bronze, lead, clay, leather, oyster shells and bones, metals, wooden and wax
tablets of the 8th century BC from all over Greece host inscriptions, suggesting
some knowledge of writing.6 The vases preserve traces of epics.7 Writing was
used in the 7th century for the composition of texts, but these texts were still
intended for performance, most of them to musical accompaniment. During
the 6th century choral, lyric, rhapsodic, and cultic poetry continued to be per-
formed in public. According to later sources, texts and copies of texts circulated
from the Archaic period.8 Although there is no evidence for a reading culture
at this time, the essential precondition for its existence, elementary literacy,
may still be posited from public inscriptions, legal scripts, coins or graffiti.9
Vase painting reveals that Homeric texts were known in Athens at least from
the 7th century BC. At the end of the 6th cent. BC a broader use of written
Homeric texts is observed. Versions of these texts, recited at rhapsodic agons,
were probably presented in various places, oral and written narration there-
fore simultaneously coexisting. These versions survived to a certain extent as
Homeric texts from Argos, Chios, Crete, Cyprus, Sinope and other places and
perhaps were used by Zenodotus and Aristarchus from Samothrace (the so-
called ).10 The Panathenaic performances which fixed the
attribution of the two epics to Homer constitute an important part of this
story. From 522 BC the Iliad and the Odyssey were recited every four years by
rhapsodes at the Great Panathenaea.11 Reading and writing was connected
with learning the epic (and sometimes lyric) texts by rote, above all those by
century BC until the 4th century AD, used for practicing the alphabet, in Lang [1976] 67;
cf. SEG XLVII 1476 (Cumae; ca. 700690 BC); SEG LVII 672 (Histria; 5th/4th cent. BC) for
an evaluation of the argument that specific abecedaria were the work of schoolboys. I am
grateful to Benjamin Millis for this reference.
6 On the first religious and magical texts of Egyptian, Phoenician, Israeli and other Old
Oriental writings, see Speyer [1992] 7085. On writing on various material message-
bearing objects, see Steiner [1994] 1099; on metaphorical representations of letters, see
Steiner [1994] 100126.
7 For more, see Platthy [1968] 7578, Robb [1994] 2326, 4559; Knox [1989] 155156.
8 On Hesiods copy written on lead see Paus. 9, 31; on Heraclitus copy in the temple of
Artemis see D. L. 9, 56.
9 See Harris [1989] 5052.
10 See Phlmann [1994] 21; Bolling [1925] 3741. On specific versions of Homeric epics,
see also Cassio [2002] and Cassio [2012] 253. On the early Greek artists perception of
Homeric epic, see a detailed analysis by Snodgrass [1998].
11 Ps.-Pl. Hipparch. 228b; Isoc. Panath. 159; Lycurg. Leoc. 102; West [2001a] 1719; for Attic
influences on pre-Alexandrian Homeric text transmission, see West [2000] 29.
greek scholarship from its beginnings to alexandria 7
Homer. The inscriptions on the rolls painted on vases are of poetic, for the
most part epic texts.12
12 Xen. Symp. 3, 5; see Immerwahr [1964]; Immerwahr [1973]; Robb [1994] 186.
13 See Yunis [2003] 89.
14 Griffith [2001] 69 with further bibliography.
15 Diod. Sic. 11, 86.
16 Ar. Eq. 188193, 12351242, see also Morgan [1999] 54; Slater [1996] 104.
17 See Slater [1996] 112.
18 Ar. Ve. 959961; on the spread of literacy in Greece and especially in Athens, see Harris
[1989] 45115; Steiner [1994] 186241; Thomas [2001]; Thomas [2009].
8 novokhatko
19 For books on vases, see Immerwahr [1964] and Immerwahr [1973], for papyrus rolls on
vase painting as a symbol of intellectual creativity, see Whitehorne [2002] 2829.
20 Homer, Hesiod, Pindar, Anacreon, some of them preserved as later Roman copies of Greek
originals from the mid 5th century BC, see Zanker [1995] 2036; for the famous picture of
Alcaeus and Sappho on an Attic red-figure kalathos around 470 BC by the Brygos painter,
see Zanker [1995] 32 with further bibliography.
21 For silent reading, see E. fr. 369 TrGF, Ar. Ran. 5253, the famous tombstone of a young
man with a book-roll from Grottaferrata; see also Johne [1991] 5354 and Johnson [2000]
593600 with further bibliography.
22 See West [2013] 82. On the content and date of this tomb, see Phlmann [2013] 1214; on
the writing of tablets and, in particular, papyrus, see Phlmann-West [2012]; West [2013].
23 West [2013] 76.
24 Them. Or. 26, 317ac; D. L. 2, 2; Clem. Al. Strom. 1, 78.
greek scholarship from its beginnings to alexandria 9
25 See Phlmann [1994] 2021; Lloyd [1991]; on the connection of literacy with Hippocratic
epistemology, see Miller [1990] and Althoff [1993].
26 On the metaphor of writing in the mind see examples from tragedy in Svenbro [1993]
180182.
27 Aesch. PV 460461; cf. Gorg. Pal. 30. On Platos interpretation of Prometheus myth and
its connection with the invention of language in the dialogue Protagoras, see Gera [2003]
127147.
28 Eur. fr. 382 TrGF, Agatho fr. 4 TrGF, Theodect. fr. 6 TrGF, Ath. 10, 454be; on intertextual
relationship between these passages, see Slater [2002] 123124.
29 Achae. fr. 33 TrGF, Ath. 11, 466ef; cf. (to have read and compared) in Cratin.
fr. 289 PCG; cf. Alexand. Com. fr. 272 PCG.
30 Soph. fr. 121 TrGF, Ath. 10, 454f.
31 The level of audience comprehension may perhaps be gauged through the use of recited
literary texts in Aristophanes Birds in 414 BC or his Frogs in 405 BC; cf. Revermann [2006]
120; Slater [1996]. On the literary consciousness of Euripides audience, see Marshall
[1996].
32 For more, see Goldhill [1999]. On dramatic and literary contests, see Wright [2012] 3169.
10 novokhatko
33 On the chorus as an educational institution in Archaic and Classical Greece, see below
2.1.1. Cf. Pl. Leg. 654ab: the uneducated man () has no chorus-training
(), whereas the educated man () is sufficiently choir-trained
( ); Revermann [2006] 107115.
34 On Euripides embodiment of the contemporary intellectual developments, see Egli
[2003].
35 Eup. frs. 4, 17, 18 PCG, see Storey [2003] 6971.
36 Eup. frs. 157, 158 PCG.
37 Pl. Com. Sophistai fr. 143, 145 PCG, cf. also Soph. P. Oxy. 1083, fr. 1.
38 On Aristophanes Frogs see below, 3.4.2. On intellectual discourses presented in the Old
comedy, see also Zimmermann [2011] 694, 696701.
greek scholarship from its beginnings to alexandria 11
5th century BC. The audience was capable not only of recognising, but also of
evaluating intellectual trends.39
Evaluation of the level of literacy in 5th century Greece requires caution,
however. As the majority of texts belong to the 4th century, the backdating
of widespread literacy to the 5th cent. remains problematic.40 From the end
of the 5th century BC reading and writing became a regular part of Athenian
education.41 The first evidence on the book trade and literacy in Athens and in
other parts of Greek world comes from the end of the 5th century BC.42 The
readership remained lite, but it was situated in a large number of cities.
39 On the relationship of Old comedy with intellectual movements of the time, see
Zimmermann [1993a], Carey [2000], Whitehorne [2002]; on the level of competence of
the audience, see Revermann [2006]. Vase painting also seems to confirm the comic or
satirical use of images of intellectuals as also of sophistic trends. Here grotesque features
and aesthetic deformities render the perversity of novel ideas visually. For the caricature
of a sophist on a red-figure askos and a so-called Aesop on a red-figure cup, both around
440 BC, see Zanker [1995] 40; see also Whitehorne [2002].
40 On a certain increase in literacy at this time, see Nieddu [1982] 235; Harris [1989] 114115;
Morgan [1999] 5051.
41 Cf. Ar. Ran. 1114; Dem. De Cor. 258; see also Kleberg [1967] 310; Revermann [2006] 120.
42 Books must have been sold on the market in late fifth-century Athens. On the useful
discussion of Old comedys evidence for literate culture, see Slater [1996]. An Eupolis
fragment refers to a place where books are for sale beside garlic, onion and incense
stands: Eup. fr. 327 PCG, Poll. 9, 47; cf. Ar. Av. 12881289; Pl. Ap. 26d; Arist. fr. 140 Rose;
see also a word for bookseller which appears in comedy of this time:
Theopomp. Com. fr. 79 PCG; Nicopho fr. 10, 4 PCG; Aristomen. fr. 9 PCG; Poll. 7, 211. On
reading of books by Anaxagoras in Athens and on the price of a book sold in the agora see
Pl. Ap. 26de; Phd. 97b; on books found in merchant cases in Thrace, see Xen. An. 7, 5, 14;
on vase painting with cyclic epic from the 5th century BC in Olbia on the Black Sea, see
Vinogradov [1997]; see further Morgan [1999] 5859; Harris [1989] 4952.
43 Theopomp. Hist. 115, fr. 155 FGrHist; Olymp. Hist. 94, 2; see more in Platthy [1968] 7;
Pfeiffer [1968] 30; on the Homeric text in the Attic alphabet, see West [2001a] 2123.
12 novokhatko
being incorporated into the play.44 According to Svenbro: The idea of such a
play could arise only in the mind of someone to whom the grammata already
seem autonomous and to whom their vocalisation no longer constitutes a nec-
essary condition for their deciphering.45 Despite the persistence of debates on
the chronology and genre of the play, as well as on its attribution to Callias,46
the readiness of the audience to accept such jokes constitutes a clear marker
of societal attitudes to literacy.47
The alphabet remained significant. At a somewhat later date the comic
character Sappho is made to ask a riddle involving a female who bears chil-
dren that are voiceless. The children can however converse with people at a
distance. The correct answer is a letter (feminine ) bearing gram-
mata within it.48
44 Call. Com. test. *7 PCG; Ath. 7, 276a; 10, 448b; 10, 453c-e; Clearch. fr. 89a Wehrli. See Gagn
[2013].
45 Svenbro [1993] 186.
46 Phlmann [1986] 5557; Slater [2002] 126129; Zimmermann [2011] 732734.
47 It should not be considered a coincidence that the words (and )
denoting knowledge (or ignorance) of the alphabet appear for the first time at this time:
Xen. Mem. 4, 2, 20; cf. Nicoch. fr. 5 PCG.
48 Antiph. fr. 194 PCG. See Konstantakos [2000] 161180, Gagn [2013] 315316.
49 See e.g. Pl. Minos 316e; Symp. 177b, Phdr. 266d, 268c; Grg. 518b; Isoc. Hel. 12; Xen. Mem. 4,
2, 10; Xen. Oec. 16, 1; Vitr. 7, pr. 1112; more in Demont [1993], Cambiano-Canfora-Lanza
[1992] 379491; Cambiano [1992]; Casson [2001] 23.
greek scholarship from its beginnings to alexandria 13
50 Cf. Isoc. In Soph. 1012; Morgan [1999] 55, 59; on Isocrates relationship with oral and
written texts and with his audience, see Usener [1994] 13138.
51 Arist. Pol. 8, 1338a1618; see Morgan [1998] 1018.
52 Epicr. fr. 10 PCG, Amph. frs. 6, 13 PCG, Theopomp. Com. fr. 16 PCG, Alex. fr. 1, 151, 185 PCG,
Anaxandr. fr. 20 PCG.
53 See e.g. Amph. fr. 6, Alex. Com. fr. 98, 13, Philipp. Com. fr. 6, and especially Epicr. fr. 10,
1215 and Olson [2007] 239241.
54 For the Pythagoreans, see Alexis comedies Pythagorizousa, frs. 201203 PCG, Tarantinoi
frs. 222227 PCG; Aristophons Pythagoristes frs. 10, 12 PCG, Arnott [1996], 579586, 624
647; Olson [2007] 243248; for the Cynics, see Eub. fr. 137 PCG, Olson [2007] 248249; in
general for representations of philosophy in Greek comedy of the 4th century BC, see
Webster [19702] 5056, 110113 with further examples.
55 Alexis plays Archilochus frs. 22, 23 PCG; Cleobouline fr. 109 PCG; Poietai frs. 187, 188 PCG;
and Poietria fr. 189 PCG; Nicostratus Hesiodus fr. 11 PCG; Amphis Sappho fr. 32 PCG;
Antiphanes Sappho frs. 194, 195 PCG; Ephippus Sappho fr. 20 PCG; Timocles Sappho fr. 32
PCG; Diphilus Sappho frs. 70, 71 PCG.
56 Webster [19783] 117; Whitehorne [2002] 30.
14 novokhatko
Aischines, the historian Thucydides, the 4th century tragic poet Astydamas
and the 5th century tragedians Aeschylus, Sophocles and Euripides appeared.57
The literary, archaeological and epigraphic evidence reveals that by the
middle of the 4th century BC some Greek authors moved from Athens to
Macedon, a factor that contributed to the prestige of tragedy, particularly
Euripidean tragedy.58 Macedon had extended its dominance over Greece and
attracted intellectual talent from Athens and elsewhere. Greek authors thus
served Macedons rise to cultural prominence and political hegemony.
62 E.g. Arist. Soph. El. 166b, 177b, 178a, see Knox [1989] 166167. Note also that these elements
belong to philological scholarship in its Alexandrian sense (see Montana in this volume).
63 Plut. Mor. 841f; Paus. 1, 21, 12. On Lycurgus and his programme, see Moss [1989], Scodel
[2007] 149152, and Hanink [2014] 6091.
64 Scodel [2007] 150. On Lycurgus in the context of the conservation of classical texts in the
4th century BC, see also Battezzato [2003] 1012, 1419 and Hanink [2014] 6074.
65 Ptolemy III later borrowed the official Athenian copy for the library in Alexandria,
and never returned it, cf. Gal. comm. 2, 4 in Hippoc. Epidem. 3; Pfeiffer [1968] 82; see
also Montana in this volume. However, it is clear that although Lycurgus legislation
guaranteed the survival of all texts referred to, it could nonetheless not prevent small
interpolations and variants. See Scodel [2007] 151152.
16 novokhatko
2.1 Education
2.1.1 Education in Archaic Greece
Little evidence for pre-classical Greek education survives. Archaic culture prob-
ably included various forms of oral education for the social lite: instruction
within the family, encounters with mythic and religious traditions in choruses
and acting at festivals, not to mention rites of passage as a form of education.66
Social forms for the association of younger and older men included drinking-
parties where poetry was recited ( and ) and athletic competi-
tions (), forms that often promoted paederastic relationships. Musical
activity in Sapphos circle could also be considered an educational expe-
rience, and choral practice remained one of the central components in the
paideia of young men.67
Elementary schooling was conducted in private houses (
or ) with secondary schooling in public buildings such as the
, which was primarily a venue for physical training and was, for the
most part, financed privately.68 As education in Greece remained private until
the Hellenistic period, and as teachers had to be paid by parents, wide-ranging
education remained an lite affair, though a considerable part of the popu-
lation of Athens (with regional differences posited for the rest of the Greek
world) received a basic training in reading and writing.69
In pre-classical Greece physical and musical education were central. Musical
training consisted of all activities overseen by the Muses, including poetry
(Pl. Resp. 376e). Cithara players and physical trainers were the two types
of teacher in this early form of education (cf. Ar. Nub. 9611023). Literacy
and the beginnings of Greek grammar (metric and prosody) were taught as
music lessons, the borders between music and letters being indeterminate.70
This educational programme was probably available only to boys; there is
66 For Alcmans Parthenion from 7th century BC Sparta, see Calame [1997].
67 Epich. frs. 13, 103 PCG; Antipho De Choreut. 11; Pl. Leg. 653ab, 654a-b, 673a; for the lyric
chorus corresponding to the concept of education in Plato, see Calame [1997] 222231;
for an overview of Archaic Greek education, see Griffith [2001] with further bibliography;
on the institution of in Athens, see Wilson [2000].
68 Lynch [1972] 3237.
69 Beck [1964] 7294.
70 Cf. the teaching scene in Aristophanes Clouds where metre and rhythm (Ar. Nub. 638
656) and also grammar are taught (Ar. Nub. 638, 658693); Morgan [1999] 5053.
greek scholarship from its beginnings to alexandria 17
71 Euripides Phaidra is the only woman who can write (Eur. Hipp. 856881); see Pl. Resp.
452ab for Platos proposal to educate girls, see further Lodge [1950] 287308, Beck [1964]
8588, Baumgarten [2006] 98100.
72 Cole [1981] 224230.
73 See Immerwahr [1964] 2428; on the earliest representation of a reading woman on
an Attic white ground lekythos dated 460450 BC, see Immerwahr [1973] 146147. On
epigraphic evidence of literate women in Archaic Greece, see Steinhart [2003].
74 Aeschin. In Tim. 912; see Beck [1964] 9294; Too [2001] 118; Knox [1989] 159.
75 Paus. 4, 15, 6; sch. Leg. 629a.
76 Diod. Sic. 12, 12, 4; 13, 34. Chronological errors point to the unreliability of Diodorus
evidence (Diodorus transferred the legendary legislator from the 6th century to the
colony of Thurii founded in 444/443 BC). See Harris [1989] 98. However, this remains valid
evidence for the significance of universal education as attested in later sources.
77 Harris [1989] 5759; Robb [1994] 183184 and 207208.
18 novokhatko
78 More in Phlmann [1989] 7679; Beck [1975] 18, 22; Harris [1989] 97.
79 Berlin 2285, Immerwahr [1964] 19; Phlmann [1989] 78.
80 Sider [2010] 552.
81 Immerwahr [1973] 143144; Robb [1994] 186187.
82 Winter [1916]; Immerwahr [1964]; Immerwahr [1973]; Beck [1975] plates 915, 6975; Robb
[1994] 185188.
83 D.-K.12 A2, 4, 6, 911, 17; D.-K.13 A1, 2, 5, 9, 10, 14a, 17.
84 Porph. Pythag. 20; D. L. 8, 10, on Pythagorean school see further hmud [1997] 7880.
greek scholarship from its beginnings to alexandria 19
Symp. 209e), thus permitting a clear separation from other forms of schooling
which, subject to payment, were open. Centres of practical medical advice and
healing and also of methodology and the formation of medical theories were
established in Cos and Cnidos at about this time.85
85 On the mistaken backdating of the term school for ancient medical writings, see Smith
[1973].
86 Morgan [1999] 4648; cf. Quint. Inst. 1, 10, 17; sch. Dion. T. Grammatici Graeci 1, 3, 490, 5.
87 The musician and music teacher Lamprus was mocked by Phrynichus (fr. 74 PCG), Konnos
was known as Socrates music teacher (Ameipsias comedy Konnos frs. 710 PCG; Pl. Euthd.
272c, Menex. 236a), Cleon in Aristophanes Knights featured a music teacher (Ar. Equites
987996), Eupolis Aiges included a teacher of music and grammar (Eup. frs. 17, 18 PCG),
and Plato apparently wrote a comedy which included Pericles music teacher (Pl. Com. fr.
207 PCG). On the function of musical education in Classical Athens, see Murray-Wilson
[2004].
88 Cf. also Euripides versus Aeschylus and Simonides in Ar. Nub. 13611376.
89 On Euripides connection with books, cf. Ar. Ran. 943, 1409.
90 Ar. Nub. 412419, Ran. 5254, 1114; Pl. Menex. 94bc, Prt. 326cd; Euthd. 276a, Resp. 376e,
Leg. 764ce, 795de, Isoc. Antid. 266267, see further Morgan [1998] 914, Morgan [1999]
5152.
91 ORegan [1992] 1117, 3839.
20 novokhatko
the schoolwork they were expected to carry out involved explanations of lin-
guistic and stylistic peculiarities and also a knowledge of mythology.103 Lines
from literature were used for some of the first writing exercises.104 Numeracy
also constituted a part of ; thus the ordinary Athenian citizen was
taught to calculate as well as to write.105
By the beginning of the 4th century the first higher educational institutions
had been established. This reflected a trend towards increasing specialisation
in the educational process.106 Schools of rhetoric, previously practiced by the
sophists, were accorded official status. Isocrates, who studied alongside the
sophists and was known as Gorgias student, founded the first school of rhetoric
in Athens, around 390 BC. Isocrates himself did not deliver his speeches orally,
but his speeches circulated as written texts (Isocr. Panath. 10). The introduc-
tion of something approaching a methodical education (age of pupils, dura-
tion of courses, quantity of material to be studied etc.) is linked to the name
of Isocrates; his school became a model for the promotion of the teaching of
rhetoric in Greece and in other states.107 He taught alone, as both a teacher
and director of a school; therefore, after his death, his school ceased to operate.
It was around 387 BC in Athens that Plato established his philosophical
school, the Academeia. This was a rival to Isocrates school of rhetoric, both
schools constituting two branches of study, although the archaeological evi-
dence on the literate or educational practices of the 5th and 4th centuries is too
poor to provide reliable evidence.108 It is known, however, that an introductory
training course was compulsory before students could enter Platos Academy.
The Academy focused on teaching but also promoted research into ethics, phi-
losophy, logic, mathematics and others fields, such as astronomy, biology, and
political theory. It formed a complex system, where alongside Plato, the head
of the school, a number of others also taught, such as Speusippos, Xenocrates
and Aristotle. At first, the Academy was financed by Platos financial resources,
then through various donations, in marked contrast to the sophists who taught
103 Ar. Ran. 10301035; Isoc. Paneg. 159; Xen. Symp. 3, 5; see Beck [1964] 117122 with further
examples.
104 Pl. Prt. 325d326a; Chrm. 159c; Isoc. Antid. 266267; In Soph. 10.
105 E.g. Ar. Ve. 656, Pl. Prt. 318d, Resp. 522bc, 536de, Leg. 809cd, 819, Men. 4, 4, 7, Alex. fr. 15
PCG, see also Morgan [1999] 5253.
106 E.g. for judicial schooling in Athens, see Aeschin. In Tim. 912; Too [2001].
107 Isoc. Paneg. 4750; Nic. Cypr. 59; Nic. 39; Antid. 180181, 266267, 271, 273; Panath. 200;
Nilsson [1955] 9; Khnert [1961] 118121; Lth [2006] 126 with further bibliography.
108 On the archaeological evidence for the Old Academy, see Huber [2008] 2597 with further
bibliography. For the so-called Academy inscriptions, which, so it was claimed, were
schoolboys writing tablets from the 5th or 4th century BC though they almost certainly
belong to the 19th century AD, see Lynch [1983] and Threatte [2007].
greek scholarship from its beginnings to alexandria 23
for a fee. Common symposia in the tradition of the Pythagorian cults aimed
at fostering a collegiate atmosphere. Plato discussed the educational princi-
ples of the Academy in many of his writings, especially in his Republic and the
Laws, in the context of the possibility that a society could be ruled by reason.109
A number of important markers for the development of criticism and scholar-
ship were laid down at this time. Platos school reinforced the sophistic predi-
lection for disputes with pro and contra, dialectic being an art that was held to
stimulate independent thought. Exercises in the distribution and definition of
material were also practiced.110
Other disciples of Socrates, many of them close to sophistic circles, also
worked on establishing schools for a secondary education. Thus, accord-
ing to ancient tradition, one of Socrates senior students Antisthenes may
have founded a school at the Cynosarges gymnasion in Athens. The school
was supposedly intended for illegitimate children.111 There is also some fur-
ther epigraphic evidence on educational practices from the middle of the
4th century BC.112
A number of minor schools founded by Socrates disciples further devel-
oped and interpreted Socrates ideas. By the beginning of the 4th century BC
Socrates student Euclides of Megara had founded the so-called Megarian
school which focused on dialectic questioning. This school is mentioned by
Aristotle (the Megarians: ), although it seems the school had no
set location (members resided in various places).113 The philosopher in the
barrel Diogenes of Sinope, influenced by Antisthenes, established the Cynic
school which emphasised the agreement of virtue with nature. Aristippus
from Cyrene, another student of Socrates, returned to the practice of teach-
ing for a fee. He and his followers, the so-called Cyrenaic school, interpreted
109 E.g. Pl. Ep. 7, 326b; Resp. 473c; 6, 499b; Leg. 801c, 804d, see Lodge [1950].
110 On Platos Academy as an educational institution, see Khnert [1961] 112121; Pedersen
[1997] 912; Mller [1999].
111 Dem. Aristocr. 213214; Ath. 6, 234e; Plut. Vit. Them. 1, 3; D. L. 6, 13; see more Lynch [1972]
4854; Billot [1993]; Dring [2011] 4245. On the question of Antisthenes teaching
activities and subsequent links to the Cynic school, see Giannantoni [1993].
112 For scanty epigraphical evidence on educational practices, cf. IG II2 1168 = I.Eleusis 70
(a mid-fourth century decree from the deme of Eleusis in honour of Damasias of Thebes
and Phryniskos of Thebes). Damasias, who seems to have been a musician, is honoured for
his support of the Eleusinian Dionysia. Damasias students (presumably music-students)
also made contributions. See Ghiron-Bistagne [1976] 9091. Cf. further an inscription
from the middle to third quarter of the 4th century BC honouring the general Derkylos,
inter alia, for educating boys, perhaps in a secondary school (see IG II2 1187 = I. Eleusis
99). I am grateful to Benjamin Millis for both references.
113 Arist. Metaph. 9, 1046b2932. For more, see Dring [1998] 207.
24 novokhatko
114 Ath. 12, 544ab; D. L. 2, 65; 2, 85, 86; Sext. Emp. Math. 7, 191199.
115 Cic. Acad. 2, 42; D. L.1, 1819; 2, 135. See Dring [1998] 238245.
116 Pl. Euthd. 271a,Euthphr. 2a,Symp. 223d, (Pl.)Eryx. 397cd, Isoc.Panath. 33, Alex.fr. 25
PCG, Antiph. fr. 120 PCG, D. L. 9, 54, see Lynch [1972] 6875 with further evidence from
later periods. On the archaeological evidence for the site of the Lykeion building, see
Ligouri [19961997].
117 Arist. Eth. Nic. 1180a; Pol. 1337a.
118 Arist. Pol. 1266b3233.
119 Arist. Part. An. 639a, 644b1520; Metaph. 993a.
120 Cic. De Or. 3, 35, 141, on differences between the Lyceum and the Academy, see Lynch
[1972] 8396.
121 Arist. Metaph. 993a30b5, contra see Pl. Phdr. 276e277a, Ep. 7, 341c.
122 Arist. An. Post. 1, 19, 81b1923.
greek scholarship from its beginnings to alexandria 25
123 For the Aristotelic influence on Alexandrian scholarship and exegetical method see
Montana, Hunter, and Nnlist in this volume.
124 Cf. D. L. 5,51; 5,62; 5,70.
125 See Lynch [1972] 96105.
126 Ath. 4, 184b; see Morgan [1998] 2126; Morgan [1999] 6061.
127 See below 3.10. For Rhodes, see also Montana in this volume with further bibliography.
128 See Lth [2006] 129.
26 novokhatko
could flourish. The establishment of the mouseion and the library, two state
institutions founded by Ptolemy I in Alexandria in 300 BC, constitute a culmi-
nation of this process.129
136 For epigraphic and literary evidence referring to the city archive and for the
establishment of archives in the context of the interaction between oral and written
culture, see Thomas [1989] 3845 with further bibliography and Shear [1995].
137 On the connection of Euripides with written texts, see above 2.1.4; cf. Ar. Ran. 943, 1409;
Ath. 1, 3a. See Pinto [2013] 89, n. 17.
138 Xen. Mem. 4, 2, 1, 10. See Jacob [2013] 5963; Pinto [2013] 90.
139 Gell. NA 3, 17; D. L. 3, 9; 8, 8485; Platthy [1968] 121124 and Pinto [2013] 90, n. 21. On the
hypothetical reconstruction of Platos library, see Staikos [2013], especially pp. 912 and
158162.
140 Recorded by the historian Nymphis from Heraclea, cf. FHG 3, 527 Mller, FGrHist 3B, 434,
frs. 1, 2, pp. 33738; see Platthy [1968] 158; Pinto [2013] 9495.
141 Alex. fr. 140 PCG, Ath. 4, 164bd, Arnott [1996] 406415; Knox [1989] 166; Casson [2001] 28;
Pinto [2013] 88.
142 Crat. Jun. fr. 11 PCG, Poll. Onom. 7, 211.
143 Isocr. Aegin. 56; Pinto [2013] 8889.
28 novokhatko
However, the first systematic library with an extensive archive was established
at Aristotles school.144 This is probably the reason why Aristotle was known in
later Antiquity as the first collector of books.145 The books of Aristotles school
were left to the physician Diocles, and this collection perhaps served as an
example for the libraries at Alexandria and Pergamum.146 However, the quality
of the schools library may have declined after Theophrastus death in 287 BC,
with the apparent loss of many of Aristotles works to Neleus of Scepsis.147
By the end of the 4th century BC the fundamentals for the creation of a
library such as Alexandrian had been met: works on a wide diversity of issues
were obtainable, archives and scriptoria existed for keeping copies and
copying in multiple ways, and the copies were sold.148 The decree of Lycurgus
ordering fixed written versions of the works of Aeschylus, Sophocles and
Euripides played an important role in the process of establishing public librar-
ies, as these texts could then be referred to as reliable versions.149
Though poets frequently commented on their own use of language, such self-
referential deliberations should not be equated with the theoretical study
of philology.150 This section will focus on textual criticism and the growth of
theoretical writing that consciously engaged with philological concepts and
methods.
Philological ideas developed for the most part in three not necessarily
clearly distinguishable directions: textual criticism, linguistics, and stylistics.
144 On Aristotles archival studies in the form of lists of the victors at the Olympic games
(), at the Pythian games (, ), of victories in the
dramatic contests of the Dionysia ( ) and of the performances of plays
at the Dionysia (, D. L. 5, 26), see Blum [1991] 2343. On Aristotles library, see
Blum [1991] 5264 and Pinto [2013] 90, n. 21.
145 Strab. 13, 1, 54: Aristotle wasto the best of our knowledgethe first to have collected
books and to have taught the kings in Egypt how to put a library together. See Jacob
[2013] 7476.
146 Platthy [1968] 89, cf. also Lapini in this volume.
147 Strab. 13, 1, 54; Jacob [2013] 6674. See also Canfora [1999] 1720, Battezzato [2003] 2225,
and Montana in this volume.
148 For an overview of the first Greek libraries up till the end of the 4th century BC, see Blanck
[1992] 133136. For the Alexandrian library, see Montana in this volume with further
bibliography.
149 See above 2.1.2. See also Battezzato [2003] 1214.
150 For early Greek observations on poetry, see Lanata [1963]; Grube [1965] 112; Nagy [1989].
greek scholarship from its beginnings to alexandria 29
157 See above 1.1. On the Homeridai see also Burkert [1987] 49; West [2001a] 1517; Graziosi
[2002] 208217.
158 Cf. Arist. Poet. 1459a910.
159 Cf. Pl. Prt. 325e326d; see Latacz [2000] 23.
160 Sch. Pind. Nem. 2, 1c. West [2001a] 1617.
161 Cf. Dieuchide 485 fr. 6 FGrHist ap. D. L. 1, 57; Her. Meg. 486 fr. 1 FGrHist ap. Plut. Vit.
Thes. 20, 2; Cic. De Or. 3, 137. On possible Alexandrian backdating for such activities, see
Pfeiffer [1968] 6. For Peisistratus or the Peisistratids edition of the Homeric poems, see
Pfeiffer [1968] 68, West [2000] 29. For the Panathenaic performances of Homer, and
their connection with the copying and archiving of the texts, see above 1.1, esp. n. 10.
162 Frs. 192, 193 PMGF; cf. Pl. Phdr. 243a; on Stesichorus relationship to Homeric texts, see
Willi [2008] 91118, Grossardt [2012] 4378, and Cassio [2012] 255259; on Stesichorean
performance, see Burkert [1987] 5253; for the further tradition of rationalising Homer,
cf. Pind. Nem. 7, 20ff.; Hdt. 2, 112120; 4, 32; Thuc. 1, 122, cf. Richardson [1992a] 3132.
163 D.-K. 21 A1, 11, 19, B 2, 1012, 1416.
greek scholarship from its beginnings to alexandria 31
164 D.-K. 21 B1011; see more Pfeiffer [1968], 89; on Xenophanes criticism in sympotic
context, see Ford [2002] 4666; on linguistic approaches in Xenophanes, see Schmitter
[1991b] 6568.
165 Theag. D.-K. 8 A2, 1314; A4.
166 Porph. Quaest. Hom. 1, 240, 14 = Theag. D.-K. 8 A2. See Janko [2009] 52.
167 Sch. Dion. T., Ars gram. GG I 3, 164, 2329 and 448, 1216 = D.-K. 8 A1a; see more in Wehrli
[1928] 8991; Pfeiffer [1968] 911; Ford [2002] 6869. For the concept of , see
Pagani in this volume.
168 Sch. A ad Hom. Il. 1, 381 (= D.-K. 8 A3); on Theagenes copies of the Homeric text, see
Cassio [2002] 118119 and Cassio [2012] 254255.
169 Hdt. 7, 6, 3; see Diels [1910] 1011; Cassio [2002] 116.
32 novokhatko
readings of Homer (cf. Ar. fr. 233 PCG). The first Homeric-Attic dictionaries,
which formed the basis of the so-called D-scholia (from their attribution to
Didymus) and as a result of later Homeric commentaries, had their origins in
these lists. The interpretations of Homeric words sometimes required a knowl-
edge of the religious and historical background as well as of the Homeric
language.170
thus cannot be considered a secure source for the perception of true being.174
Another enigmatic fragment reflects the one-sideness of a perspective under
which a name refers to an object: one thing, the only truly wise, does not
want and wants to be called by the name of Zeus (
, D.-K. 22 B32).175 Heraclitus was the first who
understood that a word or text could have several valid meanings; arguably,
this was, for Heraclitus, not a result of conventional association, but rather of
the nature of words as motivated signs.176
Heraclituss contemporary Parmenides of Elea wrote in verse, and separated
the world of reliable truth from that of opinion or judgment.177 In Parmenides
poem the goddess marks the shift from truth to opinion by distinguishing
between her own truthful speech ( ) and the fraudulent order of
her words ( , D.-K. 28 B8, 5053). In another frag-
ment, the names of things are based on convention and on the arbitrariness
of people, and the multiplicity of things is a deceit of human receptive organs.
People made up and misleadingly gave signs () to them, each
of which received a different name (
D.-K. 28 B19, 3).178
Both Heraclitus and Parmenides affirmed that mirror reality only in
part. Parmenides further maintained the unity of language and thought, posit-
ing a division within this unity based on the distinction between right and
wrong ways of thinking. For Heraclitus, however, the separation of thought
from language was a condition for attaining knowledge.179
Empedocles of Agrigentum also reflected on the meanings of names
(D.-K. 31 B8; 15; 17, 2124; 105), and his arguments are comparable to Parmenides
concept of the arbitrariness of names (D.-K. 31 B9). Empedocles adopted a cog-
nitive approach to language which he saw as fundamental to order (D.-K. 31
B115, 14). Tin his perspective, the theory and practice of language were closely
related to the perception of knowledge.180
the cultural climate in Sicily in the first third of the 5th century BC mirrored
Athens of the second half of the 5th century BC remains an open question.186
186 For a detailed analysis of Epicharmus criticism in the literary context, see in Willi [2008]
162192 and Rodrguez-Noriega Guilln [2012].
187 On transgressions of genres in Athenian theatre, analysed by Plato and Aristotle, see
Nagy [1989] 6667.
188 More on Euripides as critic, see Wright [2010].
189 See a recent thorough study on the topic in Biles [2011] and Bakola-Prauscello-Tel [2013].
190 See more in Conti Bizzarro [1999], Silk [2000], Wright [2012]. On the self-representation of
the comic poet and the allegory of comedy in Cratinus Pytine, see Rosen [2000].
191 E.g. Stesichorus in Ar. Pax 796816, Archilochus and Anacreon in Ar. Av. 967988, 1373
1374, Alcman in Ar. Lys. 12481320, Zimmermann [1993b], Zimmermann [2000]; on the
parody of dithyrambic poetry in Aristophanes, see Zimmermann [1997].
192 For the classical work on the subject, see Rau [1967]; also a recent account in Wright
[2012] 156162.
36 novokhatko
198 See Pfeiffer [1968] 280281; Fehling [1976]; Classen [1976], 218226; Sluiter [1997b] 175; Di
Cesare [1991] 100104; Rademaker [2013].
199 Arist. Rh. 1407b6 = D.-K. 80 A27.
200 Pl. Cra. 391c, Euthd. 277e.
201 Ar. Nub. 658693, note vv. 659, 679, cf. 228, 251. See Di Cesare [1991] 102104.
202 Arist. Soph. El. 173b17 = D.-K. 80 A28.
203 Sud. 3132 s.v. .
204 D.-K. 80 A1 = D. L. 9, 53; D.-K. A29 = Arist. Poet. 1456b1519.
205 See Cassio [2002] 120121.
206 Xen. Symp. 3, 56, cf. Plut. Mor. 19e. See West [2001a] 24.
207 D. L. 2, 11 = D.-K. 59 A1.
38 novokhatko
208 D.-K. 61 A5 = Porph. Quaest. Hom. ad Il. 10, 252; cf. Arist. Poet. 1461a25, see Richardson
[2006] 6668; Janko [2009] 5253.
209 Arist. Soph. El. 177b178a. For the list of possible pre-Alexandrian emendations to the
Homeric texts, see West [2001a] 2628. See also Bolling [1925] 3156. On rhapsodic
emendations, see Jachmann [1949] 207208.
210 See Richardson [2006] 72; cf. Obbink [2003]. Burkert suggested Stesimbrotus as the author
of the Derveni papyrus, see Burkert [1987] 44.
211 Il. 11, 636637; 15, 189, 193; 21, 76, see Richardson [2006] 7275 with references.
212 Arist. Soph. El. 177b178a, Poet. 5 1461a22; see Pfeiffer [1968] 45; Cassio [2002] 129131.
213 D.-K. 58 C6; 9 test. 1 FGrHist; Porph. Pyth. 42.
214 Richardson [2006] 7677 with further bibliography.
215 Richardson [2006] 7879.
216 Cf. Lapini in this volume.
greek scholarship from its beginnings to alexandria 39
words should have the same meaning (near synonyms); 2) no one word should
have more than one meaning or connotation (homonyms); 3) the etymology of
a word should match its meaning, or at least should not contradict it.217 Plato
mentions that Prodicus dealt mainly with lexicography and synonymy,218 dis-
tinguishing between synonyms by explaining differences in their semantic
load.219 Prodicus took part in the aforementioned discussion concerning the
allegorical treatment of poetry. Like Protagoras, he used myths for the popular
promotion of his ideas, and identified gods with physical objects such as bread
(Demeter), wine (Dionysus), water or fire.220
Prodicus contemporary Democritus showed an intense interest in linguis-
tic and literary matters. Like his teacher, the atomist Leucippus, Democritus
thought that language products are not a result of necessity, but of casual
invention and connection. Democritus linguistic observations should be
understood in the context of his universal philosophical system.221 In Proclus
commentary on Platos Cratylus, Democritus is said to believe that the rela-
tion between names and things is arbitrary. His argument was fourfold:
, from homonym, from the multiplicity of
names, from the change in names, and
from the deficiency of similar items.222 Democritus
apparently compiled a Homeric dictionary explaining rare and ancient words
, a D-scholia form of dictionary. This
seems to place him in the mainstream of linguistic and Homeric discussion of
his time.223
The sophist Hippias of Elis combined scholarly and scientific knowledge.
He investigated the antiquities () from mythological, historical and
geographical points of view. Cataloging and listing were his preferred forms.224
Hippias listed the poets Orphaeus, Musaeus, Hesiod and Homer, a canon
217 Pl. Prt. 337ac, 340a-341e; Hermog. Sch. Pl. Phdr. 267b; Arist. Top. 112b2126; Gal. Diff. febr.
2; Nat. Fac. 2, 9; Plac. Hipp. Plat. 8, 6, 4650; see Mayhew [2011] xv, see also Wolfdorf [2011].
218 Pl. Prt. 337ac=D.-K. A13; Plat. Euthd. 277e; Mayer [1913].
219 Cf. Alex. Aphr. Comm. Arist. Top. 181; Mayhew [2011] 130131.
220 Pl. Prt. 320c322d; D.-K. 84 B5; cf. this motif in Eur. Bacch. 274283; on Prodicus famous
allegory of virtue and vice, see Xen. Mem. 2, 1, 2134; see also Richardson [2006] 67.
221 Cf. Arist. Gen. Corr. 315a34 = D.-K. 68 A35, see Pfeiffer [1968] 43. On Democritus as a source
for Philodemus work On poems, see Janko [2011] 208215.
222 Proclus Comm. Pl. Cra. 16, 5, 25 Pasqu. = D.-K. 68 B26, see Sluiter [1997b] 172173. More on
Democritus linguistic criteria, see Schmitter [2000] 354356.
223 D.-K. 68 A33, 11; A101; B20a.
224 Pl. Hp. Mai. 285d, 382e; Hp. Mi. 368bd; Xen. Mem. 4, 4, 525, see Pfeiffer [1968] 5154;
Blum [1991] 19.
40 novokhatko
which Plato may have subsequently used.225 Arguably, Hippias introduced the
classical opposition of nature and convention (, cf. Pl. Prt. 337c).
He also discussed Homeric questions (86 B9, B18 D.-K.) and examined lan-
guage together with music, distinguishing the importance of letters, syllables,
rhythms and scales.226
Glaucus of Rhegium may have written his treatise on the ancient poets and
musicians at this time, though information is scarce.227 The traditional Greek
unity of word and music was maintained, but the emphasis appears to have
shifted from music to language.228
Combining the roles of scholar and poet, Critias foreshadowed the
Alexandrian model. He wrote elegies, but was also known as a collector of
learned material, incorporating it into his verse (D.-K. 88 B9, 44). Another pupil
of Stesimbrotus, Antimachus of Colophon, also combined these roles, and, as
a result, he is regarded as a forerunner of Callimachus. Antimachus contin-
ued the tradition of the cyclic epic in Greece in the early 4th century BC. He
made an edition of the Homeric text,229 emending and writing comments on
the Homeric poems (he composed a book on Homeric problems 107 frs. 2125
FGrHist). He also ornamented his own poetry with glosses and aetiologies.230
A number of philosophers authored material related to scholarly discus-
sions of this time. Thus Socrates disciple Antisthenes (see above 2.1.6) was
known for his Homeric criticism;231 he wrote various treatises that addressed
Homeric subjects and was concerned with Homeric interpretation, mainly
with the Odyssey, but also to a certain extent with the Iliad. He interpreted
the lines on Nestors cup metaphorically. His famous interpretation of the
word (Od. 10, 330) reveals linguistic analysis and an understand-
ing of context.232 Antisthenes dwelt on literature, ethics and politics, but his
primary interest was language. He adopted Socrates view that the process of
225 D.-K. 86 B6; Pl. Ap. 41a, Ion 536b, cf. Ar. Ran. 1030ff., Hermesian. fr. 7, 1640 Powell, see
Pfeiffer [1968] 52; Snell [1976] 486490.
226 Pl. Hp. Mai. 285d; cf. Democr. D.-K. 68 B15c, 16.
227 Ps.-Plut. De Mus. 4, 1132e; 7, 1133f; Ps.-Plut. Vit. X Orat. 833d; see Lanata [1963] 270277.
228 Cf. above on the shift from musical to literate education 2.1.4.
229 Frs. 131148, 178, 190, cf. xxixxxxi Wyss; 107 test. 5 FGrHist. See Phlmann [1994] 21. On the
obscure Homer edition of Euripides, see Pfeiffer [1968] 72, n. 4.
230 Cf. Od. 21, 390 and Antim. Lyd. fr. 57; cf. frs. 3; 53 Wyss; on Antimachus Homeric studies
see Pfeiffer [1968] 9395, Wilson [1969] 369; Matthews [1996] 4651, 373403.
231 Cf. fr. 189 Giannantoni, Sch. Od. 9, 106, see Richardson [2006] 8081 with further
bibliography.
232 Antisth. frs. 185197 Giannantoni; for more, see Apfel [1938] 247; Giannantoni [1990]
331346.
greek scholarship from its beginnings to alexandria 41
definition was fundamental for language and, more generally, for knowledge.
He further argued that a thing could not be represented in language by any
utterance other than its name.233 Following Socrates, Antisthenes examined
questions of ethics, connecting investigations to language (cf.
, the beginning of education is the study of names
fr. 160 Giannantoni).
233 Aristotle was especially interested in Antisthenes views on language, such as Antisth.
frs. 150 (=Arist. Metaph. 1043b), 152 (=Arist. Metaph. 1024b), 153 (=Arist. Top. 104b)
Giannantoni. For Antisthenes on language (frs. 149159 Giannantoni), see Brancacci
[1990] especially 4384, and Prince [2009] 8082.
234 Pl. Menex. 95c; Grg. 459bc.
235 On possible Empedoclean influences on Gorgias use of figures, see Diels [1976], Classen
[1976] 229.
236 The term goes back to Gorgias, cf. Gorg. Hel. 910, 13, 15, Pl. Phdr. 261a78;
271c-272b; Isoc. Nic. 49, see Classen [1976] 226230. See also Schmitter [2000] 359360; de
Jonge-Ophuijsen [2010] 489.
237 Isoc. Nicocl. 59, see Pfeiffer [1968] 4950.
238 For Isocrates educational aims, see above 1.5. On Isocrates praise of Homer, see Apfel
[1938] 245246.
239 Tzetz. Chil. 12, 566; Quint. Inst. 4, 2, 31; 36; 40; 52; 61.
42 novokhatko
A correct name was held to indicate the nature of the thing named (Pl. Cra.
425b), acting as a verbal representation of its referent in syllables and letters
( , Pl. Cra. 433b).245
Platos understanding of language should be approached within the con-
ceptual framework of his philosophy. Plato expanded the relationship between
model and copy to cover language, language being a copy or an imitation of
reality.246 The relationship between names, knowledge and reality is discussed
in Platos Seventh Letter (a significant text for the development of linguistics,
whether it is spurious or, more probably, original). Plato distinguished between
name (), definition () and image (), all of which contribute
to knowledge (). Name is what is uttered ( ); definition
is constituted by nouns and verbs ( ),
image is physical: it is painted and then effaced and honed and then deleted
( ). The
fourth level is knowledge, reason and true opinion regarding these objects
( ). Knowledge is neither vocal
nor physical but is something that exists in souls (
; see Pl. Ep. 7, 342a344d). A name is thus a basic
notion of philosophical analysis for Plato.
These reflections on language are further developed in Platos dialogues
Theaetetus and the Sophist. In both dialogues, thinking () is under-
stood as a conversation of the soul with itself. It involves questioning, answer-
ing, affirming and denying.247 Thus a structural link between language and
thought is affirmed.
In the Theaetetus the definition of knowledge () as true judgment
combined with rational explanation () is discussed (Pl. Tht. 201c202b).
This rational explanation is understood as verbalised thought.248
In Platos dialogues a distinction is formulated between noun ()
and phrase, verb (), a problem which seems to have been discussed in
245 On the approach to language in the dialogue Cratylus with further bibliography, see
Rijlaarsdam [1978], Baxter [1992], Williams [1994], Sedley [2003b], Del Bello [2005] 6682,
de Jonge-Ophuijsen [2010] 491492, and Diehl [2012]. On the multiplicity of names, cf.
Socrates student Euclides of Megara who claimed that the good ( ) is one but can
be given various names ( ) such as intelligence, god, or reason
(D. L. 2, 106).
246 See Sluiter [1997b] 177188.
247 Pl. Tht. 189e190a, Soph. 263e264a.
248 For contemporary research on the relationship of language and thought in the Theaetetus,
see Annas [1982], Denyer [1991] 83127, and Hardy [2001] 267288.
44 novokhatko
249 Pl. Tht. 206cd; cf. also Cra. 421de, 425a, 431bc and Aeschin. In Ctes. 72.
250 Cf. Rehn [1986].
251 On the philosophy of language in the Sophist, see Mojsisch [1986] and Borsche [1991]
152158; on the function and meaning of language mainly in Platos dialogues Cratylus,
Theaetetus, and in the Sophist, see Rehn [1982] and Bostock [1994].
252 See Nagy [1989] 2429 and Maehler [1963].
253 See Platos principal passages on the topic in Murray [1996] 235238.
254 For further discussion of inspired poets in Plato, see Bttner [2011].
greek scholarship from its beginnings to alexandria 45
255 See in detail in Halliwell [2011] 167170. On the Ion foreshadowing Alexandrian scholarship,
see Hunter [2011].
256 Pl. Grg. 501d502d, cf. Gorg. Hel. 8, 10.
257 See the distinction between the relative virtues of speech and writing in Pl. Phdr. 227ae.
258 Halliwell [2011] 155159.
259 For Platos hostility to poetry and the written word, see Pl. Phdr. 275d277a.
260 Ps.-Long. Subl. 13, 34; cf. Richardson [1992a] 3435; on Homeric criticism in Plato, see
Apfel [1938] 247250.
261 Pl. Prt. 338348a; Simon. fr. 542 Page.
262 On this methodology, see Tsitsiridis [2001].
46 novokhatko
opinion of the author, since the poet only imitates words and does not neces-
sarily mean what he says. Thus a crucial distinction is drawn between perfor-
mance per se and the proper understanding of what is performed.263
If in Platos earlier dialogues poetry is understood as theatrical performance,
in the middle and later dialogues a further important function of poetry is
considered: poetry as imitation. The theory of mimesis is explored in the 3rd
and in the 10th books of the Republic.264 The image of the mirror was used for
the description of the relationship between a work of art and nature (Pl. Resp.
596de). This characterisation of art as imitation, which is at a third degree
from reality, led Plato to judge poetry as less than serious (Pl. Resp. 602b). If
myths are to have worth, they should contain some moral truth, a truth only
philosophers can perceive. For Plato, poetry harms its audience through the
empowerment of non-rational parts of the soul, and its status as mimesis pre-
vents it from providing knowledge.265
Plato returned to the question of allegorical interpretation, which had been
discussed since the time of Pherecydes of Syros266 and was to be developed
further by Chrysippus and Crates among many others.267 Plato considered
allegorical interpretations of the Homeric theomachy.268 However, rationalis-
tic allegorical interpretation ran contrary to Platos principle idea that posited
the divine inspiration of poets.269
In the Laws, emphasis is placed on the importance of poetry in the educa-
tional and cultural life of citizens. Only one who is morally worthy should be
permitted to compose poetry.270 A number of critical theories are put forward
in the Laws. The origins of poetry and music are traced back to the effect of a
specifically human sense of rhythm; harmony is traced to the cries and motion
of infants, and the permanent desire to please the crowd spoils artistic pro-
duction and leads to theatrocracy instead of an aristocracy of taste (Pl. Leg.
700a701d). Plato stressed the need to keep the different genres fixed, objecting
to excessive appeals to the emotions (Pl. Leg. 800d).271
271 For more detail on literary criticism in Plato, see Grube [1965] 4665; Else [1986] 364;
Ferrari [1989]; Murray [1996]; Ford [2002] 209226; Halliwell [2011] 155207 with further
current bibliography.
272 See Heath [2009] 252253 with further bibliography.
273 Podlecki [1969] 115117; Wehrli [1983] 523529; Montanari [2012d] 353.
274 Frs. 96106 Schtrumpf. See Heath [2009] 264271.
275 This work perhaps treated the plots of tragedies. Cf. Dicaearchus below 3.10.
276 See Podlecki [1969] 115; Barker [2009]. On the connection of the to the writings
of Glaucus of Rhegium, see Wehrli [19692e] 112. On the links between music and the
metric quantity of poetry, see Pl. Resp. 617d.
277 Wehrli [19692e] 117119.
278 Frs. 99104 Schtrumpf, see Heath [2009] 255263. On the triviality of Heraclides
judgments, see Gottschalk [1980] 136.
48 novokhatko
279 See Heath [2009] 254. On Heraclides as a Philodemus source, see Janko [2000] 134138.
280 Fr. 1 Heinze = Sext. Emp. Adv. Logic. 1, 16.
281 Cf. test. 2 Heinze = D. L. 4, 2, 13; see frs. 10, 11 Heinze; on Xenocrates linguistic studies and
their link to later Stoic programme, see Krmer [1983] 4950.
282 Dion. Hal. Amm. 1, 2.
283 Dion. Hal. Isoc. 18; Ath. 2, 60de; cf. Ath. 3, 122b; 8, 354c.
284 Dion. Hal. Pomp. 1, 16; Num. in Euseb. Praep. evang. 14, 6, 910. Another Isocrates student,
the historian Theopompus, wrote a (Theopomp. 115
test. 7; 48; fr. 259 FGrHist).
285 For more on the methodology of criticism in the Derveni papyrus, see Henry [1986]
and Obbink [2003]; on the vocabulary of criticism in this text, see also Lamedica [1991],
Lamedica [1992].
greek scholarship from its beginnings to alexandria 49
Comic production during the 4th century BC also provides material con-
cerning poetry criticism, self-referentiality being a characteristic of the genre
of comedy. The comedy of this period inherited a range of common topoi
on criticism from Old comedy. However, to a certain extent it was an altered
form of comedy. In the 4th century BC it was the carefully structured plot of
the comedy that was emphasised. The partial differentiation of comedy from
political life should also be noted.286 An important example is Antiphanes
comedy Poiesis, in particular fr. 189 PCG (perhaps from the late 4th century
BC). The speaker, representing comic poetry and playwrights, complains about
the difficulty of writing a comedy as compared to a tragedy; the comic poets
have to look for new names and plots for each play, while the tragic poets write
down myths known to the audience before the actors utter a single word. It is
interesting to note that the poet is discussing the structure and composition of
his play, and uses vocabulary from critical analysis In the lines 1921:
, , , (what hap-
pened before, the present situation, the catastrophe, the opening of the play).
According to a later source, construction of the plot was a central focus for the
poets of Middle comedy.287 This fragment can also be discussed in the context
of contemporary and later criticism of tragedy, and can usefully be examined
in terms of the opposition between comedy and tragedy and the function of
tragedy in the 4th century literary canon.288
As in Old comedy, the interweaving of contemporary discourses on criti-
cism with the work and thoughts of the comic poet himself helped to foster
novelty. Only fragments from Middle comedy survive and thus the plot can-
not be reconstructed; fragment 189 PCG thus remains significant primarily as
a reflection of contemporary discourses inquiring into the role and function of
the comic genre.
297 On the function of language in the treatise On interpretation, see further Arens [1984]
2457; Hennigfeld [1994] 7194; Sedley [1996]; Sluiter [1997b] 188195; Modrak [2001];
Whitaker [1996]; De Rijk [2002] 190357; on Aristotles notion of meaning as discussed in
his Posterior Analytics, see Charles [2000] 23178.
298 Arist. Poet. 1456b1459a.
299 Arist. De an. 419b; Hist. an. 535a2728; 535b.
300 For more on Aristotles theory of language, see Weidemann [1991], Ax [1992], Arens
[2000], de Jonge-Ophuijsen [2010] 492493, and Lapini in this volume.
301 Frs. 142179 Rose and Ps.-Arist. frs. 20a (145), 30a (156), 38 (165) Rose. See Breitenberger
[2006] 369430.
302 See Heath [2009] 255263 on the detailed analysis of Heraclides and Aristotles
observations on Homeric problems. Heath argues that at two points Aristotle responded
52 novokhatko
to Heraclides and at one point they both responded independently to an earlier scholar.
See also Blum [1991] 2123.
303 Fr. 6 Friedlaender, see Apfel [1938] 250252; Heath [2009] 253254.
304 Fr. 175 Rose = Sch. Od. 12, 129, see also Montanari [2012d] 348349. Cf. Aristotles note on
the practice of allegory in Arist. Metaph. 12, 1074b and Montanari [1993b] 260.
305 See Apfel [1938] 253257; Richardson [1992a] 3637; Latacz [2000] 79.
306 The title On Tragedies has also survived. On the tradition of the treatises on poets,
see Janko [2011] 385386. On Aristotles treatise On poets, see Janko [2011] 313407 and
485539.
307 Halliwell [1989] 149.
greek scholarship from its beginnings to alexandria 53
logic, the theory of science and rhetoric, pupils were taught subjects suitable
for the composition and interpretation of poetry.
Aristotle considered contemplation (), action () and mak-
ing or production () to be the fundamental human actions.308 He
regarded poetry not as but rather as a form of (Arist. Poet. 1448b
417); his arguments also embodied the concept that generality is inherent in
the particular and therefore not abstract.
In discussing the historical evolution of poetry, Aristotle built on the already
existing canon while emphasising the history of separate genres.309 In his view,
tragedy and comedy originated in earlier serious and also light poetry:310 thus
Margites had the same relation to comedy as the Iliad and the Odyssey to trag-
edy (Arist. Poet. 1448b371449a2). In his assessment of the concept and
of genre, Aristotle examined the evolution of comedy and tragedy and also the
distinction between epic poetry and tragedy (Arist. Poet. 1449a71449b22). His
judgment of poetry was determined by the character and aims of a particular
genre and form (epic poetry, tragedy, comedy, dithyramb, music for aulos and
lyre, Arist. Poet. 1447a).
Poetry is placed by Aristotle among the visual arts, music and dancing, in
other words, within a more general concept of artistic imitation of life ()
as proposed by Plato and then developed further by Aristotle himself.311 The
first four chapters of the Poetics establish principles for this concept (see also
chapter 25). Imitation, which in Aristotles view came naturally to people from
their childhood onwards, is discussed as poetic , in the sense that a poet
reveals how a persons character is realised in a given situation (Arist. Poet.
1448b).
The poets task is defined by Aristotle as depicting the kinds of things that
could happen and are possible either in accordance with probability or neces-
sity ( , ,
). This is in contrast to the historian, who
depicts events which actually occur (Arist. Poet. 1451a3639).
Aristotle believed that the emotional influence of poetry comes from the
nature of the action described. He distinguished between the actions of supe-
rior characters who follow serious goals, represented in Homer, the actions of
usual characters (similar to us), represented in the works of the minor tragic
poet Cleophon, and the actions of inferior characters, represented by the
composers of parody such as Hegemon of Thasos and Nicochares (Arist. Poet.
1448a117). This constitutes for Aristotle the principle difference between com-
edy and tragedy: tragedy aims to imitate characters superior to those actually
existing, whereas comedy aims to imitate inferior figures (
, Arist. Poet. 1448a1517).
The main section (Arist. Poet. 1449b241456a31) of the Poetics discusses trag-
edy: its essence, plot, structure, characters, emotional impact ( accom-
plished through pity and fear), style and language. Aristotle further discusses
the epic genre (Arist. Poet. 1459a161460b5) and makes a comparative assess-
ment of the two genres, tragedy and epic, (Arist. Poet. 1461b261462b19).312
Aristotles Rhetoric is one of the most important works on Greek rhetori-
cal theory and stylistics (the other two extant theoretical treatises of the 4th
century BC are the anonymous Rhetoric to Alexander and Alcidamas On the
sophists). Theoretical and practical engagement with the language is central
to Aristotles Rhetoric, in particular his discussion of ways to appeal to the audi-
ence and its emotions and his discussion of style (). Aristotle refined and
developed the division of rhetoric into three genres (already found in Gorgias):
deliberative, forensic, and epideictic rhetoric (
, , , ).313 Ethics,
he argued, should bear a relation to rhetoric (countering Platos criticism
that rhetoric was uninterested in truth). Aristotle also provided a systematic
overview of stylistic virtues, his discussion of metaphor in the 3rd book of his
Rhetoric proving to be a significant influence on later stylistic and cognitive
studies.314 Metaphor in Aristotle is considered on the level of learning, its cog-
nitive function being part of the dynamic process of communication.315
Aristotelian principles of analysis were to become a seminal influence for
Alexandrian scholarshipboth in a narrow sense in terms of the scrupulous
editing of classical texts, and also in a broader sense as a methodology for lin-
guistic and literary criticism.316
312 On the main principles of Aristotles literary criticism in the Poetics, see Grube [1965]
7092, Halliwell [1989], Dale [2006], Bernays [2006], Schmitt [2008], and Halliwell [2011]
208265.
313 Arist. Rh. 1358a361358b20; cf. Rhet. ad Alex. 1421b7; Rhet. Her. 1, 2, 2; Quint. Inst. 3, 3, 14.
314 Arist. Rh. 1405a1406b; cf. Arist. Poet. 1457b79.
315 Arist. Rh. 1410b11413b2, see Levin [1982].
316 On Aristotle and his influence on Hellenistic scholarship, see Montanari [1993b] 262
264; Richardson [1994], Montanari [2012d], Montana, Nnlist, and Lapini in this volume.
greek scholarship from its beginnings to alexandria 55
Aristotles influence can be seen in later anecdotes such as the testimonies collected by
Platthy [1968] 124129.
317 Podlecki [1969] 118; Montanari [2012d] 349352.
318 Fr. 666 FHSG.
319 Frs. 708, 709 FHSG, see Fortenbaugh [2005] 351375.
320 Fr. 671 FHSG. Cf. Rhet. Alex. 1421b67; Arist. Rh. 1, 3, 1358b68. In all probability the division
was current in the rhetorical discourse of the 4th century BC. See Fortenbaugh [2005] 171.
321 Frs. 111ae, 112ac, 674 FHSG. On the discussion of early peripatetic syllogistic, see Barnes
[1985] and Fortenbaugh [1998].
322 See Pagani, this volume.
323 See fr. 684 FHSG = Cic. Orat. 79. See Fortenbaugh [2005] 266273. Cf. 3.5 above.
324 Fr. 703 FHSG = Demetr. Eloc. 41; frs. 702, 704 FHSG = Cic. Orat. 218, 192194; see further
Fortenbaugh [2005] 327335.
56 novokhatko
325 Fr. 701 FHSG = Cic. De Or. 3, 186187; Fortenbaugh [2005] 326327.
326 Siebenborn [1976] 90ff.; see also Pl. Cra. 434e4435c1 and Fortenbaugh [2005] 276,
286292.
327 See Innes [1985] 262263.
328 Frs. 32, 33 Wehrli = Ath. 8, 352c, Clem. Al. Strom. 1, 21, 131, 6. See Against the Sophists
(fr. 10 Wehrli); some treatises such as On the Socratics (frs. 3031 Wehrli) seem to be purely
biographical.
329 On the relationship of Eudemus Analytics with Aristotles and Theophrastus work, see
Huby [2002].
330 See Wehrli [1983] 530531; Fortenbaugh [2002], especially 7981.
331 Frs. 9ac Matelli. See also Montana in this volume.
greek scholarship from its beginnings to alexandria 57
four books on the Odyssey,349 a book on the comic poet Antiphanes,350 and
also on prose writers, such as his criticism of Platos style.351 He made a criti-
cal analysis of the Homeric text, discussing the problems of the line Il. 2, 409.352
This is the oldest source to have believed that this line is spurious, an opinion
that would later be supported by some of the Alexandrian grammarians.353 He
also commented on Od. 23, 296, a line which engendered further discussions in
Alexandria.354 Demetrius of Phalerum is a good example of direct continuity
between pre-Alexandrian and Alexandrian scholarship: in effect, he moved to
Alexandria after 297 BC and worked there at least until 283 BC when Ptolemy II
came to power, a period which coincided with Zenodotus in his prime.355
While this is indeed a story of increased literacy, with institutional devel-
opment and thematic differentiation culminating in Alexandrian scholar-
ship, this overall picture of progress should not be viewed as either inevitable
or triumphant. The performative criticism embodied in Old comedy, to cite
but one example, was largely lost en route. By the end of the 4th century BC,
however, institutions supporting the development of scholarship had been
largely established, together with the idea of a scholarly programme having
linguistic studies and textual, stylistic and interpretative analysis at its core.
Methodological, theoretical and practical approaches to philological analysis
had already been developed by the time of philological scholarship began to
flourish in the Alexandrian age.
Hellenistic Scholarship*
Fausto Montana
1 Preliminaries
1.1 Court Poetry and Scholarship in Hellenistic Societies
1.2 Historiographic Pattern
1.3 Scholarship and Knowledge
2 Alexandrian Scholarship to 144 BC
2.1 Traces of Scholarship Outside Alexandria in the Early Hellenistic Age
2.2 Culture and Royal Patronage in Early Ptolemaic Egypt: The Museum
2.3 Making the Universal Library
2.4 Philology for Books, Books for Philology
2.5 Librarians Diadokh and Learned Community
3 The Spread of Scholarship in the 2nd and 1st Centuries
3.1 Rise and Zenith of Pergamene Scholarship (2nd Century)
3.2 Pluralism and Exchange in Late Hellenistic Scholarship (14431)
3.3 Alexandrian Scholars in an Augustan World
1 Preliminaries
song;2 whereas the latter emphasizes the cohesion of poetic culture deriving
from its social pervasiveness, contending that this was sustained by a great
number of public festival competitions, civic or religious ceremonies and pri-
vate symposia.3 However, although appraisal of these issues is currently still
fraught with controversy, a significant portion of the extant Greek poetry of
this period can, by virtue of its characteristics of court literature, be regarded
as the learned expression of fairly narrow elites. These elites were composed
of select social groups living in urban centers of the traditional Greek world
as well as of predominantly ethnic Greek minorities ruling over Hellenized
non-Greek East-Mediterranean areas (Egypt, Asia Minor, Near East) subject to
Macedonian-rooted monocratic dynasties, and eventually came to include the
most educated circles of the Roman aristocracy. It is therefore no cause for
surprise, in terms of social history, that a derived and intellectualized activ-
ity such as the professional study of literaturei.e. philology, which sprang
up towards the beginning of the 3rd centuryfirst arose and long remained
as a minimal niche and ultimately became the self-referential expression of a
Greek elitarian culture, in striking contrast to far less cultured and widely non-
Greek backgrounds.4
Hellenistic scholarship was undoubtedly a most exceptional cradle of ideas
and culture for militant intellectuals and poets of the age (who not infrequently
were scholars in their own right); and, in the long run, such a phenomenon
inevitably had a strong impact on the poetics, reception and transmission of
Greek literary texts. Yet it cannot be overlooked that the highly specialized
2 E.g., with varying emphasis, Hardie [1983] 1536; Zanker [1987] 137; Bing [1988]; Gentili
[1988] 174176; Fantuzzi [1993] and in Fantuzzi-Hunter [2004] 141; Hunter [2003].
3 Cameron [1995], especially 44103; cf. Falkner [2002] 343344; Krevans-Sens [2006] 192194;
Pretagostini [2009]; Acosta-Hughes Stephens [2012]. For the discussion stimulated by
Camerons monograph see Knox [1996]; Griffiths [1997]; Zanker [1997]; Green [1998]; Lehnus
[1999]; Bing [2001]. Although within a reiteration of the dual view, Fantuzzi [2010] argues
for some mutual influence between genuinely ritual-performative and learned-fictionalized
(Callimachean and Theocritean) religious poetry.
4 On the multifaceted society of, for instance, Hellenistic Alexandria and the relations between
its components (Macedonian/Greek citizens, ruling in a foreign land; Egyptian natives;
selected Greek intellectuals, juridically xenoi, constituting the entourage of the court, accord-
ing to the typology suggested by Fraser [1972] 1.6092), see Lewis [1986]; Stephens [2010];
Vandorpe [2010] 171173; Del Corso [2014]. For cultural implications in the times of Ptolemy II:
Stephens [2003]; McKechnie-Guillaume [2008]. The debate on the Ptolemaic policy of inter-
cultural integration has been revived by the discovery of monuments that seem to reinforce
the idea of an Egyptizing attitude of the rulers: Empereur [2004]; Goddio-Claus [2006];
Manning [2009]; Weber [2010b].
62 Montana
in van Nijf-Alston-Williamson [2011]. The peer polity interaction between poleis, leading
to networks of peer communities, is stressed as a major fact of continuity from Archaic to
Hellenistic Age by Ma [2003].
7 Merkelbach [1981] 2930; Bing [1988] 128135; Nagy [1998].
64 Montana
and ideal relevance of the book, scholars interest and proficiency in the metri-
cal and musical features of ancient works.8
The separateness of the elite body of Greek scholars from the heterogeneous
social bodies subject to Hellenistic monarchic powers can also be perceived
by looking at their respective approaches to and reception of literature. One
stream of current debate focuses precisely on the dynamics of musical and
textual reception and transmission in the late 4th and early 3rd century, and
on the closely related question of the actual skills and methods of Hellenistic
scholars in these matters. Many scientists and learned Greek personalities
(very often poets) were selectively summoned from their cities or countries
to join small protected communities under royal patronage, as an organic
intelligentsia and court entourage that would assure ideological cohesiveness.
Such communities were positioned at the very tip of the social pyramid, far
removed from its base, to a greater extent than had ever been the case before
over the prolonged span of Greek history. The goal these scholars cultivated
in their pursuit of literature differed from the quest for enjoyment and enter-
tainment that had traditionally stirred ordinary audiences: rather, this learned
elite was mainly called upon to respond to a demand for ethnic and political
self-identification. In undertaking this task, they devoted great attention firstly
to retrieving, riddling and collecting reliable books in royal institutions, and
secondly to the aim of in-depth understanding and explaining of the actual
text: i.e. reading, unraveling obscurities, emending corruptions, and discussing
the textual surface and the value of ancient written works as such. This com-
mitment resulted in taking a new special care of text trasmission and, on the
other hand, in apparently disregarding or overshadowing aspects concerning
aural reception and performance that had played such a crucial role in the
original conception of these works.9
A similar attitude is mirrored in an anecdote referred to by the 1st century
Latin writer Vitruvius to illustrate the enormous book-oriented culture accu-
mulated roughly a century earlier by Aristophanes of Byzantium, librarian-in-
chief and one of the greatest scholars in Hellenistic Alexandria. The anecdote
also has the virtue of offering a vivid portrayal of the relation that must have
held between learned and popular reception of current poetry. Vitruvius relates
that king Ptolemy (which Ptolemy was involved remains unspecified) desig-
nated Aristophanes together with another six judges to be the assessors in a
competition of poets who were to be called upon to perform their works before
8 For a sketch of changes in poetic communication from 5th to 3rd century see Fantuzzi in
Fantuzzi-Hunter [2004] 1726.
9 Cf. Nagy [1996] 150.
Hellenistic Scholarship 65
a popular audience. While the other six took the publics reaction into account
in forming their judgment, Aristophanes expressed the opposite evaluation
(eum primum renuntiari iussit, qui minime populo placuisset) and unmasked
the fallacy of the criterion adopted by his colleagues. What Aristophanes dem-
onstrated was that precisely the poet who had not been favorably received
either by the audience or by the other judges was actually the only one to have
performed an original poem of his own, whereas the remaining competitors
had recited works that were not their own creation:
I asked moreover whether the expunged lines had been written by him,
and he answered that they were all of his own! Then I understood how
For all these reasons, the actual musical, orchestic and metric interests and
competence of the Hellenistic (Alexandrian) scholars are central subjects of
present-day critical debate. According to a recent reassessment, despite the
contrast within critical opinion between radical skepticism13 versus wide-
ranging trust14 there might still be scope for a middle way: the possibility of
episodic and non-standardized contacts and exchanges between the prag-
matic-performative and the strictly textual tradition of poetic songs (for
instance between Bhnenexemplare and Lesetexte in the theatrical field) that
may have taken place in Alexandrian scholarship from the early 2nd century
onwards, with the aim of restoring their alleged original metrical display or
colometry.15 In this perspective, the features connected with poetic perfor-
mance, although more probably in contemporary adaptations than in its origi-
nal context, seem to have interested the Hellenistic scholars, if anything, with
regard to the contribution such features were expected to give to the textual
constitution and metrical understanding and layout of a verse work. Hence
this line of inquiry likewise leads to an image of scholarship as at least partially
divorced from the concrete expectations of the surrounding society, not unlike
the anecdotal sketch provided by Vitruvius and the fictitious satirical picture
invented by Lucian.
To conclude on this point, although many aspects of continuity between
late Classical and early Hellenistic Age have legitimately been stressed with
regard to the modes of poetic performance and reception, the very different
political backgrounds, the functional distinction between high and low cul-
ture operating in the increasingly multi-ethnic and strongly hierarchical soci-
eties of the Hellenistic kingdoms, and finally, as we will see in greater detail
further on, the growing specialized professionalization of philology within and
in the service of royal institutions seem to be sufficient historical reasons to
12 Lucian, Ver. hist. 2.20: see Nesselrath [2002]. Pedantry translates , lit. cold-
ness of mind. A comparable attitude dismissing philological activities is displayed by the
Latin Stoic philosopher Seneca, Ep. 88.39: see most recently Braswell [2013] 3638.
13 Upheld by von Wilamowitz and taken up again, among others, by Pfeiffer [1968] 181,
Phlmann (most recently [2007]), and Parker [2001].
14 Fleming-Kopff [1992], followed among others by Gentili-Lomiento [2003], especially 711.
15 Prauscello [2003] and [2006], building on some positions maintained e.g. by Dihle [1981]
3738, Falkner [2002], and Fantuzzi in Fantuzzi-Hunter [2004] 2728 n. 101. Subsequent
discussion: Lomiento [2007]; Prauscello [2007]; E. Ch. Kopff in Gentili-Lomiento [2008]
1317; Tessier [2009] and [2010a].
Hellenistic Scholarship 67
confirm the view of court poetry and scholarship of the time as exclusive and
elitarian entities. And while this does not go so far as to imply that poets and
scholars were living and working in an ivory tower, it certainly seems to point
in this direction.16
16 E.g. Fraser [1972], especially 1.305312, with an extensive following, up to Rihll [2010]
410411 (concerning Alexandrian science and technology) and Strootman [2010] 4445.
Against the idea of Hellenistic poetry and scholarship as an ivory tower: Pfeiffer [1968]
9798 (a passage after whichsignificantlyin page 103 one finds the anachronistic
and exaggerated view of the Ptolemaic Library as an open-access institution); cf. Nicolai
[1992] 294296; more vigorously Cameron [1995] 2470, according to whom (29) Modern
critics have simply rationalized this [Victorian] prejudice [of alleged artificiality and lack
of inspiration] against the postclassical, arguing that Hellenistic poets composed for a dif-
ferent audience and in a different way (but, of course, claiming that a work was destined
for a narrow audience and that it was a written composition does not per se involve an
assumption of artificiality).
17 Montanari [1993b] 235259; Dickey [2007]; Porro [2009]; Montanari [2012b]; Dickey and
Dubischar, this volume. On the debated relation between ancient exegesis and Byzantine
scholia: Maehler [1994]; Montana [2011a]; Montana-Porro [2014]. For an outline of the
lexicographical and etymological tradition, from antiquity up to the Byzantine era, see
Alpers [2001].
18 For this documentation see McNamee [2007] and the issues of Commentaria et lexica
Graeca in papyris, Berlin-Boston, in progress. On education in Hellenistic and Roman
Egypt: Morgan [1998]; Cribiore [2001].
68 Montana
24 Rossi [1973] 115 identifies the core of philology as ansia di ricerca (research anxiety); and,
for instance, Russo [2004] 223 claims that the Hellenistic linguistic notions...consti-
tute an important aspect of the scientific revolution and Stoic semantics [opening up
the Hellenistic path of observation and systematic definition of linguistic phenomena] is
none other than an aspect of the same revolution in thought that led also to science.
25 Romano [1993] 377, equating Vitruvius encyclopedic horizon to the polycentrism dis-
tinctive of the Hellenistic culture, effectively describes the latter as la crisi dei grandi
sistemi di sapere in cui ogni scienza o arte riproduce al suo interno, in piccolo, una enci-
clopedia. Cf. Romano [1987] 50; Montanari [1993a] 632635; Bonanno [2000] 211212; this
volume, section III.3.
26 E.g., about scholar historians: Montana [2009c]; cf. Dettori [2000a] 49 with n. 159, 5052.
27 Fraser [1972] 1.313316; cf. Montanari [2012d]; Hatzimichali [2013b]; below, 2.2.
Hellenistic Scholarship 71
the poet and grammarian Philitas seems to have been the most important rep-
resentative, has by no means been dispelled.28 Here poets such as (possibly)
Theocritus of Syracuse and Hermesianax of Colophon also dwelt for a while,
and Theocritus himself as well as Callimachus offer deferential acts of homage
to Philitas in their most important poems.29 Ptolemy I Soter (the Saviour),
the founder of the royal house of Hellenized Egypt who reigned from 305 to
283, chose Philitas as a tutor to his son,30 who was born on the island in 309/8.31
Philitas presumably returned to Cos before the foundation of the Alexandrian
royal Library, but he was also one of the teachers of the Ephesian Zenodotus,32
who was later appointed as the first librarian by Ptolemy. Philitas owes his
place in the history of scholarship to his authorship of a word collection
known as or simply , perhaps non-(alphabetically-)
arranged unusual words, only fragments of which remain, so that the title,
content and nature of this work are still under discussion.33 A special interest
in Homeric glosses as well as dialectal (not only literary) words is undeniable.
It is not possible to state with precision whether and how Philitas conceived
of these two spheres of activity as related; but it is notable that Aristotle in the
Poetics (21.2, 1457b 17), undoubtedly referring back to a more ancient practice,
singled out among the body of (words) the subcategory of ,
unusual words (in a relative sense, diachronic as well as diatopic, in compari-
son to the common use of a defined speaking community) also adopted as a
stylistic feature in poetry (Homeric examples are quoted in 25.6, 1461a 1016),
among which dialectal words constitute a special type. This connection can
ultimately be seen as one of the premises for an interest in dialectology among
34 The rise of systematic study of dialects dates to the early 3rd century (Sosibius Laco) in
the view of Pfeiffer [1968] 202 n. 2. On the relation between literary and spoken languages
in ancient dialectology see Cassio [1993a], [1993b], [2007], and [2008] 57 and 2931. As
for dialectology in the ancient exegesis to Homer, some stimulating observations can be
read in Montanari [2012a].
35 On : Dyck [1987] (with edition of the fragments); cf. Tosi [1994b] 152155.
36 Mygind [1999]; cf. Bringmann [2002].
37 Edition of testimonies and fragments of Eudemus: Wehrli [19692b]. See Mygind [1999]
254 (No. 2); Bodnr-Fortenbaugh [2002].
38 Editions of testimonies and fragments of Hieronymus: Wehrli [19692d] 944; White
[2004]. See Mygind [1999] 255 (No. 7).
39 Editions of testimonies and fragments of Praxiphanes: Wehrli [19692c]; Matelli [2012a],
[2012b]. On his life and works: Mygind [1999] 263 (No. 33); Martano-Matelli-Mirhady
[2012].
Hellenistic Scholarship 73
40 With regard to this aspect of Hieronymus interests and activity: Martano [2004] (on the
inauthenticity of [Hesiod]s Shield); Mirhady [2004] (on Isocrates style); Matelli [2004],
especially 307309.
41 Thus in a debated source (Callim. fr. 460 Pfeiffer = Praxiph. fr. 7 Matelli) that convinces
Cameron [1995] 209213.
42 Clem. Al., Strom. 1.16.79.3: Praxiph. fr. 10 Wehrli = fr. 9A Matelli ,
, . See Matelli [2012b], especially 3140 and 248253.
43 Praxiph. fr. 20 Wehrli = fr. 25 Matelli = CPF 86 2T (Homer); fr. 22ab Wehrli = 28AB Matelli
(Hesiod); fr. 23 Wehrli = fr. 29AC Matelli (Sophocles).
44 On Aristarchean reception of Praxiphanes opinions: Matelli [2009]. Particularly on the
proem of Hesiods Works and Days: Montanari [2009a] 316322; Matelli [2012b] 306315.
On Homeric reverse order: Lundon [1999d]; Nnlist [2009a] 326337, especially 332333;
Matelli [2012b] 294298.
45 Praxiph. fr. 12 Wehrli = fr. 27 Matelli (quoted by Phld., On poems 5.2). There are some dif-
ferences of opinion as to whether Diogenes Laertius was referring to this same work when
(3.8) he cites a text in which Praxiphanes described a discussion by Plato and Isocrates
on poets ( ): Praxiph. fr. 11 Wehrli = fr. 22 Matelli; Vallozza [2012] is in favor
of maintaining the two works distinct. Overall, on literary historiography and criticism
within the Peripatus: Montanari [2012d].
46 Callim. fr. 460 Pfeiffer = Praxiph. fr. 16 Wehrli = fr. 11 Matelli. The name of Praxiphanes
apparently figures among the critics (Telchines) of Callimachus poetry who are listed
in the so-called scholia Florentina to the Aitia (PSI 11.1219, fr. 1, ad Callim., Aitia 1, fr. 1
74 Montana
Pfeiffer = fr. 1 Massimilla): Praxiph. fr. 15 Wehrli = fr. 10 Matelli = CPF 86 4T; cf. Massimilla
[1996] 6263; Manetti-Montanari [1999]; Matelli [2012b] 253259. A certain skepticism
with regard to Praxiphanes involvement in the polemic to which Callimachus refers is
expressed by Lefkowitz [1981]; Cameron [1995] 213, 220 and 376377.
47 Mygind [1999] 271 (No. 65); Di Gregorio [2008].
48 On Hellenistic Rhodes: Berthold [1984]; Rossetti-Liviabella Furiani [1993]; Gabrielsen et
al. [1999]; Bringmann [2002]. For late-Hellenistic Rhodes see also below, 3.2.
49 Pfeiffer [1968] 121 with n. 4. Aratus work is qualified as a diorthsis (i.e., as we shall see
below, a textual revision preliminary to, and performed for, the purposes of a critical
edition) in two ancient Lives: I, 8 Martin ; III, 16 Martin
.
.
50 Sources collected by Platthy [1968] 170173 (Nos. 166173), whose No. 166 = Euph. test.
1 van Groningen = test. 1 Acosta-Hughes Cusset (by Suda 3801):
,
(Euphorion) went to Antiochus the Great, king of Syria, and was put by him at the head
of the local public librarya testimony defined by Acosta-Hughes Cusset [2012] XVI
une des rares attestations qui nous soient restes de patronage de cour concernant les
Hellenistic Scholarship 75
2.2 Culture and Royal Patronage in Early Ptolemaic Egypt: The Museum
The outstanding character of Alexandrian philology during the early Hellenistic
Age was due first and foremost to its origin as the product of a successful large-
scale planned cultural policy and personal patronage by the Ptolemaic (or
Lagid) royal house.54 The latter was formally constituted in 305 by Ptolemy son
of Lagos,55 one of the members of the high Macedonian aristocracy who as
generals fought alongside Alexander the Great in his military campaigns and
who, after the death of the king, battled with one another for imperial power
(and thereafter were called , successors). The initial steps taken by the
first Ptolemies in their cultural enterprise appear to have been closely linked to
56 Selectively: Turner [1962] 140141, 144 and [1968] 106107; Bevan [1968] 124; Momigliano
[1968]; Wilson [1969] 368369; Fraser [1972] 1.314315 and 320; Rossi [1976] 111115; Blum
[1977] 27134; Arrighetti [1987]; Canfora [1993] 1116 and [1999]; Nicolai [1992] 265270;
Montanari [1993b], 259264; Richardson [1994]; Erskine [1995] 3940; Nagy [1996], espe-
cially 187206, and [1998] 189206; Montanari [2000a].
57 Pfeiffer [1968], while not totally rejecting a link (e.g. 103104), rules out the initial
Peripatetic matrix of Alexandrian philology, preferring to stress the role of Philitas.
58 An inclination of Alexander towards the foundation of libraries and translation of books
into Greek is argued by Canfora [1993] 1819.
59 Str. 13.608.
60 Diog. Laert. 5.37: Fortenbaugh-Huby-Sharples-Gutas [1992] 2021, No. 1.
61 Diog. Laert. 5.58: Strato fr. 1 Wehrly = fr. 1 Sharples, reporting the rumor of a fee of 80
talents. On Strato see Fraser [1972] 1.427428; Desclos-Fortenbaugh [2011]. Editions of tes-
timonies and fragments: Wehrli [19692a]; Sharples [2011].
62 Editions of testimonies and fragments: Wehrli [19682]; Stork-Opuijsen-Dorandi [2000]; a
good number of fragments, of mostly historical interest, have been published by F. Jacoby
as FGrHist 228. There remain a few fragments of Homeric scholarship and grammati-
cal subject-matter: Dem. Phal. frr. 190193 and 196 Wehrli = frr. 143147 Stork-Opuijsen-
Dorandi; see Montanari [2000a] and [2012d].
63 Chiefly Str. 9.398: Dem. Phal. fr. 55 Wehrli = fr. 19 Stork-Ophuijsen-Dorandi; Diod. Sic.
20.45.4: Dem. Phal. fr. 50 Wehrli = fr. 30 Stork-Ophuijsen-Dorandi; Ael., VH 3.17: Dem Phal.
fr. 65 Wehrli = fr. 40 Stork-Ophuijsen-Dorandi; Diog. Laert. 5.78 (Hermipp. fr. 69 Wehrli =
78 Montana
under the protection of Eurydice, Cassanders sister and the first wife of
Ptolemy I.64 He is said by ancient sources to have influenced the kings cul-
tural policy, assisting him in the initial constitution of a royal Library and pos-
sibly providing him with the inspiration to found the Museum, or Shrine of
the Muses, a cultural institution apparently moulded upon the Platonic and
Aristotelian schools at Athens. The testimony of the 2nd/3rd century AD writer
Athenaeus suggests that king Ptolemy II Philadelphus (283246) bought from
Neleus of Scepsis, a pupil of Theophrastus, the books constituting the private
library of Aristotle65 and it is those very works that may have made up the orig-
inal fund of books of the Alexandrian Library.66 However, again according to
Strabo, the Aristotelian esoteric (internal) writings, composed for specialized
discussion within the Peripatus, did not become known to the general public
until they were published in the second half of the 1st century.67 The two testi-
monies can be reconciled by assuming that Neleus sold to Ptolemy the books
owned by Aristotle except for the collection of Aristotles own works.68 While
the full historical reliability of these testimonies may be open to doubt, they
do at least document the ancient sensation of a strong connectionindeed a
genetic linkbetween the Peripatus and the Ptolemaic cultural policy.
In fact an Aristotelian imprint appears from the internal structure and the
activities themselves of the Museum. Describing the royal palace-complex in
the Brucheion, the northeastern quarter of Alexandria, Strabo tells us:
The Museum is part of the royal quarter and it has a cloister69 and an
arcade and a large house in which is provided the common meal of the
men of learning who share the Museum. And this community has com-
mon funds, and a priest in charge of the Museum, who was appointed
previously by the kings, but now by Caesar.70
FGrHist 1026 fr. 75): Dem. Phal. fr. 69 Wehrli = fr. 1 Stork-Ophuijsen-Dorandi. On Ptolemys
advances see Fraser [1972] 1.314315. On Demetrius in Alexandria: Williams [1987] 9091.
64 The connection is highlighted by J. D. Morgan apud Nagy [1996] 196 with n. 30 and 198
with n. 38.
65 Ath. 1.3b.
66 Blum [1977] 109134; Canfora [1999]; Tanner [2000].
67 Str. 13.609.
68 We will return in detail to the question later, 3.2.
69 In Greek , a word emphasized by Nagy [1998] 198, by arguing that [t]his physi-
cal feature is also a notional feature metonymically linking the Museum to the Lyceum.
70 Str. 17.794, translated by Fraser [1972] 1.315.
Hellenistic Scholarship 79
71 Str. 13.608609; Diog. Laert. 5.5157. See Fraser [1972] 1.312316; Canfora [1993] 1116. Lynch
[1972] 121123, on the contrary, minimizes the importance of these similarities. On the
long-lasting debate concerning the alleged religious character (thiasoi) and juridical sta-
tus of both of the philosophical schools in classical Athens see Natali [2013] 7890, whose
conclusion is that (86) the principal purpose of the establishment of the philosophical
schools was not the cult of the Muses but was something else, the implementation of the
ideal of the theoretical life as envisaged especially by Aristotle in terms of .
It was not before the 1st century that these schools became institutions juridically self-
standing, according to Maffi [2008]. Luzzatto [2008], especially 151, rather stresses the
infuence exerted on the Alexandrian institutions by the Isocratean pattern of culture
( in classical, still encyclopedic acceptation: see below, n. 76). Unfortunately the
treatise On the Museum at Alexandria by the grammarian of the Augustan Age Aristonicus
is lost.
72 As underscored by Canfora [1993] 15.
73 Fraser [1972] 1.318.
74 The quotation is from Fraser [1972] 1.305. The connection is confirmed by proven
Peripatetic influences on interests, concepts and methods of militant Hellenistic scholar-
ship: see e.g. Meijering [1987]; Richardson [1992a] and [1994]; Montanari [1994a]; Schironi
[2009b]; Cadoni [2010]; Montanari [2012d]; Hunter and Nnlist, this volume.
80 Montana
/
/
in the populous land of Egypt many are they who get fed, / cloistered
bookworms, endlessly arguing / in the bird-coop of the Muses.75
Athenaeus, introducing the quotation of these lines, says that Timon is ridi-
culing philosophers of the Museum because they are like valuable birds
fed in a coop.76 Regardless of whether the metaphor
is interpreted as in a closed birdcage, namely an entity secluded from the
external world or a zoo,77 or as in a bird-coop of a farm where delicacies for
the table are fattened,78 in either case these birds are unfledged, confined to
the nest, unable to nourish themselves, and thus dependent on their parent-
bird.79 Contentious rivalry and inept seclusion seem to be closely combined
in this confined environment. The imagery recalls the quarrelsome ambience
portrayed in some of the prominent works by Callimachus, the great poet
and scholar who likewise worked in the Museum during the first half of the
3rd century.80
It has been argued that Peripatetic influence is not sufficient to completely
explain the origin and specificity of the Alexandrian Museum. Intuitively, the
organicity of the institution with the newborn royal power may suggest that
an interest in culture was exploitable in several ways. On the one hand, in gen-
eral terms the patronage of Hellenistic kings resumed and institutionalized
the inclination of archaic and classical Greek aristocracies towards private
patronage over intellectuals and artists in order to obtain personal prestige.81
On the other, one may perceive in the Lyceum-shaped Museum a response to
the political needs of the first Ptolemy, whose main concern, immediately after
his installation, lay not only in self-legitimation both as a Greek sovereign and
as the true heir of Alexander (the relation of the latter with Aristotle being
75 Tim. Phl., Silli, SH 786 = fr. 12 Di Marco, quoted by Ath. 1.22d, here in the translation by
Fraser [1972] 1.317.
76 The term could well have represented, in its Isocratean wide (encyclopedic)
acceptation, the official denomination and cultural profile of the learned members of the
Alexandrian Museum: Luzzatto [2008] 147154, with scrutiny of sources.
77 Di Marco [1989] 142143.
78 As suspected by Fraser [1972] 2.471 n. 88, and argued by Cameron [1995] 3132; cf. Clayman
[2009] 93.
79 Bing [2001] 7677.
80 Iambi 1 and 13 and the opening lines of the Aitia.
81 Nagy [1998].
Hellenistic Scholarship 81
82 Ellis [1976] 161162, argues that the young Ptolemy son of Lagos was also a pupil of
Aristotle at Mieza in about 342, among the Macedonian royal Pages of Alexander; cf. Ellis
[1994] 4 and 61. About Pages: Heckel [1992] 237298; Strootman [2013] 4546.
83 Erskine [1995].
84 Murray [2008] 24.
85 Fraser [1972], especially 1.307309.
86 We lack any positive evidence about a philological activity of Theocritus of Syracuse at
Alexandria.
87 Erskine [1995] 45. On the thalassocratic policy pursued by the first Ptolemies: Buraselis-
Stefanou-Thompson [2013]. For an instance of poetry mirroring Ptolemies imperial
ambitions see Bing [2005] (Posidippus of Pella).
82 Montana
88 The modern critic literature on the Alexandrian Library has impressively increased.
Selectively: Parsons [1952]; Canfora [1990]; El-Abbadi [19922]; the essays collected in
MacLeod [2000], especially Barnes [2000], and in El-Abbadi Fathallah [2008]; Berti-
Costa [2010].
89 Haikal [2008], who claims (54) that the Ptolemaic royal Library must have been the
equivalent of the pr md3t pr 3 or House of Books of the [Pharaonic] Royal Palace with its
scribes. Cf. Pedersn [1998]; Potts [2000].
90 Dem. Phal. frr. 17, 66, 67, 188, 199, 201, 202 Wehrli = frr. 58A-66 Stork-Ophuijsen-Dorandi.
See especially Fraser [1972] 1.314; Canfora [1993] 1213.
91 Honigman [2003] 8891.
Hellenistic Scholarship 83
92 Ellis [1994] 5556. On the new cult of Sarapis: Tac., Hist. 4.8384; see Fraser [1972] 1.246
249; Stambaugh [1972], especially 613; Pfeiffer [2008]; Bergmann [2010].
93 On the daughter Library: Pfeiffer [1968] 102; Fraser [1972] 1.323324; El-Abbadi [2008].
94 Nagy [1996] 198.
95 Diog. Laert. 5.78 (Hermipp. fr. 69 Wehrli = FGrHist 1026 fr. 75): Dem. Phal. fr. 69 Wehrli = fr.
1 Stork-Ophuijsen-Dorandi. Demetrius may have worked together with the first Ptolemy
when the son of the latter was only associated with the royal power (285283): e.g. Canfora
[1999] 15. Collins [2000] 82114 believes that Demetrius was genuinely the first to hold the
position of head of the Library under the reign of Soter.
96 John Tzetzes, Prolegomena, Prooemium II (XIa II, 32.1617 Koster).
97 John Tzetzes, Prolegomena, Prooemium II (XIa II, 32.911 Koster): Dem. Phal. fr. 58B Stork-
Ophuijsen-Dorandi (cf. fr. 67 Wehrli). The Scholium Plautinum (Vat. Lat. 11469, f. 181r),
84 Montana
constituted by more than one roll98and the books reached the number of
about 700,000 according to the 2nd century AD Latin writer Aulus Gellius.99 In
the same period, again as described by Tzetzes, the Library in the Sarapeum
contained 42,800 rolls.100
It is legitimate to imagine that the making of such a large and remarkable
amount of books, though gradual over time, was planned by the first Ptolemies
as the common goal of a specialized team of learned personalities working on
what must, intuitively, have been at least a three-stage task: firstly, retrieval of
copies; secondly, textual checking and emendation of the material in order to
(re)establish authenticity, general reliability and correctness; and finally, criti-
cal re-editing of texts and drawing up commentaries together with study of the
collected works by successive generations of learned men within the Museum.
The second and third stages of activity will be described in the next section;
here we will address the first stage.
Of interest in this regard is some traditional information, or rather stories,
about the Ptolemies book acquisition strategy, which consisted in import-
ing, copying, and translation. Some sources, perhaps reflecting Ptolemaic
propaganda itself, depict this process as a series of anecdotes about the
kings bulimia towards books, with particular reference to the second and the
third Ptolemy. Their predilection for conspicuously flaunting the acquisition
of available books could be interpreted as an imperialist attitude.101 Many
rolls were regularly purchased at the renowned book markets of Athens and
Rhodes102 and one source attests that Philadelphus launched an impressive
call for books to all the kings and rulers of the earth in order to obtain texts of
every genre.103 Other books were acquired through what might be termed (in a
modern perspective) more questionable ways of appropriation, for example by
commandeering all books coming into the town harbors by ship from abroad,
and then returning copies to the legitimate owners instead of the originals.
a humanistic Latin annotation derived from Tzetzes, quotes Callimachus as the source
for these numbers, but wrongly, as it seems: Parsons [1952] 108112; Koster [1961]; Pfeiffer
[1968] 48 n. 19, 175 n. 86, 184, 213214.
98 Usually the adjective is taken here to mean containing several works (e.g.
Lloyd-Jones [1990] 27) or compound (e.g. Turner [1968] 102). In the opinion of Canfora
[1993] 24 it designated the roll che, insieme con altri, concorre a formare ununica opera
[which, together with others, contributes to forming a single work].
99 Gell. 7.17.3.
100 Extreme skepticism with regard to these figures is expressed by Bagnall [2002].
101 Erskine [1995] 45.
102 Ath. 1.3b.
103 Epiphan. Schol., De mensuris et ponderibus, PG 43.252.
Hellenistic Scholarship 85
These rolls formed a fund called , from ships. A similar trick is attrib-
uted to Ptolemy III, who is said to have borrowed some precious rolls from
the city of Athens, after depositing a sizeable security of 15 silver talents for
the loan: these were none other than the rolls that had been created at the
behest of Lycurgus during the political leadership of the latter (338326) in
order to establish an official and authorized text of the tragedies of Aeschylus,
Sophocles and Euripides. Ptolemy then apparently commissioned a costly
copy of the books and sent these new copies back to Athens instead of the
originals, forsaking the deposit.104 The anecdote is a telling revelation of the
value attributed to the Athenian exemplars in the eyes of Ptolemy, who evi-
dently was well aware of the process of text corruption that frequently affected
copies, and it is representative of the so-called Alexandrian ideology, eager to
possess canonical texts of ancient literature(s).105
Another important chapter in the story of early Ptolemaic voracity for
bookswhether the tradition is reliable or whether it is no more than a con-
structed mythis represented by translations into Greek of relevant works
written in different languages, such as those forming part of the Chaldaic,
Egyptian, and Roman heritage;106 a translation of the Zoroastrian corpus
is also attested to by Pliny the Elder.107 To this end, many expert men are
said to have been engaged by the Ptolemies, aware of their language as well
as of the Greek one.108 Obviously, translation into Greek can be seen as part
and parcel of the Hellenization (in the sense of the symbolic appropriation
and subduing) of foreign cultures.109 Furthermore, the philological relevance
of translation is also worth noting: the process of transposition from one
104 Both of the stories on books from ships and the Athenian rolls of the tragic poets can be
read in Gal., In Hippocratis librum III epidemiarum 2.4, 17/1.606607 Khn: Aeschylus test.
146 Radt = Sophocles test. 157 Radt = Euripides test. 219 Kannicht. See Wenkebach [1936]
7980; Platthy [1968] 118119; Fraser [1972] 1.480481; Battezzato [2003] 1922; Scodel
[2007].
105 Nagy [1996] 201205, who sees the seminal act of this Ptolemaic project as residing in the
acquisition of the alleged Aristotelian edition of the Iliad owned by Alexander (Plut.,
Alex. 8.2); cf. Honigman [2003] 4344. A skeptical opinion in this respect is maintained by
Sanz Morales [1994] 2239.
106 Georgius Syncellus, Chronographical selection 516 Dindorf.
107 H N 30.24. Callimachus pupil Hermippus commented on the (translated) Zoroastrian
corpus and provided it with indexes: FGrHist 1026 fr. 57, with the comment by Bollanse
[1999b], especially 440441.
108 John Tzetzes, Prolegomena, Prooemium II (XIa II, 33.1 Koster).
109 Erskine [1995] 43; cf. Canfora [1993] 21; Gruen [2006]; and more generally, on the attitude
of Greeks towards foreign cultures, Momigliano [1975]. On Hellenization as a translation
86 Montana
of Egyptian patterns and a means for their appropriation by the Alexandrian court and
elite see Koenen [1993].
110 Thus the most recent comprehensive study on the topic, Honigman [2003], especially
119143. On the dating of the Book see, therein, 128130. On the extensive success and
tradition of this work: Canfora [1996].
111 On the interaction between Jewish Biblical interpretation and Hellenistic scholarship:
Siegert [1996]; Niehoff [2011]. On the Jewish community of Hellenistic Alexandria and
relevant ancient representations: Fraser [1972] 1.5458; Gruen [1998], [2003], [2006] and
[2010]; Kovelman [2005].
Hellenistic Scholarship 87
Grafted into his narrative are several lengthy digressions shaped around
the plot of the biblical Exodus and regarding the history of Judaism and the
Graeco-Egyptian Jewish community up to the liberation from slavery by
Philadelphus.112 As an antecedent of the embassy, reference is made near the
opening of the work to a conversational exchangewhose historicity is in fact
mostly rejected as quite improbablebetween Ptolemy II and Demetrius of
Phalerum. Let us read the actual account given by the ancient writer.113
[9] When Demetrius of Phalerum was made head of the kings library, he
was furnished with large sums of money to collect, if possible, all the
books in the world. He started buying (them) and having (them) tran-
scribed, and he brought the kings project to completion, as far as lay in
his power. [10] In fact, when asked in our presence just how many tens of
thousands of books there were, he said: More than twenty, sire. Within a
short time, I will fill up the remainder so as to bring the total up to 500,000.
It is reported to me that the law books of the Jews too deserve to be tran-
scribed and included in your library. [11] Well, then, (the king) said,
what is keeping you from doing that? For everything you need has been
put at your disposal. Demetrius said: A translation is needed. For in the
Jews country they use their own special characters, just as the Egyptians
(use their own) writing system: accordingly they also have their own spe-
cial spoken language. They are supposed to use the Syrian language, but
that is not true; (their language is a) different type.
Invited by the king to show some suggestions for transcription of the books of
the Jewish Law, Demetrius drew up a memorandum in which he noted that
[30]...these are put in Hebrew characters and language, and have been
recorded in written signs114 rather carelessly and not as well as is possible,
as is reported by the experts. For they have not received a kings provident
care. [31] It is fitting that these books too be available to you, in an
112 The relevance of this second theme is underscored by Kovelman [2005] 131.
113 Book of Aristeas 911, 3031, 3839, 301303, here in the translation by Stork-Ophuijsen-
Dorandi [2000] 112117 (their Dem. Phal. fr. 59), except for 3839 (not included in that
edition).
114 More precisely have been transcribed (), with reference to Hebrew copies
made on the basis of the original or official Torah: Zuntz [1959] ([1972] 133135); Honigman
[2003] 48.
88 Montana
accurately established text, because this code of laws is both quite philo-
sophical and uncontaminated, being as it is, so to speak, of divine origin.
Eleazar agreed and sent the books together with the 72 selected elders, who,
as can be read near the final part of the Book, were taken to the small island
of Pharos off the coast of Alexandria. There, working with great zeal, they
achieved their task in 72 days under Demetrius supervision.
[301] Three days later Demetrius took them along with him, passed along
the seven stades dam in the sea, which led to the island, crossed the
bridge, and proceeded to the northern part, where he established work-
ing sessions in a house prepared for that purpose near the beach, excel-
lently furnished and located in a very quiet spot. There he invited the
men to accomplish the translations, anything they might possibly need
for their work being at their command. And they accomplished each (of
the translations), achieving agreement among themselves through dis-
cussion. [302] The (text which was) produced through agreement was
thus written out in a fitting manner under the direction of Demetrius.
[303] The sessions lasted until the ninth hour; after that they broke up to
take care of their bodily needs.116
An impressive critical debate has arisen concerning the ideal and religious
implications that underlie LXX and also concerning the genre, aim, audi-
ence, and historical reliability of the Book of Aristeas.117 Whatever the answer
to these questions, as far as our specific topic is concerned the Book provides
non-negligible evidence on (a later perception or recasting of) the Ptolemaic
Hellenocentric interestextending between the 3rd and mid-2nd century
in books representative of non-Greek cultures as well as in related philologi-
cal translations apparently carried out with the same accuracy and skill that
habitually characterized editions of texts.118
The life of both the Library and Museum, which continued throughout the
Hellenistic Age, is from our perspective identified with the body of knowledge
on the work of many personalities of Alexandrian scholarshipwith which
we will be concerned later.
The post-Hellenistic history of the Library and the circumstances of its
end are still under debate. According to ancient sources it was accidentally
destroyed, but more probably only diminished, by the torching of the Egyptian
fleet anchored in the Eastern Harbor of Alexandria, when the blaze spread to
the shore in the days of the Alexandrian War fought by Julius Caesar in his pur-
suit of Pompey (48/47).119 However, the scholars operating in Alexandria dur-
ing the Augustan Age such as Didymus, Theon, and Tryphon, must still have
been able to avail themselves of a fairly extensive repository of books in the
city for their studies;120 and when Strabo, who was visiting Alexandria in about
the year 25, briefly describes the site of the Museum (though without mention-
ing the Library), he makes no reference to a relatively recent fire or destruction
within the Brucheion. Therefore, it is widely agreed that the Library substan-
tially survived at least until the days of the emperor Aurelianus, who in 273 AD
attacked Alexandria where he aimed to defeat Firmus, an ally of Zenobia the
lation of the Jewish Bible through seventy-two Hebrew interpreters expert in both
languages.
117 Honigman [2003] 105118. In the steps of Erskine [1995], she argues (117) that the acquisi-
tion of the LXX and its incorporation into the Library could have served political purposes
in the 3rd century dispute between the Ptolemies and the Seleucids for the control over
(Judaea in) Coele-Syria. In her opinion, the 2nd century author of the Book is chiefly con-
cerned with the (good) quality and consequent reliability of the Greek translation of the
LXX, in order to build up a charter myth, i.e. an apologetic validation for it through a nar-
rative. Collins [2000], on the other hand, attempts a reappraisal of the overall historicity
of the Book; cf. Niehoff [2011].
118 Cf. 3031 of the Book, quoted above; see Zuntz [1959]; Honigman [2003] 4448.
119 Cherf [2008]; Bbler [2010].
120 Fraser [1972] 1.334335. Cf. Hatzimichali [2013a].
90 Montana
121 Canfora [1990]; cf. Lloyd-Jones [1990] 29; Ellis [1994] 5657; Empereur [2008].
122 Lewis [2008]; Quassem [2008]. On the contrary, Mojsov [2010] maintains the old view
that an active and crucial role in bringing about the final downfall of the Library was
played by religious fanaticism, first Christian and then Arab, between the 5th and 7th
century AD.
123 Majcherek [2008]. On the destruction of the daughter library in 391 AD, as an out-
come of the attack on the Sarapeum by the mob instigated by the Bishop Theophilus,
see El-Abbadi [2008]. The notion of the Alexandrian Library has been and is usually
exploited, and abused, as the standard myth of both universal written culture and book
burning: e.g. Polastron [2004]; Raven [2004].
124 Gal., In Hippocratis de natura hominis 15.105 Khn
,
, in fact, before the Alexandrian and Pergamene souverains
launched into the race for book acquisition, there existed no falsely attributed works.
Similar statements were made with regard to the corpus of Aristotles writings by various
late antique commentators, quoted by Fraser [1972] 2.481482 n. 151. On Echtheitskritik in
ancient philology see Bhler [1977] 4953.
Hellenistic Scholarship 91
the 2nd century (Callistratus of Alexandria, Crates of Mallos at Pergamum, and especially
Didymus in the 1st century).
129 On the aesthetic and critical category of (the convenient): Pohlenz [1933];
van der Valk [19631964] 2.1135; Schenkeveld [1970] (with subsequent assessments by
Lundon [1998], [1999a], and [1999b]); Nickau [1977] 183229.
130 Quoted by the 3rd century Porphyry, Quaestiones Homericae 1.11 (56.34 Sodano). On the
debate concerning the (Aristarchean?) authorship of the maxim, begun with Pfeiffer
[1968] 225227 and Wilson [1971], see Porter [1992] 7085; Montanari [1997o] 285286.
131 Porter [1992], especially 7071 and 7475.
132 Montanari [2004].
Hellenistic Scholarship 93
133 On the editorial features of the hypomnmata: Del Fabbro [1979]; Luppe [2002]; Messeri
Savorelli-Pintaudi [2002]; Schironi [2012a]. On the Alexandrian smeia, an important tes-
timony of which is the so-called Anecdoton Parisinum (ms. Paris, Bibliothque Nationale
de France, lat. 7530), see Ludwich [18841885] 1.1922; Gudeman [1922b]; McNamee [1992].
134 Pfeiffer [1968] 146 with n. 2. For definitions of hypomnma and syngramma see also
Dubischar in this volume.
135 Pfeiffer [1968] 198.
136 Such an interest is well attested, alongside and in connection with the work on literary
languages, in the late Hellenistic and Imperial Age: Pfeiffer [1968] 202; Cassio [1993b],
especially 81 with n. 24.
137 E.g., especially on Homer, Richardson [1975], [1992a] and [2006]; West [2001a] 332;
Cassio [2002]; Novokhatko, this volume.
94 Montana
many of the learned persons, not only physicians but also grammarians,
devoted their attention to explaining (the works of) this man and to
translating his words into a more common linguistic usage.
138 Nicolai [1992] 271275; Irigoin [1994] 50, 54, and 88 (discussion with D. M. Schenkeveld).
139 P.Amh. 2.12 (2nd century AD): we will return to this when treating Aristarchus. As regards
papyrus evidence of exegesis on Greek prose authors: McNamee [2007] 117125, Dickey in
this volume.
140 Schol. Soph., Phil. 201 (357 Papageorgius) and Suda 3753: Hellanic. fr. 5 Montanari, con-
cerning Hdt. 2.171.2; cf. F. Montanari [1988] 52.
141 Dion. Thrax, Tekhn grammatik 1 (cf. Sext. Emp., Math. 1.57).
142 Apart from fragments (Schmidt [1854]), we have the remnants of a 2nd century AD
papyrus scroll, P.Berol. 9780 (Mertens-Pack3 339), which contain parts of Didymus On
Demosthenes concerning Dem. 911 and 13: a work based on a pre-Didymean Alexandrian
edition of the Attic orator, in the view of Luzzatto [2011].
143 Editions: Schne [1896]; Kollesch-Kudlien [1965].
96 Montana
likewise, according to Dion. Hal., Pomp. 3, Herodotus composed his work in a varied man-
ner, inasmuch as he was a fan of Homer (
). See Pfeiffer [1968] 224; Boedeker [2002]; overall, on Herodotus recep-
tion in Hellenistic Age: Murray [1972]; Priestley [2014].
148 Gal., Gloss. Hippoc. explicatio 19.65 Khn. See Irigoin [1994] 93 (discussion with R. Tosi).
149 Gal., De indolentia 23b Boudon-Millot Jouanna.
150 Rosn [1962] 231; Pfeiffer [1968] 224 with n. 6; Montana [2009c] 166170. Nicolai [1992]
186197 and 265275 emphasizes the continuity of an interest in prose literature from the
Peripatus to the early Alexandrian scholars.
151 Quint., Inst. 10.1.54 Apollonius (scil. Rhodius) in ordinem a grammaticis datum non venit,
quia Aristarchus atque Aristophanes, poetarum iudices, neminem sui temporis in numerum
redegerunt.
152 Respectively P.Louvre inv. 7733 verso (firstly edited by Lasserre [1975]; SH 983984) and
P.Lille 76d+78abc+82+84+111c (firstly edited by Meillier [1976]; SH 254265 = Callim.
frr. 148, 151, 150, 152, 143, 153 Massimilla). On both as testimonies of early Hellenistic exege-
sis on contemporary poets see Montanari [2002c] 7477.
98 Montana
was not simply exploitable for purposes of explanation, and thus as an exegetic
tool involving only a second-degree interest in the quoted author, but whether,
on the contrary, it presupposed knowledge and an investigative approach
of a philological type also intrinsically pertaining to the work adduced as a
parallel.153 The Hellenistic scholarly reception of Menanders plays provides
another small series of good examples. Aristophanes of Byzantiums avowed
admiration for the theatrical mimsis of Menander is one among other con-
crete clues suggesting that he took a philological interest in this author.154 Two
generations later, in the 2nd century, the multiskilled Apollodorus, Aristarchus
pupil, included biographical notes on Menander in his Chronicle, also stating
the total number of Menanders works: the incorporation of this type of infor-
mation is representative of the typically Alexandrian interest in ascertaining
the authenticity and compiling catalogues of literary works.155 In addition, we
have some evidence concerning a commentary on Menanders Kolx com-
posed by the 2nd/1st century Rhodian Timachidaswho was also the author
of a commentary on the poem by Eratosthenes entitled Herms (we will return
to this later). Lastly, an analogous perspective of a critical-exegetical approach
on quasi-contemporary poetry is detectable with regard to an astronomical
work in three books on the Phenomena of Aratus and on his scientific source
Eudoxus of Cnidus, that was composed around the mid-2nd century by the
geographer Hipparchus of Nicaea and which has come down to the present
day intact through the direct Medieval tradition. This work was intended to
emend the conceptual errors of Aratus poem, given the wide-ranging acclaim
and diffusion of the latter, and to refute some favourable interpretations of
the Phenomena advanced by an earlier commentator, Attalus of Rhodes: this
decidedly scientific and pragmatic character of Hipparchus writing has been
the real reason of its survival.156
Thus we have reached the third stage in the scholarly task required by the
Museum and the Library: interpretation. From the time of the foundation of
these Ptolemaic institutions up to the middle of the 2nd century, Alexandrian
153 Montanari [1995a], raising the question; Rengakos [2000], answering essentially in a
negative way; Montanari [2002c], relaunching his thesis especially as far as Callimachus,
Aratus, and (in the footsteps of Fantuzzi [2000]) Apollonius Rhodius are concerned.
154 Montana [2007].
155 Apollod., FGrHist 244 fr. 43.
156 Edition of Hipparchus writing: Manitius [1894]; see Fraser [1972] 1.422423. More infor-
mation on this subject by Luiselli, this volume. In the same 2nd century, an interest in
Aratus poem is documented in the works of the Pergamene scholars Crates of Mallos (in
the fragments of his Homeric exegesis 50, 65, 131133 Broggiato) and Zenodotus of Mallos
(frr. 56* Broggiato): on both see below, 3.1 and 3.2 respectively.
Hellenistic Scholarship 99
157 West [1967]; cf. Pfeiffer [1968] 109110 and 215; Nagy [1996] 187206; Haslam [1997] 6469
and 8487; West [19982000] 1.VII.
158 On the divergent general evaluation by modern critics see Montanari [2004] and [2009b]
160 n. 32.
100 Montana
gives this by no means irreprehensible list at the end of its damaged first col-
umn and in its second column (ll. 121):
[]||[] | [] |[]
| | 5
|
[] , | |
| |10 <> |
| |
| [] [] [ ] | 15 |
|[] | [] |[]
[] |20[] [] |[].
164 Testimonies and fragments in TrGF I, respectively 100 and 101 Snell. On Alexander in
Philadelphus service: Magnelli [1999] 1011; cf. Montanari [2009c] 412. On these figures as
both poets and scholars: Lowe [2013].
165 It is wrong to intepret in the passage of Tzetzes as to put in right order, as some
have done in the wake of the scholium Plautinum (above, n. 97), which recites poeticos
libros in unum collegerunt et in ordinem redegerunt: Pfeiffer [1968] 106107.
166 Editions of the fragments: Strecker [1884] 26 and 2378 (all known fragments, includ-
ing many hypothetical or dubious); Rutherford [1905] 417; Bagordo [1998] 3536 and 150
(No. 63; only three fragments of sure attribution).
167 Lyc. fr. 85 Strecker = 63 fr. 3 Bagordo (cf. Pherecrates, fr. 101 Kassel-Austin).
168 Lyc. fr. 13 Strecker = 63 fr. 1 Bagordo: Antiphanes, test. 8 Kassel-Austin. The threefold
division arkhai, mes and ne is traceable back to Callimachus and Aristophanes of
Byzantium (Nesselrath [1990] 28187, especially 186187; cf. Sidwell [2000] 255256),
rather than to Aristotle (as in the opinion of Janko [1984] 247250, based on Tractatus
Coislinianus de comoedia 18: contra Nesselrath [1990] 147149; cf. Halliwell [1987] 87 n. 2
and [20003] 273274; Preler [1999] 161162 n. 618).
Hellenistic Scholarship 103
a poet and grammarian, as his teacher was before him.169 Zenodotus also fol-
lowed in the steps of Philitas by succeeding the latter as the tutor of the royal
family and compiling a collection of Glssai, which however, in contrast to that
of Philitas, appears to have been alphabetically ordered: this introduced a cri-
terion which, though apparently achieving some immediate success,170 would
be widely adopted in later lexicography.171 One cannot exclude that another
collection known by the title of (Dialectical words)172 attrib-
uted to a Zenodotus, hypothetically identifiable with the Ephesian, actually
coincided with a special section of the Glssai, with which, unlike Philitas, he
sought to build up a body of dialectal lexicography considered as a field in its
own right.173
Only a few fragments survive from his diorthseis of Hesiod and Pindar, and
perhaps Anacreon,174 while the Medieval scholia to Homer allow us to gain
greater insight into his activity on the Iliad and Odyssey.175 The interventions
that have come down to us illustrate quite clearly the practice of emending the
Homeric text, a practice which, one may surmise, was ineluctable when the edi-
tor found himself facing the challenge of the great amount of copies that had
made their way into the Alexandrian Library, often with fairly divergent texts.
The papyri prior to the mid-2nd century confirm the instability of the text
particularly with regard to the number of linesthat the first Alexandrian
editor found himself having to deal with.176 To address this difficulty, the philo-
logical method devised by Zenodotus on the one hand plausibly availed itself
169 He is described as , writer of verse, by Suda 74 (cf. SH 853), but this statement
which could be an autoschediasm moulded on the profile of his teacher Philitasis gen-
erally regarded as suspicious by modern critics.
170 Neoptolemus of Parion in Mysia, known as a poet and glossographer (,
test. 1 and fr. 12a Mette) possibly living in Zenodotus days, composed a work On Homers
glosses ( ) in no less than three books, that seems to have been an
alphabetical glossary containing explanations based on etymology: Mette [1980] 14 and
2122; cf. Cassio [19871988]. Parts of an alphabetical lexicon are preserved in P.Hibeh 175,
datable to ca. 260240.
171 Zenod. fr. 1 Pusch (schol. Hom., Od. 3.444b1 Pontani). Cf. Pfeiffer [1968] 115 with n. 2, 195;
Tosi [1994b] 151. In the view of Latte [1925] 154 and 162171, the collection of words in
question should not be taken as belonging to this Zenodotus, but to a younger namesake;
cf. Blum [1977] 166167.
172 Pusch [1890] 175.
173 Nickau [1972a] 4043; Tosi [1994b] 152.
174 Pfeiffer [1968] 198200; Nickau [1972a] 3839. For Hesiod: Montanari [2009a] 332335.
175 Dntzer [1848]; van Thiel [2014] (for the Iliad).
176 Cf. above, n. 157.
104 Montana
177 Strong skepticism in this respect and a radical disparagement of the Zenodotean
diorthsis are put forward by West [2001a] 3345 and [2002].
178 , , , , : cf. Nickau [1977] 130.
179 Cf. Pfeiffer [1968] 115.
180 Pfeiffer [1968] 110; Nickau [1972a] 3031; Nagy [1996] 119152; West [2001a] 3345 and
[2002] (Zenodotus working copy was a 4th century Ephesian rhapsodes text); Montanari
[2009b] 143154, [forthcoming] and this volume. That Zenodotus ekdosis consisted, on
the contrary, in a new copy of the poems, in which the chosen variants were incorporated
in the text, while doubtful lines were marked with the obelos and lines held to be false
were eliminated tout court, is the opinion of Haslam [1997] 73; Rengakos [2012] 248251.
181 As in the opinion of van Thiel [1992], [1997] and [2014] 1.8.
182 Nickau [1977]; cf. Montanari [1998a], [1998d], [2002a], and [2009b]; Rengakos [2012] 251.
Hellenistic Scholarship 105
arguments for or against some readings present in his edition were laid out in
other writings such as, for example, the Glssai.183 Furthermore, we know that
he took care to express his own interpretations in purpose-composed mono-
graphic writings: for instance, an epigraphic source seems to attest a work of
Zenodotus on the problem of the actual number of days of the Trojan War
involved in the narrative of the Iliad.184 This notwithstanding, reconstruction
of Zenodotus interventions and arguments already constituted a serious prob-
lem even in antiquity, if the 2nd century Alexandrian grammarian Ptolemy
nicknamed Epithetes (, i.e. the Opponent, for the critical stance he
adopted against Aristarchus)185 devoted an entire work to reconstructing and,
it seems, defending the readings of Zenodotus edition (
) so strongly opposed by Aristarchus.186 Ultimately, we must inter-
pret this apparent lack of a genuine comment as a silent clue pointing to the
still largely oral character of this ecdotic/exegetic practice and of the related
transmission within the Museum.187
The notable diffusion and widespread acclaim of Zenodotus editions
become clear if one reflects that they were taken into consideration by the epic
poet Rhianus of Crete in performing his own diorthseis of the Homeric poems
in the second half of the 3rd century,188 and that they exerted some influence
on the poetic works of Callimachus and Apollonius of Rhodes. Moreover,
the fact that a monograph by Apollonius called into question some editorial
choices made by Zenodotus is an incontrovertible demonstration of the high
regard enjoyed by this ekdosis among the Alexandrian poets and scholars.189 In
183 Nickau [1972a] 3940; Montanari [1993b] 266; Tosi [1994b] 151.
184 IG 14.1290 (Tabula Iliaca); cf. Lachmann [18652]; Pfeiffer [1968] 116117.
185 Ptol. Epith. test. 1 Montanari.
186 Ptol. Epith. test. 2 = fr. 1 Montanari, with F. Montanari [1988] 8385. Ptolemys Homeric
scholarship included a work On the Iliad ( : test. 3 and frr. 25 Montanari) and
a hypomnma on the Odyssey (test. 1 Montanari, cf. Pontani [2005b] 5455).
187 Montanari [2009b] 154. A similar problem arises for Aristophanes of Byzantium, who
composed numerous and important ekdoseis but, as far as we know, no hypomnmata: it
can plausibly be suggested that the explanations of Aristophanes ecdotic choices were
gathered together (and discussed) in the hypomnmata written by his disciple Callistratus
(see below, 3.2).
188 More than forty textual readings from Rhianus editions are attested to in the scholia to
Homer: edition by Leurini [2007]; see La Roche [1866]; Mayhoff [1870]; Aly [1914]; van der
Valk [1949] 107108; West [2001a] 5658; Esposto [2008]. On Rhianus poet and scholar
( according to Suda 158): Castelli [1994], with further bibliography.
189 Pfeiffer [1968] 139140 and 146148; Rengakos [1993] 4987 and 169170, [1994], [2001],
and [2002a]; Montanari [1995a] and [2002c] 5964.
106 Montana
190 Diog. Laert. 9.113 (= Tim. Phl. test. 1 Di Marco). See Pfeiffer [1968] 98; Fraser [1972]
1.450 with 2.650 n. 22. One could legitimately wonder whether this anecdote is obliquely
lampooning the diorthseis of the Homeric poems performed by Aratus himself (see
above, n. 49).
191 Clayman [2009] 213214, at the same time (107108) claiming that Timon was aware of the
Zenodotean philology from a comparison of Tim. Phl., Silli, SH 804 = fr. 30 Di Marco with
its Homeric model Il. 3.150152: in l. 152 Timon seems to have preferred the Zenodotean
and more unusual reading against , which later on the contrary was chosen
by Aristarchus.
192 Suda 3035 attests to the chain teacher-pupil Zenodotus-Agathocles-Hellanicus-Ptolemy
Epithetes.
193 FHG IV 288 (n.); FGrHist 3b Kommentar 372; cf. F. Montanari [1988] 1519. Editions: FGrHist
472; F. Montanari [1988] 2630 (text) and 3142 (comment).
194 Schol. Ap. Rhod. 4.761765b: Agathocl. fr. 8 Jacoby = test. 4 and fr. 8 Montanari.
195 Agathocl. fr. 10 Jacoby = fr. 10 Montanari.
Hellenistic Scholarship 107
196 Agathocl. frr. 9 and 11 Jacoby = frr. 9 and 11 Montanari. Montanari argues that the three
fragments of Homeric scholarship belonged to the ; furthermore he publishes,
albeit with reservation on the authorship, a possible new fragment identified by H. J.
Mette (Fragmentum dubium). On Crates see below, 3.1.
197 Montana [2006b] 208 and [2009a] 177178.
198 See above, n. 97. The assumption that Callimachus held the post of librarian is defended
by Blum [1977] 177191.
199 Suda 227.
200 The fragments are edited by Pfeiffer [19491953] 1.344349 (frr. 429453). Most detailed
study: Blum [1977]; cf. Pfeiffer [19491953] 1.349 and [1968] 127131; Fraser [1972] 1.452
453. With regard to the influence of the pinacographical (Callimachean) criteria on the
standardization of book titles in antiquity: Caroli [2007] 6179.
201 Canfora [1993] 24.
108 Montana
authenticity202 or problems with titles203 were addressed. One can assume, for
instance, that a Callimachean epigram referring to the poetic value of the epic
poem The capture of Oechalia disputably attributed to Creophylus of Samos,
and so implicitly decreeing its admissibility into the Library, could be asso-
ciated to this activity of critical cataloguing.204 Furthermore, Callimachus is
known to have written at least two other Pinakes on more sectorial fields, the
Table and inventory of the dramatic poets in chronological order and from the
beginning,205 for which the author is likely to have drawn on the Aristotelian
Didaskaliai, and the Table of glosses and writings (?) of Democritus, probably
concerning lexical innovations introduced by the philosopher.206 A specific
interest in lexicology and the inclination towards a classifying method may per-
haps have been a feature of a collection of Dialectical nouns ( ),
apparently shaped as an onomastikon, i.e. a writing that, unlike Zenodotus
Glssai, placed in succession lists of words not ordered alphabetically, but
grouped according to a semantic affinity or kinship. At least the presence of a
section concerning names of fish is adequately attested to.207 As already men-
tioned, the work To or Against Praxiphanes criticized the aesthetic theories of
this pupil of Theophrastus208 and must have contained appreciations in the
sphere of literary criticism, since Callimachus expressed within it a favorable
judgment towards the poetic quality and scientific information embodied by
the astronomical poem of his contemporary Aratus.209
It is widely known that Callimachus used to great advantage the impressive
cultural background he had acquired through his activity within the Library,
boldly opening up new perspectives in contemporary poetry as well as display-
ing erudite and antiquarian implications in his sophisticated poems, mainly
and programmatically in the Aitia.210 An instance of this tendency can be seen
in the final part of the love elegy on Acontius and Cydippe, where the digres-
sion concerning the historical-antiquarian source employed (Xenomedes, the
5th century author of a local history on Keios) is an emblematic illustration of
211 Aitia 3, fr. 75.5377 Pfeiffer = fr. 174.5377 Massimilla, with the comment by the latter
([2010] 376392).
212 Pfeiffer [1968] 134136; Fraser [1972] 1.453456.
213 The most significant extant example of Peripatetic biography is the Life of Euripides, in
dialogue form, by Satyrus of Callatis (P.Oxy. 9.1176), who belonged to the next generation
after that of Callimachus, edited by Arrighetti [1964] and Schorn [2004] (F6; commen-
tary: 181347). On the highly disputed spheres of the origins, characteristics and typology
of ancient biography see Momigliano [1993]; Arrighetti [1987] and [1994]; and the essays
collected in Erler-Schorn [2007].
214 On the three Callimacheans: Jacoby [1954] 618627; cf. Pfeiffer [1968] 150151. On
Hermippus: Bollanse [1999a] and [1999b] (edition of the fragments); the list of Aristotles
writings in Diog. Laert. 5.21 goes back to the Peripatetic Aristo of Ceos (Moraux [1951]
237 ff.) more probably than to Hermippus (Dring [1956]; cf. Tanner [2000] 8386). On
Philostephanus: Capel Badino [2010]. On Ister: Fraser [1972] 1.511512; Jackson [2000];
Berti [2009] and [2013].
215 Chiefly P.Oxy. 10.1241, quoted at the beginning of this section, and Suda 3419; Callim.
testt. 11a19a Pfeiffer.
216 Str. 14.655; P.Oxy. 10.1241; Suda 3419.
217 The biographical entry in the Suda is wrong in placing Apollonius librarianship after
that of Eratosthenes (
, being successor of Eratosthenes in the direction of the Alexandrian
Library), merging information concerning the poet with that on his homonymous nick-
named the classifier: Pfeiffer [1968] 141142; Fraser [1972] 1.330332; Blum [1977] 184187.
110 Montana
explained, in two less than accurate Lifes,218 as due to the hostile reception of
a public recitation of a draft (proekdosis) of his major literary work, the epic
poem Argonautics. Only later, according to one of these biographies, did he
return to Alexandria, achieving the long-awaited success with his final edition
of the poem. However, the reliability of this account is regarded with some sus-
picion, and it seems preferable to view Apollonius departure from Alexandria
in relation to the accession to the throne of the third Ptolemy in 246 and, pos-
sibly, also in connection with the fact that the new king favoured Eratosthenes,
a scholar who is said to have been a pupil of Callimachus. Eratosthenes was a
native of Cyrene like Callimachus, and this happened also to be the homeland
of the new queen Berenice.219
Our primary surviving evidence of Apollonius scholarship resides pre-
cisely in the Argonautics, in which historical and literary learning, undoubt-
edly encouraged by the wide availability and use of books in the Library,
coexists and intermingles with intertextual emulation and interpretation of
Homeric poetry.220 As an instance of this, we may recall that in the last line
of the Argonautics (4.1781 , you hap-
pily landed at the Pagasian coast) the poet apparently alludes to Od. 23.296
( , they (Odysseus and Penelope) hap-
pily regained full right of possession of their old bed), thus perhaps indicat-
ing that in his opinion the latter marked the true end of the Homeric poem,
as indeed later assumed by such authoritative scholars as Aristophanes of
Byzantium and Aristarchus.221 Effectively, in tackling Homeric problems
Apollonius found himself in disagreement with textual and interpretive solu-
tions previously adopted by the first librarian and editor of the poems, and he
gathered together his own observations in a writing Against Zenodotus (
). Evidently the Zenodotean edition had risen to the status of a ref-
erence text within the Library and offered much material to the lively intel-
218 Edited by Wendel [1935] 12. See Mygind [1999] 272 (No. 69).
219 Due emphasis should be assigned to the circumstance: Pfeiffer [1968] 141142; Cameron
[1995] 214219.
220 Erbse [1953]; Rengakos [1993], [1994], [2001] and [2002a]; Fantuzzi [2000].
221 Schol. MaVX Hom., Od. 23.296 Dindorf, speaking of , it is unclear
whether in the meaning of actual end (so e.g. Rossi [1968] and [1976] 114 n. 3) or in the
(Aristotelian) sense of narrative fulfillment (Gallavotti [1969]; Erbse [1972] 166177);
cf. Eust., Od. 1948.49. On the question see Pfeiffer [1968] 175177; Richardson [1994]
2122; Pontani [2005b] 36, 49. Montanari [2012d] 345347 underscores that Demetrius
of Phalerum had already highlighted the relevance of this line for the economy of the
Odyssey, because it marks the recomposition of the conjugal (Dem. Phal. fr.
193 Wehrli = fr. 145 Stork-Opuijsen-Dorandi).
Hellenistic Scholarship 111
222 Ath. 10.451d. On the interest in Archilochus works during the Classical and Hellenistic
Ages see Pfeiffer [1968] 144146.
223 Montanari [2009a] 323324 and 335.
224 Cf. Pfeiffer [1968] 146. See Novokhatko, this volume.
225 Suda 2898: Eratosth. FGrHist 241 test. 1.
226 Str. 1.15: Eratosth. FGrHist 241 test. 10. Pfeiffer [1968] 154 with n. 3, gives little credence to
this suggestion.
227 Suda 2898; Ps.-Luc., Macrobii 27.
228 A sketch of Cyrene as a center of Hellenic culture is provided by Fraser [1972] 1.786789;
Montanari [1993a] 636638.
229 Again Suda 2898.
230 Pfeiffer [1968] 154155.
112 Montana
231 Poetic fragments: Powell [1925] 5868; Lloyd-Jones Parsons [1983] 183186 (frr. 397399);
Lloyd-Jones [2005] 4849. The author of the writing On the sublime expresses en passant
a far from hostile judgement on the elegy entitled rigon (Subl. 33.5: Eratosth., Erig. test.
1 Rosokoki; see the comment by Rosokoki [1995] 76). The short poem Herms gained the
honor of a commentary by the 2nd/1st century Rhodian Timachidas (cf. Powell [1925]
5859 and below) and had a very favorable reception over the next centuries (e.g. Agosti
[2008]).
232 See Pfeiffer [1968] 156159.
233 Cf. Suet., Gram. et rhet. 10 (Eratosth. FGrHist 241 test. 9): (L. Ateius) philologi appellationem
assumpsisse videtur, quia sic ut Eratosthenes, qui primus hoc cognomen sibi vindicavit, mul-
tiplici variaque doctrina censebatur.
234 Schol. Vat. to Dion. Thrax, in GG 1/3.160.1011
, . On this definition see Matthaios
[2011a]; cf. Pagani [2011] 1718 with n. 3; Swiggers and Wouters, this volume (section II.2).
235 Suda 2898.
236 Eratosth. frr. 17 and 60 Strecker, and Strecker [1884] 13 = 43 frr. 2123 Bagordo.
237 Clem. Al., Strom. 1.16.79.3.
238 Fragments collected by Strecker [1884] and Bagordo [1998] 127136 (No. 43). In fr. 25
Strecker = 43 fr. 2 Bagordo, Eratosthenes explicitly criticizes Lycophron because the
latter is ignorant of the meaning of a neologism coined by the comic poet Cratinus
( ): see Montana [2013].
Hellenistic Scholarship 113
(to whom the comic poet Plato can be added, if a statement by Eratosthenes
about this authors career, found in a papyrus commentary, is to be assigned
to the same work),239 point to an interest in the language of the playwrights
and the features of the classical Attic dialect, explicitly also as a function of
the debate on the authenticity of the plays.240 Additionally, his writings high-
light an attention to the chronology of authors and works and to questions
involving performance and other issues generally related to the world of comic
drama. For instance, in a fragment Eratosthenes contests the authenticity of
the Miners () attributed to Pherecrates, applying the dialectologi-
cal criterion with a strictness that some have seen as a forerunner of much
later Atticism.241 In another passage Eratosthenes confutes the story about
Eupolis allegedly murdered by Alcibiades during the navigation towards Sicily
in the year 415:242 such an event, of dubious historical reliability, had risen to
the status of a fundamental node in the debate on comic parrhesia and on its
presumed limitations in the evolution of comedy between 5th and 4th centu-
ries, with consequences on reconstruction of the transformations and on peri-
odization of the comic genre.243 In yet another fragment Eratosthenes corrects
Callimachus, who in his Table on the dramatic poets had thought he perceived
a mistake in the Aristotelian Didaskaliai concerning the reciprocal chronol-
ogy of Aristophanes (first and second) Clouds and Eupolis Marikas.244 It was
possibly when discussing the poetic models underlying some comic passages
239 P.Oxy. 35.2737 (Ar. fr. 590 Kassel-Austin = Aristophanes 27 CLGP = Eratosth. 43 fr. 18
Bagordo), Fr. 1, Col. II, ll. 1017: see the comment by Montana [2012a] 174177; cf. Perrone
[2010] 91. A second quotation from Eratosthenes has been hypothesized by W. Luppe in
l. 31 of the same column of the papyrus: Montana [2012a] 179.
240 The playwrights attested to in the fragments show that of the title stands
roughly for what we precisely mean by Archaia: Pfeiffer [1968] 161; Nesselrath [1990] 176
180 and 181 n. 93; Bagordo [1998] 38.
241 Eratosth. fr. 93 Strecker = 43 fr. 5 Bagordo, cf. fr. 46 Strecker; another attribution is treated
in fr. 149 Strecker = 43 fr. 17 Bagordo. On Eratosthenes tendency to Attic purism when
treating Attic comedy see Tosi [1994b] 168171.
242 Eratosth. fr. 48 Strecker = FGrHist 241 fr. 19 = 43 fr. 12 Bagordo. Cf. Duris, FGrHist 76 fr. 73.
The sources of the episode are discussed by Storey [2003] 5660 and 379381.
243 Nesselrath [1990] 178179 and [2000] 237240: the debate set up a contrast between the
Alexandrian or literary approach to comedy (Eratosthenes) and the Peripatetic or politi-
cal approach (e.g. Platon., Diff. com. 2123 Perusino, who identified the reprisal against
Eupolis as the watershed between arkhai and mes: cf. the comment by Perusino [1989]
1415 and 4849).
244 Eratosth. fr. 97 Strecker = 43 fr. 14 Bagordo; cf. Callim. fr. 454 Pfeiffer. See Storey [2003] 61.
114 Montana
245 Eratosth. fr. 136 Strecker = FGrHist 241 fr. 44 = 43 fr. 16 Bagordo, on the genre of Archil. fr.
324 West; fr. 101 Strecker, on the popular song PMG 735 Page, assigned by Eratosthenes to
the poet Lamprocles.
246 Tosi [1994b] 187189. On the possible impact of Eratosthenes exegetical approach to
ancient comedy in later scholarship, see Montana [2013].
247 On Eratosthenes in the context of Hellenistic science: Fraser [1972] 1.409415. Especially
on his role in musical theory: Creese [2010] 178209.
248 Edition of the extant fragments: FGrHist 241 frr. 13 and 915.
Hellenistic Scholarship 115
this work Apollodorus of Athens composed his own Chronicle, which in turn
became one of the sources of the Christian chronography on which depends
a large part of our knowledge of ancient chronology. In another work, entitled
The Olympian wins (), Eratosthenes drew up a list of winners of
these competitions, achieving an improvement over earlier attempts.249
His masterwork is considered to be the Geography (), originally
in three books, many fragments of which are preserved by Strabo, the geogra-
pher writing in the Augustan Age.250 One of the novel elements in this work
as compared to analogous previous compositions certainly consisted in its
marked broadening of horizons and geographic knowledge, as a consequence
of the immense conquests made by Alexander the Great. Another innovative
aspect, so characteristic of this author, was the application of a scientific atti-
tude to the subject of his investigations. His belief, expounded in his history of
the discipline, that the setting of the Homeric poems was the fruit of imagina-
tion and was not traceable to real placesfor it was his conviction that the
aim of poetry is to capture the soul rather than to learndraws on a tradi-
tional line of thought mocked by Stoic thinkers.251 Famous in this regard, and
emblematic of his attitude, is the sarcastic statement that
a man might find the places of Odysseus wanderings if the day were to
come when he would find the leatherworker who stitched the goatskin of
the winds.252
Opinions such as these could seem to bear some affinity with Platos dis-
paragement of the educational value of poetry, especially epic poetry; but
Eratosthenes main focus seems not so much to concern philosophical or ethi-
cal truth as, rather, scientific epistemology; he aimed to clear the field of the
249 FGrHist 241 frr. 48 and 915. Hippias of Elis, in the 5th century, drew up a List of the
Olympian winners.
250 The fragments have been edited by Berger [1880]. For a reassessment and a commentary
see Roller [2010].
251 Str. 1.15 (Eratosth. fr. I A 20 Berger) ()
, . Eratosthenes assertion is quoted and criticized by Strabo in
the context of a well-structured defense, of Stoic inspiration, of the truth value of poetry:
Halliwell [2002] 269271. Cusset [2008] 126127 argues for the positive acceptation of
in this context, given the Platonic (Socratic) use of this term as a definition of
rhetoric (Phaedrus 261ac and 271c272b).
252 Str. 1.24 (Eratosth. fr. I A 16 Berger) , ,
. Cf. Eust., Od. 1645.64 (concerning
Od. 10.19).
116 Montana
253 Pfeiffer [1968] 166167. Eratosthenes, allowing the Alexandrian mould, admitted scien-
tific erudition in his own poems, but as a feature subject to the psychagogic function of
poetry: Cusset [2008].
254 E.g. Russo [2004] 273277.
255 Str. 1.62 , in the second
(book) he seeks to make a critical revision of the geography.
256 Recent commented editions of Constellations epitome: Pmias i Massana [2004]; Pmias-
Geus [2007]; Pmias i Massana-Zucker [2013]; see also Santoni [2009].
Hellenistic Scholarship 117
history of ancient scholarship should recognize in him the true emblem of the
union and reciprocal influence of (poetry,) philology and science.257
Endorsing a neo-humanist vision of the birth and nature of ancient schol-
arship, Rudolf Pfeiffer emphasized the novelty and uniqueness of the figure of
Eratosthenes as the driving force behind the convergence of science and phi-
lology. Viewing the situation from this standpoint, Pfeiffer was able to deny any
direct Peripatetic derivation or imprinting of Alexandrian learning. If, how-
ever, the perspective is broadened from single individualities to the ensemble
of Ptolemaic institutions and cultural policy, it should be recognized that sci-
entists and scholars were gathered together as a group in the Museum, and
that since its very foundation they had been working according to a carefully
planned and organic project which in many respects mirrors the multifac-
eted approach of the Aristotelian school. In such a context, Eratosthenes can
in a sense be seen as accomplishing in his own person the synthesis properly
belonging to the ordinary nature and structure of the Alexandrian institutions
themselves. Furthermore, even if one sets aside the appeal to scholarship and
science within Alexandrian poetry designed to achieve realism,258 the blend
of philology and science and the mutual exchange of practices and meth-
ods can also be perceived in a number of significant personalities within and
outside Alexandria.259 In the medical field, for instance, examples include
Eratosthenes contemporary Bacchius of Tanagra, a physician and lexicogra-
pher, who was also an editor and commentator of Hippocratic works, and,
in the next century, his colleague Zeuxis the Empiricist, who is even credited
with commentaries on Hippocrates and was active in ensuring the acquisition
of medical writings for the Ptolemaic Library.260 Likewise, the Alexandrian
scholar Ptolemy Epithetes, in the 2nd century, was the author of a work on the
wounds in the Homeric poems, which seems to stand midway between philol-
ogy and medicine.261 Mention could also be made of Apollodorus of Athens,
a pupil of Aristarchus and the true heir of Eratosthenic learning in the 2nd
century interchange between Alexandria and Pergamum (see below). Thus
257 Pfeiffer [1968] 152153, 155156, 167168; cf. Jacob [1992]; Tosi [1998a]; Geus [2002]; Cusset
[2008].
258 Zanker [1987] 113131.
259 Mette [1952] 6264; Montanari [1993a] 635.
260 About Bacchius: Von Staden [1989] 484500 and [1992]. Other Hippocratic lexica are
attributed by sources to Xenocritus of Cos, Philitas contemporary; Philinus, himself
from Cos and the founder of the empirical school of medicine; and the poet Euphorion
of Chalcis, on whom see Acosta-Hughes Cusset [2012] XVI and 100101 (frr. 4950). On
Zeuxis: Deichgrber [1930] 221 (fr. 343); Von Staden [1989] 481 with n. 4.
261 Ptol. Epith. test. 1 Montanari, with F. Montanari [1988] 8183 and [1993a] 634.
118 Montana
there is ample justification for asserting that the opportunity for exchange and
osmosis between the two spheres on the issues of concepts and methodologies
was by no means lacking in Hellenistic and Alexandrian culture: indeed, the
claim that such opportunities did exist is fully warranted and, at the very least,
it calls for more adequate and in-depth exploration.262
Aristophanes of Byzantium (ca. 265/257190/180), Eratosthenes successor
at the head of the royal Library, opens the more mature season of Alexandrian
philology. The son of an army officer who had moved from Byzantium to
Alexandria, he is said to have been a disciple of Zenodotus (although this sug-
gestion is open to considerable doubt for chronological reasons), as well as
of Callimachus and indeed of Eratosthenes himself. The comic poet Machon,
Dionysius Iambus and a certain Euphronidas are also remembered among his
teachers. Aged 62, Aristophanes became the fourth librarian-in-chief and held
this post for fifteen years, until his death. It is during this period that he is sup-
posed to have been imprisoned in Alexandriaassuming that this piece of
information genuinely concerns his life and that it is in fact truefor having
planned to desert the patronage of Ptolemy V Epiphanes in favor of that of
Eumenes II, king of Pergamum from 197, who at the time was intent on found-
ing his own cultural patronage.263 However, it would not be surprising were
the story to turn out to be a forgery, moulded on the traditional picture of a
strong rivalry between these two main centers of learning from the beginning
of the 2nd century onwards. What is certain is that it implies the idea of jeal-
ous and fierce competition between royal powers, even more than between
philologists residing in the two different capitals.
Among his activities within the Library, it is worth citing, first and foremost,
a contribution in the sphere of pinacographical studies, the work In addition to
(better than Against)264 Callimachus Tables ( ). This
was an updating, and possibly also a discussion or defense of debated issues,265
of the great catalogue arranged by Callimachus fifty years earlier, in which a
number of problems concerning the genuineness, attribution and cataloguing
of books now needed to be addressed.266 As an instance, Aristophanes doubts
concerning the authenticity of the Shield attributed to Hesiodin agreement
262 Overall, on science in Ptolemaic Alexandria and in the Hellenistic world: Fraser [1972]
1.336446; Lloyd [1973]; Giannantoni-Vegetti [1984]; Russo [2004].
263 Suda 3936, s.v. : Ar. Byz. test. 1 Slater.
264 Nauck [1848a] 245247; Pfeiffer [1968] 133; Slater [1986] 134.
265 Slater [1976].
266 Ar. Byz. frr. 368369 Slater; cf. Callim. fr. 439 Pfeiffer and Pfeiffer [19491953] 1.349 (after
Callim. fr. 453); Bagordo [1998] 4445 and 88 (No. 15 frr. 13).
Hellenistic Scholarship 119
271 Ar. Byz. test. 7 and 9 Slater = 15 frr. 1112 Bagordo = Men. test. 83 and 170c Kassel-Austin. See
Pfeiffer [1968] 190191; Cantarella [1969] 189194.
272 Montana [2007].
273 E.g. A.P. 9.184. The definition of the Alexandrian canon was strongly influenced by
Peripatetic studies on lyric poetry, in the view of Carey [2011] 453.
274 Pfeiffer [1968] 203208; cf. Nicolai [1992] 251265 and 275296.
275 This assumption, chiefly based on Dion. Hal., Comp. 156 and 221, has been contested by
Tessier [1995] 134.
276 Pre-Aristophanean colometric arrangements of melic texts, although rare, are positively
attested to by extant 3rd century papyri (a list of five items including texts of Sappho,
Stesichorus, Sophocles and Euripides is drawn up by Phlmann [2007] 105 n. 9) or
hypothesizable (e.g., with regard to Pindaric poems, Tessier [1995] and DAlessio [1997]).
In particular, some colometric discrepancies in the same verse of Pindar edited in two
Greek papyri of Imperial Age, P.Oxy. 5.841 (Pindars Paeans) and P.Oxy. 15.1792 (Pindars
Prosodia), respectively as Paean 6 triad 3 (Rutherford [2001] 301302 with his comment,
especially 336338) and as a prosodion, have been interpreted by DAlessio as different
arrangements hardly both included in the Alexandrian (Aristophanean) edition (as, by
contrast, believed by Rutherford [2001] 148), but rather going back to divergent coliza-
tions arranged by prior Hellenistic scholars or to performers texts with musical nota-
tion (a suggestion owed to Liana Lomiento), or else to the 5th century and possibly to
Pindar himself, namely as a traditional reflection of two distinct original performative
occasions. The latter solution is shared by Prauscello [2006] 84 n. 260 and is not ruled out
by Battezzato [2008] 145, while in the view of Tessier [2010a] 1316 the colometric diver-
Hellenistic Scholarship 121
this kind of treatment apparently when working on the Pindaric poems, pro-
viding them with purpose-designed smeia for metrical-textual scansion: the
kornis (, namely the stylized picture of a little crow), inserted in
the left margin, divided the compositions from one another; the paragraphoi
(), horizontal strokes written to the left of the text, between the
lines, distinguished the strophes of a given composition; and, as attested to
for the edition of Alcaeus poems, the asteriskos (instead of the kornis) was
introduced to mark the distinction between two poems of different meter.277
Although there are some critical arguments suggesting that when dealing with
the colometry of poetic texts the Alexandrian scholars resorted directly to
musical scores designed for performance, we are still far from being able to
document that such an approach was actually implemented in a standardized
and systematic manner.278 What we can positively observe, and at present this
must be presumed even for Aristophanes, is some episodic interaction or inter-
lacing between the musical/performative and the textual/bookish tradition.279
As for the text layout and colometry (as well as musical notation?) of the melic
sections of classical Attic tragedy, one cannot exclude that Aristophanes may
have been significantly aided by the Lycurgan copies of Aeschylus, Sophocles
and Euripides held in the Alexandrian Library: and this would provide an even
more convincing explanation of the irresistible attraction exerted by these
books over king Ptolemy III, who sacrificed a fortune to obtain them.280 But at
present this remains a mere hypothesis.281
gences could more plausibly give evidence of scores provided with different musical and
rythmical notation which were available to Hellenistic scholars.
277 Heph., Enkheiridion 73.1674.14 Consbruch. See Negri [2004]; further, Porro [1994], espe-
cially 222226.
278 See above, nn. 14, 15, and 276. A skeptical attitude in this respect is maintained by
Battezzato [2003] 19 with n. 52; Prauscello [2003] and [2006]. Prauscello has reexamined
the scanty available papyrological evidence, focusing on the only two papyri carrying
musical notation which, as for the text is concerned, are comparable with the medieval
tradition: P.Vind. G 2315 (Eur., Or. 338344) and P.Leid. inv. 510 (Eur., IA 1500?1509 and
784793), both dating to the first half of the 3rd century. It is worth remembering, with
Lucia Prauscello, the telling circumstance that the (fragmentary) musical scores known to
date normally display a general non-colometric disposition of the lineation and there-
fore this category seems unlikely to have constituted the usual source of Alexandrian
colometry (183). Phlmann [2007] 106 mentions some additional epigraphic evidence.
279 Overall, Prauscello [2006].
280 Cf. Fleming-Kopff [1992] 763.
281 Prauscello [2006] 10. Phlmann [1991] highlights the lack of evidence of musical notation
in the Lycurgan text of the tragic songs. Scodel [2007] points out that the character of
122 Montana
the Athenian official copy was plausibly above all symbolic and ideological, more than
authoritative from a textual and performative point of view.
282 Porro [1994] 56, with previous bibliography.
283 Liberman [1999] XLVIIILXI, especially LVIIILX.
284 Ar. Byz. frr. 380A and 381 Slater. See Irigoin [1994] 4549.
285 This possibility, however, with respect to Aristophanes work on the Homeric poems is
inductively argued by Slater [1986] 205210.
286 Ar. Byz. fr. 367 Slater.
287 Ar. Byz. fr. 363 Slater = 15 fr. 4 Bagordo.
288 Ar. Byz. frr. 364366 Slater = FGrHist 347 fr. 1 = 15 frr. 59 Bagordo. See Montana [2006b] 214.
289 Ar. Byz. fr. 376 Slater = 15 fr. 10 Bagordo = Men. test. 76 Kassel-Austin. See Montana [2007]
257258.
Hellenistic Scholarship 123
product one may cite his succinct introductions to the plays, or hypotheseis
(, subjects), containing information on the first performance, the
setting, characters and plot of each work, some elements of which remain, var-
iously transmitted by papyri and Medieval manuscripts.290 The most signifi-
cant precedent, according to a (debatable) ancient testimony,291 is represented
by some (subjects of plots of
Euripides and Sophocles) composed by the late 4th century Peripatetic
Dicaearchus of Messana.292 The importance of this genre of literary erudi-
tion cannot be underestimated, as it is the outcome of the use and combi-
nation of two different and converging critical approaches: on the one hand,
Peripatetic antiquarian inquiry into theatre performances (the Aristotelian
Didaskaliai) as well as into the plots of the plays (, to which Aristotle, in
the Poetics, assigns a fundamental role for the aesthetic evaluation of works);
and on the other, the spirit of reconstruction and erudite rearrangement that
characterized early Alexandrian (Callimachean) pinacography.293 This two-
fold approach is recognizable in other Aristophanean areas of research, such
as studies on the literary use of proverbs and proverbial features (Aristophanes
collected four books of Non-metrical and two of Metrical proverbs), which were
rooted in Aristotelian interest in the tradition of paroemiographic wisdom,294
or the paradoxographic work On animals ( ), to a large extent an epit-
ome of various Peripatetic sources.295
The experiences of, among others, Lycophron, Callimachus and Eratosthenes
provided a good illustration of the fact that critical study devoted to attribu-
tion, text constitution and interpretation of literary works derived great ben-
efit from in-depth observation of the authors vocabulary. Aristophanes by no
means neglected this feature, which was closely linked to his editorial activity,
and he collected the results of his research in this respect in the (Words),
a broad-based lexicographic compilation. It opened with the section On the
290 Pfeiffer [1968] 192196; Meijering [1985]; van Rossum-Steenbeeck [1998]; Montanari
[2009c]. This learned type of didascalic hypotheseis linked to the name of Aristophanes
must be kept distinct from the more popular narrative hypotheseis of Euripides plays,
which circulated widely in Hellenistic and Graeco-Roman Egypt, as attested to by many
papyri, on which see e.g. Bing [2011]; Meccariello [2014].
291 Sext. Emp., Math. 3.3: Dicaearch. fr. 112 Mirhady.
292 Status quaestionis by Montanari [2009c] 384390.
293 On this convergence: Pfeiffer [1968] 192193; Montanari [2009c] 399401.
294 Ar. Byz. frr. 354362 Slater. On Peripatetic paroemiography: Tosi [1994b] 179182. On
Aristophanes studies in this field and their relation with the Peripatetic tradition: Tosi
[1993] and [1994b] 182187.
295 Ar. Byz. fr. 377 Slater.
124 Montana
words which are suspected of not having been used by the ancients, apparently a
collection of glosses whose antiquity (i.e. their belonging to the vocabulary of
classical authors) was disputed.296 Some of the sections were arranged accord-
ing to different semantic areasprobably on the example of Callimachus
297systematically listing Greek words (e.g. Nouns indicating
age, Kinship nouns)298 and providing them with an explanation based on form,
dialect, evolution, meanings, literary attestations and so forth. Aristophanes
interest in lexical correctness, in dialectsfocus on the latter being traceable
back at least to Philitas and possibly Zenodotus, and to Aristophanes teachers
Callimachus and Dionysius Iambus as well299and especially in the Attic lan-
guage (although apparently with more moderation and a more open approach
than Eratosthenes)300 undoubtedly furnished solid support for the later appre-
ciation of the Attic dialect. This trend would ultimately develop into the pref-
erential and prescriptive literary use of this dialect, or Atticism, from the late
1st century onwards.301
His familiarity with lexical phenomena and linguistic uses and anomalies
in literature as well as in the vernacular302apparently attracted Aristophanes
attention to the recurrence of certain patterns of inflection. As is known, the
first reflections on language can be found in the context of 5th and 4th cen-
tury philosophy and rhetoric. At the beginning of the Hellenistic Age, Stoic
296 Ar. Byz. frr. 136 Slater. According to Slater [1976] 236237, in this part of the Lexeis
Aristophanes assumed an antipuristic attitude, directed against those who considered
as modern (and therefore less pure) some of the strangest and very rare Greek words.
Slaters statement is shared but partially reassessed by Tosi [1994b] 155162 and 202205
(discussion with D. M. Schenkeveld).
297 Wendel [1939a] 508; cf. Tosi [1994b] 166167.
298 Ar. Byz. frr. 37336 Slater.
299 Dionysius was the author of a work On dialects ( ) in which he made use of
parallels drawn from spoken languages (Ath. 7.284b).
300 Ar. Byz. fr. 36 Slater, with regard to the Attic use of , shows a less rigid position than
that displayed by Eratosthenes in frr. 46 and 93 Strecker also in relation to the attribution
of the comedy Miners to Pherecrates: Slater [1976] 240; cf. Tosi [1994b] 169 with n. 47.
301 Ar. Byz. frr. 337347 Slater are labeled as . For an assessment and up-dated
bibliography on the debated question of the alleged influence of Alexandrian lexicogra-
phy on later Atticism see Ascheri [2010] 127128 with n. 10; Tosi, this volume. In particular,
on the close relation between Aristophanes lexicography and the lexicon of the Imperial
Age known as Antiatticista (so called inasmuch as, although within an Atticistic frame, it
offers a moderate view in opposition to the rigorous purism such as that of Phrynicus) see
Alpers [1981] 108; Tosi [1994b] 162166 (who also discusses some objections advanced by
Slater [1986] 120); Tosi [1997b].
302 Pfeiffer [1968] 202.
Hellenistic Scholarship 125
303 E.g. Blank [1982] 110; Frede [1987b]; Taylor [1987]; Schenkeveld [1994], also emphasizing
the Peripatetic influence; cf. Ax [1993]; Matthaios [2001a] and [2002f].
304 Gutzwiller [2010] 354359 and, for further bibliography, 365.
305 Most recently: Sluiter [2000b]; in this volume, Novokhatko, and Swiggers and Wouters
(section III.2).
306 See Ar. Byz. frr. 370375 Slater. We have no positive proof of a supposed Aristophanean
treatise On analogy ( ): Pfeiffer [1968] 202203; Callanan [1987] 107; Ax
[1990] 12 and [1991] 282. Overall, on analogy: Callanan [1987]; Ax [1990]; Schenkeveld
[1990] 290297; Pagani, this volume.
307 Matthaios [2001a] and [2002f].
126 Montana
308 For opposite evaluations of Aristophanes philology see Slater [1976], [1982] and [1986]
205210 (negative); and Blank-Dyck [1984] (positive).
309 Ar. test. 113 Kassel-Austin. In the scholia to Aristophanes, Euphronius is quoted 27 times
(of which 14 in the scholia to the Frogs, 9 in those to the Wasps). See Pfeiffer [1968] 160161.
The fragments of Euphronius exegesis to Aristophanes plays are collected by Strecker [1884].
310 Eratosth. frr. 72, 105 and 114 Strecker. See Tosi [1994b] 189.
311 Suda 3933 (Ar. Byz. test. 1 Slater) testifies that Aristophanes was a pupil
, of Euphronidas the Corinthian or Sicyonian, a passage which
has been corrected (Slater [1986] 1, cf. Schmidt [1848] 327 n. 53; Nauck [1848a] 2 n. 3)
to < > , of Euphronius <and Machon>
the Corinthian or Sicyonian on the basis of a further correction by Bergk in Choerob.,
Comm. in Heph. 241.1517 Consbruch (Ar. Byz. test. 14 Slater)
,
Hellenistic Scholarship 127
325 On the problems posed by the list and on Apollonius place after Aristophanes see above,
at the beginning of this section.
326 Etymologicum Genuinum AB s.v. = Etymologicum Magnum 295.52 ff. This man-
ner of classification was perhaps applied to the Pindaric odes, according to schol. Pind.,
Pyth. 2 inscr. and 2.31 Drachmann.
327 Thus Fleming-Kopff [1992] 762; Fleming [1999] 25.
328 Most [1985] 100 n. 26.
329 Prauscello [2006] 2933, summarizing some assumptions of Irigoin [1952] 50, and Fraser
[1972] 2.666 n. 126. She also derives from the testimony of Aristox., Harm. 2.39 (49.1 ff. Da
Rios) that being able to grasp and recognize the different musical modes does not seem
necessarily to entail the ability to write down and then consequently decipher musical
diagrams (31).
330 Prauscello [2006] 33: her argument would be particularly strong if we could succeed
in demonstrating that Apollonius was appointed as librarian before Aristophanes of
Byzantium, who arranged many authoritative editions of poetry such as the Pindaric odes
130 Montana
(this possibility is contemplated, together with the opposite case, by DAlessio [1997] 53
with n. 178). On Hellenistic eidography, especially with respect to the classification of the
Pindaric songs, see Rutherford [2001] 90108 and 152158, who assumes that (107) most
cases of eidographic indeterminacy arose because Hellenistic classifiers tended to neglect
the performance scenario of songs in favour of formal features, and to the extent that they
were concerned with performance, they may sometimes have misinterpreted it.
331 The following dealt with Homeric problems: Against Philitas, Against Comanus, Against
Xenons uncommon opinion, On the Iliad and Odyssey, On the ships at the anchor.
332 The biographical entry in the Suda ( 3892) attributes to Aristarchus more than 800
books (i.e. rolls), only as far as his commentaries. Nonetheless, von Aristarch selbst
scheint es keine publizierten Kommentare zu Homer gegeben zu haben in the opinion
of van Thiel [2014] 1.8.
Hellenistic Scholarship 131
333 Standard editions of the Homeric scholia maiora containing Aristarchean scholarship
are, for the Iliad, Erbse [19691988]; for the Odyssey, Dindorf [1855] and Pontani [2007
2010] (in progress). Partial collections or editions of Aristarchus fragments include Lehrs
[18823]; Ludwich [18841885]; Matthaios [1999]; Schironi [2004]; van Thiel [2014].
334 Venezia, Biblioteca Nazionale Marciana, gr. 822 (olim 454), of the 10th century.
335 The fragments of these works are edited separately by Schmidt [1854] 112179 (Didymus);
Friedlnder [1853] and Carnuth [1869] (Aristonicus); Lentz [18671870] 3.2.2.1 22165
(Herodian); Friedlnder [18572] and Carnuth [1875] (Nicanor). P.Oxy. 8.1086, fragments of
an anonymous hypomnma on the Iliad (comments on 2.751827 survive; another scroll,
P.Oxy. 65.4451, with comment on Il. 1.5658, is written by the same hand and displays sim-
ilar features), dated to the 1st century, therefore composed shortly after Aristarchus life-
time and perhaps earlier than Didymus and Aristonicus, is an intriguing testimony, since
it explicitly quotes and apparently endorses a lot of Aristarchean methods (e.g. smeia)
and opinions, whose entity, provenance and authenticity are disputed: see Lundon [2001]
(with objection by M. Haslam at 839 n. 46), [2002b] and [2011a] 172174.
336 Bibliography quoted above, n. 157. But the phenomenon may have been due to the fact
that Aristophanes and Aristarchus... followed another source or sources more similar
to the vulgate, in the belittling opinion of West [2001a] 36. On the debated concept of
vulgata of Homers text: Haslam [1997] 63.
132 Montana
337 Cf. Pasquali [19522] 216217. Status quaestionis by Haslam [1997] 8487.
338 Pfeiffer [1968] 215217; West [2001a] 6167.
339 For this interpretation: Montanari [1998a], [1998d], [2000b] and [2009b] 156159.
Different views have been put forward by West [2001a] 6263 (both Ammonius and
Didymus knew two ekdoseis by Aristarchus) and by Nagy [2004] 86 and [2009] 2133
(there were two Aristarchean diorthseis according to Ammonius, two ekdoseis according
to Didymus). It is worth noting, however, that in at least two instances Didymus in his
was referring to Aristarchus work on the text of the Iliad
by the plural : schol. A Hom., Il. 1.522a1 ... , and schol.
A Hom., Il. 2.192b1 (scil. ). See Montana
[2014]. According to Helmut van Thiel, Aristarchus by no means wrote hypomnemata and
all his interventions were shaped as Randnotizen in his working copy of the poem (see
above, n. 332); futhermore, (van Thiel [2014] 1.8) die beiden aristarchischen Ausgaben
( ) wurden nach seinem Tod von einem Schler kombiniert und publiziert.
Der Herausgeber benutzte Aristarchs ursprunglichen Handtext und vermutlich Aristarchs
persnliches Exemplar der Ekdosis, in das er zustzliche Notizen eingetragen hatte, die
wiederum nur fr ihn verstndlich sein mussten; and, so doing, the Herausgeber misun-
derstood and contaminated the genuine Aristarchean tradition and induced Ammonius
to give a clarification (cf. 14). Finally, van Thiel (192, cf. 140) takes in both of the
Hellenistic Scholarship 133
within these writings those texts were preserved which, book by book,
had been written or transcribed for their own personal use by those of
whom the same books bore the names.341
Thus one may perhaps surmise that these lost copies contained addenda and
corrigenda dating back to the very eponyms of the rolls and therefore, with
regard to the (), to Aristarchus in person. If so, this could
confirm that at the time of Galen a single Aristarchean edition of the two
Homeric poems was in circulation, enriched by contributions and corrections
that had been added by the philologist himself at a date subsequent to his orig-
inal edition: therefore, in Ammonius words, a re-edited diorthsis (
).342
Didymean scholia mentioned above as vermutlich Randnotizen der Hrer und Schler
Aristarchs in ihren Homertexten.
340 Editions of (De indolentia): Boudon-Millot [2007b]; Boudon-Millot Jouanna
[2010]; Kotzia-Sotiroudis [2010]; Garofalo [2012a]. See also Vegetti [2013] 249303 and
Nutton [2014].
341 ( cod.) ,
, , according to the text given (fol-
lowing in the footsteps of Roselli [2010]) by Garofalo [2012a].
342 See Montana [2014]. On the rich debate sparked by Galens passage seein addi-
tion to the contributions by Roselli and Garofalo cited in the previous noteat least
Boudon-Millot Jouanna [2010] 5254; Stramaglia [2011] 120129; Manetti [2012b] 1416;
Mazzucchi [2012] 252253.
134 Montana
343 On the Aristarchean smeia, see the references quoted above, n. 133, and further at least
Pfeiffer [1968] 218; Montanari [1997o] 274281; van Thiel [2014] 1.2930.
344 The circumstance is attested to with respect to Alcaeus edition by Heph., Enkheiridion
74.1114 Consbruch; see Porro [1994] 34.
345 See above, n. 130.
346 Ath. 1.12e , .
, the poet uses
Hellenistic Scholarship 135
the word only for human beings, never for beasts. But Zenodotus, ignoring the mean-
ing of this term, in his own edition writes: etc.. In actual fact, occurs in our text of the
Iliad (24.43) in reference to the meal of a lion.
347 Overall, on Aristarchean interpretation of Homeric vocabulary: Nnlist [2012b].
348 In his edition, Erbse takes the quoted words to be part of the excerptum from Aristarchus
hypomnma; this position is shared, among others, by Rengakos [2012] 245. The oppo-
site opinion is maintained by West [2001a] 7072 (cf. Fhrer-Schmidt [2001] 56), who is
inclined to attribute these words, and the act of collation of different Iliadic testimonies,
to the source of the piece of information, namely Didymus, rather than to Aristarchus.
See also above, n. 128.
349 Nagy [2004] 87109.
136 Montana
350 ,
. The same is repeated, more succinctly, in schol. 222b2 and b3. This case is also
adduced, with other instances, by Rengakos [2012] 245 and 247.
351 Schol. A Hom., Il. 1.2931 (Aristonicus). On Aristarchus athethseis: Lhrs [1992]. For the
decoding of the dialectics between Aristarchus and Zenodotus athethseis the ancient
grammarians availed themselves of Callistratus and other similar
works: see Montana [2008] and below.
352 Reassessments by Cucchiarelli [1997], underlining the actual admission of rhetorical alle-
gory () by Aristarchus and the Alexandrian scholars (with regard, for instance,
to the interpretation of Alcaeus intrinsically allegorical imagery); and by Nnlist [2011],
stressing the actual tepidity of this alleged Aristarchean controversy.
Hellenistic Scholarship 137
poems, should not waste time on anything else than what is actually said by
the poet.353
Through his analysis of language together with an exploration of the histori-
cal and cultural aspects of the works, Aristarchus thus formulated a definition
of a specific Homeric quid ( ), distinct from that of later cyclic poets
(called , the younger ones)354 and Hesiodic poetry. On the basis
of this organic historical-literary vision, he defended the Homeric author-
ship of both poems, thus adopting a position which contrasted with a view
held by other scholarsamong whom Xenon and Hellanicus355known as
(khrizontes, those who separate), an appellative reflecting their
belief that the Iliad and the Odyssey were works by two distinct authors, with
only the first poem belonging to Homer and the second to a younger poet.356
Aristarchus is known to have composed a commentary on Hesiod, where in
all probability he shared Praxiphanes opinion of the non-authenticity of the
proem of Works and Days.357 Within the canon of lyric poets that was gradu-
ally being determined, he himself produced an edition of Alcaeus and possibly
of Anacreon, which superseded those by Aristophanes of Byzantium,358 and
drew up a commentary on Alcman, Pindar (with regard to whom the Medieval
scholia testify to seventy-odd Aristarchean interventions) and Bacchylides; in
addition, he concerned himself with Archilochus, Semonides, Hipponax and
353 Schol. D Hom., Il. 5.385 ; cf. Eust., Il. 561.28 ff.
, , ,
, . See Porter [1992] 7071.
354 Severyns [1928].
355 This Hellanicus, Agathocles of Cyzicums pupil as well as Ptolemy Epithetes teacher, was
already mentioned earlier for an intervention consisting of word division in a passage of
Herodotus Histories: F. Montanari [1988] 5253; above, n. 140.
356 Proclus, Vita Homeri 7374 Severyns: Hellanic. test. 2 Montanari: ()
, , , Homer
wrote two poems, Iliad and Odyssey, the latter of which Xenon and Hellanicus strip from
him. On the khrizontes: Kohl [1917]. On Hellanicus and Xenon: F. Montanari [1988],
respectively 4373 and 119121. Aristarchus wrote a treatise against Xenon (see above,
n. 331).
357 See above, n. 44.
358 Hephaestio on the one hand (74.12) contrasts the edition of Alcaeus by Aristophanes of
Byzantium with that by Aristarchus, qualified as the present Aristarchean (scil. edition)
( ; see Porro [1994] 34), on the other (68.22) mentions the pres-
ent (scil. Aristarchean) edition ( ) of Anacreon, implicitly referring to one
older (Aristophanean?) edition of this poet. See Pfeiffer [1968] 185.
138 Montana
359 Raffaelli [1992]. Some points of view put forward by Aristarchus, when discussing histori-
cal questions issued from two Pindaric passages, were refuted by the Pergamene scholar
historian Artemon, frr. 12 Broggiato. On the negative evaluation of Aristarchus histori-
cal proficiency, widespread in antiquity as well as in modern times, see Broggiato [2014] 16
with n. 21; partial reassessment by Muzzolon [2005] 5960 (scholarship on Aristophanes)
and Vassilaki [2009] (scholarship on Pindar).
360 In schol. Theoc. 10.18e (p. 229 Wendel) ,
the reading (mss. UEAGPT) is preferred to (ms. K): Wendel, l.c.;
Pfeiffer [1968] 222223 n. 7; Wartelle [1971] 165 n. 1; Montanari [2009c] 416417.
361 Muzzolon [2005].
362 Montanari [1995a] and [2002c]; skeptical, Rengakos [2000].
363 Editions: Grenfell-Hunt [1901] 34; Viljoen [1915] 1722; Paap [1948] 3740. A new edition
by myself is forthcoming within the corpus of Commentaria et lexica Graeca in papyris
(CLGP).
364 Montana [2012b].
Hellenistic Scholarship 139
365 As recognized by Vannini [2009], should now be considered as part of the lemma
and not of the comment, and therefore as a veritable Herodotean lectio upheld by
Aristarchus.
366 I I 1316 [] | []: []|
| , they do not use iron nor silver: Sophocles in the Shepherds (fr. 500 Radt)
(says) neither bronze nor iron enter into contact with the flesh.
367 See Pfeiffer [1968] 224; McNamee [1977] 141; Radt [1977] 396 (app.); Del Fabbro [1979] 94
n. 78; Montanari [1993b] 248 and [2013] 67; Messeri Savorelli-Pintaudi [2002] 43 with
n. 1; Vannini [2009] 93 n. 1; West [2011] 8081; Montana [2012b]. Paap [1948] 3940, suggested
the jump from 1.194 to 1.215 may be due to a mechanical cause (a lacuna in the model of
the papyrus).
368 Six scrolls apart from P.Amh. 2.12: Bandiera [1997] 52.
369 For the Oxyrhynchite area see Lama [1991] 112113.
140 Montana
370 Nicolai [1992] 265275; Montanari [1997o] 282288 and [2013] 3032; Montana [2009c];
Priestley [2014] 223229. That Aristarchus also commented on the second book of
Herodotus Histories is inferred by Matijai [2013] from St. Byz. 466.1213 Meineke
. (2.163.2, Dat. -; 2.169.1, Acc. -).
, . The supposition that he realized an ekdosis of the Histories
(Jacoby [1913] 515; cf. Rosn [1962] 211, 231; Hemmerdinger [1981] 154164) is rejected by
Pfeiffer [1968] 225 n. 3; Alberti [1983] 195; Baldwin [1984] 32; West [2011] 7980; reticent
Irigoin [1994] (see 50, 54, 88). The attractive hypothesis that Aristarchus may have com-
mented on Platos works from a philological and linguistic point of view, as maintained
by Schironi [2005] with interesting but not incontrovertible arguments, awaits more solid
confirmation.
371 Cf. Proclus, Vita Homeri 5962 Severyns; Vitae Homeri 244.13 and IV, 247.78 Allen; schol.
A Hom., Il. 13.197 (Aristonicus). This assumption was put forward in the work
, scil. : Davison [1955] 21; Pfeiffer [1968] 228; Janko [1992] 32 n. 53 and 71;
Nagy [1996] 151; Heat [1998] 2728; Cassio [2002] 110; Ascheri [2010] 133134 n. 31) and it
stood in contrast to the belief of a derived Pisistratean (Attic) recension of the Homeric
poems in the 6th century, whether this recension was a rather ancient construct (Nagy
[1998] 227), or whether it was a recent invention of the Pergamene scholars (West [1983]
249). Some importance must be attached to the observation that in Aristarchus view the
dialectal words different from the Ionic present in the Homeric poetry could be treated
as glssai in the Aristotelian sense (words felt as eccentric for diachronic or diatopic rea-
sons): Montanari [2012a].
372 The grammarian Ptolemy nicknamed the Pindarion, one among the many pupils of
Aristarchus, devised a theory that merged together syntheia, linguistic correctness and
Hellenistic Scholarship 141
remains open, but it can now base its arguments on a substantial body of criti-
cally collected and studied evidence.378
This profile of Aristarchus can be completed with a few concluding consid-
erations. In the context of a general belittling of Alexandrian philology, some
critics have expressed an overall negative assessment of this scholar, empha-
sizing the defects inherent in his conjectural and analogist approach to the
Homeric text and pointing to a number of visibly mistaken and debatable
textual and interpretive choices. This is a judgment which contains elements
of truth, but it fails to award due credit to other aspects, such as Aristarchus
awareness of the importance of the textual tradition or, in more general terms,
the historical and cultural background of the scholar. Modern disregard for
such aspects of Alexandrian scholarship may lead to an underestimation both
of the philological reliability of Aristarchus and his pupils and also of the intel-
lectual and methodological progress they achieved.379 On the other hand,
one should guard against an overestimation of Aristarchus attainments, great
though the merits of this scholar may have been. For it is a moot point whether
he could have reached the same results without the benefit, as a preliminary
starting point, of the enormous work conducted by Aristophanes on literary
texts (edition, colometry, antiquarian documentation), vocabulary and lan-
guage. Admittedly, with regard to the fields of learning to which both of these
scholars devoted their efforts, we have far less documentation for Aristophanes
than for Aristarchus, but in many cases this is to be ascribed to the fact that
their opinions were in agreement on many issues: in short, the assumption
that the disciple apparently overshadowed the teacher is possibly a mislead-
ing distortion to be imputed to the tradition.380 Therefore, the assessment of
the historical role of the two erudites and of their reciprocal relation should
be addressed with extreme caution. Finally, it is important to bear in mind a
further aspect that ultimately helps us to comprehend the true greatness of
Aristarchus: namely, the inspirational nature of his teachings, which produced
a wealth of (direct and indirect) disciples who excelled in their fields, to the
point that it became customary to speak of an Aristarchean school, in refer-
ence to at least two generations of scholars after his own.381 Some of these
378 Overviews of the question: Swiggers-Wouters [2002c] and [2005]; Pagani [2010a] 105107
and [2011]; Matthaios [2012] and [forthcoming/b].
379 On this discussion, started in modern times by van der Valk [1949] and [19631964], see
most recently Montanari [2009a] 318319, [2009b] 160 with n. 32, and [forthcoming].
380 Richardson [1994] 21.
381 Aristarchus pupils reached the number of roughly 40, according to Suda 3892. See Blau
[1883].
Hellenistic Scholarship 143
scholars, continuing along the lines laid out by their teacher, played a funda-
mental role in the development of a number of sectors and disciplines. Since
their biographic and intellectual vicissitudes became intertwined with the
severe political crisis of 145/144 that risked almost irremediably compromising
the learned activity within the Museuma crisis which we have for this very
reason taken as a chronological watershedwe will address these personali-
ties in a later section of this chapter.382
382 Our main source about the dynastic crisis in 145/144 is Ath. 4.184b, who, apparently
basing himself on the testimonies of Menecles of Barca (FGrHist 270 fr. 9) and Andron
of Alexandria (FGrHist 246 fr. 1), points out that, with the accession of the cruel new
Ptolemy, somewhat paradoxically there was a second renewal of all sorts of learning
(... ): the king, partly killing and partly banishing
the Alexandrian intellectuals loyal to his predecessor, in fact filled the islands and towns
(scil. outside Egypt) with grammarians, philosophers, geometers, musicians, painters,
trainers, physicians and many other men of skill in different fields; who, compelled by
poverty to teach what they knew, produced a great number of celebrated pupils. On this
testimony see Pfeiffer [1968] 252253; stimulating discussion in Luzzatto [2008] 151154.
144 Montana
383 On the Attalids overtures and close relations established with the Academy and the
Peripatus, see Hansen [1971] 396397. Nagy [1998] 214 prefers to emphasize the competi-
tive attitude of Pergamum towards contemporary Athens more than towards Alexandria.
384 E.g. Kosmetatou [2003] 166173.
385 Collection of ancient testimonies on the Library: Platthy [1968] 159165 (Nos. 138153).
Penetrating observations concerning the ideology underlying the Pergamene Library are
put forward by Nagy [1998]; cf. Nicolai [2000b].
386 Plut., Ant. 58.9, speaking of two hundred thousand , single rolls, i.e. with-
out taking account of their reciprocal relation as or (see above, n. 98).
While the figure can be credited with some likelihood, the purported gift to Cleopatra has,
in contrast, been devoid of credibility since as early as Plut., Ant. 59.1; cf. Pfeiffer [1968]
236237.
Hellenistic Scholarship 145
391 Alexandrian personalities who had, or are thought to have had, affinities or relations with
the Pergamene cultural milieu were, among others, Agathocles of Cyzicum, Aristophanes
of Byzantium (for both scholars see above, 2.5), Demetrius Ixion, Apollodorus of Athens
(see below, 3.2).
392 Overall, on ancient Pergamum: Evans [2012]. On the Attalid dynasty, ideology, and cultural
policy: Hansen [1971]; Virgilio [1993]; Gruen [2000]; Shipley [2000] 312319; Kosmetatou
[2003]. On the Attalid Library as a classical model, namely a means of acquiring prestige:
Nagy [1998]. On Pergamene scholarship in Imperiale Age: Matthaios, this volume.
393 Hansen [1971] 397407.
Hellenistic Scholarship 147
394 This Antigonus should be regarded as distinct from the 1st century poet(s) by the same
name: Dorandi [1999] XVIIXXIII.
395 Edition of the fragments: Dorandi [1999]. On Antigonus idea of biography: von
Wilamowitz [1881]; Pfeiffer [1968] 246247, cf. 134; Hansen [1971] 397400; and Dorandi
[1999] XXXIIILXXXI, who argues for the pertinence of Antigonus lives to littrature de
mmoires rather than to biography in the strict sense (LXXX).
396 Herodicus of Babylon fr. 9 Broggiato, apud Ath. 6.234d.
397 Polem. frr. 4748 Preller, cf. 76 frr. 35 Bagordo.
398 Polem. fr. 45 Preller = 76 fr. 1 Bagordo, quoting Hipponax, fr. 126 Degani = fr. 128 West.
In general on Polemo: Preller [1838] (study and edition of the fragments); Pfeiffer [1968]
247249; Hansen [1971] 400403.
148 Montana
against his contemporary, the Pergamene scholar Crates,399 and it was widely
used later as a source by the Aristarchean Apollodorus in composing his On
the catalogue of the ships ( ).400 And to conclude
this brief survey of the culture that flourished under the reign of Attalus I it
is worth recalling the Great Geometer Apollonius of Perge, who had received
his training in Alexandria from the successors of Euclid and under the influ-
ence of Archimedes, and then had come into contact with the Pergamene
environment through Eudemus of Pergamumalthough the suggestion that
the dedicatee, by the name of Attalus, of several books of Apollonius funda-
mental treatise on Conic sections should be identified with the ruling king is
very unlikely.401
Pergamene scholarship reached its acme in the first half of the 2nd century,
as one of the planned achievements of king Eumenes II.402 Unlike Alexandrian
culture, which at that time had also risen to its highest degree, the cultural
approach of Pergamum shows an almost exclusive inclination to literary stud-
ies and significant receptiveness to philosophical inputs. Eumenes invited
and welcomed as a guest in the capital the scholar who would become the
most illustrious figure in the learned circle of Pergamum, Crates from Mallos
in Cilicia (southern Asia Minor). The biographic entry devoted to him in the
Suda states that he was a Stoic philosopher, nicknamed the Homeric and
the Critic because of his grammatical and literary studies, and a contempo-
rary of Aristarchus the grammarian in the time of Ptolemy [VI] Philometor,
who reigned from 180 to 145.403 An anecdote concerning Crates life is men-
tioned by the Roman biographer Suetonius: sent by the Attalids on a diplo-
matic mission to the Senate in about 168, in Rome he accidentally suffered a
leg injury and put to use his forced stay in the city by holding a series of lec-
tures, thus effectively transmitting the germ of philology to the Roman cultural
establishment404or, more precisely, enhancing a branch of studies which
had already for some time been experiencing the development of a tradition
of its own in Rome.405 Although no definite evidence is available, it is pos-
sible that Crates may have actively contributed to the setting up of the Attalid
Library, which during that very period was undergoing a phase of considerable
expansion and had recently been equipped with a catalogue () on the
model of the Alexandrian Library.406
In defining his philological activity, Crates distinguished his position from
that of the , grammarian, whose task he viewed as embodying
the limited perspective of an expert in glssai and the prosody of literary texts
(a definition that would seem to refer to the Alexandrian scholars of his
day). To describe his own activity Crates preferred to use the denomination
of , critic,407 which he saw as expressing a wider and more in-depth
organic approach to language and literature, ultimately seeking to provide a
critical appraisal of works rather than focus purely on textual details:
he stated that the kritikos is different from the grammatikos and the for-
mer must be an expert in all philosophical knowledge concerning lan-
guage (... ), the latter instead simply having to
explain glosses and give account of the prosody and be knowledgeable
about such questions: so that the kritikos can be likened to a master
builder, the grammarian to a workman.408
406 The Pergamene pinakes are attested to by Dion. Hal., Din. 1 and 11 (respectively 297.1516
and 317.34 Usener-Radermacher), and Ath. 8.336e.
407 The use of this word in a technical meaning is prior to that of grammatikos: Gudeman
[1922a] 1912; Schenkeveld [1968]. Philitas of Cos, as mentioned above (n. 28), was called
poet as well as critic ( ) by Str. 14.657.
408 Sext. Emp., Math. 1.79: Crates fr. 94 Broggiato. The metaphor exploits the polarization
between master builder () and workman (), or higher (intellectual)
and lower (practical) tekhnits (expert), which is typical in ancient debate on tekh-
nai, beginning at least with Plato and Aristotle: Romano [1987] 4849. In the view of
Crates, therefore, the kritikoi of Pergamon stand for a more holistic approach to schol-
arship than the grammatikoi of the Library of Alexandria (Nagy [1998] 187). Crates
definition of is comparable to that of grammar ascribed to the Stoic philosopher
Chrysippus (3rd century) in a passage, previously unknown and recently discovered
in the ms. Riccardianus gr. 62, of the Prolegomena to the scholia Vaticana to Dionysius
Thrax Grammar (Meliad [2013]): grammar is homeland of those who learn (
), inasmuch as it is mother of every form of education involving language (
).
150 Montana
nature of tekhnai: the logical part (), namely concerned with diction
and grammatical figures; the practical or empirical (), regarding dia-
lects and styles; and finally, the historical (), dealing with what cannot
be methodically organized, namely myths and historical facts.409
A long-standing traditional line of interpretation has attributed great
importance to the Stoic influence on Crates scholarship, maintaining that in
the field of linguistic theory he shared the assumption that superiority should
be awarded to custom (), which champions the dignity of different
uses or anomaly (), against the Alexandrian tendency to prefer regu-
larity, i.e. abstract normative rules or analogy. However, in the current debate
the portrayal of the two different views in terms of a sharp theoretical contro-
versy between Alexandrian analogists and Pergamene anomalists, or dogmatic
vs empirical thought,410 is mainly regarded as devoid of genuine historical
reliability, ultimately to be seen as the outcome of the dichotomic reconstruc-
tion of the question provided by the Roman erudite Varro (11627) in his work
De lingua Latina. Instead, such a conception is giving way to the picture of a
more mobile and intertwined discussion arising from (pragmatic) difficulties,
such as how to single out and apply linguistic regularities for the constitution
of literary texts and to determine the correctness of language (): this
eventually achieved the first steps towards the foundation of grammar as a self-
standing science or (that means, in our present concern, free in essence
from philosophical purposes).411 Given this sphere of interestshared both
by Alexandrian and Pergamene scholarsin the definition of , one
may ascribe some plausibility to the highly hypothetical attribution to Crates
of Mallos, instead of his namesake from Athens, the student of antiquities,412
of the work On the Attic dialect ( ), some fragments of
which quoted by Athenaeus show a moderately Atticistic attitude.413
409 Sext. Emp., Math. 1.248249: Tauriscus fr. 1 Broggiato = Crates test. 20 Broggiato. On Tauriscus,
according to the quoted testimony, and his fr. 1 see Broggiato [2014] 145153.
410 See Mette [1952]; most recently, Calboli [2011], especially 322325.
411 Fehling [1956] 264270; Pinborg [1975] 110112; Taylor [1987] 68; Blank [1994] and
[2005]; Schenkeveld [1994] 283287; Broggiato [2001] XXXIIIXL (with her frr. 102105).
Overviews with further bibliography: Dickey [2007] 6 n. 15; Pagani [2011]. In the opin-
ion of Ax [1991] 289295, viewing the conflict between analogy and anomaly as a purely
academic dispute is an oversimplification, as the issue involves a fairly broad-ranging
cultural question which had a bearing on philology (establishing the correctness of the
classical texts), instruction (endowing language with rules) and rhetoric.
412 Cf. FGrHist 362.
413 See Broggiato [2001] XLIIXLVI (with her frr. 106121*).
Hellenistic Scholarship 151
The ancient sources agree in attesting that in the field of literary criticism
Crates applied allegorical interpretation. According to a modern common-
place, this was a Stoic device which paved the way for the use of philosophical
thought to aid the explanation and understanding of poetry. However, despite
Pfeiffers authoritative statement,414 it is quite uncertain whether allegoresis
was genuinely a standard set up by the first generations of Stoic thinkers, who
rather appear to have been interested in the study and interpretation, also
in terms of etymology, of divine names and myths transmitted by archaic
poetry.415 Of Crates activity on the Homeric poems two titles remain, to which
a large part of the surviving fragments are to be attributed: the (or
), in eight or nine books, were devoted above all to textual criti-
cism, while the in all likelihood addressed exegetical questions of a
more general character, including aspects of a cosmological and geographical
nature.416 In the Suda the first work is defined as diorthsis (in acc. , a
correction of the transmitted ), but an edition of the poems by Crates
in the Alexandrian acceptation is excluded by almost all the modern critics.417
It is also debatedbut in a sense it may be an idle questionwhether these
writings were hypomnmata or rather syngrammata.418 The extant fragments
frequently show views contrasting with the position of Aristarchus, especially
if they are considered in the light of Crates different exegetical premises,
namely the assumption that Homeric poetry can be the basis for cosmological,
astronomical and geographic knowledge and investigations, as illustrated in
the following instance. In explaining the narrative of Hephaestus falling down
from the sky to earth after having been flung down by Zeus, in book 1 of the
Iliad (ll. 590593), Aristarchus interpreted the words (l. 592), express-
ing the time of the gods fall, literally as for all the remaining time of the day,
till evening. Crates, on the other hand, took these words as giving a precise
indication of the overall duration of the fall, in the sense of during a whole
day, i.e. the entire span of time required for the sun to cross the sky; moreover,
he considered them as useful evidence to calculate the size and the spherical
419 See Crates fr. 3 Broggiato with her comment (142144). On Cratetean cosmology: Mette
[1936].
420 Crates fr. 12 Broggiato. In contrast to Mette [1936] 3041, Pfeiffer [1968] 240241, and Porter
[1992] 9194 (and additionally Halliwell [2002] 274275; Gutzwiller [2010] 356), Broggiato
[2001] 157164 adopts a cautionary position with regard to the Cratetean authorship of
the two allegorical explanations.
421 Respectively Crates frr. 8284 (lyric poets) and 8689 Broggiato (Euripides). See Broggiato
[2001] XXIVXXV.
422 Maass [1892] 167203; cf. Broggiato [2001] XXII.
423 Crates fr. 78 Broggiato, with related comment (and further Broggiato [2001] XXIII).
424 Most recent discussion: Montanari [2009a] 316322.
425 Broggiato [2001] XXVIIXXXIII (with her frr. 94101). On Crates discussions scrutinized
by Philodemus in his first book On poems (frr. 96*-98* Broggiato) see Janko [2000] 120
189. Cf. Porter [1992] 112113; Halliwell [2011] 317319.
Hellenistic Scholarship 153
It appears quite natural to assume that Crates, as a kritikos, directed the results
of his textual scholarship towards the aim of genuine criticism of literary
works ( ), which, in the next generation, the Alexandrian
Dionysius Thrax, the pupil of Aristarchus, would recognize as the final part,
and the highest achievement, of the grammatik tekhn. In contrast, we must
resign ourselves to the lack of any positive evidence about the possibility that a
Crates repeatedly quoted by John Tzetzes for opinions relevant inter alia to the
parts () of comedy and the parabasis, subjects which probably formed
part of a treatise on ancient drama, may be identified with the Pergamene
scholar.426
426 Broggiato [2001] XXVXXVII (with her frr. 90*93*). Bagordo [1998] 61 and 116118 (No. 28),
subscribes, albeit cautiously, to the identification of this Crates with the Athenian
Academic philosopher of the 1st century, the author of a work On comedy ( ).
427 E.g. Montanari [1993c] 648649. On scholars mentioned by ancient sources as Crates
pupils or Kratteioi, see Crates testt. 2027 Broggiato; Hansen [1971] 418422; Broggiato
[2001] XVIIIXIX and 137138; Broggiato [2014]. On Tauriscus, , see
above, n. 409. Broggiato [2014] also includes among Crates pupils the scholar historian
Artemon of Pergamum (FGrHist 569; Pitcher [2007] in Brills New Jacoby).
154 Montana
gradual absorption of the latter under Roman rule. Among the effects of this
epoch-making historical transition one may cite, as of particular interest here,
a greater circulation of intellectuals and scholars of different imprint and, con-
sequently, greater opportunity for cultural exchange and influence, which to
all intents and purposes brought an end to the apparent monopoly, or duopoly,
of Hellenistic scholarship.432
Overall, acknowledgment of the opportunities for close-meshed cultural
interaction between the major seats of learning in the late Hellenistic Age, as
well as of the expansion of the horizons of scholarship on the eve of and during
Romanization, calls for an act of epistemological frankness. Specifically, there
should be a willingness to overcome once and for all in our historiographic
description the Manichean preconception implicit in the pattern Alexandria
vs Pergamum, which hinders the possibility of fair recognition that, alongside
differences of vision and polemical approaches, there were undoubtedly also
cultural links and convergences, reciprocal influences, more wide-ranging fre-
quentations and more pluralistic strands of belonging.
A grammarian who apparently adopted the critical line traced by Crates is
his fellow countryman Zenodotus of Mallos (2nd or 2nd/1st century), plausibly
to be identified with a Zenodotus qualified as Kratteios, disciple of Crates,
in a Homeric scholium.433 A work entitled Against Aristarchus expunction
of lines of the Poet ( ) is ascribable
to him:434 this lost monograph, of alleged polemical aim, took shape in the
context of a lively debate about the Aristarchean athetseis in the Homeric
texts.435 Before Zenodotus, the debate had already seen the contributions of
at least two Alexandrians: Callistratus, Aristophanes pupil, and Demetrius of
Adramyttium (in Mysia), called Ixion, one of the many disciples of Aristarchus,
the author of two works of Homeric scholarship entitled Against or On the
explanations ( ) and Against or On the athetized lines (
, scil. ).436 In the ancient tradition, this Demetrius was
branded as a betrayer of his teacher (this is the sense of the nickname Ixion,
432 Jolivet [2010]. On the political and economic background of this cultural transition:
Monson [2012].
433 Schol. ex. Hom., Il. 23.79b: Crates test. 24 Broggiato = Zenod. Mall. fr. 5 Pusch = fr. 3 Broggiato.
434 Suda 275, s.v. Zenodotus of Alexandria: Zenod. Mall. test. 2* Broggiato.
435 On Zenodotus of Mallos: Nickau [1972b]. Editions of the scanty fragments, predomi-
nantly pertaining to Homeric scholarship: Pusch [1890] 149160; Broggiato [2005] and
[2014] 107140.
436 On Callistratus see above, 2.5. Demetrius also was concerned with the comic poet
Aristophanes and possibly Hesiod (for the latter see Montanari [2009a] 341). The frag-
ments have been edited by Staesche [1883] and those pertaining to Homeric scholarship
156 Montana
by Ascheri [2003] and, far more inclusively, van Thiel [2014], who (1.10 and 2022) readily
identifies Ixion as the main author of the so-called exegetical scholia to the Iliad.
437 Suda 430 (= Dem. Ix. test. 1 Ascheri) . ... ,
... , Demetrius called Ixion. ...He received this
surname, ...because he opposed his teacher Aristarchus. See Blau [1883] 1920.
438 That Demetrius may have been open to solutions of an anomalist type is argued by
Ascheri [2010], especially 149150, according to whom the grammarian upheld a mod-
erate Atticism congruous with the Ptolemaic ambition to assimilate Alexandrian and
Athenian cultures and languages.
439 Ascheri [2003] XXVI, [2004] 337338, and [2010].
440 For divergences see e.g. scholl. Hom., Il. 1.364b2, 1.554c, 2.262b. On Dionysius: Montanari
[1997i]; Pontani [2005b] 56.
Hellenistic Scholarship 157
tradition, for the purposes of literary explanation.441 The same categories (real
persons satirized in comedy, among whom parasites and high society courte-
sans) were taken into consideration during the same period by the Pergamene
Carystius, author of a work on stage productions ( ),442 and the
Kratteios Herodicus of Babylon in a work entitled , two surviving
fragments of which, drawn from the sixth book, concern the hetaerae Sinope
and Phryne;443 a mention of a second and different staging of Aeschylus
Persians, inferable from a statement attributed to Herodicus in a scholion to
Aristophanes, may originate from the same work.444 The Cratetean definition
of the kritikos in opposition to the grammatikos can be detected by reading
between the lines of a satirical epigram by Herodicus designed to stigmatize
what he felt to be the pedantic limitedness of views held by the Aristarkheioi
in the field of linguistic correctness.445
The secessio doctorum which affected Alexandria in 145/144 undoubtedly
acted as one of the involuntary factors of further fruitful intellectual exchange
in the late Hellenistic Age. By the irony of fate, at the very time of the
Aristarchean apogee many Alexandrian scholars were forced to seek refuge
and a new cultural homeland in Pergamum or in other centers of learning
beyond Ptolemaic control. Such was the destiny, for example, of the Athenian
Apollodorus (ca. 180110). After receiving his training in Athens at the school
of the Stoic philosopher Diogenes of Babylon (or of Seleucia), also one of the
presumed teachers of Crates, he moved to Alexandria where he was a disciple
and co-worker of Aristarchus. It appears that at the time of Ptolemies dynastic
crisis he fled to Pergamum and then eventually returned to Athens. This intel-
lectual experience provides eloquent evidence of the actual possibility that
multiple cultural inputs from major centers of learning of the time (Athens,
441 FGrHist 350; Bagordo [1998] 50 and 7476 (No. 3). See Steinhausen [1910].
442 Bagordo [1998] 57 and 111 (No. 25).
443 Ath. 13.586a and 591c: 55 frr. 12 Bagordo = Herodicus frr. 67 Broggiato (see her comment,
[2014] 7880). On Herodicus as a Kratteios see Crates test. 25 Broggiato = Herodicus test.
2 Broggiato; Dring [1941]. The identification of Herodicus the Kratteios with his name-
sake of Babylon, author of the epigram quoted below in the text (see n. 445), is almost
generally accepted: Broggiato [2014] 42 with n 3. Herodicus chronology is debated, but
the second half of the 2nd century is a plausible inference: Broggiato [2014] 4243.
444 Schol. Ar., Ran. 1028e = Herodicus fr. 10 Broggiato, a note conceived in order to explain the
reference in l. 1028 of the comedy to king Darius as a character in the Persians. Herodicus
frr. 25 Broggiato concern Homeric scholarship; it is uncertain to which work they
belonged.
445 Ath. 5.222a: SH fr. 494 = Herodicus fr. 1 Broggiato; see De Martino [1997]; Manetti [2002];
Broggiato [2014] 5968.
158 Montana
Alexandria, Pergamum) may converge in the same person, thus belying the
conventional picture of sharp conflict and rivalry among these communities
of scholars.446
An ancient source defines Apollodorus as a philologist (),
in the same manner as Eratosthenes,447 of whom he was ideally a succes-
sor, partly by virtue of his comparable intellectual curiosity, and partly also
because he had cultivated fields of inquiry bearing some similarity to those in
which Eratosthenes was active.448 One of these common fields is chronology.
Apollodorus Chronicle (), dedicated to the Pergamene king Attalus II
and composed in iambic trimeters for the sake of memorization, was evidently
inspired by Eratosthenes work and intended to improve on it. In a chrono-
logical grid that extended from the capture of Troy (1184/3) up to the authors
own days and which was based on the list of the Athenian archonts, political
and military events were recorded alongside information concerning several
branches of human activity and knowledge such as philosophy, art, and lit-
erature. It is worth underscoring that the information Apollodorus provided
included biographical data on Menander, with an indication of the total num-
ber of his plays (105), drawing on the tradition of the didaskaliai widely plun-
dered by the Alexandrian scholars.449 As far as can be gathered from the very
scanty material that has come down to us, the work On the gods ( )
constituted a perfect synthesis of Apollodorus multifaceted personality and
a mirror of his composite cultural-biographic background. It was in essence a
study of religious history, carried out in a contextual setting of Homeric schol-
arship and endeavoring to conduct an in-depth analysis of divine names and
epithets, even resorting to etymology. Thus the Alexandrian predilection for
lexical inquiry and literary interpretation blended with a historical-antiquar-
ian interest and with a methodology that applied hermeneutics by etymol-
ogy, possibly under Stoic/Pergamene influence.450 The work On the catalogue
of the ships ( ), concerning problems posed by the
446 Cf. Fraser [1972] 1.470, though admitting contacts only on a purely personal level and
excluding a general dilution of the hostility between the two schools.
447 Ps.-Scymnus, Periegesis 1649 (with the comment by Marcotte [2000] 151152): FGrHist
244 test. 2.
448 Edition of the fragments in FGrHist 244, to which should be added Theodoridis [1972] and
[1979], and Mette [1978] 2023.
449 Apollod., FGrHist 244 fr 43. See Pfeiffer [1968] 257.
450 FGrHist 244 frr. 88153. Mention can be made here of the Library (), a mytho-
graphic handbook of the 1st or 2nd century AD, wrongly assigned to Apollodorus by
medieval manuscripts (although it may be not completely extraneous to his research on
religion and myths). See Wagner [19262], Carrire-Massonie [1991], Scarpi [1996], Fowler
Hellenistic Scholarship 159
catalogue of the Achean army at Troy in the second book of the Iliad and,
more generally, by the puzzling Homeric geography, offered contributions to
this highly specialized and challenging sphere of ancient scholarship, which
Eratosthenes, Demetrius of Scepsis, and Aristarchus had previously grappled
with.451 The scope of Apollodorus literary study also extended to theater:
he composed monographic studies on Doric authors of comedy and mime
(On Epicharm, On Sophron) and on the Athenian hetaerae introduced as char-
acters into Attic comic plays ( ).452 This had become
a typical subject of Hellenistic erudition, having been treated in the previous
century by Aristophanes of Byzantium and his pupil Callistratus, and, in the
times of Apollodorus himself, by his Aristarchean fellow-disciple Ammonius
and the Kratteios Herodicus. Let us conclude this profile by mentioning
Apollodorus studies on vocabulary (Glssai) and etymology, the latter being
treated not only in the On the gods, but also in a dedicated work of Etymologies,
which apparently represented an original fusion of Stoic and Alexandrian
inputs.453
Slightly younger than Apollodorus, the Alexandrian Dionysius Thrax
(ca. 17090) was able to complete his training at Aristarchus school just before
the diaspora of 144, which led him to move to Rhodes, where he obtained a
teaching position.454 That he adopted a polemical attitude towards the most
representative Pergamene scholar is openly testified to by his syngramma
Against Crates ( ),455 of which a precedent can be perceived
in a work bearing the same title composed by the roughly contemporary
Alexandrian Parmeniscus.456 Dionysius philological and exegetical activ-
ity on the Homeric poems, partly in agreement and partly in contrast with
[2000], Drger [2005], Meliad, this volume; and the bibliographical database ABEL
(Apollodori Bibliotheca Electronica) at http://abel.arts.kuleuven.be/.
451 Str. 8.339 attests to the extensive use of the works of Eratosthenes and Demetrius by
Apollodorus. The fragments are edited in FGrHist 244 frr. 154207.
452 FGrHist 244 frr. 208218; Bagordo [1998] 4546 and 8084 (No. 10).
453 Frede [1977] 52; Schenkeveld [1984] 348.
454 Mygind [1999] 263264 (No. 34). Dionysius was the author of a historiographical work on
Rhodes (FGrHist 512).
455 Dion. Thrax fr. 15 in the edition by Linke [1977]; cf. Crates test. 29 Broggiato.
456 Parmeniscus fr. 2 Breithaupt; Crates test. 28 Broggiato. By contrast, we cannot establish
with precision what kind of topics were addressed in the discussions forming the object
of the mentioned work On the doctrine of Crates composed by Ptolemy of Ascalon: see
above, n. 431.
160 Montana
457 Dion. Thrax frr. 147 Linke. For instance, Dionysius (fr. 47) endorsed Aristarchus analogist
criterion and the assumption of the Athenian origin of Homer.
458 For Hesiod: Montanari [2009a] 341.
459 Dion. Thrax frr. 5355 Linke.
460 Sext. Emp., Math. 1.57.
461 Edition by Uhlig [1883] 1101; then Pecorella [1962], with commentary. Greek text repro-
duced in Lallot [19982], with French translation and commentary; Krschner [1996], with
German translation; Swiggers-Wouters [1998], with Dutch and German (= Krschner
[1996]) translation; Callipo [2011], with Italian translation and commentary.
462 The controversy started with Di Benedetto [19581959].
463 .
Sext. Emp., Math. 1.57 gives the non-innocuous variant
, empirical knowledge as far as possible of
what is said by poets and prose writers. On the much debated interpretive problem posed
by the discrepancy between these passages see the comment by Lallot [19982] and more
recently Ventrella [2004].
464 Overview of the question in Pagani [2010b] and [2011], with exhaustive assessments of the
overwhelming bibliography on the topic.
465 The supplements were edited by Uhlig [1883] 103132, the scholia by Hilgard [1901]; for a
recent addition see above, n. 408.
Hellenistic Scholarship 161
of the greater part of the Tekhn have also been used to deny the existence of
grammatical reflections and interests on the part of the Alexandrian scholars
belonging to the earlier age (Aristophanes and Aristarchus).466 On the other
hand, this argument has been countered by pointing out that the question of
the genuineness or otherwise of the Tekhn in no way impairs the basic fact that
Dionysius was the author of a grammatical treatise: the definition of grammar
genuinely attested to by the Tekhn and by Sextus and some documented posi-
tions held by Dionysius in matters concerning parts of speech (which reveal
signs of Stoic derivation, in particular from Diogenes of Babylon),467 as well as
the similar achievements traceable to Aristophanes and Aristarchus, are suffi-
cient to provide a sound foundation for the statement that the science of gram-
mar began to burgeon in Alexandria starting at least from the 2nd century, in
the form of linguistic observation closely linked to / aimed at constituting and
interpreting literary texts.468 The new perspectives opened up by research into
the actual grammatical knowledge of the 2nd century philologists, jointly with
the current state of the discussion concerning Dionysius Thrax, ultimately
allow a picture of the achievements of these generations of scholars in terms
of first steps and evolution in the gradual process of translating philosophical,
rhetorical and philological concepts about language into technical grammar.469
Attempts to ascertain the chronology and cultural framework of the gram-
marian and historian Asclepiades of Myrlea (later Apamea, in Bithynia)470
leave us with an aporia, partly because the related biographical entry in the
Suda is of only limited usefulness as it is clearly corrupt or contaminated.471
The available data allow him to be placed roughly between the second half of
the 2nd and the first half of the 1st century. Knowledge of opinions put forward
by Dionysius Thrax detectable in the fragments of Asclepiades demonstrate
that the latter was either contemporary with or shortly later than Aristarchus
466 This line of reasoning has been upheld by Di Benedetto, first and foremost in his study of
19581959 and then repeatedly ([1973], [1990], [1998], [2000]), and it has been endorsed
by e.g. Pinborg [1975]; Siebenborn [1976]; Frede [1977]; Taylor [1987]; Law [2003]. The
opposite position includes Erbse [1980]; Ax [1982] and [1991]; Matthaios [1999], [2001a]
and [2002f].
467 On the influence of Stoic grammar on the Alexandrian post-Aristarchean scholars
(Apollodorus and Dionysius), see e.g. Schenkeveld [1994] 280281; Matthaios [2009a] 399.
468 Pagani [2010b].
469 Matthaios [2001a], [2002f] and this volume (with further bibliography). An outline of the
theoretical definitions and development of ancient Greek systematic grammar is traced by
Pagani [2011]; Seppnen [2014]; see further Swiggers-Wouters, this volume (section II.2).
470 He wrote a History of Bithynia (): FGrHist 697.
471 Suda 4173: Asclep. Myrl. test. 1 Pagani.
162 Montana
pupil. It cannot be ruled out that Asclepiades may have spent part of his life
in Rome and in Spain (Baetica). There is no clear-cut evidence, although it
is plausible, that he was a pupil of Crates or that he stayed in Pergamum.
Effectively, a Stoic and/or Pergamene influence seems difficult to deny, if one
considers the cosmological-allegorical methodology applied to Homeric exe-
gesis in his syngramma On Nestors cup ( ), concerning the
form, function and sense of this bowl described in Iliad 11.632637, which is
interpreted as an allusive image of the sky and of the Pleias constellation in
the heavens.472 A commentary on the Odyssey is explicitly attested to by the
sources473 and another on the Iliad is arguable from some indirect evidence.474
It is particularly significant that, in addition to his work on Homer, he also
concerned himself with Pindar, Theocritus and perhapsthough it is uncer-
tain whether in purpose-composed hypomnmata or in monographs devoted
to other topicsalso with Aratus and Apollonius of Rhodes, thus becoming
one among the first exegetes of Hellenistic poets.475 Of equally fundamental
importance, in the current critical debate on the origins of Greek grammati-
cal science, are the remains of Asclepiades Grammar ( ), in
particular his definition of grammar, clearly polemicizing against Dionysius
Thrax, not as empirical knowledge but as tekhn of what is said by poets
and prose writers. By this statement he effectively proclaimed the character of
grammar as, in a sense, both scientific and exhaustive, against an idea of this
discipline as a conjectural and imperfect intellectual activity. In another frag-
ment Asclepiades proposed a threefold subdivision of grammar, one technical
(i.e., systematic description of language), one concerning the historiai (namely,
philological study of realia), and one strictly philological, which to some
extent recalls the Pergamene partition conceived by Tauriscus and traceable to
Crates.476 Finally, we know of a work On the grammarians ( ),
472 Asclep. Myrl. frr. 410 Pagani; see Pagani [2007a] 1823 and her comment, 149225;
Gutzwiller [2010] 356357.
473 Asclep. Myrl. test. 12 and fr. 3 Pagani; see Pagani [2007a] 1618.
474 Asclep. Myrl. frr. 12 Pagani; see Pagani [2007a] 16.
475 On Asclepiades actual or presumed work on all of these poets: Pagani [2007a] 2431;
especially concerning Theocritus: Belcher [2005] 192194 and 199200 (texts); Pagani
[2007c] 287288 and 298 (texts). A treatise On Cratinus by Asclepiades, apparently attrib-
uted to him by Athenaeus (cf. Bagordo [1998] 60 and 102103, No. 20), is generally ruled
out: Pagani [2007a] 40 and 218219.
476 Asclepiades definition of grammar: Sext. Emp., Math. 1.7274. His subdivision of grammar:
Sext. Emp., Math. 1.252: see Slater [1972]; Blank [1998] 146148 and 264266.
Hellenistic Scholarship 163
latter was on the island,482 and Posidonius of Apameia.483 Both of these fig-
ures were the expression of an eclectic idea of culture, which seems to have
constituted an intellectual trend in Rhodes at the time, even for minor per-
sonalities. This can be observed in the case of the versatile Timachidas of
Lindos, the author of an erudite poetic work in hexameters entitled Banquet
(, 11 books or more), of a collection of , and of commentaries on
Euripides Medea, Aristophanes Frogs andtaking his place among the pio-
neers in the field of the exegesis on Hellenistic literatureMenanders Kolx
and Eratosthenes Herms (in at least four books). In addition, he has been
identified as one of the citizens of Lindos entrusted with drawing up the highly
erudite historical inscription known as the Chronicle of the temple of Lindos.484
Given the political relations officially established between Rome and
Rhodes towards the mid-2nd century, one can easily understand why numer-
ous members of the Roman elite would frequently visit the island to refine
their philosophical and rhetorical training. Thus after its by no means insignifi-
cant earlier period as an economic hub in the Eastern Mediterranean, when it
passed under Roman control Rhodes achieved considerable status as a lively
and highly attractive focus of cultural cross-fertilization. It is important to bear
in mind that this was indeed the context which Dionysius Thrax adopted as
the seat of his work and his chosen milieu of cultural learning after abandon-
ing Alexandria. Rhodes became a unique venue that allowed the blending of
482 See Pfeiffer [1968] 232, 245, 270; cf. Nagy [1998] 222223; Mygind [1999] 256257 (No. 10).
Ath. 14.634c testifies to Panaetius admiration towards Aristarchus, whom he defined as
a prophet () capable of penetrating into the real meaning () of Homeric
poetryunless the sentence had an ironical overtone, as Porter [1992] 70 is inclined to
think. As far as concerns the Plato of Panaetius testified by Gal., De indolentia 13 Boudon-
Millot Jouanna, rather than being an edition (as argued by Gourinat [2008] 141), it is
more likely to have been an exemplar of the Platonic Dialogues possessed and annotated
by Panaetius (thus Dorandi [2010b] 171; cf. Stramaglia [2011] 125).
483 See Mygind [1999] 257 (No. 12). Edition of Posidonius testimonies and fragments:
Edelstein-Kidd [19892]; cf. FGrHist 87.
484 : frr. 14 Blinkenberg; SH 769773. : frr. 1832 Blinkenberg. Commentaries:
on Euripides Medea, frr. 1516 Blinkenberg; on Aristophanes Frogs, frr. 513 Blinkenberg;
on Menanders Kolx, fr. 14 Blinkenberg = Men. test. 77 Kassel-Austin; on Eratosthenes
Herms, fr. 17 Blinkenberg (Ath. 11.501d, where the reading has been cor-
rected into by Susemihl), cf. Powell [1925] 5859. For the Chronicle: SIG3 725;
Blinkenberg [1915]; FGrHist 532. On Timachidas see Mygind [1999] 264 (No. 35); Montana
[2009c] 179180. A new edition of all of the extant fragments is being prepared by Thomas
Coward.
Hellenistic Scholarship 165
485 Di Benedetto [1958] 202, argues that great importance should be attached to the contacts
in 2nd century Rhodes between rhetoric and grammar, with regard to establishment of
the respective tasks and boundaries.
486 Anecdoton Parisinum (ms. Paris, Bibliothque Nationale de France, lat. 7530), edited
by Bergk [1845] = Bergk [1884] 580612; see also Suet., Gram. et rhet. 10, quoted above
(n. 233).
487 Theodorus and Tiberius: Mygind [1999], respectively 261262 and 289 (Nos. 29 and 154);
Bringmann [2002] 77. Aristocles: Mygind [1999] 264 (No. 36); Bringmann [2002] 77.
A short sketch of the influence of Hellenistic (especially Homeric) scholarship on Roman
intellectuals and poets of late Republican Age can be read in Pontani [2005b] 5759.
488 Suda 1184.
166 Montana
Diocles and on the basis of some alleged affinities of Latin with Aeolic, that
Latin is a Greek dialect.494 This trend in the dialectological inquiry fits well
with the more general growing tendency towards cultural integration of the
Roman and Hellenistic elites during the last decades of the Republic; a process
which emblematically also included learned-ideological discussions about the
descent of Rome from Greece (in competition with the Trojan thesis, particu-
larly upheld by scholars close to Pergamene patronage) and culminated in the
assumption of the Roman origin of Homer.495
As a matter of fact, the centrality of Rome as a political and military capi-
tal of the Mediterranean on the eve of the Principate was rapidly undergoing
a transformation into centrality as a new cultural capital, which prided itself
on its libraries and its formidable array of scholars in addition to a wealth of
opportunities for study, edition, copy and contamination of ancient textual
traditions. Indeed, Rome could now justifiably claim to be no less a presti-
gious protagonist of scholarship than the ancient seats of learning. The most
emblematic episode of this new rule of Rome, which eloquently illustrates the
intermeshing of political-military and cultural factors, is the earlier mentioned
question of the Aristotelian private library, the fate of which we can now
examine in its essential traditional lines.496 Athenaeus of Naucratis497 draw
up a long list of figures from Greek history who possessed rich private librar-
ies (a list also including Polycrates of Samos, the Athenian tyrant Pisistratus,
Euclid of Athens, Nicocrates of Cyprus, the kings of Pergamum, and the poet
Euripides), at the end of which he also mentions Aristotle the philosopher,
<Theophrastus>, and Neleus, who kept watch over the books of both the
latter. Athenaeus adds that
our king Ptolemy [II], called Philadelphus, after purchasing all of them
[i.e. those books: ...] from him [i.e. Neleus], had them
transported to the fine city of Alexandria together with the books pur-
chased in Athens and Rhodes.
494 Testimonies and fragments concerning Philoxenus are edited by Theodoridis [1976].
The main extant testimony is Suda 394: Philox. test. 1 Theodoridis. On Philoxenos as a
Homerist: Pagani [forthcoming]. On his linguistic theory and its reception: Lallot [1991b].
For On the dialect of the Romans see frr. 311329 Theodoridis.
495 This opinion is attributed by the Vita Homeri VI, 251.1823 Allen, to Aristodemus of Nysa,
perhaps the 1st century grammarian and rhetor, who was a son of Aristarchus pupil
Menecrates and among the teachers of Pompey the Greats sons in Rome and of Strabo in
Nysa (Str. 14.650). See Heat [1998]; Ascheri [2011].
496 Reference-study: Moraux [1973] 394.
497 Ath. 1.3ab.
168 Montana
506 Irigoin [1994] 5053. However, significant exceptions indicate that some of Aristotles
esoteric works were known to the Alexandrians: for instance, as has been seen, erudite
Aristotelian writings acted as the source for some of Eratosthenes inquiries and for
Aristophanes of Byzantiums hypotheseis of the dramatic plays. Furthermore, Aristophanes
composed an epitome of Peripatetic sources on zoological problems (On animals); nor
should one omit to mention the concepts of poetics, rhetoric and Aristotelian literary
criticism of which concrete traces remain in the fragments of the Hellenistic exegesis to
classical authors (cf. above, n. 74). See Moraux [1973] 1215 with n. 36.
507 Some symbolic implications are stressed by Nicholls [2011] 131, particularly with reference
to the trasfer and storing of Greek books from Alexandria and Athens to Roman libraries
during the first imperial age.
508 Str. 1.38; Suda 3036 (a biographical entry on Aristonicus father Ptolemy, he himself a
grammarian who lived in Rome). See Jolivet [2010].
509 Firstly Lehrs [18823] 28; Ludwich [18841885] 1.51; Schmidt [1854] 277; West [2001a] 4950;
cf. Pontani [2005b] 62; Razzetti [2010] 6061.
Hellenistic Scholarship 171
521 P.Oxy. 9.1174: Soph. fr. 314 Radt. See McNamee [2007] 366370.
522 FGrHist 633 fr. 1; Bagordo [1998] 6465 and 8788 (No. 14).
523 Suda 3924.
524 Razzetti [2010].
525 Suda 872, 399, 29.
526 Demetrius of Troezen (ap. Ath. 4.139c) and Suda 872 for the first figure; Sen., Ep. 88.37,
for the second. See Braswell [2013] 3639.
Hellenistic Scholarship 173
527 The rather old edition of Didymus fragments by Schmidt [1854], although in need of
thorough revision and up-dating, has not yet been supplanted as a whole. A reassessment
and a critical catalogue collecting 69 titles of attributed works are provided by Braswell
[2013] 40103.
528 Suda 872.
529 Apart from Schmidt [1854], these fragments have been collected by Ludwich [18841885]
1.175631.
530 Mentioned in the so-called Ammonius commentary on the Iliad P.Oxy. 2.221 (2nd cen-
tury AD), col. 15, ll. 1617, on Il. 21.290: . On
Seleucus, whose works include a and a , see Mueller [1891];
Reitzenstein [1897] 157166; FGrHist 341; Matthaios, this volume. The fragments have been
edited by Duke [1969]; cf. West [2001a] 4748.
531 Cf. schol. Hes., Theog. 126.
532 Schol. Pind., Ol. 5 inscr. a; Lactant., Div. inst. 1.22.19.
174 Montana
the scholia to which he is quoted roughly eighty times), Paeans,533 and proba-
bly Hymns, and Bacchylides Epinicia.534 As far as drama is concerned, his work
on Sophocles is well attested,535 traces of which also remain in the Medieval
scholia; his name appears in the scholia to Euripides tragedies and a subscrip-
tion to the Medea in a manuscript points to a commentary by Didymus on
this author;536 he also concerned himself with Ion and perhaps Achaeus. The
composition of commentaries on Aristophanes comedies is extensively testi-
fied to by Medieval scholia, which preserve more than sixty open mentions of
Didymus; it is likewise possible that Didymean hypomnmata were the source
from which derived, through reworkings and interventions of a compendiary
nature, two 2nd century fragmentary papyrus commentaries on unidentified
Aristophanean comedies.537 Furthermore he composed hypomnmata on
Cratinus,538 Menander,539 and possibly Eupolis.540
Among the prose writers, Didymus probably dedicated attention to
Thucydides, if one is to give credence to the citations of this scholar in Marcellinus
Life of Thucydides (4th? century AD).541 It would not be altogether surprising if
one were to find that Didymus, following in the footsteps of Aristarchus, took
an interest in Herodotus, as is conceivably suggested by an admittedly faint
trace of evidence detectable in an anonymous commentary attested to by a
papyrus scroll.542 There can be no doubt, on the other hand, that he concerned
himself with the Attic orators, above all Demosthenes, Aeschines, Hyperides,
Isaeus. Our main indirect source in this respect is the rhetoric lexicon of
Harpocration (2nd century AD), but we are lucky enough to have an exten-
533 Ammon., Diff. 231 Nickau. New edition of the fragments of Didymean exegesis to Pindar:
Braswell [2013]; cf. Braswell [2011].
534 Ammon., Diff. 333 Nickau; cf. Eust., Od. 1954.5.
535 Ath. 2.70c.
536 Ms. Paris, Bibliothque Nationale de France, gr. 2713.
537 P.Oxy. 35.2737 (commentary on Anagyros?) and P.Flor. 2.112, reedited by Montana [2012a].
On Didymus in P.Flor. 2.112 see Montana [2009d] 4454.
538 Schol. Ar., Vesp. 151.
539 Etymologicum Gudianum 338.25 Sturz.
540 Schmidt [1854] 308310; Storey [2003] 35.
541 Mazzarino [1966] 2/2 466 (hesitatingly scrutinized by Piccirilli [1985] 6768 and 8990)
argued that the Didymus quoted in 3, 16 and 32 of Marcellinus Life could well be
Didymus Claudius, the Greek grammarian presumably active in Rome in the 1st century
AD and author of a writing On the mistakes of Thucydides relating to analogy, rather than
the Chalcenterus. This assumption has been rejected by Arrighetti [1968] 97 n. 94 and
[1987] 226227 n. 202; cf. Porciani [2001] 45 n. 106.
542 P.Oxy. 65.4455 (3rd century AD). See Montana [2009a] 253254; Braswell [2013] 40 n. 48.
Hellenistic Scholarship 175
sive papyrus fragment (P.Berol. 9780 of the 2nd century AD) of a work bearing
in the colophon the title Didymus On Demosthenes ( ),
which lists a series of lemmata drawn from the Philippicae (Speeches 911 and
13) accompanied by explanatory notes.543 The peculiar arrangement of the
material and its manner of selectively discussing the literary text, with marked
emphasis on historical sources and antiquarian erudition (Sachphilologie), has
raised questions concerning the integrity and the nature of the text that has
come down to us: it is unclear, firstly, whether it is the original or an epitome
and, secondly, whether we are dealing with a hypomnma or a syngramma.544
However, these doubts can be dispelled if gaps and presumed omissions in the
exegesis are explained with the assumption that Didymus used an Alexandrian
edition of Demosthenes speeches wherein contrast to the collection that
has come down to us through the Medieval manuscriptsthe Philippicae
occupied a very advanced position. If this is the case, then it would hardly
be surprising that, as he proceeded with the commentary, he avoided reiter-
ating explanations, some of which lengthy and complex, already provided in
earlier parts of his own hypomnma.545 That is to say, we would be dealing
with a Didymean hypomnma in its original facies. The unusual circumstance
of being able to examine a textual exhibit belonging to the direct tradition
explains why it has been used extensively in (even highly divergent) assess-
ments of the quality of Didymus philology.546
Throughout the prolonged period of Alexandrian scholarship, literary exe-
gesis was accompanied by the study of vocabulary, or lexicography. Didymus
was no exception in this regard, as he was the author of a comic and a tragic
lexicon ( and ), both quoted by the lexicographer
Hesychius (5th/6th century AD) in the epistle to Eulogius prefatory to his own
Lexicon. We may gain an idea of the extensive circulation and literary fortune
543 Edited by Pearson-Stephens [1983]; cf. Gibson [2002]; Harding [2006], with commentary.
Cf. Braswell [2013] 80 (No. 38).
544 The view that it is (a collection of excerpta from) a hypomnma has been espressed by
the editores principes Diels-Schubart [1904] and, more recently, by West [1970]; Arrighetti
[1987] 203204; Gibson [2002] 1325 and 5169. In contrast, the opinion that it is a syn-
gramma was put forward by Leo [1904], and it is also supported by Harding [2006] 1320.
545 Thus Luzzatto [2011], who takes to its extreme consequences a neglected suggestion put
forward by Blass [1906] 284292 (No. 231) and who regards this hypomnma as a tool
designed for erudite consultation.
546 The papyrus text, in the opinion of West [1970], is very likely to be the original Didymean
hypomnma and thus testifies to the sloppiness of its author. By the same premise, cor-
rected as said above, the work, on the contrary, illustrates the qualified scholarship of
Didymus in the view of Luzzatto [2011]; cf. Montana [2009c] 163166.
176 Montana
of these Didymean tools during Graeco-Roman antiquity from the fact that
Galen (2nd century AD) made use of the fifty books of Didymus comic lexis
and drew up an epitome of it in six thousand lines, maintaining that
558 West [1970]; Harris [1989]. Well-balanced assessment of Didymus philology can be read
in Montanari [1997f]; Braswell [2011] 196197; cf. Braswell [2013] 11. Surveys on divergent
appraisals: Gibson [2002] 5169; Harding [2006] 3139; cf. Montana [2009c] 163166.
559 P.Oxy. 9.1174: Soph. fr. 314 Radt = Theon frr. 1935 Guhl. See McNamee [2007] 366370.
560 Etymologicum Magnum 696.712: Theon fr. 14 Guhl; Etymologicum Gudianum 376.1920
De Stefani: Theon fr. 15 Guhl. See Guhl [1969] 1113; Pontani [2005b] 63.
561 P.Oxy. 24.2390, a hypomnma on Alcman where a Tyrannion (the grammarian of Amysus?
fr. 61 Haas) is also quoted: cf. Alcm. fr. 5 Page = fr. 5 Davies.
562 P.Oxy. 37.2803: Stesich. frr. S133147 Page = frr. S133147 Davies. Cf. McNamee [2007]
373375.
563 Exegetical work at least on Paeans, Olympians and Pythians is attested to by references
respectively in P.Oxy. 5.841 (McNamee [2007] 315343), schol. Pind., Ol. 5.42a, and P.Oxy.
31.2536 (fragments of a hypomnma with subscription: Theon fr. 38 Guhl). Some aspects of
congruence between papyrus annotations and corresponding medieval scholia to Pindar
could indicate that Theons commentary should be dated earlier than that of Didymus,
from which the scholia are generally believed to have derived: McNamee [1977] 6465;
Benelli [2013] 619. The reverse opinion is maintained by Deas [1931] 34.
564 Marginal notes in P.Oxy. 25.2427, containing fragments of Epicharms Pyrrha and
Promatheus ( ), fr. 113 Kassel-Austin; cf. McNamee [2007] 245247.
565 Bongelli [2000] 281.
566 Attested to in the epistle to Eulogius prefatory to Hesychius Lexicon.
567 Phryn., Att. 355, on ; Hsch. 1031, s.v. . Theon is, in addition to his father
Artemidorus and Didymus, among the candidates for authorship of some items included
in the 2nd/3rd century AD papyrus lexicon P.Oxy. 15.1801, containing mainly comic
Hellenistic Scholarship 179
for the poets of the early Hellenistic era, which perhaps even became his pre-
dominant focus of interest. This interest was not an innovation, if it is true
that Aristophanes of Byzantium perceived issues of philological relevance in
the plays of Menander and as early as the 2nd century some scholars occa-
sionally drew on their familiarity with the texts of Hellenistic poetry, at least
as a reservoir of useful parallels, in commenting on more ancient authors
and, finally, Asclepiades of Myrlea devoted some of his work to Theocritus
and hypothetically to Aratus and Apollonius of Rhodes. However, Theons
assiduous exegetical production is not only better attested, but it also seems
to be distinguished by its broader scope and by the greaterindeed epoch-
makinginfluence it exerted on the later textual and interpretive tradition.
There is some positive evidence concerning Theons work on Theocritus,
including plausibly an edition, subsequent to the bucolic corpus drawn up by
his father Artemidorus,568 and definitely a commentary;569 it is plausible to
assume that significant remains of the latter are preserved in several marginal
annotations in Theocritean papyri of Imperial Age and Late Antiquity.570 A few
fragments survive of Theons commentary to Callimachus Aitia571 and there is
reason to believe that he also composed a comment on Hecale572 and Hymns.573
The well known subscription at the end of the Argonautics of Apollonius of
Rhodes in the manuscript Firenze, Biblioteca Medicea Laurenziana, Plut.
32.9 (10th century) attests that the scholia copied there are drawn from (the
lemmata: Naoumides [1969] (No. 4); Esposito [2009] 291; re-edition of Aristophanean
entries: Esposito [2012].
568 AP. 9.205 (Theoc., Epigr. 26).
569 Orion by Etymologicum Gudianum 323.1821 De Stefani: Theon fr. 1 Guhl.
570 In particular P.Oxy. 2064 (published by Hunt-Johnson [1930]) + 50.3548 (MacNamee
[2007] 427442; see also Meliad [2004b]) and P.Antinoe s.n., the so-called Antinoe
Theocritus (MacNamee [2007] 109112, and 376427 for the text; cf. Montana [2011b];
Meliad [forthcoming]). See Belcher [2005] 194 and 200 (texts); Pagani [2007c] 288290
and 298299 (texts).
571 Etymologicum Genuinum AB 1198 Lasserre-Livadaras: cf. Callim. fr. 383 Pfeiffer = fr. 143
Massimilla, and Hecale fr. 45 Hollis = Theon fr. 2 Guhl; Etymologicum Genuinum AB 207
Lasserre-Livadaras: cf. Callim. fr. 42 Pfeiffer = fr. 49 Massimilla = Theon fr. 5 Guhl;
Etymologicum Genuinum AB 1316 Lasserre-Livadaras: cf. Callim. fr. 261 Pfeiffer = Theon
fr. 6 Guhl.
572 Callim. fr. 261 Pfeiffer, quoted by Theon in his commentary on the second book of the
Aitia (see previous n.), actually belongs not to the Aitia, but to the Hecale (fr. 71 Hollis);
and Callim. fr. 383 Pfeiffer = fr. 143 Massimilla (also quoted in the previous n.) is cited in a
couple with Hecale fr. 45 Hollis, possibly by the same Theon. This would denote Theons
familiarity with the Callimachean epyllion: Bongelli [2000] 284.
573 Bongelli [2000] 284290.
180 Montana
dialects were believed to be derived),586 then the main ethnic dialects (cer-
tainly Doric and, possibly, Asian Aeolic), and finally local languages subsum-
able under the latter (Laconian, Argolic, Himerian, Rheginum, Syracusan, and
so forth). A second treatise, or series of treatises collectively recorded under
the title The dialects in Homer, Simonides, Pindar, Alcman and the other lyric
poets, plausibly focused on a blend of local language varieties (this appearing
to be the meaning of the plural dialects) in works of archaic epic and choral
poetry.587 In all likelihood these writings also touched on questions of style.
Tryphon devoted other special treatises, later used and frequently cited
by the 2nd century AD grammarian Apollonius Dyscolus, to formal descrip-
tion of the individual parts of speech and to a number of relevant flectional
aspects: noun, verb, participle (considered as an independent part, intermedi-
ate between noun and verb), article (a definition including relative and indefi-
nite pronouns), preposition (in its own right and also as a nominal and verbal
preverb), adverb (including some interjections and particles), conjunction.
The explored noun formation and the
dealt with comparatives. Tryphons lexicographic research found expression
in a collection of explained names ( ), some fragments of which,
referring to musical instruments, are preserved by Athenaeus; other titles of
which we have knowledge, (Names) of plants () and On animals (
), could refer to different sections of the latter.588
What can be reconstructed of Tryphons doctrine from this great mass of
indirect and fragmentary information is that by developing his approach in
the framework of a firm belief in analogy, he on the one hand played a role
in redefining the overall body of knowledge concerning the fields of prosody
and orthography,589 while on the other he was the first to conduct an organic
and systematic study of dialectology and the parts of speech, thus earning
repute among the beginners of ancient normative grammar and making a con-
tribution which would prove to be useful, in the early Imperial Age, for the
advance of syntaxis as an autonomous branch.590 His influence on these fields
591 Anecd. Gr. Ox. 3.269.28 Cramer; schol. Dion. Thrax, in GG 1/3.356.22. See Matthaios, this
volume.
592 The present outline has benefited from the contents of the digital project Lessico dei
Grammatici Greci Antichi, directed by Franco Montanari and co-directed by Lara Pagani
and myself (http://www.aristarchus.unige.it/lgga/). I wish to express my thanks to all of
the colleagues and friends who are contributing to the implementation of this resource
in progress.
chapter 3
Stephanos Matthaios
When faced with the task of writing the history of Greek philology and
grammar1 in the Imperial era and Late Antiquity, a period stretching over
almost six centuries,2 a number of difficulties must be addressed even in a
summary outline such as this. The greatest problem, which also arises through-
out the entire field of ancient philology and grammar, is the preserved source
material and the state of its transmission. With very few exceptions, the origi-
nal writings of the ancient philologists and grammarians have been lost. In
their stead, we have to be content with occasional or, in the best cases, a small
number of fragments and testimonies from later secondary sources. Therefore,
our picture of the contents and developments of ancient scholarship is incom-
plete and based upon the accidental transmission of evidence. Moreover,
many of the philologists and grammarians in this period remain shadowy fig-
ures, whose place of activity or the contents and purpose of their writings are
largely unknown.
In addition to the insufficient and scant transmission of the source mate-
rial, the state of current research on the historical and theoretical aspects can-
not be described as satisfactory. Study of the philology and grammar of the
Imperial era and Late Antiquity has profited relatively little from the expan-
sion and intensity of investigations which, over recent decades, have focused
on the Hellenisticabove all the Alexandrianperiod, largely motivated
by the interest of scholarship in the learned character and the philological
background of Hellenistic poetry.3 With regard to the Imperial era and Late
1 When here and below I speak of philology and grammar, I am in no way referring to two
different sciences, but to the two essential areas of one and the same discipline called
; on the notion , see below in 2.3 and Swiggers-Wouters
(section II.2) in this volume. To avoid misunderstandings and to distinguish the expression
from the present-day meaning of the term grammar, the ancient form will be rendered in this
article as philology and grammar. But occasionally, the equivalent expression grammar,
enclosed in single quotation marks, is also used. The same applies to the representatives of
the ancient discipline, the , who are referred to herein mainly as philologists and
grammarians, but if at times I often speak simply of grammarians, this characterization is to
be understood as encompassing both areas, namely philological interpretation and linguistic
analysis.
2 The period to which the present paper refers is specified and explained below in 2.1.
3 On the philological background of Hellenistic poetry, which finds its expression clearly in
the double identity of its representatives, the so-called poets and philologists (
186 Matthaios
Antiquity, on the other hand, the existing collections of fragments and text
editions, insofar as complete works have been transmitted, involve a very lim-
ited number of philologists and grammarians from this period, with most of
the studies dating back to the 19th and early 20th century. At that time, such
works were seen as accompaniments to research that was essentially stimu-
lated by other achievements in the field, especially by the monumental edition
Grammatici Graeci [18671901] and the (first) editions of ancient commentar-
ies and scholia to the classical authors.4
Interest in the history of Imperial and Late Antique philology and gram-
mar in its own right began to increase as from the mid 20th century, basi-
cally due to the work of H. Erbse on the edition of the scholia on the Iliad and
W. W. Kosters project of a new edition of the Aristophanes scholia.5 These proj-
ects breathed new life into the whole field of ancient scholarship and thrust it
into the foreground of vivid and intensive research activity. Yet with the excep-
tion of a small number of dissertations and studies on specific Imperial gram-
marians, mainly inspired by H. Erbse,6 the period in question has benefited
only sporadically. This outcome can be inferred from the program that formed
the basis for the series Sammlung griechischer und lateinischer Grammatiker
(SGLG): founded expressly for the publication of editions of ancient philo-
logical and grammatical writings, a large proportion of the editions published
in this series so far have consisted of lexicographic works from the Imperial
), see Matthaios [2008] 549569 and [2011a] 8185. On this question, see also the
contribution of Montana in this volume, especially 1.1, 2.4.
4 On the editions and studies in the field of scholia up to the beginning of the 20th century, see
the survey by Gudeman [1921]. The lexicographic works that were published in this period,
such as the edition of Suidas Lexicon by Adler [19281938] and of Pollux Onomasticon by Bethe
[19001937], are of central importance. These editions basically resulted from Reitzensteins
investigations into Late Antique lexicography, especially Byzantine Etymologica [1897].
5 See Erbse [19691988] and also his special study on the transmission of the Iliad scholia
[1960]. In 2007, the final volume published by D. Holwerda completed the project of a new
edition of the Aristophanes scholia initiated by W. W. Koster in the year 1969.
6 I mention here as an example Guhls collection and commentary on Theons fragments
[1969]on Theon, see below in 3.3, and Montana in this volume, Nickaus studies on
the so-called Ammonius lexicon [1960] and the edition of this lexicon that followed [1966]
on this lexicon, see below in 6.2as well as the study of Alpers [1964] on Theognostus
orthographical work. Under H. Erbses supervision, Fischer [1974] edited Phrynichus Ecloga
and Neitzel [1977] collected the fragments of Apions work (Homeric
glosses). H. Erbse also stimulated Blanks edition [1988] of Lesbonax treatise
(On figures) and Dycks edition of the Homeric epimerisms [19831995].
Greek Scholarship in the Imperial Era and Late Antiquity 187
period and Late Antiquity,7 whereas other philologists and grammarians from
the same time span are poorly represented. This is in sharp contrast to the edi-
tions and collections of fragments concerning the Alexandrian predecessors.8
Unlike the philologists, the language theoreticiansthe grammarians in
the modern sensefrom the Imperial era and Late Antiquity have received
intense scholarly scrutiny since the mid 20th century. This phenomenon has
mostly been prompted by studies in the area of the philosophy of language and
the historiography of modern linguistics. A similar interest, however, can effec-
tively also be seen within classical philology, especially in the context of the
very intensive debate that began around the mid 20th century on the author-
ship and authenticity of the grammatical manual attrib-
uted to Dionysius Thrax.9 Discussion of the development of ancient linguistic
theory has, as expected, also involved the Imperial and Late Antique periods
of ancient grammar, whose main representatives, Apollonius Dyscolus and
Herodian, are the two figures that have profited most significantly from this
renewed interest. Apollonius Dyscolus has even been honored with new edi-
tions and modern commentaries as well as with a number of special investiga-
tions into his doctrine.10 At the same time, our knowledge on the development
of ancient linguistic theory has grown considerably, thanks to the evidence
from a series of grammatical papyri, dating mostly from the Imperial era and
Late Antiquity. Thus one of the achievements of recent research has involved
examination of papyrus sources in order to investigate their theoretical rela-
tionship with authored works; accordingly, the contents of these grammatical
papyri can now be placed in the context of the development of grammatical
doctrine.11
7 In general, the field of Atticist lexicography has met with special interest in recent research
through the studies on rhetoric during the Imperial era and the Second Sophistic; cf.
below in 6.4.
8 See for instance the edition of pseudo-Herodians On figures by Hajd [1998]. Apart from
this work published in the series SGLG, a new edition of the fragments of the grammarian
Epaphroditus was published by Braswell-Billerbeck [2008].
9 For an overview of the discussion concerning the authenticity of the Techne ascribed
to Dionysius Thrax, see Pagani [2011] 3040 with reference to older bibliography on this
subject; cf. also Matthaios [2009a] and Pagani [2010b].
10 Editions, commentaries and translations of Apollonius works are listed below in 5
n. 376; cf. also the bibliography in the section on Apollonius Dyscolus ( 5).
11 Wouters [1979] and [1988] edited and commented a great amount of the grammatical
manuals transmitted in papyri.
188 Matthaios
The field of philology, i.e. the interpretation of literary texts through com-
mentaries and the investigation of their language through lexicographic col-
lections, has likewise profited from the increase in new papyri fragments and
the insights to be gained from them.12 Yet despite these encouraging advances,
particularly in a field that is often considered as marginal against the backdrop
of mainstream classical philology, important aspects of advanced research
still await urgent attention. These include background work, in particular the
renewal and actualization of the textual basis underlying the production of the
Imperial philologists and grammarians. Such advances can only be achieved
through new editions and commentaries, capable of closing the many gaps
that still exist.
In light of the above, it is evident why general or comprehensive over-
views of the history of philology and grammar in the Imperial era and Late
Antiquity are still awaited. Compared with the complete documentation and
investigation of the Hellenistic epoch of ancient philology, the contrast with
the works available for the subsequent periods is sobering. This is mainly due
to the regrettable fact that there has been little effort to replicate the scope
and quality with which Pfeiffer analyzed the Hellenistic period in his monu-
mental History of Classical Scholarship [1968].13 Readers interested in Imperial
and Late Antique scholarship as well as specialists are generally dependent on
portrayals in older reference works, especially in histories of ancient Greek lit-
erature. Such overviews are primarily the relevant chapters in Schmid-Sthlins
Geschichte der griechischen Literatur.14 While not intended to be comprehen-
sive, they offer the most informative survey of the developments in the field
during the whole period.
It would be remiss in this context to pass over Grfenhans four-volume
Geschichte der klassischen Philologie im Alterthum [18431850] without even a
mention, as is unfortunately often the case in recent literature.15 Grfenhan,
12 Related to the study of the new papyri material is the project Commentaria et lexica
Graeca in papyris reperta (CLGP). Special studies on this body of texts, such as that by
Trojahn [2002] on the papyri containing commentaries on ancient comedy, are especially
welcome.
13 The first volume of the work deals with the 1st century, the period of the so-called Epigoni
in the history of the Alexandrian philology, whereas the second part [1976] treats the
Renaissance and early modern period.
14 For Imperial and Late Antique philology and grammar the following sections are relevant:
SchmidSthlin [19201926] I 425446 (from 146 BC to 100 AD), II 866896 (100300 AD)
and II 10751094 (300530 AD).
15 Philology and grammar during the Roman Empire are treated in the third volume of
Grfenhans Geschichte.
Greek Scholarship in the Imperial Era and Late Antiquity 189
and later Sandys, are so far the only scholars to have provided a systematic
analysis of the field. However, Grfenhans portrayal only extends to the 4th
century. Furthermore, although it represents an admirable achievement for his
time, it must also be admitted that he presents the rich material in a confusing
manner, often providing a tangled web of details, so that the systematic aspect
suffers, and at the same time the explanation of the historical and cultural
context is neglected or poorly clarified. Sandys, on the other hand, treated the
entire period in question in the first volume of his work A History of Classical
Scholarship [19213]. His presentation, however, is no real history of scholarship
itself, as Pfeiffer conceded,16 but rather an incomplete catalogue of figures,
whose philological and grammatical achievements are vaguely assessed.17
Beyond these comprehensive overviews, we are by and large dependent
upon shorter surveys either in specialized works, such as Gudemans Grundriss
der Geschichte der Klassischen Philologie [1909], or in recently published
histories of literature18 and introductory works on the study of classical
philology.19 N. G. Wilsons Scholars of Byzantium [1996] and the second
volume of H. Hungers history of Byzantine literature [1978], which addresses
the philological and grammatical activities of the Byzantines in a separate
chapter, are also relevant in this context. The interested reader can draw cur-
sory information from these works concerning the 5th and 6th centuries of
ancient scholarship. On the ancient education system and the role it assigned
to philology and grammar, the studies of R. Cribiore [1996] and [2001] as well
as T. Morgan [1998] are particularly important,20 even if they cover only parts
of the period in question here. In his Guardians of Language. The Grammarian
and Society in Late Antiquity [1988], R. A. Kaster analyzes the profile of the
grammarian from a socio-cultural perspective, while at the same time provid-
ing a brilliant contribution to the prosopographic analysis of the philological
and grammatical field. His presentation, however, goes no further than the
4th and 5th centuries of philological-grammatical activity. For the reconstruc-
tion and analysis of the writings and the scholarly profile of the philologists
and grammarians active throughout the entire period under consideration, the
relevant articles in Paulys Realencyclopdie are still indispensable.21
While the above discussion presents the current situation of research in
detail and highlights the lack of editions and commentaries, especially the
lack of a comprehensive survey of the philological and grammatical studies
during the Imperial period and Late Antiquity, the emphasis on its failings is
not intended to act as the defining characteristic of the task involved in this
contribution. Rather, its aim is to underline the challenging limits of this
undertaking and to demarcate the boundaries of the present attempt. Thus
the paper does not intend to provide an overview encompassing all aspects of
a temporally protracted, and geographically, culturally and politically varied
period in the history of ancient philology and grammar. Instead, in line with
its nature as a presentation designed for a Companion, it is historically oriented
and built around the main points of the ancient , aiming above all
at providing basic knowledge concerning the development of the discipline in
the Imperial era and Late Antiquity by focusing on its contents and contexts.
In order to clarify the demarcation boundaries of this undertaking, I will
begin by mentioning the main concessions and compromises made for the
treatment of the subject, with regard to which I count on the readers forbear-
ance. The most important of these restrictions concerns the scope and per-
spective of the intended presentation. In a study that makes ancient philology
and grammar its subject, the attribute ancient indicates the character of
the analysis and constitutes its methodological starting position. If a scholar
takes to heart Pfeiffers definition of philology as the art of understanding,
explaining and restoring the literary tradition ([1968] 3), and also embraces
the ancient philologists and grammarians understanding of their area of
responsibility,22 then the contents of the present study can be expressed more
precisely: the focus falls on the interpretation of ancient literature and refers
to those persons who have accomplished this specific task. This results in the
total exclusion of Christian literature, which arose during Late Antiquity, but
also of its exegesis, which, according to the modern understanding, is con-
sidered to belong to philology. However, given the nature and objectives of
ancient scholarship, the exclusion of interpretive activity on Christian authors
21 Actualized bibliographical references for each philologist and grammarian are supplied
in the relevant entries in Der Neue Pauly and Brills New Pauly. An essential contribution
to prosopographical investigation of ancient philologists and grammarians is provided
by the online databank Lessico dei Grammatici Greci Antichi (LGGA), established and
directed by F. Montanari (University of Genoa).
22 The contents of the ancient and the tasks of philologists and grammarians
during the Roman Empire and Late Antiquity are discussed below in 2.3.
Greek Scholarship in the Imperial Era and Late Antiquity 191
23 This subject area is described by N. Wilson in Nesselrath [1997] 99101. Wilson, however,
did not take it into account in his study of Byzantine philology [1996]; see Wilson [1996]
in his Note to the revised edition, 1996: The coverage is admittedly not complete,
and possibly Classical Philology in Byzantium would have been a more precise title;
I recognise the possibility that someone may show that the Byzantines reached a high
level in theological scholarship. On this point, see the criticism put forward by Alpers
[1988] in his review of Wilsons study. Hunger [1978 likewise restricts his presentation of
Byzantine philology to the interpretation of ancient literature.
24 See Sext. Emp. Math. 1.76; on the wording of Chairis definition, cf. Sch. Dion. T. 118.10
12. Sextus cites this definition mistakenly under the name of the grammarian Chares;
see Blank [1998] 137 with n. 105. For an interpretation of Chairis definition, see Blank
[1998] 137140 and Matthaios [2011a] 7273; cf. also the contribution of Swiggers-Wouters
(section II.2) in this volume.
25 For an overview of the technical and scientific literature in the Roman Empire and Late
Antiquity, see Schmid-Sthlin [19201926] II 446455, II 896925 and 10941100.
192 Matthaios
here, and it cannot be taken into account in this presentation.26 In effect, these
texts were rarely studied by grammarians; moreover, they were not perceived
by the latter as forming part of the literary tradition.
The Imperial era and Late Antiquity also saw the growth of strong con-
tacts between philology and rhetoric. But the difference lies primarily in the
research perspective and objectives of each discipline. Rhetoric defined the
program for contemporary literary production and was interested in ancient
prose texts and oratory, insofar as they served the purpose of educating the
speaker and provided the benchmark for assessing the new texts being com-
posed. This explains the rhetoricians engagement with ancient prose authors,
especially with the Attic orators, whose activity was becoming strongly linked
with the domain of rhetoric at that time. Literary criticisma subject which
by definition belonged to the tasks of ancient philology27was almost exclu-
sively cultivated by rhetoricians during this period. Consequently, poetry was
regarded as the sole area of philological-grammatical activity.28 With the
exception of the Sophists and rhetoricians who were active in a traditionally
philological area, that of lexicography, Imperial rhetoric and literary criticism
will only be considered here to a limited extent.29
Finally, a history of philology and grammar in the Imperial era and during
Late Antiquity should also take into account the relationship between Greek
and Latin speakers in the Roman Empire. In the overall context under consid-
eration here, the teaching of Latin and the Latin grammarians active in the
eastern provinces would comprise special aspects of such an analysis. Such a
study, however, is beyond the scope of this survey.30
The present contribution is divided into two main sections, an introduction
and a prosopographical section. The introductory part discusses the character
and contents of Imperial and Late antique philology and grammar by plac-
ing them within their historical, cultural and scientific background as well as
in the context of the ancient educational system. The prosopographical sec-
tion takes into account both historical and systematic aspects. The grammar-
ians of each period are classified according to the major focal points of their
26 On the interpretation of philosophical texts during this period, see N. Wilson in
Nesselrath [1997] 101102; on the commentating tradition of the Platonic and Aristotelian
philosophy, see Dickey [2007] 4651; cf. also the contribution of Lapini in this volume.
27 See below in 2.3.
28 Concerning the points of contact between rhetoric and philology and the ancient view on
them, see Wolf [1952] 3639.
29 This area is presented in the contribution by de Jonge in this volume.
30 This and other aspects are dealt with in the study by Rochette [1997]. On bilingual
language instruction, the introductory remarks in Dickeys edition [2012] of Ps.-Dositheus
Colloquia are significant.
Greek Scholarship in the Imperial Era and Late Antiquity 193
31 Several grammarians from the 4th and 5th century AD who were mainly active as teachers
of grammar, but left no writings are discussed by Kaster [1988]. On the other hand,
grammatical manuals and other material on papyri have been transmitted anonymously.
The grammatical treatises are edited by Wouters [1979]; several school exercises deriving
from the school and education activity in Greco-Roman Egypt are listed by Cribiore
[1996] 173284.
32 Immediately following the initial mentioning of the person to be treated, reference
bibliography, but also editions and collections of fragments of their works are cited in a
footnote.
33 For a definition and specification of the term Late Antiquity, see Inglebert [2012]; cf.
E. Pack in Nesselrath [1997] 435436 as well as R. Klein in Christes-Klein-Lth [2006]
2327.
194 Matthaios
symbolic significance, in the sense that the prolonged struggle between pagan
and Christian culture was decided to the advantage of the latter.
Approximately half way through the period lying between these two mile-
stones, a second event took place, which had a lasting effect on the course
of the Greek cultural history. This was the overthrow of Rome, which was
ousted from its position as the imperial capital, a process that had begun with
Diocletian and was sealed by Constantine in 330 AD with the establishment of
Constantinople as the new capital of the Roman Empire. Overall, the reigns of
Diocletian (284305) and Constantine (306337) shaped an important histori-
cal turning point, which became evident above all in the history of literature
with the advent and constant growth of Christian literature. In response to the
crisis-ridden experiences of the 3rd century, which were due partly to the bar-
barian invasions and partly also to the continuing riots and unrest within the
empire, Diocletian drew up an extensive plan to reform both state and society
in order to strengthen the internal and external unity of the empire. Diocletians
project was then further developed and ultimately brought to completion by
Constantine. Under the reign of Constantine the transition from Principate to
Dominate was brought about, eventually resulting in the transformation of the
Roman Empire into an absolute monarchy. Constantine began with a policy
of tolerance toward Christians, with a number of moves that led to a decisive
favoring of Christianity: in so doing, he prepared the conditions that gave rise
to Justinians Caesaropapism. In effect, the historical break between the two
periods dealt with here, and thus the beginning of so-called Late Antiquity,
can essentially be dated to the reigns of Diocletian and Constantine. The
first centuries of Late Antiquity, however, are claimed by both Classicists and
Byzantinists as their own domain. Depending on the respective perspective
from which these early centuries are viewed, they are regarded by Classicists as
the end of the ancient world or, by Byzantine scholars, as the beginning of the
Byzantine era and consequently as belonging to the early Byzantine period as
far as its (literary) history is concerned.34
Turning now more specifically to the history of literature,35 classicism and,
in a strengthened and clearer form, the atticistic movement, were predominant
in the first sub-period, i.e. during the Imperial era. Classicism in its role as a
34 On the question of the beginning of Byzantine literature, see Shepard [2008] 2126; cf.
Haldon [2000] 1532, who regards the 4th and 5th centuries as the period of transformation
of the Roman world and assigns it to the Byzantine millennium. See also James [2010] 18.
35 The main characteristics of the Roman Empire and Late Antiquity from a literary and
historical perspective are sketched by Schmid-Sthlin [19201926] II 663671 and 943
956; see also the introductory chapter of Dihle [1989] 1374. The Imperial era and Late
Antiquity are systematically presented by Cambiano-Canfora-Lanza [1994].
Greek Scholarship in the Imperial Era and Late Antiquity 195
poets and scholars, without significantly sharing the contribution of the latter
in injecting fresh energy into contemporary poetry.
Now, if one inquires into the position of philology and grammar in the his-
torical and literary context that has been briefly sketched here, searching at
the same time for criteria that could lead to the periodization of scholarship
during the Imperial era and Late Antiquity, it becomes clear that philology and
grammar were strikingly coherent in their disciplinary character throughout
the entire period. That is to say, one cannot speak of a specific self-awareness
of an epoch, on the basis of which Imperial and Late Antique philology and
grammar could, from a theoretical point of view, be separated from the pre-
ceding and following periods. On the contrary, philology and grammar dur-
ing this period display a marked continuity that links the field seamlessly with
the earlier Hellenistic period and leads it just as seamlessly to the Byzantine
epoch. Change affected only the external conditions, resulting in an expansion
and reorganization of philology and grammar in the new centers of scholar-
ship and culture, simultaneously also strengthening the position of the disci-
pline in educational and intellectual life. The main criterion that forged strong
bonds in the philological and grammatical discipline throughout the extended
period of the Roman Empire and Late Antiquity and determined their history
was exclusively their institutional status and integration into the ancient edu-
cational system, which also guaranteed the position of the field within politi-
cal and socio-cultural life.
Accordingly, study of the history of philology and grammar in the Imperial
era and Late Antiquity must take as its basis the terms school and institution
which refer to this discipline. Any differentiations and specifications within
the history of the subject during this extensive period arise only from the crite-
rion institution and are due to the development of the philological and gram-
matical subjects. This aspect will be considered in further detail below.
systematically during the last three centuries BC, especially in the scholarly
center of the above city.40 The final period of the history of Hellenistic philol-
ogy and grammar is marked by the accomplishments of two figures: Didymus
and Tryphon. Both scholars were active at about the same time in Alexandria
and by virtue of their extensive works they are justifiably associated with the
perfection of the field. Didymus significance lies primarily in his wide range of
philological studies. Tryphon, on the other hand, is important for his extensive
studies in the area of grammar in the strict sense, i.e. the study of language, its
structure and the conditions of its correct application.41 But what is of particu-
lar interest here is to inquire into the origin of this division of areas of work
within ancient philological and grammatical doctrine, and its significance for
the further development of the discipline during the Imperial era and Late
Antiquity.
Philology and grammar in Antiquity did not constitute two different aca-
demic fields, but two areas of responsibility in one and the same discipline,
which claimed for itself the designation (grammar).
However, whereas grammar in the narrow sense, i.e. linguistic theory, initially
stood at the service of philology, textual criticism and the interpretation of
literature, it later gained its autonomy, insofar as language began to be stud-
ied and systematically described independently of the interpretation of the
literary text at hand. Given this shift in focus, a distinction is usually drawn
between two periods in the history of Hellenistic, especially Alexandrian phi-
lology. The first period refers to philologists such as Zenodotus, Aristophanes
of Byzantium and Aristarchus, while the second includes language theoreti-
cians such as Tyrannion and Tryphon in the 1st century BC and, in the AD era,
figures such as Apollonius Dyscolus and Herodian from the 2nd century. The
temporal boundary between these two main periods of ancient scholarship
and grammar is approximately the turn of the 2nd to the 1st century BC. This
is mainly due to the situation mentioned above: whereas grammar during the
first period supplied the instrument for philological activity and was tied to
the explanation of literary texts, in the second stage a demand for theoretical
40 The history of Hellenistic philology and grammar, especially that of Alexandria and
Pergamum, are treated in the monumental work of Pfeiffer [1968] 87279. The same
period is discussed in the contribution of Montana in this volume; see also Matthaios
[2008].
41 On Didymus accomplishments, see Pfeiffer [1968] 274279. Both scholarsDidymus and
Tryphoare presented by Montana in this volume.
198 Matthaios
elaboration and systematization of the linguistic theory was felt, and this soon
found expression in special monographs and textbooks.42
The contents and tasks of the ancient began to be systematized
very early.43 The available evidence suggests that the systematization pro-
cess was initiated by Dionysius Thrax, whose grammatical manual entitled
(Instructions),44 defined the and determined its
(parts).45 According to his definition, grammatical science aimed to
analyze what has been said by poets and prose writers, i.e. to explore the liter-
ary contents and the manner of their expression.46 As emerges from the list of
specific tasks, the philological interpretation includes some categories aiming
at a linguistic approach, such as finding etymologies ( ), and
calculating analogies for the purpose of assigning the words to their correct
inflectional paradigm ( ). The linguistic approachgram-
mar in the modern sensewas intended to be merely an instrument aiding
philological comprehension and the interpretation of literary contents.
The language description fulfils this same function in Asclepiades of Myrlea,
a contemporary of Dionysius, and later in Tyrannion, even when these two
scholars tried to present the parts of philology and grammar in a rather more
systematic form than that of Dionysius. According to the testimony of Sextus
Empiricus (Math. 1.9196; cf. Math. 1.252), Asclepiades divided philology
into three parts (); a) a technical part ( )the systematic
description of language, b) a peculiar part ( )philological
and textual criticismand finally c) a historical part ( )the
47 On Asclepiades classification of grammar, see Glck [1967] 1723, Blank [1998] 146149
and [2000] 407413.
48 This so-called four-part system is mentioned in the Sch. Dion. T., GG I/3, 10.810, 123.1315,
164.911 and 170.1820; the four parts are all listed in Sch. Dion. T., GG I/3, 12.38 and
115.89; see Blank [2000] 408. The fact that the four parts (lectio, enarratio, emendatio,
iudicium) were known to Varro speaks for the age of this system; see Var. fr. 236 Funaioli
and Glck [1967] 19 and 21 with n. 3. On the position of this system within grammatical
education, see below in 2.4.
49 The concrete contents of the grammatical theory are listed by Sext. Emp. Math. 1. 92;
on Sextus testimony, see Ax [2000] 9798 and 128129. See also the contributions of
Swiggers-Wouters (section II.2) and Pagani in this volume.
50 On Demetrius identity and lifetime, see Blank [1998] 144 with n. 112.
51 For a discussion of Demetrius definition, see Blank [1998] 144146.
200 Matthaios
included knowledge of language in its everyday form, i.e. in its usage outside of
the literary contexts.
This new approach was considered to be of no lesser value than the origi-
nal task: indeed, it fairly rapidly became the mainstream perspective, and
Demetrius view was soon shared by other philologists, as Sextus attests (Math.
1.84: ). However, modern
scholarship has awarded little attention to the fact that Quintilian did not
define the grammatica in his Institutio oratoria in a substantially different man-
ner than Demetrius (Inst. 1.4.2): haec igitur professio, cum brevissime in duas
partes divitatur, recte loquendi scientiam et poetarum enarrationem, plus habet
in recessu quam fronte promittit.52 Even Quintilian postulated the twofold sub-
ject matter of grammar, which encompassed the doctrine of the proper use of
language and the interpretation of poetry. Quintilians view is presumably of
Greek origin. In Inst. 1.9.1 he designates the two parts of grammarhere they
are called ratio loquendi and enarratio auctorumby using the Greek terms
methodice and historice. As is to be expected from a rhetorician, Quintilian still
viewed literature as the means to an end, in the sense that literary language
was the model and benchmark by which language use in general was to be
judged.
Nonetheless, the twofold structure of grammatica, especially the equal
status of the grammatical doctrine and the interpretation of literature, became
the essential characteristic of the philological discipline in the 1st century AD.
But the history of then passed through a further stage, which led to
a clear narrowing of the scope of the term. In the transition from Antiquity to
the Middle Ages, the term grammar, initially designating the philological dis-
cipline, became restricted to grammar in the modern sense. This can be seen
in the example of Isidor of Sevilles definition of grammatica, when he defines
the field as follows (Orig. 1.2.1): grammatica est loquendi peritia and (Orig. 1.5.1):
grammatica est scientia recte loquendi.53
The above observations prompt the question of whether the development of
the term and scope of that emerged in the period from Quintilian
to the beginning of the Latin Middle Ages also led to a change of perspective
among philologists and grammarians from the Greek Imperial era and Late
Antiquity. According to our sources, philologists and grammarians active dur-
ing that period worked intensively on defining and describing their scientific
52 On Quintilians definition of grammar, see Ax [2011] 9596 and 405 with further
references on this subject; cf. Glck [1967] 2122, who traces this development back to
Quintilians grammar teacher Remmius Palaemon.
53 See Glck [1967] 2223 and Ax [2011] 96.
Greek Scholarship in the Imperial Era and Late Antiquity 201
54 On this type of monograph, see the contribution of Valente (section II.4) in this volume.
55 The preserved papyrus- are edited by Wouters [1979] 33210.
56 On the individual figures and their works, see the relevant section in the prosopographical
part of this article, 3.3, 3.4.
57 See below in 3.3. A further theoretical statement found in the context of grammatical
treatises of the period in question is that of Lucillus of Tarrha, who deals with the question
of the status of the within the ancient system of sciences; see below in 4.2.
202 Matthaios
58 The different emphases can be clearly seen in the prosopographical section of the article;
see also the introductory remarks ( 1).
59 Interestingly, the expression appears as the title of the grammatical treatise of
Lucillus of Tarrha; see below in 4.2. On the work ascribed to Draco with the title
see below in 4.1.
60 For a discussion of the terms and see Lallot [1997] I 1418. On the
meaning of the term during Late Antiquity, see Wolf [1952] 3141.
Greek Scholarship in the Imperial Era and Late Antiquity 203
61 On the history and development of the terms and in the meaning
philology and philologist, see H. Usener in Susemihl [18911892] II 663665, Pfeiffer
[1968] 157158, Blank [1998] 110111 and 113115 as well as Matthaios [2008] 560562; cf.
also Matthaios [2011a] 66 with n. 43, where references to further literature on this subject
are cited, and Montana in this volume.
62 A differentiation of grammar into a perfect (, or ) and a
inferior () is attested, apart from Sextus, also in Phil. De congr. quaer. erudit.
gr. 148, III 103.24103.3 and De somn. 1.205, III 249.1416. According to the Sch. Dion. T.,
GG I/3, 114.2334 and 164.2329, grammar is divided into a (great) or
(younger) and a (small) or (old); on this differentiation, see Matthaios
[2011a] 6067 with further bibliography on this topic.
204 Matthaios
of words into syllables, familiarity with metrical principles for the determi-
nation of quantities, and punctuation as well. The second task consisted in
explanation of the linguistic and stylistic peculiarities of literary texts. This
was accompanied by explanation of challenging words and interpretation of
mythical and historical material, and also by two specific linguistic tasks: the
explanation of etymologies and the calculation of analogies for the purpose
of determining the grammatical peculiarities of the vocabulary of the literary
text under consideration. Finally, overall aesthetic and critical evaluation of
the text completed the task.
Quintilian describes the structure of a grammatical lesson in a similar
fashion,67 and it is correctly assumed that his account represents the teaching
tradition originating from the Alexandrian school. In the so-called Tyrannion
system, also known from its Latin counterpart, which is attested by Varro
(fr. 236 Funaioli),68 the various steps are compressed into the four parts ()
of grammar: first the (lectio), second the
(enarratio), which consists in the interpretation of a text by focusing on its
mythological and historical peculiarities as well as in providing the correct
explanation of words and the determination of stylistic and other means, third
the (emendatio, the task concerning textual criticism, and
fourth the (iudicium), the expression of an opinion concerning
the aesthetic values, but also addressing the issue of the authenticity of the
literary text.69
Courses at the level of general education aimed primarily to enhance
learning, and secondarily to encourage development of interpretive skills,
which were reserved for professional studies. Collections of maxims (),
didactic anecdotes () and proverbs () served to provide an
overview of ancient literature and values, and were especially popular in the
Roman Empire and during Late Antiquity. Since Roman education was, after
all, bilingual, the grammaticus Graecus was taught alongside the grammati-
cus Latinus.70 The teaching staff responsible for the Greek grammar courses
67 See Quint. Inst. 1.8.112 (lectio) and 1321 (enarratio poetarum). For a commentary on
these passages, see Ax [2011] 350404.
68 See above n. 48.
69 On the description of the contents of grammar instruction, see Glck [1967] 1724 and
Ax [2011] 9497; cf. also R. Baumgarten in Christes-Klein-Lth [2006] 9596, D. Bornmann
in Christes-Klein-Lth [2006] 104110 and Chr. Krumeich in Christes-Klein-Lth [2006]
115123; see also Montana and Swiggers-Wouters (section II.2) in this volume.
70 Cf. Quint. Inst. 1.4.1; see also the commentary of Ax [2011] 95 on this passage. On the
bilingual education in Rome, see Marrou [19656] 374388 and the references quoted
above in n. 30.
206 Matthaios
were recruited from the Greek-speaking world and often consisted of slaves
and freedmen.71
In addition to its permanent position in general education, grammar as
an autonomous scientific field experienced an astonishing geographic expan-
sion and growth at the level of higher education during the entire period of
the Roman Empire and Late Antiquity. As a result, the discipline acquired a
more cosmopolitan dimension in comparison to its previous almost exclusive
presence in Alexandria and Pergamum during the Hellenistic period.72 At the
beginning of the Imperial era, Alexandria still maintained its leading position
as a center of research and cultivation of science and, as such, also of schol-
arship and grammar. The extant biographical sources concerning philologists
and grammarians from Alexandria during the predominance of the Roman
Empire testify that these scholars maintained the chair for grammatical stud-
ies established by the librarians and the members of the Alexandrian Museum;
moreover, they preserved the tradition that linked them to their Hellenistic fore-
runners. As far as we can reconstruct the , Theon, Apion, Chaeremon,
Dionysius of Alexandria, Pamphilus and Vestinus succeeded one another as the
head of the Alexandrian chair of grammar.73 In general, Alexandria continued
to be the most influential center of philological and grammatical education
and research right up to the 6th century.74 Although other important scientific
centers developed in this period, Alexandrian scholarship retained its impor-
tance precisely due to the strong attachment to its rich tradition. At the begin-
ning of the Imperial era and throughout the entire 1st century AD, most of the
Greek philologists and grammarians who were active in Rome had originally
been trained in Alexandria. However, Rome achieved an independent position
in the discipline of Greek scholarship during Hadrians time, in concomitance
with Hadrians founding, in 135 AD, of the Athenaeum as a landmark of Greek
erudition in Rome, following the model of the Alexandrian Museum.75 Thus
in the 1st century AD philology and grammar oscillated between Alexandria
71 Slaves and freedmen, who were active as grammarians in Rome from the period of the
waning Republic until the Principate, are presented by Christes [1979].
72 On Alexandrian and Pergamenian scholarship, see Montana in this volume.
73 On the individual figures, see the relevant section in the prosopographical section below
in 3.3, 3.4.
74 Schemmel [1909] studied the Alexandrian university and other educational institutions
of the 4th and 5th centuries; see also Wilson [1996] 4249 and Bowersock [1996]. The
library of Dioscorus of Aphrodito provides a significant picture of the rich intellectual
and cultural life in Egypt in the 6th century; the preserved material has been edited
and commented on by Fournet [1999]. Philologists and grammarians of the 4th and 5th
century active in Alexandria and Egypt, are listed by Kaster [1988] 469473.
75 See Schmid-Sthlin [19201926] II 664 and 866.
Greek Scholarship in the Imperial Era and Late Antiquity 207
and Rome, but the discipline of scholarship became integrated from the 2nd
century onwards throughout the Roman Empire, by virtue of the fixed sala-
ried positions provided for the universal and higher education system of the
Imperial period. Hadrians regulations stated that public grammarians were to
have the same freedoms as other scholars and scientists such as philosophers,
rhetoricians and physicians. In addition to public teaching posts for grammar,
teaching chairs at the Alexandrian Museum and at the Athenaeum in Rome
were also available. During the Imperial era, Athens also received a significant
stimulus, as professorships were established there and endowed by the Roman
Emperor. Athens, however, was the recognized center for philosophical and
rhetorical studies during the Roman Empire and throughout Late Antiquity,
whereas grammatical studies played a relatively minor role in Athens. For
instance, during Late Antiquity the Emperor financed one chair in rhetoric
and the city itself funded two chairs, but only one Greek chair was provided
for philology and grammar.76
An important turning point in the history of philological and grammatical
studies was the founding of Constantinople and the nomination of the city as
the new capital of the Roman Empire, which gave a new impulse to Greek cul-
ture in the direction of the east. The old educational institutions in Alexandria,
Rome and Athens were now supplemented by a number of new or formerly less
well-known centers in other cities. Most apparent is the wealth of educational
centers in the 4th century in Asia Minor and Syria. Foremost among these
were Nicomedia, Ancyra, Tarsus, Nicaea, Cyzicus, Smyrna, Sardis, Pergamum,
Caesarea, Seleucia; in Syria and Palestine especially Antioch, Sidon, Tyre,
Berytus, Apameia, Emesa, Gaza; finally, in Egypt, Pelusium and Hermupolis in
addition to Alexandria.77 Philological and grammatical studies strongly ben-
efited from this extraordinary geographic and cultural expansion. Several of
these places achieved recognition as the birthplace of Late Antiquity scholars.
Antioch achieved a special status in the 4th century as a result of Libanius
school of rhetoric,78 and several grammarians are known from Libanius cor-
respondence.79 Gaza was also home to an important school of rhetoric, which,
thanks to Procopius, gained an exceptional reputation in the early 6th century.
76 On the position of Athens in philological and grammatical studies, see Schmid-Sthlin
[19201926] II 664 and 947948. See also Schemmel [1908]; Wilson [1996] 3642; Lapini,
and Pontani in this volume.
77 See Schmid-Sthlin [19201926] II 948949.
78 On Libanius school of rhetoric in Antioch, see Cribiore [2007]; cf. Wilson [1996] 2830.
79 On these figures, see the prosopographical study of Seeck [1906]. Grammarians known
from the letters of Libanius are listed in Schmid-Sthlin [19201926] II 1075 n. 4.
208 Matthaios
Various grammarians from this city are known as well.80 Furthermore, philo-
logical and grammatical studies achieved considerable status and were stimu-
lated during Late Antiquity in Constantinople itself, and this development was
essentially related to the re-organization of the University by Theodosius II.81
According to Theodosius decree in 425 (cod. Theodos. 14.9.3), there were to
be five chairs of Greek rhetoric, three chairs of Latin rhetoric, one chair for
philosophy, two chairs for law and ten chairs (!) each for Greek and Latin gram-
mar. In the 4th and 5th centuries, the major figures of the intellectual world
were active in Constantinople.82 The outstanding position Constantinople
had already reached by this time as an educational, scientific and cultural cen-
ter was crucial for the nature and the further development of the history of
Byzantine philology and grammar.83
One may wonder whether and to what extent it would be reasonable and
appropriate to arrange and present the history of philology and grammar in
the Imperial era and Late Antiquity according to the different scholarly cen-
ters and places of activity of the grammarians who were active in each such
location. This would certainly be possible, although the distribution of philo-
logical and grammatical scholarship in the various research and educational
centers provides a merely external criterion for presentation of the subject:
since there are no fixed boundaries between the individual schools, and it is
also characteristic that scholars who were active in one place later later moved
to a different institution. On the other hand, the unity and cohesiveness of
the disciplinary contents, quite independently of the place of activity of each
grammarian, clearly speaks against imposing the geographic criterion as the
principle for periodization of the history of scholarship in the Imperial era
80 On the significance of Gaza as an educational and cultural center, see Wilson [1996]
3033. On the so-called school of Gaza, see Seitz [1892]. The grammarians who came
from the eastern centers of the Empire and were active there are listed in Kaster [1988]
475478; on each personality, see Kasters prosopographical entry.
81 The intellectual, cultural and educational history of Constantinople during Late Antiquity
and the early Byzantine period from the 4th to the 6th century is described by Lemerle
[1971] 4373; on the University in Constantinople, see ibid., pp. 6364; cf. also the works of
Schemmel [1908] and [1912] on this topic as well as Wilson [1996] 4960. On the question
whether the University of Constantinople was to be thought as a high school providing
higher education or as secondary school, as Speck [1974b] 387 and, following him, also
Wilson [1996] 50 maintained, see the objections of Alpers [1981] 95 n. 43; on this topic see
also Pontani in this volume. On the educational system in Constantinople, see Schlange-
Schningen [1995] and [1999].
82 The philologists and grammarians from the 4th and 5th centuries, who were active in
Constantinople, are listed in Kaster [1988] 464467.
83 On the history of scholarship in Byzantium see the contribution of Pontani in this volume.
Greek Scholarship in the Imperial Era and Late Antiquity 209
84 The social status of grammarians in Late Antiquity is examined in the first part of
Kasters study [1988] 9230. The professional and social status of slaves and freedmen
active in Rome as grammarians, is examined by Christes [1979].
85 See Schmid-Sthlin [19201926] II 1075. On the poets and philologists of this period, see
the study of Cameron [1965].
210 Matthaios
86 The last chapter of Pfeiffers History, in which he treats the generation of grammarians
after Aristarchus ([1968] 252279), is accordingly titled The Epigoni.
Greek Scholarship in the Imperial Era and Late Antiquity 211
dull excerption and compilation of older works, and any creations of their own
were little more than abridgments of older works, often contaminated with
younger materialsin a word, a distortion and corruption of the Alexandrian
philological commentaries.87
Given this perspective and such pejorative judgments, neither the back-
ground nor the circumstances of the philological and grammatical achieve-
ments of this period are correctly recognized, nor are the potential and
autonomy of this epoch cast in the light they deserve. Apart from the fact that
the older works from the Hellenistic period of Alexandrian philology were
already not easily available at this time, excerpting and compiling were in
no way senseless or academically unnecessary procedures. For the first time,
the Imperial and Late Antique philologists had at their disposal the rich
research, source and material basis of the earlier scholars, which not only
needed to be collected and studied thoroughly in order to be transmitted
but also represented the basis for critical analysis and, moreover, for comple-
tion, extension, renewal and actualization of the philological past. These were
essentially new features, constituting the unique characteristics of the history
of philology and grammar in the Imperial era and Late Antiquity.
Instead of the narrow and one-dimensional perspective adopted by current
research, which places the Alexandrian period at the center of the history of
ancient philology, and grounds the importance of the entire discipline solely
on the value and achievement of this single era, an attempt should be made to
understand the history of philology and grammar in the Imperial Period and
Late Antiquity on its own terms and to judge its own potential and momen-
tum. Seen in this manner, while the Alexandrian period of ancient scholarship
and grammar can correctly be credited with the foundation of the discipline,
the following epochs constitute the period which led to the systematization,
expansion and renewal of philological and grammatical contents and theories.
It will be shown here that the process in question not infrequently involved
opposition and a critical attitude towards the Hellenistic predecessors.88
As already indicated in the first section of this study, the prosopographic
section is structured according to the main focal points of ancient philological
and grammatical doctrine. First, the philological, interpretative and editorial
activity during the Imperial Period and Late Antiquity will be set in relation
to the and , but also to the of the
ancient . This will be followed by a presentation of the figures and
works pertaining to the field of linguistic theory, the so-called ,
i.e. grammar in the modern sense. Studies concerning ancient theories on
metrics will also be dealt with in this context. Finally, in a third part the lexi-
cographic activity of the entire period will be presented. Subsections of each
chapter will be based either on the place of activity of the scholars and gram-
marians under discussion or, as in the lexicography section, on the special
emphases of each area.
88 On this model of the development of ancient scholarship and grammar, see Matthaios
[2009b] 152. From this perspective, in contrast to both of the first periods, the Byzantine
period is seen as connected with the process of the adaptation and transformation of
traditional knowledge and studying material.
Greek Scholarship in the Imperial Era and Late Antiquity 213
Antique period. In so doing, philologists who were active in this period did
not merely collect and compile older works into new ones, but also engaged
in critical appraisal of older research positions. By the end of these periods,
above all by the end of the Imperial era, the ancient literary tradition had been
explored and made available in its entirety both in terms of literary genres and
of the various historical stages. Its comprehension had been made available
and furthered through editions, commentaries and specialized monographs,
which dealt with the interpretive aspects and explored the source material.
This paved the way for the emergence of Byzantine commentaries and collec-
tions of scholia.
As mentioned earlier, philologists and grammarians of the 1st and 2nd cen-
turies represented both areas of , namely philological interpreta-
tion and as well as linguistic theory. The emphasis of their writing activity,
however, focused above all on textual interpretation and philological issues. It
was not until the 2nd century that philological interests increasingly gave way
to linguistic matters.
3.3 Between Alexandria, Rome and the Educational Centers of the East
Already in the waning years of the Ptolemaic dynasty and during the reign
of Augustus, two prominent scholars were active in Alexandria, Theon and
Seleucus. In spite of the fact that chronologically speaking, they belonged
to the Hellenistic era, their accomplishments reveal that they both represented
the link between tradition and innovation within Alexandrian scholarship
while, at the same time, signaling the turning point from the Hellenistic to the
Imperial period.
Theon89 was the son of the grammarian Artemidorus of Tarsus,90 who
belonged to the school of Aristophanes of Byzantium, and seems likely to
have been the father of Apollonius, the commentator of Homer mentioned
91 Grfenhan [18431850] III 6162 equated the Apollonius who is called Theons son in this
scholion with Apollonius Anteros; on Apollonius Anteros see below in this paragraph. On
Grfenhans supposition, see Susemihl [18911892] II 217 note 400.
92 See Suid. 3215 (= test. 1 Guhl): ... .
.
. For the interpretation of Suidas statement in the
sense that Theon and not Apion was a contemporary of Dionysius of Halicarnassus, see
Susemihl [18911892] II 217 note 40 and Guhl [1969] 2. The grammarian Apion is discussed
below in this paragraph.
93 See Guhl [1969] 2. Guhl based his view concerning the chronological relationship
between Didymus and Theon and the priority of the latter over the former on the
assumption that Didymusder Endpunkt der antiken Grammatikererklrung (so Guhl
[1969] 15)integrated Theons hypomnema on Pindar in his own Pindaric commentary.
The same view concerning the relationship of the two works to each other has recently
been followed by McNamee [2007] 95. Both works, however, can and should be regarded
independently of each other. Theons commentary, of which at least the part related to
the Pythian odes was already available in the 2nd c. AD in P.Oxy. XXXI 2536 (= fr. 38 Guhl)
in a reworked or abbreviated form, builds the basis for the second branch of sources of
the Pindaric scholia, which Irigoin [1952] describes as a Schulkommentar; cf. Maehler
[1994] 114119. On the history of the transmission of the Pindaric scholia, see Dickey
[2007] 3840.
94 See fr. 1418 Guhl; on the character of Theons hypomnema, see Guhl [1969] 13. The
existence of a commentary on the Iliad was accepted by Wendel [1934c] 2055 and also by
Grfenhan [18431850] III 251 with n. 19; the opposite opinion is expressed by Guhl [1969]
1213.
95 See fr. 3638 Guhl; cf. n. 93. Knowledge of Theons commentary on Pindar has been
substantially increased by the papyri findings. In addition to the only scholion previously
known (Sch. Pind. Ol. 5.42a [= fr. 36 Guhl]), a marginal note from Theons commentary
on Pind. Pae. 2.378 in P.Oxy. V 841, iv 37 (fr. 37 Guhl) and also an extensive fragment
Greek Scholarship in the Imperial Era and Late Antiquity 215
from his commentary on Pindars Pythians in P.Oxy. XXXI 2536 (cf. fr. 38 Guhl) have been
added; in the subscription preserved in the latter papyrus, Theons commentary on Pindar
is mentioned expressis verbis.
96 As the expression indicates, P.Oxy. IX 1174 preserves remnants of
Theons critical edition of Sophocles satyr play Ichneutai (fr. 1935 Guhl). It is correctly
assumed that Theons was not just limited to this satyr play, but encompassed
all the works of Sophocles. It remains unclear, however, whether Theon also wrote a
commentary on Sophocles; see Wendel [1934c] 2055 and Guhl [1969] 13.
97 Hesychius testifies Theons lexicographic activity in the dedication letter of his own
lexicon; see Hsch. Epist. ad Eulogium 1.3: ...
. The question of whether Theons lexicon covered
the vocabulary of both tragedy and comedy or was restricted to only one field has been
answered merely on the basis of the transmitted material. Because both of the surviving
fragments (fr. 3940 Guhl; cf. Bagordo [1998] 166167) deal with the language of comedy,
it has been concluded that the lexicon was dedicated exclusively to the ;
see Susemihl [18911892] I 216 with n. 394, Wendel [1934c] 2057, Guhl [1969] 1516 and,
recently, Bagordo [1998] 6364. The lexicon by Epitherses, however, on the language of
both comedy and tragedy (on Epitherses see below in this paragraph) indicates that
works encompassing both areas did in fact exist.
98 For Theons commentary on Theocritus, see fr. 12, for that on Nicander fr. 34a, on
Callimachus fr. 57, on Lycophron fr. 810 and on Apollonius Rhodius fr. 1113 in the
edition of Guhl [1969]. On the contents and scope of Theons commentaries, see Wendel
[1934c] 20552057 and Guhl [1969] 311.
99 See Sch. A.R. 329.8:
. Cf. Wendel [1932a] 105115 and Dickey [2007] 6263.
100 The philological activity on Hellenistic poetry actually started with Theons father
Artemidorus (see Montana in this volume). We know of a commentary of Artemidorus
on Callimachus Aetia. Artemidorus probably produced also an edition of Theocritus; see
Susemihl [18911892] I 185186 and Wentzel [1895c] 1332.
216 Matthaios
101 On the evaluation of Theons philological activity and its after-effect, see Guhl [1969]
1824; cf. Wendel [1934c] 20582059; see also Montana in this volume.
102 Literature on Seleucus: Schmid-Sthlin [19201926] I 269, Mller [1921b], Sandys [19213]
296297, Baumbach [2001b] and Razzetti [2002b]; cf. also Schmidt [1848] and Jacoby in his
commentary on FGrHist 341, pp. 9293. Seleucus fragments from his antiquarian works
are collected in FHG III 500 and FGrHist 341; the philologicalin the furthest sense
fragments were firstly collected by Schmidt [1848] 445452 and then by Mller [1891].
New fragments deriving mainly from Seleucus grammatical studies have been presented
by Reitzenstein [1897] 157211. An edition of the fragments comprising the entire oeuvre
of Seleucus is still pending.
103 See Suid. 200 (= FGrHist 341 T 1): , , ,
. On the nickname , see D.L. 3.109 and 9.12.
104 Seleucus move to Rome is, in addition to Suidas (cf. n. 103), also attested by Suetonius
(Tib. 56 = FGrHist 341 T 2), who testifies Seleucus position as court philologist in the
circle of Tiberius. On the identification of Seleucus named by Sueton with the Homeric
Seleucus, see Mller [1891] 13 and Mller [1921b] 12521253 with reference to older
literature on the matter.
Greek Scholarship in the Imperial Era and Late Antiquity 217
105 On the contents of this treatise, see Mller [1921b] 12551256; cf. Mller [1891] 23. Doubts
on the attribution of this writing to the Homeric Seleucus were expressed by Jacoby in
his commentary on FGrHist 341, pp. 9293.
106 See Mller [1921b] 1255; Jacoby in his commentary on FGrHist 342, p. 93, is skeptical about
the attribution of this work to the grammarian Seleucus.
107 Cf. Mller [1921b] 1254.
108 On the possible contents of this writing, see Mller [1921b] 1254. Reitzenstein [1897] 188
recognized Apollodorus as Seleucus main source. In his commentary on FGrHist 341,
p. 93, however, Jacoby denies that the work stems from the grammarian
Seleucus.
109 See Mller [1921b] 1255.
110 Plutarchs collection of proverbs is often regarded by older scholarship as an excerpt from
Seleucus work; see Mller [1921b] 1252 and Jacoby in a commentary on FGrHist 341, p. 92,
with references to older literature.
111 According to the testimony of P.Oxy. II 221 (= P.Lond.Lit. 178 = Pap. XII Erbse), xv 2425,
Seleucus commentary was titled . Based on fr. 34, Mller assumed the
existence of a commentary of Seleucus on the tragedians. His view, however, has been
met with criticism; see Reitzenstein [1897] 165166 and Mller [1921b] 1254.
218 Matthaios
112 See Suid. 201 s.v. . On the Emesian Seleucus, see below in 3.4.
113 See Schmidt [1848] 444 and Mller [1921b] 1255; cf. also Porro [1994] 1617.
114 See Reitzenstein [1897] 161, Nr. 61: ... .
Cf. Mller [1921b] 1255.
115 Cf. Mller [1921b] 1255.
116 On these works, see Mller [1921b] 1255. Seleucus fr. 73 Mller stems, as it can be seen
from the testimony mentioned by Reitzenstein [1897] 165.15 and his commentary in the
textcritical apparatus, not from Seleucus work .
Greek Scholarship in the Imperial Era and Late Antiquity 219
only by the character of his treatise, which was in effect a polemical disparage-
ment of Aristarchus critical signs and their explanations,117 but even more so
by the arguments he put forward to counter the Aristarchean interpretation of
Homer.118
Seleucus thus proves not to be a mere compiler of old exegetical material,
but an independently working scholar who critically addressed the philolog-
ical tradition of his own school. The same picture can be reconstructed for
Seleucus grammatical works in the strict sense. In the field of etymology,
Seleucus is regarded, along with Philoxenus, as the founder of a scientific ety-
mology, which ran counter to the philosophical explanation, especially the
stoic etymological explanation based upon arbitrary methods.119 Seleucus
etymologies are attested in the material provided by Reitzenstein from the
Byzantine Etymologica.120 In place of an etymological explanation that
assumes the composition of a word from segments of other words, both
Philoxenus and Seleucus developed a theoretical framework that was mainly
based upon derivation. In their conception, a word is to be traced back to its
stem by taking into consideration and explaining phonological and morpho-
logical changes that have affected the word structure. Seleucus believed that
etymology serves to determine the correct use of language, and according to
the testimony of Athenaeus (9.367a) he devoted a special study to this topic,
entitled .121 The etymologies attributed to Seleucus presum-
ably derive from this work. Finally, Seleucus lexicographic activity includes
the collection (Glosses; fr. 3668 Mller) and also a work bearing
the title On different meanings of synonymous words (
), which is testified only by Suidas.
The grammarian Ptolemaeus also belongs to the Alexandrian tradition.122
The period of his life can be set at the first half of the 1st century AD, if it can be
assumed that Ptolemaeus, in agreement with the testimony of Athenaeus and
117 On Aristarchus critical signs and also on Aristonicus writing which served to explain
them, see Montana in this volume.
118 On the philological principle , which Seleucus assumes for the
interpretation of Il. 21.290, see Nnlist [2009a] 157164, especially 169170. According to
Nnlist, Seleucus modified and expanded the traditional exegetical principle, which was
widely utilized in early Alexandrian interpretation of Homer.
119 On Seleucus etymological approach, see Reitzenstein [1897] 184188.
120 Cf. n. 102.
121 On this special type of grammatical treatises, see Siebenborn [1976] 3235; Valente
(section II.4), and Pagani in this volume.
122 Literature on Ptolemaeus: Susemihl [18911892] II 215, Schmid-Sthlin [19201926] I 438,
Dihle [1962], Matthaios [2001b] and LGGA s.v. Ptolemaeus (3) Aristonici (A. Boatti).
220 Matthaios
Herodian, was the son and not the father of Aristonicus.123 The later dating is
supported by the fact that Ptolemaeus, according to the testimony of Suidas,
was active as a grammarian in Rome, like his father Aristonicus. The most
probable date for Ptolemaeus stay in Rome was during the reign of Augustus.124
Suidas also provides a list of his writings. The few fragments of Ptolemaeus
works that have survived until today derive from his Commentary on Homer
( ), which consisted of 50 books.125 Ptolemaeus work titled
dealt, as the title suggests, with the recurrent themes in
tragedies.126 Two further works of Ptolemaeus, known only by their titlethe
treatises (Strange stories in Homer) and
(On the Muses and Nereids)were presumably
concerned presumably with mythographical and antiquarian issues.
The grammarian Apollonides of Nicaea lived during the reign of Tiberius.127
As testified by Diogenes Laertius, Apollonides dedicated his commentary
on Timons Silloi ( ) to the Emperor.128 Diogenes
Laertius quotes from the first book of Apollonides commentary, providing
biographical information on Timon as well as a synopsis of the Silloi.129 In
addition to his work on Timons Silloia somewhat strange choice for his phil-
ological activityApollonides wrote a commentary on Demosthenes speech
On the false embassy ( ).130
He also composed a work consisting of at least 8 books with the title
123 See Ath. 11.481d and Sch. Hom. Il. (A) 4.423 a1: . Due to
a misunderstanding, Suidas regards him as the father of Aristonicus; see Suid.
3036: , ,
. On Aristonicus, see Montana in this volume.
124 See Susemihl [18911892] II 215 n. 386 and Wendel [1920] 7778; cf. Matthaios [2001b].
Dihle [1962] does not believe it is possible to come to a more certain conclusion regarding
the relationship between both grammarians, due to the fact that both grandfather and
grandson often shared the same name.
125 Wendel [1920] 78 also assigned the Sch. Theoc. 1.110ac to the Homeric commentary of
Ptolemaeus.
126 See Bagordo [1998] 65 and 162.
127 Literature on Apollonides: Grfenhan [18431850] III 250, Schmid-Sthlin [19201926]
I 435, Wentzel [1895a], Di Marco [1989] 5455, Montanari [1996e] and LGGA s.v.
Apollonides (A. Ippolito). Testimonies and fragments from Apollonides works are listed
in FHG IV 310 and in LGGA.
128 See D.L. 9.109: , ,
, .
129 See D.L. 9.109111 = Timo Phliasius test. 1, fr. 1 Di Marco.
130 The work is attested in Ammon. Diff. 366 s.vv. .
Greek Scholarship in the Imperial Era and Late Antiquity 221
131 In Ammon. Diff. 253 s.vv. the third book of this work is cited; the
eighth book is cited in the Vita Arat. I 10.1619 Martin; Mller FHG IV 310 also attributes
the quotation of Harp. 27 s.v. to this work.
132 The work ascribed by Suidas ( 3422) to Apollonius Dyscolus
is unjustly regarded by Mller in FHG IV 310 as a work either of Apollonides or of the
historian Apollonius; see R. Schneider in GG II/3, 140. It is also unclear whether the
Apollonides-fragment transmitted by Priscian in GL III 406.22407.4 is to be ascribed to
this work; Apollonides is quoted here together with Lucillus Tarrhaeus for the view they
shared on the use of letters as numerals.
133 Literature on Apion: Grfenhan [18431850] III 5859, Cohn [1894b], Schmid-Sthlin
[19201926] I 437438, Sandys [19213] 295296, Montanari [1996d] as well as Neitzel [1977]
189190. The fragments from the historical writings of Apion are collected in FHG III 506
516 and FGrHist 616. On Apions glossographical work, see below in this paragraph.
134 See Suid. 3215 (= FGrHist 616 T 1); cf. the testimonies mentioned in FGrHist 616.
135 On Theon, see above in this paragraph. On Apollonius Archibiu, see below in this
paragraph.
136 Cf. Suid. 3215 (= FGrHist 616 T 1) and the testimonies listed under FGrHist 616 T 5.
137 See Suid. 3215 (= FGrHist 616 T 1); cf. Plin. N.H. 30.18.
138 See Sen. Ep. 88.40 (= FGrHist 616 T 7).
222 Matthaios
149 The papyrus commentary transmits two remarks by Apion: a metrical observation (l. 3)
and a varia lectio (l. 18); see Porro [2004] 77 and 134.
150 See Apol. Dysc. Synt. 1.154, GG II/2, 124.9125.3.
151 See Sch. Dion. T., GG I/3, 183.2531 (= FGrHist 616 F 27); the fragment refers to the question
of whether the Greeks used the Ionic alphabet.
152 Literature on Apollonius Anteros: Grfenhan [18431850] III 6162, Cohn [1895c], Schmid-
Sthlin [19201926] I 438, Montanari [1996f] and LGGA s.v. Apollonius (7) Anteros.
153 See Suid. 463: , , , ,
. ,
, , ,
.
154 See Westermann [1845] 369 (apparatus on l. 52); cf. Schmidt [1854] 910. Bergk [1845]
125, on the other hand, corrects the transmitted form into and interprets it as
the genitive of the name of a Roman grammarian named Asper. Grfenhan [18431850]
224 Matthaios
III 6263 believes that this Asper () is the Latinized Anteros, and that the name
is primarily a description of a characteristic, such as , or .
Furthermore, Grfenhan [18431850] III 63 is of the opinion that Asper did not turn
against Didymus Chalkenteros, but against Didymus the Younger or Didymus Claudius,
who was actually the son of Heraclides Ponticus. That is merely an assumption, which
cannot be proven. The same is also the case for the changed form in Suidas
testimony and the identification of the grammarian referred to there with the already
discussed Apion (see above in this paragraph), as Hertz [1862] suggested. In the view of
Hertz, Apions possible Ausfall gegen seinen Erzieher Didymos is understandable. But
this is also pure speculation, though this suggestion was endorsed by Cohn [1894b].
155 Literature on Heraclides: Grfenhan [18431850] III 6465, Schmidt-Sthlin [19201926]
I 322 and 330, Daebritz-Funaioli [1912], Fornaro [1998g] and LGGA s.v. Heraclides (6).
156 See Suid. 463 and 2634; the text of the first passage named is quoted in the previous
n. 153.
157 On this Didymus, see Daebritz-Funaioli [1912] 487.
158 The few surviving fragments of this work are collected and discussed by Meineke [1843]
377381; some hints of other possible fragments are given by Daebritz-Funaioli [1912] 487.
See also Heitsch [19631964] II 41 (S 1).
Greek Scholarship in the Imperial Era and Late Antiquity 225
168 See Apol. Dysc. Conj., GG II/1.1, 247.30248.13. Concerning the significance of expletive
conjunctions in the ancient word class system, see Matthaios [1999] 582584.
169 Literature on Dionysius of Alexandria: Grfenhan [18431850] III 54 and 67 and Cohn
[1903g].
170 See Suid. 1173 s.v. the text of this testimony is quoted in the
previous n. 166.
171 On Parthenius, see below in 6.1.
172 See Cohn [1903g] 985.
173 Literature on Soteridas: Grfenhan [18431850] III 91, 106, 227228, 258259, Schmid-
Sthlin [19201926] I, 437 and 440, Gudeman [1927e] and LGGA s.v. Soteridas (A. Ippolito).
On Soteridas lifetime, see Gudeman [1927e] 1233.
174 Pamphile is mostly known from her , which consists of 33 books
and is a collection of literary and historical material of the most varied nature, as well
as from an epitome of Ctesias historical works. She probably lived during Neros reign;
cf. Grfenhan [18431850] III 397 and 402, Schmid-Sthlin [19201926] I 437, Sandys
[19213] 295 and Regenbogen [1949]. There is contradictory information about Soteridas
and his relationship to Pamphile. In Suid. 875 Soteridas is mentioned as the husband
of Pamphile, in 876, however, as her father. The complex question surrounding the
genealogical relationship of Pamphile with Soteridas and with Socratidas, who also
appears as her betrothed, is treated by Gudeman [1927e] 12321233; cf. Regenbogen [1949]
309312. Gudeman, loc. cit., finds it more probable that Soteridas was Pamphiles father.
Greek Scholarship in the Imperial Era and Late Antiquity 227
175 Both of Suidas articles on Soteridas ( 875 and 876) as well as the article on Pamphile
herself (Suid. 139) attribute to him the authorship of Pamphiles work; the information
in 876 derives from Dionysius of Halicarnassus the Musician (on whom, see below in
this paragraph). On the authorship question of the , see Gudeman
[1927e] 12321233.
176 See Suid. 875.
177 It is improbable that Sch. Hom. Il. 4.412b1 stems from this work. A grammarian named
Soteras is indeed cited here, but Dindorf probably incorrectly changed the name to
Soteridas; on this question, see Erbse in the testimonies apparatus to Sch. Hom. Il. 4.412b1.
178 See Bagordo [1998] 65 and 165166.
179 Literature on Pamphilus: Grfenhan [18431850] III 5657 and passim, Schmid-Sthlin
[19201926] I 435436, Wendel [1949b], Tosi [2000c] and LGGA s.v. Pamphilus (1). There is
now a thorough study on Pamphilus with an edition and commentary on his fragments
by Hatzimichali [2006].
180 On Apion, see above in this paragraph.
181 See Ath. 14.642e = fr. 4 H.; cf. also Hatzimichali [2006] 5759.
182 The biographical information about Pamphilus has been presented and extensively
discussed by Hatzimichali [2006] 1114.
183 See below in 6.1.
228 Matthaios
184 For a discussion of the possible contents of this work, see Hatzimichali [2006] 1516.
185 A description of the work with edition and commentary of the fragments belonging to it
is to be found in Hatzimichali [2006] 150194.
186 See Hatzimichali [2006] 1617.
187 See Hatzimichali [2006] 1718.
188 On the terms and , see Pfeiffer [1968] 157158.
189 On this question, see Hatzimichali [2006] 1820.
190 On prosodic studies during the Roman Empire, see below in 4.
191 The fragments from Pamphilus studies on Homeric prosody have been edited and
discussed by Hatzimichali [2006] 107149. On Herodian, see below in 4.1.
Greek Scholarship in the Imperial Era and Late Antiquity 229
69114; on this work, see however the review of Martin [1929]. The identification of
Lucillus with the epigram poet Lucillus is groundless.
210 The fact that according to the subscription of the scholia on Apollonius Rhodius (see in
the main text) Lucillus belongs to the same scholarly tradition as Sophocleus and Theon
speaks for locating his activity in Alexandria.
211 See Steph. Byz. 604.5 s.v. ...
.
See also Steph. Byz. 311.6 s.v. ...
. (= FHG IV 440 fr. 1).
212 See Wendel [1932a] 108110 and Gudeman [1927a] 17871788.
213 See fr. VIIIXII Linnenkugel ([1926] 8896) as well as the literature in the previous note
209.
214 See fr. IIV Linnenkugel ([1926] 7483). On Lucillus collection of proverbs, see Gudeman
[1927a] 17881790.
215 See fr. VVII Linnenkugel ([1926] 8388).
Greek Scholarship in the Imperial Era and Late Antiquity 233
, an , a , an and a . A spe-
cial section of his was the treatise , which dealt with
the letters and history of the Greek alphabet (fr. XIVXVI Linnenkugel).
In the field of ancient scholarship, the grammarian Astyages216 was par-
ticularly important because of his commentary on Callimachus poetry (
).217 We know have no knowledge of pre-
cisely when or where Astyages lived. However, since he was concerned with
the interpretation of Hellenistic literature, it can probably be assumed that he
shared the new characteristic of contemporary scholarship and belonged to
the generation of the 1st century AD. The list of works cited by Suidas suggests
that Astyages wrote several grammatical works in addition to his Callimachus
commentary. Among his grammatical writings we know of a
(Manual on grammar), a work entitled (On dialects), a
treatise (On Versification) as well as a work named
(Rules of nominal inflection). A treatise on the number of cases,
in which Astyages argued that there are six nominal cases, is mentioned in
the part of the scholia on Dionysius Thrax in Cod. Lond. Add. 5118 that derives
from Choeroboscus. The work dealing with the rules for declensions seems to
indicate that Astyages lived after Herodian. The quantity of Astyages gram-
matical works, as compared with his philological activity, also speaks in favour
of a later datation, most probably at the beginning of the 2nd century AD. But
nothing can be assumed with certainty in this question.
The end of the 1st century AD was marked by a geographic expansion of the
philological discipline beyond the borders of Alexandria and Rome. Several
cultural centers in the eastern part of the Roman Empire were thus integrated
into the field of scholarship. A number of significant philologists, whose work
is described below, philologists stand out in this period:
Herennius Philo from Byblos in Phoenicia218 was a learned antiquarian,
doxographer and grammarian of the second half of the 1st century AD. He also
216 Literature on Astyages: Grfenhan [18431850] III 83, Schmid-Sthlin [19201926] I 439,
Cohn [1896] and LGGA s.v. Astyages (G. Uzziardello).
217 See Suid. 4259 s.v. , . , ,
, , .
218 Literature on Herennius Philo: Grfenhan [18431850] III 4445 and passim, Schmid-
Sthlin [19201926] II 867868, Gudeman [1912e], Christes [1979] 105106 and 137138,
Fornaro [1998k], where further references on Philo are given, and LGGA s.v. Herennius
Philo (2). Philos fragments, especially those from his antiquarian and cultural-historical
writings, are edited in FGrHist 790.
234 Matthaios
219 On Philos life, see Suid. 447 (= FGrHist 790 T 1); see also Suidas testimonies quoted under
FGrHist 790 T 2, which refer to Philos contemporaries. On the biographical information
regarding Philo, see Gudeman [1912e] 650651 and Christes [1979] 106.
220 See Christes [1979] 105196 and 137138.
221 On Hermippus of Berytus, see below in the next paragraph.
222 On the contents of Philos historical studies, see Gudeman [1912e] 659661.
223 See Gudeman [1912e] 654659. On Orus and Stephanus of Byzantium, see below in 6.2.
Greek Scholarship in the Imperial Era and Late Antiquity 235
224 On the synonymica, see below in 6.2. The works Deverbatives ( ) and On
the Language of the Romans ( ), which are cited in Etym. Magn.
under Philos name, should be attributed to the grammarian Philoxenus, as is evident
from the relevant entries in Etym. Gen.
225 Literature on Hermippus: Schmid-Sthlin [19201926] II 868, Heibges [1912], Christes
[1979] 137140, Sartori [1981], Montanari [1998b] and LGGA s.v. Hermippus (2) (A. Ippolito).
The fragments from his work are collected in FHG III 5152 and FGrHist 1061; see also
Mller in FHG III 3536. The main biographical testimony concerning Hermippus life is
Suid. 3045 s.v. , , ,
, ,
. ; cf. Suid. 375 s.v. , , ,
, , .
226 See Christes [1979] 139140.
227 See Suid. 375 s.v. the text is printed in the previous n. 225.
228 Literature on Nicanor: Grfenhan [18431850] III 6768, 9495 and 189, Schmid-Sthlin
[19201926] II 868869, Sandys [19213] 322, Wendel [1936], Bagordo [1998] 6869,
Matthaios [2000b] and LGGA s.v. Nicanor (3) (F. Montana). The fragments from his
writings and are edited in FGrHist 628; on the collections
of Nicanors fragments pertaining to his punctuation theory, see below n. 373.
236 Matthaios
the end of the 2nd century, perhaps after the reign of Marcus Aurelius.257 Pius
work is mentioned by its title in Etym. Gen. (= Etym. Magn. 821.55) s.v. .
Since a to the 16th book of the Odyssey is quoted there, it can be
assumed that Pius wrote a commentary on the Homeric poems, probably
on each book of the Iliad and the Odyssey. The fragments that have survived
from this work (fr. 115 Hiller), show that Pius commentary treated matters of
the Homeric language as well as questions of content and style. According to
Sch. Hom. Il. (bT) 12.17581b (= fr. 5 Hiller), Pius turned against Aristarchus
athetesis of the Iliad lines in question. This led Hiller to believe that the main
concern of Pius commentary was to oppose Aristarchus athetesis. It seems
plausible to assume that Pius stood in critical opposition to the old Alexandrian
exegesis, especially against that of Aristarchus. What is somewhat implausible
is Hillers conclusion that all the Homeric scholia directed against Aristarchus,
including those where Pius is not mentioned by name, are traceable back
to the latter.258 It has also been postulated on the basis of Sch. Soph. Aj. 408
(= fr. 16 H.) that Pius wrote a commentary on Sophocles; the exact context of
this scholium, however, remains uncertain.259
A grammarian named Irenaeus260probably not identical with the Atticist
Irenaeus, and also known by the Latin name Minucius Pacatus261lived
around the middle of the 2nd century. Irenaeus activity in the field of philol-
ogy included commentaries on the individual books of Apollonius Rhodius
Argonautica. His contribution becomes tangible through the fact that the
grammarian Sophocleus inveighed vehemently against his views.262 According
to a testimony in Lex. Rhet. Cant. 22.2323.18 (s.vv. / /
/ [= fr. 16 Haupt]), Irenaeus also wrote a commentary on
Herodotus and, on the basis of the Sch. Eur. Med. 218 (fr. 17 Haupt), presumably
on Euripides Medea as well.
257 The counterposition says that Pius was a contemporary or immediate predecessor of
Didymus, a thesis that is not very convincing. On Pius life, see Strout-French [1950] 1891
with further references.
258 See Lhrs [1992] 269 note 376.
259 See Hiller [1869] 9091 and Strout-French [1950] 1892.
260 Literature on Irenaeus: Grfenhan [18431850] III 249, Schmid-Sthlin [19201926] II 870,
Cohn [1905b] 2121, Wendel [1932a] 106107, 111 and 115 and Fornaro [1997a] 919. Irenaeus
fragments have been collected by Haupt [18711876].
261 On the Atticist Irenaeus, alias Minucius Pacatus, see below in section 6. Schmidt-Sthlin
[19201926] II/2, 870 and Wendel [1932a] 106107 speak against the identification of the
two figures; Cohn [1905b] 2121, however, is in favor of this view.
262 Irenaeus is quoted four times in the scholia on Apollonius Rhodius; see Sch. A.R. 1.1299
(= fr. 18 H.), 2.123129e (fr. 19 H.), 2.992 (fr. 20 H.) and 2.1015 (fr. 21 H.). On the relationship
between Irenaeus and Sophocleus, see Wendel [1932a] 106.
Greek Scholarship in the Imperial Era and Late Antiquity 241
The grammarian Palamedes263 also belongs to the (late) 2nd century AD.
Both Athenaeus (9.397a) and Suidas ( 43) label him by his origin, .264
The circumstances of his life and the place of his activity remain unknown.
Athenaeus lists him as a participant at Larensius banquet, and this is the only
available information on his life. A Commentary on Pindar mentioned by
Suidas ( ) attests to Palamedes philological
activity, but no fragments from this work have survived. This reasonably leads
to the conclusion that Palamedes commentary had no effect on the develop-
ment of the corpus of the Pindaric scholia.265 The major part of Palamedes
philological activity consisted in lexicographic works, which included a
(The language of comedy) and a (The language of
tragedy)266 as well as an , which was apparently a dictionary
arranged on the basis of synonymous expressions concerning specific seman-
tic fields such as Pollux Onomasticon.267
Salustius,268 a grammarian who is probably identical with the scholar of the
same name cited by Stephanus of Byzantium ( 75 s.v. ), lived according
to Wilamowitz in the 4th or even 5th century,269 though it is not inconceiv-
able that he lived somewhat earlier. Salustius is the author of a commentary
on Callimachus Hecale, which was still in use during Suidas time.270 It is very
263 Literature on Palamedes: Grfenhan [18431850] III 178 and 261, Frster [1875], Schmid-
Sthlin [19201926] II 870, Wendel [1942b], Bagordo [1998] 69, Matthaios [2000c] and
LGGA s.v. Palamedes.
264 The attempts to set Palamedes in relation with the Eleatic Zeno are discussed by Frster
[1875]; see also Wendel [1942b] 2512.
265 Wilamowitz [1889] 185 n. 126 presumed that Palamedes was actually the redactor
responsible for the compilation of the Pindaric scholia. Such a work, however, could
hardly have been designated as a ; cf. Wendel [1942b] 2513. Irigoin [1952] 75, esp.
9394 is also skeptical of Wilamowitzs view.
266 Only seven fragments are extant from these works; they are listed in Bagordo [1998]
153155.
267 Schmidt-Sthlin [19201926] II 1039 n. 6 suspect that this work was a collection of
proper names and biographies in the sense of the of Hesychius Milesius;
Wendel [1942b] 2513 argues convincingly against this assumption; he also mentions
passages that could be assigned to this work. According to Grfenhan [18431850] III 178,
the cannot have been a separate work by Palamedes; Suidas must have
misunderstood a passage of Athenaeus (9.397a), in which Palamedes himself is called an
.
268 Literature on Salustius: Schmid-Sthlin [19201926] II 870, Pfeiffer [19491953] II XXVIII
XXX, Baumbach [2001a] and LGGA s.v. Sal(l)ustius (2) (G. Ucciardello).
269 See Wilamowitz [18931941] 31.
270 See Etym. Gen. () 1279 Lass.-Liv. s.v. ...
(ad fr. 240 Pf.); () 1224 Lass-Liv. s.v. ... (ad. Callim.
242 Matthaios
Peace.278 He is also mentioned separately five times, but only in the scholia on
Aristophanes Knights.279 Contrary to an earlier research position, it is more
realistic to assume that Phaeinos did not live much later than Symmachus.280
Phaeinos produced a form of Schulkommentar on at least the eleven surviv-
ing comedies of Aristophanes, which was most likely a revision of Symmachus
work, partially enriched with additional material.281 However, there is little
ground for believing Phaeinos to be the redactor of the corpus of scholia on
Aristophanes, in contrast to the suggestion put forward by Wilamowitz, among
other scholars.282
After Theon, philological studies dealing with Hellenistic poetry continued
to flourish. As well as Astyages and Salustius,283 the grammarian Archibius,284
the fatheror more likely the sonof Apollonius Sophista285 was one
of the commentators of Callimachus. Suidas ( 4105) mentions Archibius
as the author of a commentary on Callimachus epigrammatic poetry
( ).286 Another grammarian by the
name of Sophocleus,287 who lived in the late 2nd century AD, is known for his
278 See the passages quoted in the previous n. 274. Gudeman [1921] 675 presumes that
Phaeinos name was also present in the subscription to Sch. Ar. Av.. Together with
Symmachus, Phaeinos is quoted once again in Etym. Gen. (AB) 146 Lass.-Liv. s.v.
(Etym. Magn. 200.3749); the explanation transmitted in this testimony refers to Ar. Av.
530 (cf. Sch. Ar. Av. ad loc.).
279 The extant fragments from Phaeinos commentary are discussed in detail by Gudeman
[1921] 676.
280 Wilamowitz [1889] 181 places Phaeinos in a period after the 4th century; this view is also
shared by Trojahn [2002] 142. Gudeman [1921] 676 argues for placing Phaeinos closer in
time to Symmachus.
281 See Gudeman [1921] 675677; cf. also Dunbar [1995] 41.
282 See Wilamowitz [1889] 181; Grfenhan [18431850] III 266267 already argued against this
view.
283 See above in this paragraph.
284 Literature on Archibius: Grfenhan [18431850] III 58, Schmid-Sthlin [19201926] I 138
note 5, II 870, Cohn [1895h] and LGGA s.v. Archibius (1) (A. Ippolito).
285 On Apollonius Sophistes, see below in 6.1.
286 This Archibius is not to be confused with the grammarian of the same name originally
from Leucas or Alexandria; the later Archibius was the son of a Ptolemaeus, who taught
during the reign of Trajan in Rome; see Suid. 4105 with Grfenhan [18431850] III 58,
Cohn [1895i] and LGGA s.v. Archibius (2) (A. Ippolito).
287 Literature on Sophocleus: Grfenhan [18431850] III 253, Schmid-Sthlin [19201926]
I 146 and II 870, Gudeman [1927d], Wendel [1932a] 105107 and 110116, Matthaios [2001e]
and LGGA s.v. Sophocleus (L. Pagani). On the form of the name (Sophocles or rather
Sophocleus), see Wendel [1932a] 90 n. 1.
244 Matthaios
and summaries, but dealt additionally with prosodic issues as well as word
explanations and questions of content. A further interpreter of Theocritus,
the grammarian Theaetetus, whose identity cannot be closely determined,
rejected the views of Munatius.297 Finally, Dionysius Leptos ( )298 is
mentioned by Fronto as a grammarian and rhetorician and also as his own
teacher.299 Accordingly, Dionysius must have lived during the 2nd century AD.
He was probably identical with the so-called Dionysius , who,
according to a testimony in Etym. Magn. 278.15, was so called either because
he was constantly quoting Il. 9.82, or because he was so tall, thin and pale that
he looked like an . His philological writings are now known only
from a single quotation in Athenaeus (11.475f), which refers to a passage from
Dionysius commentary on the Hellenistic poet Theodoridas.
According to Porphyrys testimony (Plot. 7.11.1216 H.Schw.), Zoticus,300 a
friend of Plotinus, wrote a work of textual criticism on Antimachus (
). In the same testimony, Zoticus is described as a
(both a critic and a poet). Whether Zoticus claimed this character-
ization for himself is unknown, but it is in any case reminiscent of the profile of
Antimachus and the Hellenistic poets and scholars.301 Nothing has survived
from his poetic work bearing the title , which was a versification of
Platos Critias, and no fragments from his philological activity are extant.
When dealing with prose authors, the boundaries between philology,
rhetoric and scientific literature are especially fluid. This is in marked contrast
to the interpretation of poets, which was always regarded as the domain of
philology. Attic oratory thus constituted the primary subject of rhetoricians;
lexicographers of the Imperial era who dealt with the language of the classical
orators were basically rhetoricians.302 Ancient philosophy, first and foremost
Plato and Aristotle as well as the corpus Hippocraticum, were mainly inter-
preted by philosophers and physicians.303 Ancient historiography presented a
297 Regading Munatius commentary on Theocritus, see Wendel [1920] 7478 (with additional
information on Theaetetus) and 8890.
298 Literature on Dionysius Leptos: Cohn [1903h] and LGGA s.v. Dionysius (13) Tenuior.
299 See Fronto Ad M. Antonin. de eloq. 5.152.2 van den Hout: in eos quoque meus magister
Dionysius Tenuis arte compositam fabulam protulit de disceptatione vitis et arboris ilicis and
Ad Caes. II 1.17.8 van den Hout: . ...
.
300 Literature on Zoticus: Grfenhan [18431850] III 333, Schmidt-Sthlin [19201926] II 870,
Ziegler [1972] and LGGA s.v. Zoticus; cf. also Matthews [1996] 75.
301 See above 2.1. On Antimachus as a model of Alexandrian poets and scholars, see
Matthaios [2008] 640642, cf. also Novokhatko and Montana in this volume.
302 On the lexicographic analysis of classical Attic rhetoric, see below in 6.2.
303 See also our discussion above, 1 and 2.
246 Matthaios
special case, inasmuch as rhetoricians to a certain extent took over its interpre-
tation. Antyllus304 is attested by Suidas ( 2770) as a rhetorician without any
further information on his activity. According to the little that is known about
him, Antyllus focused on the work of Thucydides, composing a biography of
Thucydides305 as well as a commentary cited in just a few scholia.306 The rhet-
orician Sabinus,307 who is known only through Suidas, lived during the reign of
Emperor Hadrian and wrote a commentary on Thucydides.308 Another rheto-
rician, Hero,309 who probably lived at approximately the same time, is stated
by Suidas ( 552) to have composed not only biographical studies on the Attic
orators but also an on Dinarchus and commentaries on Herodotus,
Thucydides and Xenophon. His writings incude a work consisting of three
books with the title (Words that are attested in recognized
authors), which was probably an Atticistic glossary. Additionally, the above
mentioned grammarians Irenaeus and Salustius also dealt with Herodotus.310
The rhetorician Zeno of Citium,311 who probably lived in the 2nd century AD,
wrote a series of technical treatises on rhetoric but likewise composed com-
mentaries on Xenophon, Lysias and Demosthenes.312
The 36 books (History of music), written by the grammar-
ian Dionysius of Halicarnassus,313 who lived in Hadrians time, also belong to
the area of scholarship. To distinguish him from the rhetorician Dionysius of
Halicarnassus as well as from the Atticistic lexicographer Aelius Dionysius,
the grammarian was called .314 Several articles by Suidas point
304 Literature on Antyllus: Grfenhan [18431850] III 287, Schmid-Sthlin [19201926] II 870,
Brzoska [1894], Weienberger [1996] and LGGA s.v. Antyllus (L. Pagani).
305 See Marcelin. Vit. Thuc. 22, 36 and 55; in the last quotation Antyllus credibility is
highlighted.
306 See Sch. Thuc. 3.95, 4.19 and 4.28; Antyllus fragments are discussed by Goslings [1874]
5457.
307 Literature on Sabinus: Grfenhan [18431850] III 261, Schmid-Sthlin [19201926] II 870,
Gerth [1920], Weienberger [2001] and LGGA s.v. Sabinus.
308 Biographical details about Sabinus and list of works are mentioned in Suid. 11.
309 Literature on Hero: Grfenhan [18431850] III 222, Schmid-Sthlin [19201926] II 870,
Kroll [1912] and LGGA s.v. Heron.
310 See above in this paragraph.
311 Literature on Zeno: Grfenhan [18431850] III 140 and 269, Schmid-Sthlin [19201926] II
870, Grtner [1972], Weienberger [2002] and LGGA s.v. Zeno (5).
312 Biographical details on Zeno and a list of his writings are mentioned in Suid. 81.
313 Literature on Dionysius: Grfenhan [18431850] III 437, Westphal [1883] 248250, Scherer
[1886], Schmid-Sthlin [19201926] II 870871, Cohn [1903j], Montanari [1997l] and LGGA
s.v. Dionysius (11) Musicus (E. Rocconi).
314 See Suid. 1171.
Greek Scholarship in the Imperial Era and Late Antiquity 247
318 Ludwich [1895] and [191213 and 1914] ascribed the fragment of a commentary on Il. 1.1
560 transmitted in Cod. Vindob. Philol. Gr. 49, fol. 8r12r to Demo; objections against the
attribution to Demo were made by Kroll [1918b] 332333.
319 Literature on Seleucus of Emesa: Grfenhan [18431850] III 42 and 218, Schmid-Sthlin
[19201926] II 1076, Mller [1921a], Seeck [1906] 272273, Seeck [1921], Kaster [1988] 428
429 (Nr. 253), Matthaios [2001d] and LGGA s.v. Seleucus (2) (G. Ucciardello).
320 This work is presented as a historical prose treatise in FGrHist 780 T 1; cf. Schmid-Sthlin
[19201926] II/2, 1077. On Seleucus poetical work, see Schmid-Sthlin [19201926] II/2,
959.
321 On Seleucus from Cilicia, see Kaster [1988] 250252 (Nr. 25).
322 See Seeck [1906] 272273 and [1921]. Seecks identification has been accepted by Schmid-
Sthlin [19201926] II/2, 1075 n. 2; Mller [1921a] 1251 and Kaster [1988] 428429, however,
expressed objections against Seecks view.
323 Literature on Dionysius: Susemihl [18911892] II 11 n. 54, Schmidt-Sthlin [19201926]
II 1076, Cohn [1903i], Montanari [1997k], Bagordo [1998] 62 and LGGA s.v. Dionysius (8)
(L. Pagani). Wilamowitz [1889] 134 n. 21 presumed that Dionysius was identical to
Dionysius of Halicarnassus, the author of the ; on Dionysius, see above
3.3. Cohn [1903i] 986 opposed the identification.
324 See Sch. Eur. Med. subscr.:
and Sch. Eur. Or. subscr.:
.
325 On these figures and the various attempts to identify them, see Bagordo [1998] 6162 with
further references on this question.
Greek Scholarship in the Imperial Era and Late Antiquity 249
parts of tragedy and comedy as well as of the satyr drama.326 Cohn ([1903i]
986) argues for the possibility that the fragments of Dionysius stem from the
prolegomena to his commentary on Euripides, though it cannot be ruled out
that they may derive from a separate work dealing with drama.
The grammarian Horapollo,327 who came from the village Phenebythis in the
Egyptian Panopolites,328 first taught in Alexandria as grammarian and then in
Constantinople under Theodosius II. (408450). He is also known as a teacher
of Timotheus of Gaza.329 Suidas indicates that Horapollo wrote a grammatical
work with the title (Temple Names), in which he discussed the mor-
phology of temple names.330 His philological activity includes commentaries
on Sophocles ( ) and on Alcaeus ( ) as
well as a monograph (or commentary) on Homer ( ). If Horapollo
is identical with the grammarian mentioned by Photius in his Bibliotheca
(cod. 279, 536a1517), then he was also active as a poet. Photius credits him
with the composition of dramatic poetry () as well as an antiquarian
work about Alexandria, which was probably written in verse (
).331 As already mentioned, poetic activity fits well with the schol-
arly profile during the first period of Byzantine philology.
Diogenes of Cyzicus,332 described by Suidas ( 1146) as grammarian, is
known for his antiquarian and topographical works, which would, however,
more appropriately characterize him as a historian. One such work, according
to Suidas, is the (On the homeland Cyzicus), which Stephanus
Towards the end of the Imperial era and throughout entire Late Antiquity, phi-
lologists and grammarians dealt almost exclusively with the of
the ancient , i.e. with the study of language. Earlier in this paper,
this was tracked back to a theoretical development within the discipline of
philology and grammar, which as from the 2nd century AD led to a clear split
between the two subject areas of ancient scholarship, namely the interpreta-
tion of literature and the study of language independently of its realization in
literary contexts. By the end of Late Antiquity and during the early Byzantine
period, this split was intensified, and the concept embodied by the term
was almost exclusively restricted to linguistic theory.334 This can
be seen clearly in the writings of philologist-grammarians of the 4th and espe-
cially of the 5th century, which now focused almost solely on linguistic studies.
As far as the history of grammatical studies in the narrow sense is con-
cerned, two figures of the 2nd century stand out: Apollonius Dyscolus and
his son Herodian. Their works were central in the linguistic theory of Late
Antiquity and the entire Byzantine period, and their importance goes some
way to explaining why many of their writings have survived, although mainly
preserved indirectly, as is the case above all for Herodian. However, it would
be a disservice to the diverse and profound studies in the area if this survey
of the development of ancient linguistics were to be restricted solely to these
two scholars. As the following overview will show, the field of linguistic theory
333 Since Suidas appears to confuse Diogenes with the grammarian and lexicographer
Diogenianus (on Diogenianus, see below in 6.3), these works are ascribed to Diogenianus;
this view is followed by Jacoby in FGrHist 474 T 1. Cf. also Kaster [1988] 399. Diogenes
oeuvre, however, as it is transmitted by Suidas, fits well to the profile of a learned scholar
and grammarian active during Late Antiquity.
334 See above 2.3.
Greek Scholarship in the Imperial Era and Late Antiquity 251
4.1 Between Alexandria, Rome and the Educational Centers of the East:
The Imperial Era
The grammarian Habron335 stands at the beginning of the Imperial period
and marks the transition from the Alexandrian era to the new epoch. The
only available biographical source on Habron is Suidas article ( 97), which
is explicitly derived from Hermippus biographical work.336 According to this
testimony, Habron was a slave of Phrygian origin, who studied (or also taught)
in Rhodes. Apparently, he became greatly esteemed there and was sent to
Tryphon in Alexandria to further his studies. He presumably went to Rome
as a freedman, where, as can be seen from his chronological relationship to
Tryphon, he worked until the reign of Tiberius.337 Judging from the preserved
fragments, Habron dealt exclusively with the so-called technical part (
) of the philological discipline and not with the interpretation of litera-
ture. Habrons work (On pronouns; fr. 18 B.) is cited with its
title by Apollonius Dyscolus, who transmits all of Habrons fragments belong-
ing to this treatise. The work (On denominatives; fr. 1118 B.)
was known to Herodian and Stephanus of Byzantium. Finally, the study
On possessive names ( ; fr. 910 B.) is attested in the scholia on
Dionysius Thrax.
As far as his research interests are concerned, Habron essentially followed in
the footsteps of his teacher Tryphon. His theory of the word class system shows
him to have been closely linked with the Alexandrian tradition initiated by
335 Literature on Habron: Grfenhan [18431850] III 57, 111 and 115, Susemihl [18911892] II
213214, Schmid-Sthlin [19201926] I 435, Funaioli [1912], Christes [1979] 9293, Fornaro
[1998c] and LGGA s.v. Habron. Habrons fragments have been collected and discussed by
Berndt [1915].
336 On Hermippus, see above 3.3.
337 For the details of Habrons biography, see Christes [1979] 9293.
252 Matthaios
338 On the development of the Alexandrian theory on the word class system, see Matthaios
[2002f].
339 See fr. 1 and the commentary of Berndt [1915] 14831485 on this fragment; cf. Matthaios
[1999] 481482 and Lallot [19982] 209210.
340 See fr. 2; on Habrons views, see Berndt [1915] 14851486 and Matthaios [1999] 447457.
341 On Apion, Lucillus and Chaeremon, see the relevant sections above 3.3.
342 On Seleucus, see above 3.3.
343 On Ptolemaeus grammatical activity, see below in this paragraph.
344 See above 3.3.
Greek Scholarship in the Imperial Era and Late Antiquity 253
345 Literature on Ptolemaeus of Ascalon: Grfenhan [18431850] III 3738, Blau [1883] 2537,
Susemihl [18911892] II 156158, Schmidt-Sthlin [19201926] I 439 and 444, Dihle [1959],
Matthaios [2001c] and Razzetti [2003d]. The preserved fragments from Ptolemaeus
writings have been collected by Baege [1882].
346 See Steph. Byz. 476 s.v. , ....
.
347 This is Lehrs view [1882] 26 n. 8 on the content of this treatise, from which, however, no
fragments have been preserved. See also Baege [1882] 1415.
348 See Baege [1882] 2122.
349 See Baege [1882] 1112.
350 With regard to the structure and scope of Ptolemaeus prosodic studies on Homer, see
Ammon. Diff. 436:
(p. 42 Baege)
. ... . See also Baege [1882] 911.
254 Matthaios
sources of his own prosodic studies and quoted abundantly from it,351 this
is the best-known composition by Ptolemaeus.352 Herein, he proved to be a
strict analogist and was especially concerned with creating prosodic rules in
accordance with the principle of analogy. This often brought him into conflict
with the accentuation of Homeric words proposed by Aristarchus, which was
instead mainly grounded on philological arguments.353 Ptolemaeus grammat-
ical studies included a treatise (On meters),354 while a lexicon
of synonyms with the title (On expressions with differ-
ent meanings) forms part of his lexicographic activity; however, the epitome
transmitted under his name is spurious.355
The grammarian Heraclides of Miletus356 appears to have lived around
100 AD,357 and is presumed to have been active in Alexandria.358 Only two
of his works are known by their title, (General
prosodic rules; fr. 115 C.) and (On irregular verbs;
fr. 1655 C.).359 With his work , Heraclides appar-
ently became the first grammarian to produce a coherent presentation of a
prosodic theory for the Greek language as a whole. He thus is a forerunner of
Herodian, who in fact used Heraclides in his own prosodic study, but often
cited him without mentioning his name. However, even where Herodian did
cite Heraclides by name, it was always with polemical intentions. Heraclides
work dealt with the field of inflection and was an
investigation of verbs with irregular conjugations verbs. Since Heraclides also
351 See below in this paragraph. On accentuation theories during the Roman Empire and the
studies accomplished during that time on this subject, see Hatzimichali [2006] 109119.
352 The fragments from this work are collected by Baege [1882] 3964.
353 A representative example for this is the explanation given by Aristarchus and Ptolemaeus
on the accentuation of the pronoun in Il. 1.396397 transmitted in Sch. Hom. (A)
Il. 1.396b1; for an interpretation of this testimony, see Matthaios [2012] 265272.
354 For a description of this work, see Baege [1882] 1213; the extant fragments are edited
ibidem, p. 64.
355 See below in 6.2.
356 Literature on Heraclides Milesius: Schmidt-Sthlin [19201926] I 439, Schultz [1912b],
Fornaro [1998h] and Razzetti [2003c]. His fragments were collected first by Frye [1883]
and then by Cohn [1884a].
357 Heraclides lifetime can be determined by the fact that he used Aristonicus, but was
himself quoted by Apollonius Dyscolus; see Frye [1883] 97 and Cohn [1884a] 611.
358 Heraclides is identical with the person often quoted by Eustathius with the nickname
Alexandrian; see Frye [1883] 102 n. 1.
359 Fr. 5660 C. are of unknown provenance; fr. 6162 C., on the other hand, are regarded by
Cohn as dubia.
Greek Scholarship in the Imperial Era and Late Antiquity 255
considered dialectal aspects in this research, his study was important for the
field of dialectology as well. The work is often cited by Eustathius; several frag-
ments can also be found in lexicographic and etymological works from the
Byzantine period.360
An epitome on Didymus (Miscellanea)361 was composed
by Alexion,362 a grammarian from the second half of the 1st century AD,363
who was also active above all in the field of orthography, prosody and ety-
mology. The titles of his grammatical writings are unfortunately not known.
The preserved fragments of his works, containing grammatical explanations
that concerned primarily Homer and the Homeric language,364 are mainly
transmitted by Herodian and in the Byzantine Etymologica. They contain pas-
sages in which Alexion often criticized the views put forward by his forerun-
ners, such as Tyrannion, Didymus, Ptolemaeus of Ascalon and Heracleon, who
favored the Aristarchean explanations of the problems under discussion.
The grammarian Draco from Stratoniceia in Caria,365 whose lifespan was
a little earlier than that of Apollonius Dyscolus, probably lived during the
2nd century AD. His writings deal largely with the technical part of ancient
scholarship:366 for instance, Dracos writing (On pronouns)
is concerned with the word class theory. Apollonius Dyscolus (Pron., GG II/1.1,
17.15) cites Dracos terminological suggestion from this work, according to
which possessive pronouns should be designated as (pronouns
which imply two persons).367 Draco also dealt with issues pertaining to
360 Heraclides theory and his accomplishments are presented in detail by Cohn [1884a]
2036; cf. Schultz [1912b] 492493.
361 See Ammon. Diff. 117: ... [fr. 1 B.] (p. 378
Schmidt); cf. Eust. 1788.52 and Etym. Gud. 124.2.
362 Literature on Alexion: Grfenhan [18431850] III 98 and 404, Schmid-Sthlin [19201926]
II 439, Wentzel [1894b], Montanari [1996b] and LGGA s.v. Alexion Cholus (L. Pagani).
Alexions fragments have been collected by Berndt [1906].
363 The testimony concerning Alexions epitome of Didymus (see n. 361) provides
the terminus ante quem for Alexions lifetime; see Wentzel [1894b] 1466.
364 For a discussion of the extant surviving fragments, see Wentzel [1894b].
365 Literature on Draco: Grfenhan [18431850] I 444, 454, 468, 478 and 502, Susemihl [1891
1892] II 193, Schmid-Sthlin [19201926] II 893, Cohn [1905a], Montanari [1997n] and
LGGA s.v. Draco (L. Pagani).
366 A list of Dracos writings is attested in Suid. 1496. The first work listed there under
the title is probably a description of the character of his writings, rther than a
particular work concerning the technical part of philology, the so-called ;
see Cohn [1905a] 1662.
367 Since in the Techne Grammatike attributed to Dionysius Thrax it is said that the possessive
pronouns are also designated as ([Dion. T.] Ars Gram. 17, GG I/1, 68.34), it
256 Matthaios
has been assumed that Draco was indeed a contemporary or even older than Dionysius
Thrax; see Susemihl [18911892] II 193 with n. 257 and Cohn [1905a] 1662. In view of the
doubts regarding the authorship and authenticity of this grammatical manual, the early
dating of Dracos lifetime should be disregarded; see Matthaios [1999] 485 with n. 260.
368 The treatise (On poetic meters), ascribed to Draco in Cod. Paris.
Gr. 2675 (the work is edited by Hermann [1812]), is a forgery of the 16th century; see Cohn
[1905a] 16621663. The appeal to Dracos name, however, speaks for his authority in
questions of metrical theories that reached into the Early Modern Age.
369 Nicanor is discussed above, 3.3, in connection with his philological works.
370 See Eust. 20.12; cf. Suid. 375.
371 The titles of these works are attested in Suid. 375.
372 The Viermnnerkommentar is the name of commentary on the Iliad that stems probably
from the early Byzantine period. It is a compilation of works of Aristonicus, Didymus,
Herodian and Nicanor. It is attested in the subscription to the individual books of the Iliad
in Cod. Venetus A; see Matthaios [2002e]; Dickey [2007] 1819; and Dickey in this volume.
373 The fragments from Nicanors studies on the Homeric punctuation have been collected,
as far as the Iliad is concerned, by Friedlnder [1850] and, in relation to the Odyssey, by
Carnuth [1875]. Both collections should now be compared with the new editions of the
Homeric scholia, that of Erbse [19691988] on the Iliad and that of Pontani [20072010]
on the Odyssey. On the quotations in the corpus of the Homeric and especially the Iliadic
scholia, which derive from Nicanors workd, see Schmidt [1976] 3539.
Greek Scholarship in the Imperial Era and Late Antiquity 257
380 See Blank [1993] 709, Lallot [1997] II 1213 and Brandenburg [2005] 14.
381 See Blank [1993] 709710 with reference to several allusions to teaching contained in
Apolonius writings.
382 For an exhaustive list of Apollonius writings on the basis of self-quotations of the
grammarian and in other sources, see Schneider in GG II/3, VIIX.
383 Edited by Schneider in GG II/1.1, 1116; the treatise On the pronoun has been translated
into German by Brandenburg [2005].
384 Edited by Schneider in GG II/1.1, 117210.
385 Edited by Schneider in GG II/1.1, 211258; a French translation of the treatise with an
exhaustive commentary has been provided by Dalimier [2001].
386 The standard edition of Apollonius Syntax is that of Uhlig in GG II/2; a new edition with
a French translation and extensive commentary is provided by Lallot [1997]. Apollonius
Syntax was translated into German by Buttmann [1877], into English by Householder
[1981] and into Spanish by Becares Botas [1987].
387 On the chronological order of the preserved writings, see Blank [1993] 710 with n. 13 and
Brandenburg [2005] 1617.
388 The extant fragments have been collected, arranged and commented by Schneider in
GG II/3.
Greek Scholarship in the Imperial Era and Late Antiquity 259
emphasis on the system of the word classes. Overall, the whole writing activ-
ity of Apollonius, not only his Syntax, but also his treatises On nouns (
) and On verbs ( ) exerted
particular influence on the later grammarians.389
The theoretical foundation of ancient grammatical doctrine is rightly acco-
ciated with Apollonius Dyscolus, as is also the case with Herodian. Apollonius
and Herodian were the two major figures whose activity resulted in the sys-
tematization and completion of the contents of the Alexandrian tradition.390
In the area of word class theory,391 Apollonius picked up the thread of the old
Alexandrian tradition that dated back to Aristarchus and which was repre-
sented by Tryphon at the end of the Hellenistic period. Essentially, Apollonius
submitted the theories formulated in the earlier era to a new and thorough
investigation. The available evidence shows that he shared the views of his
Alexandrian predecessors on certain basic questions, though reliable knowl-
edge of Stoic linguistic theory is more evident in Apollonius. Thus it is clear in
many parts of his writings that he argued directly with the Stoic sources, but
whereas in the case of his predecessors, especially Tryphon, the dispute with
Stoic views resulted in polemical rejection, in the case of Apollonius, it was cre-
atively integrated into his systematization of the word class theory. Apollonius
can thus be regarded as the scholar through whose accomplishments the
Alexandrian and Stoic linguistic theory were combined into a meaningful syn-
thesis and. Furthermore, he stands as the authority under whom the process
of formulating the canon of the word class system that stemmed from the
Alexandrian philological-grammatical tradition was brought to completion.392
As well as his word class theory, Apollonius essential contribution
to ancient linguistics is, from a diachronic point of view, the foundation of
the grammatical theory of syntax.393 In assessing ancient views on syntax,
389 A short overview of Apollonius writings transmitted in fragments is given by Blank [1993]
712713.
390 Among the special studies on Apollonius Dyscolus, which have meanwhile increased
considerably, mention should be made of Blank [1982], Sluiter [1990] 39140 and Lallot
[2012a].
391 On Apollonius accomplishments concerning the theoretical foundation of the ancient
word class system, see Matthaios [2002f] 197198.
392 See Lallot [1988]; cf. also the contribution of Swigger-Wouters (section III.2.1) on word
class theory in this volume.
393 On Apollonius theory of syntax see, in addition to the references given in n. 390, also
Steinthal [18901891] II 339347, Lallot [1997] I 2973 as well as the short, but highly
informative overview by Blank [1993] 714727. See also the contribution of Lallot in this
volume.
260 Matthaios
398 On the relationship of Dionysius of Halicarnassus syntax theory to that of Apollonius, see
De Jonge [2011] 475477.
399 On Apollonius after-effect, see Blank [1993] 728729; on syntax in Byzantine grammar,
see Robins [1993] 149233 and Pontani in this volume.
400 Basic literature on Herodian: Schmid-Sthlin [19201926] II 887888, Schultz [1912c],
Sandys [19213] 321322, Dyck [1993a], Montanari [1998c], Dickey [2007] 7577 and LGGA
s.v. Aelius (2) Herodianus (L. Pagani).
401 The testimonies on Herodians life have been presented by Lentz in GG III/1, VIVII, and
include the Apollonius Dyscolus-Vita (see above in this paragraph) and the short article of
Suidas ( 546). For a discussion of Herodians biographical testimonies, see Dyck [1993a]
772774.
402 Herodians entire oeuvre has been edited by Lentz in GG III/12.
403 The authentic writings by Herodian, which were edited by Lentz, are listed in the order
of the edition by Schultz [1912c] 962963; an excellent overview of Herodians writings is
provided by Dyck [1993a].
262 Matthaios
404 For a description of the work and its transmission, see Dyck [1993a] 776783; cf. Schultz
[1912c] 963965. The work has been edited by Lentz in GG III/1, 1547.
405 See above in this paragraph.
406 See above in this paragraph.
407 The collections of fragments provided by Lentz in GG II/2.1, 22128 (
) and II/2.1, 129165 ( ) are now supplemented by
Erbses edition of the Iliad scholia [19691988] and of the Odyssey scholia by Pontani
[20072010]. For a description of both writings, see Dyck [1993a] 783786; cf. Schultz
[1912c] 966967.
408 The codex has been edited by Hunger [1967a]; see Dyck [1993a] 780782.
409 Edited by Lentz in GG II/2.2, 908952; newly edited by Papazeti [2008]. For a description
of the work, see Dyck [1993a] 790791.
410 The works falsely ascribed to Herodian are listed by Schultz [1912c] 971973.
411 On Herodians accentuation theory, see Laum [1928]; see also the contribution of Probert
in this volume.
Greek Scholarship in the Imperial Era and Late Antiquity 263
background of the Alexandrian school: at the time of its foundation the early
Alexandrians, Aristophanes of Byzantium and Aristarchus, were involved in
the famous analogy-anomaly controversy with the Pergamenian philological
tradition and its exponent Crates of Mallus.412 The analogical method is based
on the idea that a grammatical accident, be it inflectional or prosodic, will
occur in similar word forms in the same fashion. Therefore, the words com-
pared in the process of analogy should be subjected to a prior similarity test.
Conditions of similarity, which must be heeded in the implementation of
analogical deductions, were already proposed by Aristophanes of Byzantium
and Aristarchus. Herodian significantly expanded the preconditions for
correct application of the analogical process, and created a canon of condi-
tions that took into account the following criteria:413 genus, species, type of
composition, number, accent, case, same ending in the nominative singular,
nature of the penultimate syllable, quantity of the vowel of the penultimate
syllable, number of syllables and the nature of the consonants before the
ending. After this rigorous application of the principle of analogy, Herodian
formulated prosodic, but also morphological and orthographical rules.
The problem of exceptional cases in a language, and more importantly, how
Herodian dealt with such exceptions, was treated separately in his work
(On lexical singularities). Here he examined words that have
anomalous character in the senseaccording to Sluiters formulation ([2011]
292)that they look normal enough and are in frequent use, but do not con-
form to the rules that would most obviously seem to apply. Yet even for these
exceptional cases, Herodian devised a grammatical explanation. Deviation
from a rule was to be explained on the basis of phonological and morphologi-
cal changes in the basic word form, the so-called .414 In this way the chasm
between the two most distant criteria for assessment of linguistic correct-
ness, analogy and common usage (, usus),415 was bridged. From this
412 On the history of the so-called analogy-anomaly-controversy and on the essential
theses of this feud, see Matthaios [2013b] with further references on this subject; see also
Montana and Pagani in this volume.
413 This catalog is attested in Herodians (On the nominal declination)
in GG II/2.2, 634.624. On this, see Siebenborn [1976] 7275. On the character and the
transmission of this specific work of Herodian, see Dyck [1993a] 789 and Dickey [2007]
7677.
414 This was evidently the subject of Herodians writing (GG III/2.1, 166388),
which was identical with the work (GG III/2.1, 389); see
Dyck [1993a] 786787. On the ancient -theory see Lallot [2012] 2136; Nifadopoulos
[2005]; and Pagani in this volume.
415 The criteria taken into consideration for evaluation of linguistic correctness are explored
by Siebenborn [1976]; see also the contribution of Pagani in this volume.
264 Matthaios
perspective, exceptions and problematic cases that depart from a rule are
brought into the system and can thus be regulated; hence common usage is
legitimized as a criterion for the normation of linguistic correctness.416
Lupercus and Cassius Longinus emerged in the area of grammar in the
middle of the 3rd century. According to Suidas ( 691), Lupercus,417 a native of
Berytus, was active during the reign of emperor Claudius Gothicus (268270).
He probably stood in a close relationship with the emperor.418 Suidas ascribes
to Lupercus a historical work with the title
(l. ) (The foundation of Arsinoites in Egypt) as well as the
treatise (On the rooster in Plato), which
was probably an interpretation of Socrates final words in Platos Phaedo
(118a). The largest portion of his oeuvre, however, was made up of works on
the so-called technical part of the philological discipline. This included a
, as well as a work consisting of 13 books on the grammatical
genera ( ).419 This work surpris-
ingly earned Suidas praise; Lupercus is said to have exceeded even Herodian
with this text.420 If Suidas refers to the 6th book of this work (
), which is still quoted in a marginal note by the hand of Maximus
Planudes in Plut. Mor. (De inim. util. 10) 91e in Cod. Ambr. C. 126 inf., fol. 27v,421
Lupercus writing probably survived or was known until the 13th century.
Lupercus also dealt with other grammatical questions, explaining the use
of the particle in his writing , which consisted of three books,
also addressing the very controversial accentuation of (or ) in the
treatise , as well as the quantity of the iota in the word in
his work .422 Additionally, Lupercus was active in the field of
lexicography: his collection (Attic vocabulary) is mentioned by
416 A thorough and insightful study of Herodians analogistic position, as it is stated in the
writing , is provided by Sluiter [2011].
417 Literature on Lupercus: Grfenhan [18431850] III 4243, 83, 121 and 200, Schmid-Sthlin
[19201926] II 889, Gudeman [1927b], Kaster [1988] 305 (nr. 91), Baumbach [1999b] and
LGGA s.v. Lupercus (G. Ucciardello).
418 See Gudeman [1927b] 18391840.
419 The few extant fragments from Lupercus work are attributed to this writing; see Gudeman
[1927b] 18401841. Gudemans assumption (ibid.) that this work was part of Lupercus
Techne grammatike is not compelling.
420 See Suid. 691: ... ; cf. Gudeman [1927b] 1841.
421 See Paton [1912].
422 The two last mentioned works are incorrectly regarded by Gudeman [1927b] 1840 as not
being independent writings, but as questions Lupercus had dealt with in his lexicon
.
Greek Scholarship in the Imperial Era and Late Antiquity 265
Suidas in the biographical article on Lupercus and is also counted among the
material enumerated in the Preface to Suidas (praef. 1.8) as forming part of the
Suidas sources.423
With a profile encompassing philosophy, rhetoric and philology, Cassius
Longinus424 emerged as an outstanding representative of the classicistic
movement during the 3rd century. He achieved great fame and veneration for
his comprehensive erudition. The names (living library)
and (wandering museum) attributed to him denote
admiration for his great knowledge.425 Longinus lived in the period from
about 210 to 272/273, and according to Suidas ( 645), he was active during the
reign of Emperor Aurelian. His mother was a sister of the sophist Fronto of
Emesa, whose fortune Longinus inherited. He owed his philosophical educa-
tion essentially to Ammonius Saccas and Origen. After completing his studies,
he settled in Athens as a teacher of philosophy, where Porphyry became his
student. Around 267 Longinus left Athens for the court of Queen Zenobia, the
widow of Odenathos in Palmyra. After suppression of their rebellion against
the Romans, Longinus and other advisors of the queen were executed at the
end of 272 or the beginning of 273.
It is to his philological activity that Longinus owes the nickname ,426
which separates him, to a certain extent and probably intentionally, from the
, inasmuch as his works emphasize the field of literary aesthetics
and criticism. Of Longinus extensive and diverse philological-grammatical
oeuvre, more titles than fragments have been preserved.427 His works include
several writings related to Homer such as Homeric questions (
), Is Homer a philosopher? ( ) etc. A large
proportion of his work falls within the area of lexicography, including two
editions of Attic Words ( ), special lexicons on
Antimachus428 and Heracleon, but also on Homer with the title
(On ambiguous words in Homer). Longinus
423 On Suidas Preface and its authenticity, see Matthaios [2006] 4 with n. 15 and Kaster [1988]
282.
424 Literature on Longinus: Grfenhan [18431850] III 352356, Schmid-Sthlin [19201926]
II 889891, Sandys [19213] 338339, Aulitzky [1927], Brisson-Patillon [1994] and [1998],
Baltes [1999] and LGGA s.v. Cassius Longinus (C. Castelli).
425 See Eunap. Vit. Soph. 4.1.15.
426 Thus according to Suid. 645; Porphyrius in Plot. 20 even refers to him as .
427 Longinus philological writings are presented in a detailed manner by Brisson-Patillon
[1998]. The relevant fragments have been collected, translated and commented by
Patillon-Brisson [2002].
428 On Longinus lexicon , see Matthews [1996] 75.
266 Matthaios
435 The extant fragments are quoted and discussed by Cohn [1907c].
436 Literature on Theodosius: Schmid-Sthlin [19201926] II 10781079, Gudeman [1934b],
Hunger [1978] II 1112, Kaster [1988] 366367 (Nr. 152), Wilson [1996] 4243 and 6971,
Matthaios [2002a], Dickey [2007] 8384 and LGGA s.v. Theodosius (L. Pagani).
437 See the inscriptio to Theodosius :
(GG IV/1, 3.12)
438 See Syn. Ep. 4.310316; for the identification of this Theodosius with the grammarian of
the same name from Alexandria, see Oguse [1957] 8586; cf. Kaster [1988] 152 and ibid.
(Nr. 151).
439 The work is edited by Hilgard in GG IV/1, 399.
440 Choeroboscus commentary is edited by Hilgard in GG IV/1, 103417 and GG IV/2, 1371.
The commentary of Charax has survived in the excerpted version of Sophronius; it is
edited by Hilgard in GG IV/2, 375434.
441 They are edited by Hilgard [1887] 1622 and 2224.
442 See Laum [1928] 2728; cf. Kaster [1988] 367.
443 In three manuscripts of the epitome of Herodians workin the Codd. Matrit. 38, Barocc.
179 and Haun. 1965Theodosius is named as the author; see Kaster [1988] 367.
268 Matthaios
468 Both orthographical writings of Eugenius are regarded by Cohn [1907d] 988 as sections
of his lexicographic work ; but the fact that this lexicon is
arranged alphabetically makes Cohns assumption questionable.
469 See Stephanus quotation cited in the previous n. 465.
470 On Eugenius poetical work, see Schmid-Sthlin [19201926] II 974.
471 Literature on Timotheus: Seitz [1892] 3032, Schmid-Sthlin [19201926] II 974975 and
1077, Wellmann [1927], Steier [1937], Hunger [1978] II 13, 1819 and 265, Kaster [1988] 368
370 (Nr. 156), Wilson [1996] 31, 44 and 143 as well as Matthaios [2002c].
472 See also Tzetz. H. 4.166, which names Timotheus as being among the authors who,
according to Aelian and Oppian, wrote about zoology.
473 See Reitzenstein [1897] 296 and the material cited on p. 296297 and p. 312316. On the
question of which Horapollo is meant here, see Kaster [1988] 369370.
474 The work was edited by Cramer in Anecdota Graeca, vol. 4, 239244.
475 On this work, see Kaster [1988] 369; on Timotheus orthographical studies, see Egenolff
[1888] 68 and Schneider [1999] 1571.
476 The extant excerpts are collected by Haupt [1869]; an English translation is provided by
Bodenheimer-Rabinowitz [1949].
477 On these works, see Kaster [1988] 368369 with additional references.
272 Matthaios
495 See Longin. Proleg. Heph. 81.1215 Consbruch. On Longinus commentary on Heliodorus
metric manual, see above in this paragraph.
496 See Schmid-Sthlin [19201926] I 445.
497 Literature on Hephaestion: Grfenhan [18431850] III 69, 106107, 228, 327 and 340,
Hense [1912b], Schmid-Sthlin [19201926] II 891893, Sandys [19213] 328, Fornaro
[1998f], Dickey [2007] 104105 and LGGA s.v. Hephaestion (A. Ippolito); see also Dickey in
this volume.
498 See Suid. 659.
499 See Hist. Aug. Verus 2.4. On Telephus, see above 3.3.
500 For the history of the formation of Hephaestions work, see the information provided by
Choeroboscus, Scholia in Hephaestionem 181.1113 Consbruch. The work is edited together
with Choeroboscus commentary and the scholia on it by Consbruch [1906]; an English
translation with commentary is provided by van Ophuijsen [1987].
Greek Scholarship in the Imperial Era and Late Antiquity 275
exercises; they were taken from lyric poets and comedians but not from trage-
dians, while choral lyric poetry was likewise disregarded.
Choeroboscus wrote a commentary on Hephaestions manual, which was
accompanied by a rich collection of scholia from the Byzantine period; a com-
mentary on Hephaestion was also composed by Longinus. The commentaries
and scholia are important because they are partially based on the original and
extended version of the work. Several other philological-grammatical writ-
ings of Hephaestion are known through Suidas ( 659), but only by their titles.
In one passage transmitted by Porphyry (ad Il. 12.127132, 177.3135 Schrader),
remains of Hephaestions interpretation of Homer have been preserved.501
Draco of Stratoniceia was active at about the same time as Hephaestion. In
the list of his works as recorded by Suidas several metrical treatises are attrib-
uted to Draco, though none have survived.502 Eugenius of Augustopolis likewise
dealt with metrics in the 5th century. It was mentioned earlier503 that Eugenius
wrote a colometry on the lyrical passages of the tragedians, thereby continu-
ing and expanding Hephaestions work, which was restricted to Aristophanes.
A special study on the palimbaccheus is also attributed to Eugenius.
Lexicographic exploration of the Greek language was one of the most pro-
ductive areas of philological-grammatical activity in the Imperial era and
Late Antiquity. Several authors rendered service in this field and their names
have been connected with great achievements both in terms of quantity and
effort.504 Looking back at the overall evolution of ancient lexicography, it
should be noted that the area reached its widest development in this period
not just in relation to the contents of the dictionaries produced and the lexico-
graphic goal of each one, but also in terms of the typology of genres. All levels
of language, including literary language, dialectal vocabulary as well as scien-
tific language were analyzed. Regarding the principle of systematization, the
501 This fragment is ascribed by MacPhail [2011] 201 n. 136 to Hephaestions work
(On confusions that occur in poems), which is attested in Suidas.
502 On Draco, see above 4.1.
503 See above in this paragraph.
504 For an overview of Greek lexicography from Antiquity and the Middle Ages, see Cohn
[1913], Tolkiehn [1925], Serrano Aybar [1977], Degani [1995], Alpers [1990], [2001] and
[2008], and Tosi in this volume. See also Matthaios [2010a], who, in addition to history,
treats methodological aspects based upon the relations of each lexicographer to his
sources.
276 Matthaios
different from todays lexicography, namely the ability to synthesize and sys-
tematize, but also to accurately select the relevant data from the available
material and to critically analyze the respective sources.
In effect, the linguistic, historical, ideological and cultural conditions of
the era are reflected in an especially lively manner in the lexicography of the
Imperial era and Late Antiquity. This is true in particular for that branch of lex-
icography characterized by the term Atticistic lexicography, which became a
crucial feature of the period of ancient scholarship discussed here:508 indeed,
Atticistic lexicography marked a turning point in the entire field of lexicogra-
phy. Lexica, especially the Atticistic collections, no longer merely served the
purpose of explaining literary language, but were also intended for new text
production. The productive function, however, involved a closely connection
with the normative and prescriptive character of lexicography, as the lexi-
cographer attempted with his work to instruct the user in the correct use of
language, at the same time also exercising influence in questions of language
development.
The following overview is divided according to the individual genres of
lexica compiled in the Imperial era and Late Antiquity, and discusses the most
important representatives and works in each area according to a historical
order.
511 The lexicon was edited by Bekker [1833]; an edition of the first four letters is provided
by Steinicke [1957]. Papyri containing fragments from Apollonius lexicon have been
studied by Haslam [1982], [1992] and [1994], who also evaluated them for the history of
the transmission of the lexicon.
512 On Apion and his Homeric glossary, see above 3.3.
513 On the Homeric scholar Heliodorus, who was probably not identical to the metrician
of the same name (on this Heliodorus, cf. above 5), see Dyck [1993b] 16, who has also
collected and commented on his fragments.
514 A detailed report on the sources of Apollonius Sophistes is presented by Schenck [1974];
cf. Erbse [1960] 407432.
515 On this work, see Cohn [1913] 691 and Tolkiehn [1925] 2445.
516 On this work of Longinus, see Tolkiehn [1925] 2460; on Longinus, see above 4.1.
Greek Scholarship in the Imperial Era and Late Antiquity 279
525 The work is quoted by Orion 134.34 and 170.29 as well as in Etym. Magn. 552.2 and 722.22.
On Apollonius glossary, see Cohn [1913] 691 and Tolkiehn [1925] 2446.
526 See Tolkiehn [1925] 2446 with references to further literature and the editions of the
glossary.
527 Literature on Claudius Didymus: Grfenhan [18431850] III 64 and 148, Cohn [1903d],
Schmid-Sthlin [19201926] I 434 and 438439, Christes [1979] 104105, Montanari [1997g]
and LGGA s.v. Didymus (2) Claudius (F. Montana).
528 On this work, see Cohn [1913] 691 and Tolkiehn [1925] 2446.
529 The Platonic philosopher Harpocration of Argos lived in the 2nd century AD and was
a pupil of the Platonist Atticus. According to Suidas statement ( 4011), in addition to
a commentary on Plato in 24 books, Harpocration authored a lexicon on the Platonic
vocabulary consisting of two books with the title (Platonic expressions);
see Cohn [1913] 691 and Tolkiehn [1925] 2446.
530 Literature on Boethus: Grfenhan [18431850] III 181, Cohn [1903a] and [1913] 691, Schmid-
Sthlin [19201926] I 442, Tolkiehn [1925] 2446, Montanari [1997c] and LGGA s.v. Boethus
(2).
531 Boethus lifetime can be determined from an entry in the Lexicon of Hesychius ( 1201),
where it is evident that Diogenianus used Boethus and quoted him; see Cohn [1903a] 254.
Greek Scholarship in the Imperial Era and Late Antiquity 281
Bibliotheca (cod. 154, 100a1317 and cod. 155, 100a1824)532 and were used by
the Patriarch himself as sources for the compilation of his own lexicon.533
We also have knowledge of another two, fully preserved Platonic lexica, by
different authors. The first, stemming from Timaeus534 is a lexicon bearing the
title an alphabetical selection of Platonic expressions (
), which can probably be dated to the 3rd century.535 This lexicon was
available to Photius (Bibl. cod. 154, 100a1516), and was in his judgment far sur-
passed by that of Boethus. The second is entitled
(On Questionable Expressions in Plato) and is falsely ascribed
to Didymus Chalkenteros.536 Its date of origin is not easily determined: it was
compiled sometime in the period between the 3rd and the beginning of the
10th century.537
Medical language, especially the language of the Hippocratic corpus, was
already a special subject of lexicographic research, both from the point of view
of philology and of medicine.538 In the middle or towards the end of the 1st
century AD, the grammarian Erotian539 presented an extensive Collection of
expressions occurring in Hippocrates ( ),
utilizing a large amount of source material. He dedicated his work to
Andromachus, a physician at the Imperial court in Rome. In his arrangement
of the lexicographic entries, Erotian followed the text of the Hippocratic cor-
pus according to the order of the individual writings and explained the glosses
of each writing in such a manner that whenever a word appeared for the first
time, he made reference to all other occurrences of the same term. However,
Erotians lexicon has not been preserved in its original form, as the work began
532 On the character of Boethus lexica, see Cohn [1884b] 783786, 794813 and 836852; cf.
also Dyck [1985].
533 See Theodoridis [19821998] I, LXXIIILXXIV with references to other literature.
534 Literature on Timaeus: Schmid-Sthlin [19201926] II 1081, von Fritz [1936], Alpers [2001]
200 and [2008] 12541255, Matthaios [2002b] and LGGA s.v. Timaeus.
535 The lexicon has been edited by Valente [2012]; on Bonellis edition [2007], see the review
of Alpers [2009a]. On the datation of the compilation of the lexicon, see Valente [2012] 57.
536 On this Platonic glossary, see Cohn [1913] 691 and Tolkiehn [1925] 2447. The work has been
newly edited by Valente [2012].
537 On the dating of the glossary, see Valente [2012] 250.
538 On lexicographic activity on the field of ancient medicine, see Tolkiehn [1925] 24502451
and Manetti in this volume. Information on this matter is provided by Erotian in the
Preafatio to his Hippocratic lexicon (19 Nachmanson).
539 Literature on Erotian: Grfenhan [18431850] III 178179, Cohn [1907b] and [1913] 691
692, Tolkiehn [1925] 24512452, Nutton [1998a], Alpers [2001] 199200 and [2008] 1255 and
LGGA s.v. Erotianus.
282 Matthaios
540 The origin and transmission of the work was investigated by Nachmanson [1917]; the
standard edition of the lexicon (Nachmanson [1918]) is based upon this work.
541 Literature on Caecilius: Grfenhan [18431850] III 136, 192 and 348350, Brzoska [1897],
Schmid-Sthlin [19201926] II 463466, Weienberger [1997] and LGGA s.v. Caecilius.
After Ofenloch [1907], the fragments of Caecilius have been collected and commented on
by Augello [2006]; see also Augello [2006] XIIIXX for an overview of Caecilius life and
work.
542 See e.g. Brzoska [1897] 1185; cf. Cohn [1913] 696 and Tolkiehn [1925] 2447.
543 Literature on Vestinus: Grfenhan [18431850] III 52 and 187, Kroll [1918a], Ziegler [1958],
Matthaios [2002d] and LGGA s.v Iulius (4) Vestinus (A. Ippolito).
Greek Scholarship in the Imperial Era and Late Antiquity 283
fact that the wording of the lexicon preserved on the papyrus agrees with the
lexicon of Diodorus that was described by Photius.
Thanks to this finding, it is not only Diodorus lexicon that becomes tangible,
but also its prototype, the lexicon by Julianus. By following Alpers reconstruc-
tion, we can identify the source upon which Diodorus, Harpocratio, the 5th
Bekker-lexicon, the Lexicon Cantabrigiense and, last but not least, Pollux were
based. This source was not, as Wentzel maintained,548 an Attic Onomastikon,
but the lexicon of Julianus, who had first alphabetized his primary onomastic
sourcesglosses on names of months, representations of the Athenian court
system as well as Attic localities and authoritiesand unified them with the
work of Didymus Chalkenteros on the Attic orators.549
The only surviving Rednerlexikon from the Imperial era is that of Valerius
Harpocration.550 Harpocration is characterized by Suidas ( 4014) as .
He came from Alexandria and, if he is to be identified with the Harpocration
named in Hist. Aug. Verus 2.5, he lived in the second half of the 2nd century AD551
and was a teacher of Emperor L. Verus. His lexicographic activity is associated
with the (Lexicon of the ten orators), which is not only
of particular importance for the language of Attic orators, but also for the Attic
juridicial and political system as well as for the cultural history and topography
of Athens. Following, again, Alpers reconstruction (see above), Harpocration
relied significantly on Julianus, but it cannot be ruled out that he also used
and incorporated other sources. The lexicon has been transmitted through
two routes: one complete version, which has survived in younger manuscripts,
stands against an epitome, which is verifiable552 already in the 9th century,
by virtue of quotations preserved in the lexica of that time, such as Photius
and Suidas.553
548 See Alpers [1981] 117 and the reference made in n. 77 to Wentzels studies.
549 That is the conclusion reached by Alpers [1981] 123.
550 Literature on Harpocration: Schultz [1912a], Schmid-Sthlin [19201926] II 876877, Cohn
[1913] 696, Tolkiehn [1925] 2462, Tosi [1998b], Alpers [2001] 197, Dickey [2007] 94 and
LGGA s.v. Valerius (3) Harpocration (F. Montana).
551 On Harpocratios lifetime, see Alpers [1981] 116 n. 74 and the references discussed therein.
552 Harpocratios lexicon was edited by Dindorf [1853] and Keaney [1991].
553 See Theodoridis [19821998] II, XLIXLVI.
Greek Scholarship in the Imperial Era and Late Antiquity 285
554 On the lexicographic works of Callimachus and Aristophanes of Byzantium, see the
contribution by Montana in this volume. On Callimachus, cf. Matthaios [2008] 600601
with references to further literature.
555 An excellent overview of the ancient Onomastica is provided by Tosi [2000a]; cf. also Tosi
in this volume.
556 Literature on Dionysius Tryphonos: Grfenhan [18431850] I 538, Schmid-Sthlin [1920
1926] I 435, Cohn [1903f], Montanari [1997m] and LGGA s.v. Dionysius (15) Tryphonius
(V. Novembri).
557 The relevant passages are listed by Cohn [1903f]. From a passage in Athenaeus (14.641a),
where an explanation of the word is cited under Dionysius name, which,
however, repeats the same explanation given by Tryphon (Ath. 14.640e = Tryph. fr. dubiae
sedis 136 von Velsen); it appears somewhat premature to conclude that Dionysius work
was only a reworking or an epitome of Tryphons work (Tryph. fr. 109115
von Velsen), as Cohn [1903f] maintains, following Rohde [1870] 66; cf. also von Velsen
[1853] 101.
558 On Seleucus work, see above 3.3.
559 To what extent these works are identical, as Mller [1891] 21 assumed (cf. Razzetti
[2002b]), remains uncertain. On the lexicographic work of Seleucus, see Cohn [1913] 688
and Tolkiehn [1925] 2453.
286 Matthaios
he certainly lived before Athenaeus, who quotes him. Simaristos wrote a work
(On synonyms) in at least 4 books.560 Ptolemaeus of Ascalon
compiled a lexicographic work titled (On expressions
with different meanings).561 Pamphilus work On Plants ( ) was
also arranged onomastically,562 as were the lexicographic works of Telephus, in
which expressions for objects of everyday use were collected and explained.563
The only completely preserved representative of the onomastic genre is a
work which Pollux himself entitled in 10 books. Since this work
can be assigned to Atticist lexicography due to its time of origin and particu-
larly its ideological and linguistic-historical background, it will be discussed in
the relevant section.564 The collection On Ethnics by the grammarian Orus
still falls within the area of Onomastica: it is a geographic-etymological lexicon,
dealing with the adjectival form of place names.565 It was also one of the main
sources for the lexicon by Stephanus of Byzantium, which consists of
more than 50 books.566 Stephanus, active as a grammarian at the Imperial
university of Constantinople, dedicated this work to emperor Justinian. The
work, probably composed around 530, contains detailed linguistic, geographic,
historical and mythological explanations of a large number of place names
and the ethnica belonging to them. It has survived in a drastically epitomized
version along with some excerpts from the complete work in the excerpts by
Constantine Porphyrogennetos.567
Herennius Philo from the Phoenician city of Byblos is known for a work
that examined a number of synonymous expres-
sions and also devoted considerable attention to Onomastica.568 Several long
560 On Simaristos, see Cohn [1913] 688 and Tolkiehn [1925] 2453.
561 On Ptolemaeus of Ascalon, see above 4.1.
562 On Pamphilus, see above 3.3.
563 On Telephus and his lexicographic work, see above 3.3.
564 See below in 6.4.
565 On Oros and his lexicographic work, see above 4.2.
566 Literature on Stephanus of Byzantium: Schmid-Sthlin [19201926] II 10841085, Sandys
[19213] 379, Honigmann [1929], Hunger [1978] II 3637, Kaster [1988] 362363 (Nr. 144),
Grtner [2001], Dickey [2007] 101, LGGA s.v. Stephanus Byzantius, and Pontani in this
volume.
567 On Stephanus lexicon, see also Cohn [1913] 702, Tolkiehn [1925] 24692470, Alpers [2001]
201 as well as the informative Prolegomena in the edition of Billerbeck [2006]. The work
was edited by Meineke [1849]; a new edition is in preparation by M. Billerbeck. So far
three volumes have been published: Billerbeck [2006], Billerbeck-Zubler [2011] and
Billerbeck [2014].
568 On Herennius Philo and his Synonymicon, see above 3.3.
Greek Scholarship in the Imperial Era and Late Antiquity 287
passages or different versions from this collection have survived, which forms
the basis of the ancient synonymica.569 One branch of the transmission is
the lexicon ascribed to a certain Ammonius.570 Others are known under the
name of a certain Ptolemaeus, probably because of the lexicographic work of
Ptolemaeus of Ascalon, still others under the name of (H)Eren(n)ius Philo and
Eranius Philo.571 The Byzantine dictionaries of synonyms were built on these
collections, especially the so-called Etymologicum Symeonis from the 12th cen-
tury, entitled . In the first half
of the 5th century the grammarian Orus compiled a lexicon bearing the title
(On Ambiguous Expressions), from which very few frag-
ments have survived.572
The special etymological studies from the philological-grammatical tradi-
tion of the Hellenistic and Imperial era were likewise reworked in lexicographic
form. A comprehensive presentation of the etymological theory was provided
in the form of a lexicon by the grammarian Orio from Thebes in Egypt in the
5th century.573 Orio first taught in Alexandria, later went to Constantinople
and finally to Caesarea. Orio is known for his Etymologicon, which in the
manuscript tradition is titled , or .
The dictionary has survived in three different versions, each having been epito-
mized to a varying degree.574 The entries are arranged alphabetically in such
a way that the sources used recur in a specific order within each entry. The
structure of the individual lemmas shows the following sequence of sources:575
commentaries on Homer and other poets, Soranus work
, Herodians and his orthographic writings,
569 On this branch of ancient lexicography, see Cohn [1913] 702703, Tolkiehn [1925] 2453,
Nickau [1966] LXIIILXVI and Alpers [2001] 200201.
570 The work was edited by Nickau [1966]. Nickau, ibid., LXVILXVII, discusses the question
regarding the authorship of the collection of synonyms; cf. also Nickau [1990].
571 On the various epitomes, see Nickau [1966] XXVIIXLIV; cf. Erbse [1960] 295310, Nickau
[2000] and Dickey [2007] 9496. The Ptolemaeus-Epitome was edited after Heylbut
[1887] by Palmieri [19811982], that of Eranius Philo by Palmieri [1981] and that of (H)
Erennius Philo by Palmieri [1988].
572 On Orus and his lexicographic work, see above 4.2.
573 Literature on Orio: Schmid-Sthlin [19201926] II 1081, Sandys [19213] 377378, Wendel
[1939b], Hunger [1978] II 45, Kaster [1988] 322325 (Nr. 110), Tosi [2000b], Dickey [2007]
99100 and LGGA s.v. Orion (A. Ippolito).
574 Edited by Sturz [1820]. On Orios Etymologicon, see Cohn [1913] 697698, Tolkiehn [1925]
24642465 and Alpers [2001] 201.
575 See Kleist [1865] 1538 and Theodoridis [1976] 1641.
288 Matthaios
vocabulary from all areas of nature and human life as well as from various lin-
guistic, dialectal and literary levels. In order to achieve this lexicographic goal,
which was substantially broader than that of his predecessors, Pamphilus took
the various sources of Hellenistic and early Imperial philological exegesis and
lexicography into account and integrated them into his work.
The work in its original form, however, was short-lived, presumably due to its
immense size. L. Julius Vestinus582 made an epitome from Pamphilus lexicon
in the first half of the 2nd century, to which he gave the title
(Greek designations). Shortly thereafter, the grammarian Diogenianus
abridged the work of Pamphilus (or the Epitome of Vestinus) to five books,
giving the new work the title (Expressions of any kind). As
Hesychius testifies (Epist. ad Eulogium 1.523 Latte), Diogenianus then extended
the and called this new collection (Manual
for those without means).583 In this form, Diogenianuss
was the basic source of a lexicon written around 500 AD and attributed to
Hesychius, a grammarian from Alexandria.584 According to his own statement
(Epist. ad Eulogium 1.232.37), Hesychius expanded his prototype by including
proverbs, Herodians prosodic rules and other material. The transmission of the
lexicon continued throughout the entire Byzantine millennium, and during
this period the original material experienced changes, cuts, but above all addi-
tions, which have come down to us in the only available manuscript, the Cod.
Marcianus Graecus 622, written around 1430. Of these alterations, the main
change involved the interpolation of the so-called Cyrill-lexicon. Indeed, the
degree of interpolation of entries from Cyrill is so high that one can speak of an
amalgamation of the two lexica. According to new estimates, the interpolations
from Cyrill comprise one third of the total material. The Cyrill lexicon585 is a
glossary transmitted under the name of the Patriarch Cyrill of Alexandria. It
stems from approximately the same time as the Hesychius lexicon and rep-
resents the foundation of all the great Byzantine lexica, as can be seen from
its wide-ranging manuscript tradition. The sources of the lexicon consist of
elementary glossaries on Homer and Euripides, glosses on Josephus, Plato,
Demosthenes, on physicians, on the Bible, on Hellenistic poets and on Christian
authors, as well as glosses of Diogenianus and Atticistic lexicographers.
585 On the Cyrill-lexicon, see Drachmann [1936], Latte [19531966] I, XIIXIII and XLIVLI
and Cunningham [2003] 4349; a brief presentation of the lexicon is given by Cohn [1913]
698699, Tolkiehn [1925] 24652467, Serrano Aybar [1977] 101, Hunger [1978] II 3739,
Alpers [1990] 2425, [2001] 201202 and [2008] 12631266 and Dickey [2007] 100101; see
also Pontani in this volume.
586 For an overview of Atticistic lexicography, see Cohn [1913] 693695, Tolkiehn [1925] 2454
2463, Serrano Aybar [1977] 9396, Degani [1995] 519522 as well as Alpers [1990] 2024,
[2001] 196199 and [2008] 12551256.
587 The history of Atticism is the subject of the foundational study by Schmid [18871896];
see also Frchtel [1950] as well as Dihle [1977] and [1992]. On the effect of Atticism on
the development of the Greek language, see Triantafyllidis [19381981], Browning [1969]
4955 and Horrocks [1997] 7986.
588 The demands placed upon a speaker of Greek at that time are correctly described by
Dihle [1977] 162: Jeder, der berhaupt zur Feder griffund sei es nur, um die Einladung
zu einem Abendessen zu formulieren, stand nunmehr unter der Forderung, sich fr
den schriftlichen Gebrauch einer lngst obsolet gewordenen, mhsam zu erlernenden
Sprachform zu bedienen. On the significance of the Atticist movement from a socio-
political and ideological perspective, see Anderson [1993] 86100, Swain [1996] 1764,
Schmitz [1997] 6796 and 110127 as well as Whitmarsh [2005] 4156.
Greek Scholarship in the Imperial Era and Late Antiquity 291
Quite early on, lexicography was placed at the service of the Atticistic move-
ment. Thus when, during the Second Sophisticespecially during the reign
of HadrianAtticism reached its peak, rhetoricians compiled lexica for the
purpose of collecting the entire range of expressions found in the canon of
classical authors, in such a manner as to distinguish the accepted vocabulary
from the , i.e. from expressions regarded as corrupted because
they did not belong to the Attic idiom.589 A characteristic of the approach
adopted by the Atticistic lexicographers was the use of derogatory terms to
refer to the Koine forms, which were to be avoided in favor of Attic vocabu-
lary: thus descriptions such as , , or , ,
, and explicitly discouraged the Koine use. The cor-
rupted expressions were furthermore attributed to a community of speakers
labeled with attributes such as , and , or, in direct contrast
to the , as .590 Current research has shown that such expres-
sions served for the designation of corrupta Graecitas.591 The were no
longer speakers of classical Greek, but spoke a later form of language, which
represented the mean and low level. After the Second Sophistic, the Atticistic
dictionaries served as source material for the later lexicographic tradition and
formed the ground for ideological battles concerning the development of the
Greek language.
The earliest of the Atticistic dictionaries derives from the grammarian
Irenaeus, the pupil of Heliodorus,592 who taught in Rome under the Latin
name Minucius Pacatus.593 Among other studies, Irenaeus compiled a work
consisting of 3 books with the title (Attic expressions), an
alphabetically arranged lexicon
(On the Attic language use concerning vocabulary and pros-
ody), and also a treatise (On Atticism). Irenaeus is sub-
sumed under the field of Atticistic lexicography, not only because he is often
589 On the expression , see Lucian. Sol. 11 and Lex. 24, as well as Philostr. Vit.
Soph. 2.8.1. On the menaing of the expression in this context, see Alpers [1981] 45 and
[1990] 22.
590 On these expressions by which Koine-Greek and its speakers were degraded as
uneducated, see Matthaios [2010a] 187; cf. also Matthaios [2013a] with a study of the use
of such expressions in Pollux Onomasticon.
591 See Maidhof [1912] 4364.
592 On Heliodorus, see above 5.
593 Literature on Irenaeus / Minucius Pacatus: Grfenhan [18431850] III 58, 192 and passim,
Cohn [1905b], Schmid-Sthlin [19201926] II 873, Christes [1979] 104105, Fornaro [1997a]
and Regali [2007b]. On his life and writings, see Suid. 190 and 29. Irenaeus fragments
were collected by Haupt [1876].
292 Matthaios
604 Literature on Julius Polydeukes: Grfenhan [18431850] III 166169, Bethe [1918], Schmid-
Sthlin [19201926] II 877878, Tosi [1999], Dickey [2007] 96 and LGGA s.v. Iulius (2)
Pollux (L. Pagani). The Onomasticon has been edited by Bethe [19001937]. Two collected
volumes have recently been published on Pollux and his activity: Bearzot-Landucci-
Zecchini [2007] and Mauduit [2013].
605 On Pollux life and activity, see the Vita in Philostratus (Vit. Soph. 2.12) as well as the
biographical article in Suidas ( 1951); cf. Bethe [1918] 773774, Rothe [1989] 153154 and
Tosi [1999] 51.
606 On the time of origin of the Onomasticon, see Matthaios [2013a] 7073 with references to
other literature on this question.
607 See Alpers [2001] 198: es stellt immer noch eine der wichtigsten lexikographischen
Arbeiten des Altertums dar; cf. Alpers [1990] 24. On the significance of the Onomasticon,
see also Bethe [1918] 779: Das Onomastikon gibt uns neben Athenaios wohl die beste
Vorstellung von der umfassenden, geduldigen lexikographischen Arbeit der antiken
Philologen. On the position the Onomasticon holds in the history of ancient lexicography,
see Cohn [1913] 692693, Tolkiehn [1925] 24572458, Serrano Aybar [1977] 92, Degani
[1995] 521522, Tosi [1999] 5253 and Alpers [2008] 1256.
608 For a brief overview of the subjects of the Onomasticon and their division into the
individual books, see Bethe [1918] 776777.
Greek Scholarship in the Imperial Era and Late Antiquity 295
609 See Matthaios [2013a] 70 with the references cited there in n. 18.
610 See Matthaios [2013a] 7178.
611 This is shown in an analysis of Pollux attitude toward the use of language by the ,
and somewhat by the and to a certain extent by the -speakers attested in the
Onomasticon; see Matthaios [2013a] 81114. Although the boundaries between the way of
expression of these particular groups of speakers was fluid and often conditional upon
the variously accented arguments of Pollux, Pollux shares the same basic view of the
other Atticistic lexicographers of what is actually considered linguistically admissible.
612 See Matthaios [2013a] 127128.
296 Matthaios
613 On a definition of Atticism and its relationship to classicism, the distinctions made by
Gelzer [1979] 2441 are especially significant.
614 Literature on Moeris: Wendel [1932b], Baumbach [2000], Dickey [2007] 98 and LGGA
s.v. Moeris. Moeris glossary has been edited by Hansen [1998], who also analyzed the
transmission and the sources of the lexicon.
615 The lexicon was edited by Dain [1954]. On its attribution to Herodian, see Dyck [1993a]
791792.
616 On Alexander and his lexicographic work, see above 3.3.
617 On Lupercus, see above 4.1.
618 On Orus and his Atticistic lexicon, see above in this paragraph.
619 See Alpers [2001] 199.
620 See Pontani in this volume.
chapter 4
* The most reliable and exhaustive general treatments of this topic are Wilson [19962] and
Hunger [1978] 383. Also very useful are the briefer treatments by A. Pontani [1995a], Wilson
[2004] and Flusin [2008]. On written culture see the papers collected in Cavallo [2002a]. On
special periods see Lemerle [1971], Constantinides [1982], Mergiali [1996], Fryde [2000], and
partly Wilson [1992] (with A. Pontani [1995b]).
1.1 Beginnings
Sept. 22nd, 529 has been regarded by some scholars as the symbolic coupure
marking the end of antiquity in the Eastern Roman Empire:1 the date cor-
responds to Justinians closing of the philosophical school at Athens, a fact
whose historicity is beyond doubt, and whose effects on the cultural life of the
Greek East have been variously assessed.2 Whether it was a local move directed
against the teaching of philosophy and astronomy, i.e. against the school of
Damascius,3 whether it marked the definitive end of Platonic studies on Greek
soil (some teachers, from Damascius to Simplicius, apparently moved to Persia
or Syria, and there is hardly any evidence of philosophy being taught in Athens
thereafter),4 or the end of curricular instruction altogether (in Zonaras words,
being in need of an enormous amount of money, following the prefects advice
<Justinian> cut the salaries that in every town had always been given by law
to the teachers of the liberal arts, and thus the schools were left unattended
1 This periodisation is reported (and countered) in the first sentence of Krumbacher [1897] 1;
see also Hadas [1950] 273, who takes precisely 529 as the final limit of his survey.
2 Beaucamp [2002].
3 Watts [2004], who downplays the effects on other fields of teaching such as rhetoric; see
already Cameron [1969].
4 Blumenthal [1978]; on the peregrinations of the Athenian philosophers see Chuvin [2012]
138146 and Thiel [1999], to be read however with Luna [2001]; Watts [2006] 138142.
Scholarship in the Byzantine Empire ( 5291453 ) 299
see Fuchs [1926] 4; Lemerle [1971] 6871. A different context for this law is posited by
Trombley [2001] 8194.
12 Watts [2006] 8087; Frantz [1988]; Frantz [1975]; doubts on this identification have been
cast by Chuvin [2012] 251.
13 Allen-Jeffreys [1996]. The extent of the change can be seen if one compares the perception
of space and geography before and after Cosmas Indicopleustes Christian Topography
(ca. 547549): scientific commentaries on Ptolemy did not cease to exist and circulate,
but the focus of this discipline shifted to a religious dimension that did not belong to it
earlier: see Wolska-Conus [1968] 3643.
14 Cavallo [1978] and [1987]; Cavallo [2002a] 163165; Cameron [1985] 1932; a more optimis-
tic verdict on the circulation of pagan culture under Justinian is given by Kaldellis [2005];
but there is an ongoing debate between those who see a discontinuity in the Dark Ages
e.g. Cavallo [1995] and Kazhdan-Cutler [1982]and those who do notTreadgold [1990].
Scholarship in the Byzantine Empire ( 5291453 ) 301
deemed at least partly representative of the wider cultural trends of every sin-
gle age. The study of this happy combination of direct tradition and indirect
sources may be very effective in discrediting the obsolete but persistent view of
Byzantium as a substantially hostile or at best indifferent soil for the preserva-
tion of Classical texts.15
To be sure, the interest for ancient authors rested on a sensibility largely
at odds with our own, i.e. not on their moral, civic or political meaning, but
rather on formal issues connected with their stylistic and rhetorical excellence,
in a cultural frame in which mastering Greek, and particularly Attic Greek,16
was an indispensable skill for every high-ranking civil servant or member
of the ecclesiastical hierarchy.17 And it was chiefly this need for a linguistic
instruction, and for training in the production and adornment of rhetorically
elaborate texts such as the prooimia that introduced official documents,18 that
called for a certain continuity of the school curriculum. The latter was struc-
tured into a primary (hiera grammata: reading, writing and spelling, based
on Christian texts and elementary poets), a secondary (enkyklios paideia, the
necessary linguistic tools for the aspiring civil servants or ecclesiasticals, con-
sisting mainly in grammar [Dionysius Thrax, Theodosius Canones, Apollonius
Dyscolus and Choeroboscus Epimerisms on the Psalms], rhetoric [Aphthonius
and Hermogenes] and some logic [Porphyrys Eisagoge]), and a higher level
(more elaborate poetry such as Attic drama, Pindar and Theocritus, higher
mathematics and astronomy, rhetoric, philosophy).19 The third stage was nor-
mally accessible to a very restricted elite, which aspired to leading positions at
court or in the ecclesiastical hierarchy (in this framework, poetry had the role
of providing the rhetor with myths and quotations, as well as unusual gram-
matical rules and eccentric vocabulary). A paradigmatic case, because of the
high social rank and the comparatively modest scholarly stature of the man, is
represented by John Mesarites (11621207), whose long prose epitaph written
by his brother Nicholas indulges in a lengthy description of his schooldays and
15 Maas [19273] 70: Eine Wissenschaft von der klassischen Literatur hat es in Byzanz nicht
gegeben, a position substantially shared by Pasquali (see Tessier [2010b]), and resting
on a Western prejudice that has its deep roots in the prejudices of Petrarch and of other
humanists: see Bianconi [2008a] 453455, and Kaldellis [2009a] 12 on the danger of
Orientalism lurking behind this view.
16 Wilson [2007] 46. On Atticism in Byzantium, see e.g. Rollo [2008a]; Horrocks [2004];
Browning [1978b].
17 Mango [1965] 32; Fryde [2000] 810.
18 Hunger [1964].
19 Fuchs [1926] 4345; Hunger [1978] 1011; Constantinides [1982] 1; Efthymiadis [2005];
Markopoulos [2006] and [2008]; Browning [1964] 5.
302 Pontani
intellectual training, starting from reading and writing all the way up to his full
syllabus of rhetoric and philosophy.20
This didactic continuity, which relied partly on the persistence of educa-
tional institutions, but left considerable scope for the personal curiosity of
advanced scholars throughout several different epochs of the Byzantine mil-
lennium, should be regarded as an essential and by no means banal element
leading to the preservation and circulation of ancient texts in a Christian
milieu, in the very heart of a theocratic regime.21 Byzantine classical schol-
arship was, therefore, the study of an admired but foreign society, for the
Byzantines were the first culture to consume classical literature from such
a detached albeit respectful perspective:22 in this, they fulfilled perfectly the
prescriptions of the 4th-century Greek Fathers (Gregory of Nazianzus and
Basil of Caesarea, against the perspective of Julian the Apostate) who wished
to separate linguistic and rhetorical Hellenism from religious issues.23
If we keep in mind this general attitude, as well as the number (and impact)
of the personalities who made a decisive contribution to the transmission
and study of texts from the 6th through the 15th century, we will realise that
any reductive or disparaging verdict against the process and performances of
Byzantine scholarship is largely unfair.24
20 Flusin [2006].
21 Fabricius [1967] 187188.
22 Kaldellis [2009a] 7.
23 Agapitos [1998] 171174; Bowersock [1990] 113; Kennedy [1983] 180264; Chuvin [2012].
Simonetti [1983] 6996.
24 Kaldellis [2009a] 16.
25 At the outset of his Institutions, praef. 1, p. 3.710 and 2325 Mynors.
Scholarship in the Byzantine Empire ( 5291453 ) 303
26 See the summaries on the late years of these schools in Wilson [19962] 2860; Cavallo
[2002a] 6075; Chuvin [2012] 106122.
27 Lemerle [1971] 8285.
28 Westerink [1976] 23; Watts [2006] 257261.
29 Saffrey [1954]; Sorabji [1987].
30 Ed. Daly [1983].
31 For its problematic manuscript tradition and for its editions, see Consani [1991] 5559.
304 Pontani
.
,
, .
,
,
...
32 Ed. Dindorf [1825]. See Dyck [1993a] 777778; Dickey [2007] 8182.
33 See Probert in this volume.
34 Watts [2006] 232256 on the entire issue.
35 Westerink [1976].
36 Rouech [1974].
37 See Montana in this volume.
38 Elias, On Aristotles Categories (CAG 18.1), pp. 122123 Busse; see Wilson [19962] 47.
Scholarship in the Byzantine Empire ( 5291453 ) 305
to the desert of Negev. A recent census has shown that while the copying of
evangelic, liturgical or patristic books was by far predominant, and almost
exclusive from the late 7th century onwards, as far as pagan authors are con-
cerned, together with popular schooltexts such as Aristophanes, Demosthenes,
Homer and Euripides one finds a persistence of epic poetry (from Apollonius
Rhodius down to the more recent Nonnus of Panopolis), of some novelists
(Chariton, Heliodorus) and historians (Thucydides, Malalas), as well as of
various technical texts, from lexicographers to medical authorities such as
Hippocrates and Galen.44 Menander is no longer attested after the late 6th
century (PBerol 21199), and notoriously did not make his way into the medi-
eval transmission in minuscule script.45 To our eyes, perhaps the most surpris-
ing phenomenonand the most illuminating one concerning the interests of
some cultivated individuals in the late antique Egyptian chora46is the sur-
vival of papyrus codices carrying the poems of Sappho47 and Callimachus,48
authors whose existence in Byzantium was endangered by their early exclu-
sion from the educational curriculum.49
Virtually no clear sign of the practice of textual scholarship is known to us
from the so-called Dark Ages, but this does not mean that philological studies
had disappeared altogether. Anastasius of Sinai, a Cypriot monk who settled in
the monastery of St. Catherine around 660, composed a work called Hodegos
(Guide), concerned with a series of moral, dogmatic and theological issues:
in his text he also criticised a governor of Alexandria named Severianus for
sponsoring a scriptorium of 14 copyists in order to deliberately alter and falsify
the text of the Church Fathers, and especially of St. Cyril:50
,
,
<>
[Flor. Cyrill.
109.22 Hespel],
,
. ,
.
[Ambros. De
fide 2.9.77], .
,
, ,
, [Procl. Const. Or. 2, p. 104.46 Leroy],
- , ,
( ),
, .
For several years he had on his side fourteen calligraphers, who sat under
his supervision and falsified the dogmatic books of the Fathers, above
all those of St. Cyril: when we arrived at the sentence against Succensus
saying that We say that the natures are two, we did not find it soundly
copied in any book in Alexandria, but some copies carried We say that
two natures are united, others We say that the natures are understood
as two. And since we were in great distress, Isidore the librarian of the
Patriarchate took out for us a book that had the aforementioned sentence
without alterations. They corrupted in the same way the saying of the
holy Ambrosius: instead of Let us observe the difference between divin-
ity and the flesh, they said Let us observe the difference of readings.
And again, in the saying of holy Proclus about Christ, that he was born
incorruptibly, he who entered without obstacles when the doors were
shut, he whose combination of natures Thomas saw, the Gaianites read
incorruptible instead of incorruptibly with long -os. And instead of
acknowledging the relative pronoun hou, whose duplicity of nature
(namely of Christ), they take that ou as a negative, reading Thomas did
not see the combination of natures etc.
Hierosol. Taphou 36.56 And even if the Palestinian origin of these and other
textual witnesses remains to a large extent hypothetical, our sources give us
evidence of the Palestinian training of several outstanding figures of Greek lit-
eracy between the late 6th and 8th century.
Setting aside for the sake of brevity great hymnographers such as Romanos
the Melode (the author of the cruel pun against Aratus thrice
damned and Demosthenes feeble and Homer empty
dream in cant. 33.17.36 Maas-Trypanis),57 and the great homiletes such as
Theophanes and Theodorus Grapti (who studied under Michael Syncellus at
the monastery of St. Saba), one can assume a certain continuity in the trans-
mission of Greek learning from the remarkable Damascene rhetor, scholar and
Anacreontic poet Sophronius (ca. 550ca. 640), who travelled to Egypt with
the monk and ascete John Moschos and later became patriarch of Jerusalem,
down to the outstanding theologian Maximus Confessor (ca. 580662,
although other sources relate he was born in Constantinople) and to John the
Damascene (ca. 675749), the greatest Greek-speaking scholar of the 8th cen-
tury, whose works such as the theological summa called Pege gnoseos or the
florilege called Sacra parallela presuppose a wide availability of books58and
we will skip here altogether, precisely after mentioning John the Damascene
and Maximus Confessor, the thorny issue of the forms and origin of the Greek
gnomological tradition, of which the Dark Ages marked an important turning-
point in its very complicated history, made of fluctuating, non-authorial mate-
rials and desperately uncertain dates.59
The percentage of Classical Greek doctrine and quotations in the manu-
script evidence and in the literary activity of the aforementioned authors is
admittedly thin, and yet Palestine must have been a fertile soil for grammatical
instruction, for it gave birth to one of the most outstanding grammarians of
the 9th century, namely Michael Syncellus from Jerusalem (ca. 761846), the
author of a fundamental treatise On Syntax composed in 812813 in Edessa
(a city where translations from the Greek and the interest in Greek culture
56 Cavallo [2002a] 196197; Crisci [2000]; andwith a special emphasis on Christian texts in
the middle Byzantine periodPerria [2003].
57 A pun Maas [1906] 21 related specifically to the closing of the Athenian academy in 529;
but this is unlikely, see Speck [1986] 617.
58 See on these authors, and on the broader picture of Greek learning in Palestine, Cavallo
[2002a] 165166 and 198202.
59 Searby [2007] 5059. Odorico [2004]. Ihm [2001] ixxix, esp. iii n. 14.
310 Pontani
,
, .
,
.
, [ 52324] /
,
,
.
, , ,
, , ,
.
66 Perria [19831984].
67 Irigoin [1950].
312 Pontani
was definitely not the venue of the most significant scholarly or scholastic
activities.68
Philosophy and History in praise of the patriarch for his activity of restoring
studies and culture.73 It is namely unclear whether Theophylactus gained this
knowledge and skill, also apparent from his private correspondence, through
his own, personal study, or thanks to some sort of curricular instruction. Later
in the same century, with the remarkable exception of George of Pisidia and
of his flawless iambic trimeters on the campaigns of Heraclius, there surfaces
hardly any trace of familiarity with ancient models among high-brow literati.
In fact, no profane work in Greek survives from the age of Theophylactus
History down to the Historical epitome written by patriarch Nicephorus shortly
before 787.
An important symptom of the discontinuity in this context is the definitive
linguistic separation between Latin and Greek: after the age of Priscian (the
author of the most ambitious extant work on Latin grammar, the Institutiones
grammaticae, completed in the early years of the 6th century), Latin slowly
disappeared from both official use74 and from cultivated communication in
Byzantium: the Latin panegyrics written by Corippus in 566567, and later
the entire age of Emperor Heraclius (610641), appear as the swan song of a
bilingual literate society.75 In the Latin West, the decades following the great
season of Graecomania de Macrobe Cassiodore,76 witness a parallel obliv-
ion of Greek, a phenomenon which resulted inter alia in the progressive dis-
appearance of Greek manuscripts from the regions of Western Europe, with
some exceptionsof uncertain sizein the areas of Sicily and Rome, where
some translations of technical and ecclesiastical texts were undertaken.77
Cassiodorus library at Vivarium lists mathematical, geometrical and medical
texts in cupboard 8 (ubi sunt Graeci codices congregati) and some transla-
tions may have taken place in his coenobium,78 but neither this reference nor
the rare and mostly uncertain traces of the production of profane Greek books
in Ravenna or Byzantine Italy,79 nor other traces of the likely Italian circulation
73 Olajos [2000].
74 John the Lydian On magistracies 3.68 tells about this gradual oblivion of Latin in official
documents of the Empire: , the process of reduction began
to advance.
75 Jeffreys [2011]; Rochette [1997] 141144. On a specific case, see De Stefani [2006].
76 Courcelle [19482].
77 Berschin [1980] 97108; Cavallo [2001]; Chiesa [2002].
78 Berschin [1980] 100102.
79 E.g. ms. Neap. Gr. 1 of Dioscorides (7th century), but ms. Par. suppl. Gr. 1362 of Aristotles
Sophistical Confutations and ms. Bruxell. IV.459 of Paul of Aegina also belong to this age.
314 Pontani
80 The scattered palimpsest leaves of Theophrastus and Strabo in Vat. Gr. 2306 (+Vat. Gr.
2061A+Crypt. Z..43) were perhaps rewritten in Southern Italy, and some folios of Cassius
Dio were re-used in a Calabrian menologion (Vat. Gr. 1288): see Irigoin [2006a].
81 Irigoin [1969] 4345; Cavallo [2002a] 2035; Berschin [1980] 113118.
82 Lemerle [1971] 1021.
83 Dionisotti [1982]; Dickey [2012].
84 Cameron [1970]; Mattsson [1942]; Valerio [2014].
85 Kaldellis [2005] esp. 385389, see also Dickey in this volume.
86 Rare passages are quoted ad verbum and in full by Constantine VII Porphyrogenitus
in his De administrando imperio: see Billerbeck [2006] *3*9. The letters not yet included in
Billerbecks edition (Billerbeck [2006] and [2014] and Billerbeck-Zubler [2011]) must be
consulted in Meineke [1849].
Scholarship in the Byzantine Empire ( 5291453 ) 315
, , .
(1.55.1). (fr. 1009 Radt).
. (fr. 916 Radt)
. ,
,
.
Above and beyond Stephanus work, lexicography and grammar were cer-
tainly practiced in Constantinople between the 6th and the 8th century, even
if the extant evidence is meagre. As a lexicon, Stephanus Ethnika belongs to a
well-represented category in the centuries ranging from the 6th to 8th. The so-
called lexicon of Cyril, whose origin probably belongs to the age of Justinian,
underwent many different redactions and reshapings, and various minor lex-
ica were added to its original bulk;88 more importantly, before the end of the
9th century it was interpolated into the lexicon of Hesychius, thus giving rise
to one of the favourite lexica of the Byzantine period, alas still unedited to our
own day.89 More specific grammatical lexica on spirits90 and on accents91 may
belong to the later part of the 8th or early 9th century.
87 Steph. Byz. 305 Billerbeck. The occurrence of the adj. (non punctuated, hence
not corrected) should be noted.
88 Alpers [2001] 202. Drachmann [1936]; Cunningham [2003] 4349; Hagedorn [2005] vxiv
(with a preliminary edition of one of the versions); Hunger [1978] 3738.
89 Reitzenstein [1888]; Alpers [1991b].
90 , ed. Valckenaer [1739] 207242.
91 Ed. Koster [1932].
316 Pontani
In the field of grammar proper, there was little theoretical advance, but the
systematisation of the inherited doctrine continued e.g. in the works of John
Charax, a scholar who may belong to the late 6th century.92 We may recall
here his treatise On enclitical words (a compilation from Herodians General
Prosody book 21), his commentary on Theodosius of Alexandrias Canons, later
abridged by Sophronius patriarch of Alexandria in the 9th century,93 and a still
unpublished Orthography.94 The field of orthographical studies was covered
by several other specialist handbooks, partly unedited and partly preserved in
shorter rsums dealing with the distinction of homophone vowels and diph-
thongs, such as / / / / , o / , / .95
The mechanisms of textual transmission do not appear to be especially pop-
ular among writers of the Dark Ages. We hardly ever hear, either in literature
or in subscriptions to manuscripts of Classical authors, of a systematic emen-
datory work performed by learned individuals on single texts. Some attention
is devoted to this issue by John the Lydian, the author of three antiquarian
books On magistracies, On months, On omens, in his account of the history of
the oracles of the Chaldaean Sibyl:96
,
,
.
That her lines are found to be incomplete and unmetrical, does not
depend on the prophetess but on the scribes, who did not keep up with
the speed of the dictation, and were ignorant and unexperienced in
grammar.
92 Kaster [1988] 391392 argues for the 8th9th century, but see Alpers [2004] 19.
93 Mazal [2001] 477.
94 Alpers [2004] 78 and 19.
95 See Hunger [1978] 1819 for all the references, and above all Egenolff [1888]; Alpers
[2004]. The popularity of orthographic questions is attested by their frequent appearance
in manuscript notes and short treatises down to the 15th century: Hunger [1978] 2122;
Ronconi [2012a] 7280 for the general context of grammatical and orthographical teach-
ing on Dionysius Thrax and other texts.
96 On months 4.47 Wnsch.
Scholarship in the Byzantine Empire ( 5291453 ) 317
,
,
<>,
,
,
,
.
The most pious emperor together with the most illustrious magistrates
and some god-beloved bishops of the holy Synod, inquired, opened, and
examined the book, and found out that three quires had been added at the
beginning, which did not bear the numbering in calce usually apposed on
quires: the first quire-number appeared on the fourth quire, and then on
each following quire. Furthermore, the handwriting of the three quires
interpolated at the beginning, which carried the so-called letter of Menas
to Vigilius, was different from the one originally employed in the book
under review.
(scil. ) , , ,
, . ,
, ,
,
<>. ,
,
, . ,
.
100 Lamberz [2000]; Speyer [1971] 199 and 277; Lemerle [1971] 109112.
101 Long believed to belong to the age of Justinian, his chronology has been partly fixed only
recently: Theodoridis [1980] 341345; Alpers [2004] 19.
102 Ed. Consbruch [1906] 175254; see Hunger [1978] 5051.
103 Ed. Hilgard [1901] 67106.
104 Ed. Spengel [1853] I, 244256; see Besharov [1956].
105 Ed. Cramer, An. Ox. II [1835] 167281; see Valente [2010a] 639650. Alpers [2004] 3135.
106 Valente [2010a] 644645.
Scholarship in the Byzantine Empire ( 5291453 ) 319
say ammes pronouncing the epsilon present in the word. With etymology,
when we spell epeiros (land) with eta in the first syllable, and with the
diphthong epsilon-iota in the second and I say because it has no limits,
being aperos. With history, when I spell chilioi (a thousand) with iota
and I say the paradosis has it spelt this way. [transl. S. Valente]
But Choeroboscus wrote above all two remarkable works, both honoured by
prolonged success in Byzantine grammatical practice,107 and both clearly
intended for the instruction of pupils in need of elementary training in
accentuation, punctuation, prosody and grammar. The first one, indebted to
Apollonius Dyscolus, Herodian and of course Dionysius Thrax, is a long com-
mentary to the Canons of Theodosius of Alexandria,108 chiefly concerned
with the establishment and observation of a clean linguistic standard, and
of rules designed to prevent barbarism and solecism.109 The other work is
Epimerisms on the Psalms,110 a pedantic exercise of word-for-word parsing
of the Septuagints text with the aim of singling out the morphological and
grammatical category of each term, independently of its syntactic function
but with an eye to its semantics.111 A nice, if very synthetic example of how
Choeroboscus interweaves different sources in his grammatical work is his
note on Ps. 75.7:112
, .
, .
, .
, [Y 162],
[Cosm. Hieros. canon pro magna quinta feria,
PG 98.480B].
enystaxan: verb, first aorist, third person plural. Nystazo derives from
neustazo, itself deriving from neuo. What is the difference between to
move (kinein) and to shake (neustazein)? The words are different, but
the meaning is identical, for kinein and neustazein mean the same thing,
as in Homer shaking his mighty helmet, and Judas shaking his head.
he himself states in the biography of his teacher Tarasius,119 but also to use
them in his own poetical activity,120 and to show a special sensitivity for Greek
tragedy in his trimeters on Biblical subjects and in his paraphrases of Babrius
fables.121 All this does not necessarily configure him as a first-rate Classical
scholar, but his claim to a high rank in the realm of poetry and to the merit
of rescuing from oblivion the art of grammar may be at least partly justified.122
The first half of the 9th century is also the common dating of the book on
orthography dedicated by a grammarian named Theognostus to emperor Leo,
traditionally taken to be Leo V (813820), although a recent study has made
a case for Leo VI (886912), thus shifting the chronology to the middle of the
century.123 The work itself,124 with its 1006 kanones dealing first with the begin-
ning and body of words (1142) and then more broadly with their endings
(1431006), has the lexicon of Cyril as one of its main sources, and it declaredly
follows the ordering of Herodians General prosody, while emending and restor-
ing some of its items. This Orthography thus offers a comprehensive study of
Greek grammar going far beyond the mere spelling of words,125 which is why it
was chosen, together with Choeroboscus Orthography, a treatise On quantity
( ), and a series of alphabetical epimerisms, as the source of a
grammatical corpus used by the redactor of the Etymologicum Genuinum, and
copied tel quel in the 10th-century manuscript Bodl. Barocc. 50.126
If we turn for a moment to rhetorical studies, their continuity in Byzantium
is apparent (in the so-called Dark Ages, writers such as Paul of Aegina and
Theophylactus Simocatta display in their works a remarkable familiarity with
119 See Lemerle [1971] 128129 and Mango [1997] 8 for the relevant passage of the Life of
Tarasius (p. 69.710 Efthymiadis)
The former [scil. education] I enjoyed in the prime of my youth when I was initiated by
you in the best examples of the trimeter and the tetrameter, both trochaic and anapaestic,
and in dactylic verse (transl. C. Mango).
120 Lauxtermann [2004] 314318; Speck [2003] n. xii.
121 Lampakis [2001]; Browning [1968].
122 Greek Anthology 15.39.23 This is the work of Ignatius, who brought to light the art / of
grammar, buried in the depths of oblivion ( , /
).
123 Antonopoulou [2010].
124 Ed. Cramer, An. Ox. II [1835] 1165, but see above all the partial edition by Alpers [1964].
125 See Alpers [2004] 2931, and particularly Theognostus dedicatory epistle (Alpers [1964]
p. 69.410).
126 On the corpus, see Reitzenstein [1897] 19293 and Alpers [1964] 2324. On the glorious
Bodl. Barocc. 50, known to Bentley, Villoison, and Ritschl, see Ronconi [2007] 91131.
322 Pontani
the principles of artistic prose), though often hard to trace in detail due to the
lack of primary witnesses and the physiological, continuous replacement of
handbooks and treatises. As a matter of fact, the first scientific writing on this
subject after Late Antiquity is a commentary on Aphthonius Progymnasmata
by John the archbishop of Sardis (first half of the 9th century),127 Aphthonius
being the major schoolbook of rhetorical training throughout the Byzantine
era.128 Based on a series of earlier treatises (from Davids commentary on
Porphyry down to Sopaters own progymnasmata), this work not only attests
to the continuity of rhetorical teaching throughout the Dark Ages and early
Macedonian period, but it was also to provide fruitful reading for rhetors of
the later Byzantine age.129 John of Sardis definition of myth, for instance, may
depend on Sopater, but it also undoubtedly displays a certain originality:130
, ,
, ,
,
( 197) .
,
,
.
,
,
.
,
,
.
We say that there are two kinds of myth: one is allegorical, the other
political. Allegorical myth is when its outer appearance is different from
127 Ed. Rabe [1928], see xvixix for the context of this work.
128 On Byzantine rhetorical instruction in general, and the role of Aphthonius in particular,
see Hunger [1978] 75120; Kustas [1972] 526; Conley [1986]; Constantinides [2003]; Kraus
[2013]; Valiavitcharska [2013].
129 See most recently Pontani [2014].
130 pp. 10.311.3 Rabe.
Scholarship in the Byzantine Empire ( 5291453 ) 323
the message it conveys, like the myths fashioned by Homer, she seized
the Peliad by his blond hair (for in myth, Athena means the mind). The
myths of the poets are mostly of this nature, wherefore they are useless
for rhetoric, and because their falsehood is substantial they are believed
to be myths: allegorical myths are thus alien to rhetors. Political myths
are subdivided into invented and historical ones: invented myths give
a clear sense of their invention, like the myth of the old lion pretend-
ing to be ill or that of the horse and the tortoise: that these and similar
myths are invented is easy to perceive. Historical myths are those that
seem to have been found by searching and then seen in their develop-
ment, and have their declared share of falsehood, but which, on account
of the nature of their subject, resemble sculpture, like the present myth,
and that of the dog seizing the meat, or of the bird-hunter deceived by
the verse of the cicada: such myths, so to say, procure superficially an
expectation of truth, while concealing the falsehood they contain.
1.4 Iconoclasm
The most important cultural phenomenon in the period under review is
doubtless the battle against the holy images known as iconoclasm, a theologi-
cal and ideological movement stretching from the late 7th century well into
the 9th.131 Due to the lack of reliable sources, different views have been taken
concerning the impact of iconoclasm on the transmission of books and par-
ticularly of Classical culture:132 we now incline to believe that this impact must
have been rather modest, for while the iconoclasts were ready to destroy illu-
minated Bibles and patristic books, the iconodules, in their orthodoxy, showed
an even greater hostility towards paganism and ancient culture. We also have
to reckon with the consequences of propaganda: major episodes such as the
action of setting fire to the oikoumenikon didaskaleion, allegedly perpetrated
by the initiator of iconoclasm Leo III Isaurikos in 726, are probably the fruit of
iconophile propaganda rather than of historical truth.133
In fact, important grammarians and experts in ancient Greek language and
literature, such as the aforementioned Theognostus (perhaps a Sicilian)
and Choeroboscus, were trained under iconoclastic emperors, and it seems
that this did not adversely affect their instruction; the same can be said
for Theodore Stoudites (759826) and the organisation of the Stoudios
, ,
.
,
,
.
Theophilus, the last iconoclastic emperor, had been trained in his youth by
John the Grammarian (born ca. 770), who later became a patriarch and suf-
fered confinement upon Theophilus death in 842. John had been one of the
most highly regarded scholars of the first half of the 9th century,137 and in 814
he had promoted on behalf of Leo V a famous hunt for ancient manuscripts
in churches and monasteries,138 aiming at the discovery of ecclesiastical and
patristic texts rather than of pagan literature.139 A bibliophile and a reader of
pagan authors, John was criticised by the learned monk Theodore Stoudites for
being a new Pythagoras or a new Plato, a charge that is partly understandable
in the light of his interests in pagan philosophy, magic, and poetry.140
It is plausible, albeit impossible to demonstrate, that he may have pro-
moted an ambitious lexicographical compilation, largely derived from the so-
called lexicon of Cyril, other sources including the D-scholia to Homer, the
glosses of Apollonius Sophistas Homeric lexicon, and Atticist glosses to his-
torians (Thucydides, Xenophon, Arrian, Dio Cassius), orators (Demosthenes,
Aeschines etc.) and philosophers (Plato, Aristotle). This dictionary, which
proved highly influential in the coming centuries for the lexica of Photius and
Suidas, is known as the Synagoge lexeon chresimon and was transmitted above
all by the 10th-century ms. Par. Coisl. 347, andin an enlarged version that
took onboard further Atticist glosses from Harpocration, Phrynichus, Aelius
Dionysius, Pausanias, Orusin ms. Par. Coisl. 345.141
1.5 Transliteration
The name of John the Grammarian has sometimes been connected, though
without any firm basis, with one of the most important developments in the
history of Greek texts, which occurred between the last decades of the 8th
century and the first decades of the 9th: namely, transliteration. The technical
evolution of Greek handwriting from majuscule into minuscule is of uncer-
tain origin and debated chronology, but it certainly affected in a decisive man-
ner not only the layout of the individual codices, but above all the chances
of survival of ancient Greek texts.142 Although a recent, provocative study has
attempted to trace the genealogy of minuscule script back to the standard of
Byzantine notaries from as early as the 6th century,143 a scientifically more
acceptable view identifies the origin of the minuscule in 7th- and 8th-century
acts and documents of the imperial and patriarchal chanceries.144
The calligraphic evolution of this minuscule as a book script145 has often
been connected with the monastery of Stoudios in Constantinople, apparently
the place of origin of the first dated minuscule codex (the famous Uspenskij
Gospel Book of 835, ms. Petrop. GPB gr. 219), as well as of other, early books in
the new handwriting (Leid. BPG 78; Laur. 28.18).146 The astronomical content
of the latter books reminds one of the well-known ms. Vat. Gr. 1291 of Ptolemys
Handy Tables, a richly illuminated codex which has been dated to the late
8th or more realistically shortly after 811, but is still written in uncial letters.147
Now, the proof of a pivotal role of Stoudios in the transliteration movement is
admittedly thin, but we do know that in the rules of the monastery in question
the work of scribes was very strictly organised, with fines and punishments
for those who committed mistakes in matters of punctuation or accents, for
those who broke their pen out of rage or irritation, for those who followed their
dicte intrieure rather than the manuscripts text, and for those who dared
interpolate something in the text they were copying.148
The advantages of the new script, smaller and faster than the majuscule, are
evident in terms of space-saving and practicality, and particularly, due to the
more systematic presence of reading aids such as accents, spirits and punc-
tuation marks, in terms of reader-friendliness: this explains its comparatively
fast spread throughout the empire, with the majuscule remaining confined
to ecclesiastical or liturgical books, such as the famous book of Ps.-Dionysius
the Areopagite (now Par. Gr. 437) offered by the Byzantine ambassadors
of emperor Michael II to king Louis le Pieux in Compigne in 827, and there-
upon exhibited to public veneration in the abbey of St. Denis. This official
144 A unique example of this evolution is PVindob G 3, whose handwriting reproduces the
writing standard of 681: De Gregorio-Kresten [2009] 339344.
145 See esp. De Gregorio [2000a]; Mazzucchi [1991].
146 Fonki [19801982]; Perria [1997].
147 Janz [2003] vs. Brubaker-Haldon [2011] 3740.
148 Monastic penalties 5360 (PG 99.1740) see esp. 55;
, if someone learns by heart what is written in
the book from which he is copying, let him be relegated for three days; and 56:
, if someone reads more than
what is written in the book from which he is copying, let him eat dry food. These rules
probably do not stem from Theodore himself but certainly reflect his views: Leroy [1958]
210212. See in general Lemerle [1971] 121128; Cholij [2002] esp. 3133; Eleopoulos [1967].
Scholarship in the Byzantine Empire ( 5291453 ) 327
ecclesiastical use may be one of the reasons why this script was adopted by
Cyril and Methodius, the apostles of the Slavs, when shaping a new alphabet.149
On the other hand, the once fashionable idea that most Greek texts must
have undergone no more than one transliteration in the course of their his-
tory, and that every stemma codicum should eventually lead to a single 9th-
century archetype of the entire paradosis, is no more than a scholarly myth
with little firm basis in either historical evidence or common sense. However
complicated the transliteration process may have been, due to the plurality of
libraries and of Greek-speaking cultural centres in the entire Mediterranean
area, copies of texts in majuscule script certainly survived for a long time, and
the switching from one handwriting to the other must have been a more plural,
multifaceted, and at times also slow and belated process.150
2.1 General
Several recent studies have warned against considering Byzantine culture
along the parameters of renaissances (or indeed humanisms) and subse-
quent dark ages: not only do these terms anachronistically refer by way of an
ambiguous comparison to a unique experiencethe Italian Quattrocento
belonging to a totally different time and place, but above all they obfuscate the
substantial, if uneven, continuity of Classical instruction in the educational
process of all Byzantine elites.151 Ancient Greek authors may have experienced
various degrees of popularity among the members of the educated class,
but they never disappeared altogether, nor did they cease to be, if in varying
degrees, a primary component of the Greek identity. Therefore, when we talk
about revivals we reject all ideas of a new beginning, and simply adopt this
terminology in order to highlight historical moments in which a higher num-
ber of scholars devoted their efforts to the production and study of books and
texts, often in the wake of a special increase in educational policies and cul-
tural investments.152
In particular, the Byzantine Renaissance of the 9th century has been author-
itatively described as the moment when the Romans of the East started to
peruse and exploit the ancient Greek heritage in the view of their own iden-
titarian needs.153 This, however, does not necessarily imply that we believe in
the existence or the creation of an imperial academy or a patriarchal school,
two institutions whose role has too often been taken for granted on a very
slight basis: the latter is not positively attested until the early 12th century;154
as for the former, we only know that some time in the second quarter of the
century emperor Theophilus granted a room for teaching at the church of the
Forty Martyrs to Leo the Philosopher (or the Mathematician, ca. 790post
869), the cousin of John the Grammarian.155 Leo, who was later promoted to
the archbishopric of Thessalonica, had apparently studied at Andros in his
youth,156 and before being appointed by the emperor he taught privately in
Constantinople: thanks to his deep knowledge of astronomy and astrology he
won the admiration of the Arab caliph Al-Mamun, who apparently invited him
to come and teach in Baghdad.157 A much-debated passage of Theophanes
Continuatus relates that some time between 843 and 855856 the kaisar
Bardasa high-ranking civil servant, who dominated the political scene of
Byzantium around the middle of the centurycreated a new school, directed
by Leo himself, in which the , the pagan learning, finally revived after
decades of silence:158
(
)
.
Then he [scil. Bardas] took charge of pagan learning, which after such
a long time had fallen in decay and boiled down almost to nothing due
to the ignorance and boorishness of the people in power: establishing a
school of science close to the Magnaura, he made ambitious efforts in
order to reinvigorate that learning and make it flourish again.
,
,
<> .
.
159 These are by and large the opposing views of Speck [1974a] and Lemerle [1971]. A fresh
look on these topics will be provided by the proceedings of the conference la suite de
Paul Lemerle, organised by J.-C. Cheynet and B. Flusin in Paris in October 2013.
160 Theoph. Cont. 4.29, p. 192.1620 Bekker [scil. ]
,
, ,
.
161 Greek Anthology 15.38.
162 Lemerle [1971] 166167.
163 Cortassa [1997a].
164 See the different opinions in Ronconi [2003] 5659.
330 Pontani
In any case, it is hard to think that this sort of textual scholarship should have
remained without some kind of relationship with the new philological facies
of the Iliad and its scholia as we perceive it from the archetype a, the direct
ancestor of the glorious ms. Venetus A (Marc. Gr. 454).165
Whatever the implications of Cometas activity, two issues remain open: first
of all, behind the four subjects taught at the Magnaura we can hardly discern
the shape of a consistent curriculum;166 secondly, there is no evidence that this
school should be regarded as having been a public institution rather than an
elitarian system of education with no legal profile and no official recognition.167
The director, Leo the Philosopher, whether or not he had among his pupils
Constantine/Cyril the apostle of the Slavs,168 is known for a high-brow schol-
arly activity that is difficult to reconcile with his teaching at the Magnaura, for
he devoted his efforts to various authors of philosophy, e.g. emending Platos
Laws up to 5.743b (as we learn from a note preserved in three Platonic manu-
scripts, among which the venerable Vat. Gr. 1 and Par. Gr. 1807, once attributed
to Arethas patronage)169 and writing an epigram on Aristotles Categories.170
Epigrams in the Greek Anthology also inform us that copies of such dispa-
rate and demanding authors as Apollonius of Perga, Proclus, Theon, Euclid,
Archimedes, Ptolemy all featured in his library: he might have indeed owned
or made a copy of Vat. Gr. 1594 of the Almagest, and he certainly commissioned
the earliest Byzantine recension of Archimedes, as we gather from colophons
in mss. Laur. 28.4 and Par. Gr. 2360;171 his studies of profane authors, which
included even a novel such as Leucippe and Clitophon (Greek Anthology 9.203),
earned him the lively and stern reproach of Constantine the Sicilian172 in a
famous epigram condemning his proclivity to pagan deities, philosophers and
attitudes.
Leo, also nicknamed Hellen,173 represents a good example of a wider trend
in Byzantine culture and scholarship, one we will encounter often from now
on, and more so in later periods of enhanced scholarly activity: rather than
(and sometimes along with) public teaching or institutionalised curricula, the
element that fuelled the recovery and the study of classical authors through-
out the Middle Ages was the constant activity of individual members of the
cultivated elite who had received secondary education, a very restricted group
often assessed at no more than 300 people altogether174as opposed to a situ-
ation of relatively wide literacy at an elementary level.175 These people often
gathered in groups of two, three, or more, and joined their efforts in reading,
discussing, interpreting ancient pagan or christian texts, in a sense continuing
the late antique practice that had come to an abrupt end in the Latin West.176
It is thus thanks to these literary salons and coteries, to these learned gath-
erings of educated civil servants, ministers, officers, metropolitans, priests,
teachers and amateurs, that many ancient texts carved their way out of the
Dark Ages.177 This said, we should never forget that even the scholars who
were most keen on the study of pagan books exercised themselved even more
often on Christian texts, namely those which displayed a more immediate and
evident opheleia (usefulness) for everyday life and ethics:178 this is confirmed,
inter alia, by the sheer number of pagan authors to be found in book-lists of the
Byzantine period.179
2.2 Photius
One of the members of the cultivated elite, and doubtless one of the most
remarkable men of learning in the entire Byzantine age, is Photius (ca.
810893): perhaps a pupil of Leo the Philosopher, perhaps himself a private
teacher for a short while,180 he was a high-ranking civil servant, first as a
protasekretis,181 later switching to the ecclesiastical career and becoming patri-
arch of Constantinople (he held this post twice, in 858867 and in 878886),
and the man who tried to bring about the radical schism with the Roman
Church by attacking and refuting the authority of the Pope (863867). Some
modern critics have regarded Photius as the first oustanding representative of
174 This was the estimate of Lemerle [1971] 257; Treadgold [1984] 81; but see Markopoulos
[2006] 8687.
175 M. Jeffreys [2008]; Cavallo [2007a]; Mango [1975] 45.
176 Cavallo [2003]; Bianconi [2008a].
177 Cavallo [2002b] 432440 and [2010].
178 Cavallo [2002b] 441; Maltese [2003].
179 Bompaire [1979]; Wilson [1980] 285294 and 300303.
180 Fuchs [1926] 21.
181 Lemerle [1971] 183185.
332 Pontani
,
(
),
, , ...
182 On this ideal, see Podskalsky [2003]; Fryde [2000], 1113; Bossina [2003].
183 Different views are held on this e.g. by Wilson [19962] and Alpers [1988].
184 Theodoridis I [1982] lxxiilxxvi.
185 Tosi [2007].
186 Matthaios [2010a] 193197.
187 Amphilochia 21.13236 Westerink.
Scholarship in the Byzantine Empire ( 5291453 ) 333
One would compile a long book, not only if one wished to embrace all the
polysemic words (a laborious task, next to impossible), but even if one
wished to collect in one place the most common of them, those surfac-
ing more often in literature: precisely this I did, as you know, when I was
quitting the age of childhood.
188 In the list of manuscripts provided by Eleuteri [2000], just three antedate the 15th century.
189 Lemerle [1971] 189196; Ziegler [1941]; Wilson [19962] 93114; Wilson [2002].
190 Markopoulos [2004] no. xii.
191 Canfora [1998].
192 Another hypothesis has it that a special lot of books arrived from Alexandria, which would
perhaps explain the prominence of Egyptian authors, and the very form of the intro-
ductory letter, modelled on Aristeas narration about the translation of the Septuagint:
Canfora [1995] 3858.
193 Ronconi [2012a]. See on the topic Canfora [1995] 3043.
334 Pontani
194 Treadgold [1984] 96; Schamp [2011] insists on his specific paedagogical project.
195 See e.g. Jeffreys [1979]; A. Pontani [1995a] 339341.
196 With the exception of the folios in the Archimedes palimpsest: see Easterling [2008].
Scholarship in the Byzantine Empire ( 5291453 ) 335
Let us now turn to the contents: while most of Photius chapters contain
some biographical data on the author,197 some kind of rsum of the work at
issue, and some stylistic observations, three different typologies can be dis-
cerned for the pagan works:198 the short notices (Kurzreferate), the analyti-
cal notices (with in-depth analysis of the work) and the excerpts (with entire
passages from the book at issuethese instances represent an invaluable
source for us in the modern age, when the work is lost or poorly attested, e.g.
in the case of Agatharchides On the Erythraean Sea, or of Conons stories, or
of Proclus Chrestomathy). As mentioned above, stylistic observations creep in
more or less overtly, and they relate to both ancient and new categories:199 this
explains for instance the favour accorded to Arrian (cod. 9193) as the princeps
historicorum and to Aelius Aristides (cod. 24648) as the princeps oratorum,
as well as the praise bestowed on Plutarch, Philostratus and Damascius; in
some cases Photius seems to posit an interrelation between the stylistic quali-
ties and the ethical value of the single authors.200 On the whole, the constant
interaction between profane and Christian learning has a bearing on Photius
skilful technique of abridging and evaluating the books he is talking about.201
As a representative case, we choose to present a large part of codex 164 on
Galens On sects, bearing in mind that Galen is one of several medical writers
Photius proves familiar with (Oribasius, Aetius, Paul of Aegina etc.).202
.
,
, ,
, ,
, ...
, ,
.
. ,
, .
,
, ,
.
I have read Galens On sects. It is about the sects that have originated
in the art of medicine. He says that three great medical sects have seen
the light: the so-called logical one (which he also names dogmatic and
analogistic), the so-called empirical one (which is also defined obser-
vationist and recording), and thirdly the methodical one...
This book must clearly stand before all other readings in the domain of
medicine, if one needs to make out which sect is the best, and then follow
it. The book itself might not seem to be a book of medicine stricto sensu,
but it plays the role of a prologue, and belongs rather to philosophy. It is
clearly pure and neat as far as style and syntax are concerned, for Galen
is always careful in these matters, even if in several of his treatises, by
loading the text with infelicitous additions, digressions, and prolix phras-
ings, he blurs and obscures the meaning of what he writes, interrupting
the context, so to speak, and pushing the reader towards indifference
through his verbosity: but the present book is free from these faults.
203 See cod. 1 on Dionysius the Areopagite; cod. 230 and 274; in cod. 201 and 219 Photius him-
self puts forth the issue.
204 In cod. 88 a book on the history of the Council of Nicaea is presented as follows: The name
of the author is not given in the title; but in another manuscript of the same text I found
the work attributed to Gelasios, bishop of Caesarea in Palestine
. ,
(Bibl. 66b 3033). In cod. 77 Photius refers to
a double edition of Eunapius which he has found . See also cod. 98
and 111.
Scholarship in the Byzantine Empire ( 5291453 ) 337
,
, ,
,
.
, ,
,
...
In the neighbouring field of grammar, it can be added that around the mid-
dle of the 9th century, at the request of the bishop of Damietta, Sophronius
the patriarch of Alexandria (848860) drew up a highly abridged version of
John Charaxs commentary on Theodosius Canons (on which see above 1.3)
trying to obtain the useful brevity but renouncing a longer discourse, as you
thought would be suitable, not by ignorance of the rhetorical means, but rather
in the attempt to appear friendly towards beginners;215 Sophronius must
be the same man who also produced a paraphrasis of the Iliad, known to us
in the folios of ms. Sinai 26.216 The special favour enjoyed by grammatical
and linguistic studies throughout this period is proved by the wealth of extant
manuscripts dating to the later part of the 9th century: I am particularly refer-
ring to the D-scholia to Homers Iliad (Matr. 4626 + Rom. Bibl. Naz. Gr. 6),217 to
the lexica preserved in mss. Par. Coisl. 347 (see above 1.3 about the Synagoge),
and to the Leipzig palimpsest containing an uncial copy of Oros commentary
on Herodians General Prosody (Lips. Gr. 2)218Herodians work itself is noto-
riously lost, and only a minuscule copy dated ca. 900 is fragmentarily preserved
in ms. Vind. hist. Gr. 10.219 It is thus no wonder if around 895 even a learned
amateur such as the ambassador and minister Leo Choerosphactes (ca. 920)
was ready to display, in a diplomatic correspondence, the sophistication of his
grammatical training, and to take his cue from issues of syntax and punctua-
tion when addressing an official complaint to Symeon the king of Bulgaria
about the release of some prisoners:220
, ,
, ...
the scribes (scribe I), who beside copying various parts of the manuscripts also
played the role of coordinator and corrector; however, the identity of this man
remains obscure. It has even been suggested, particularly in view of the many
Alexandrian authors represented in the collection, that this sylloge of texts
(i.e. the antigrapha of these manuscripts) may have arrived in Constantinople
from the Egyptian capitalas we mentioned above ( 1.2), Stephanus of
Alexandria has been credited with the role of mediator. Nowadays there is a
tendency to enlarge the scope of the provenance of these codices, renouncing
any unitary solution, taking into account the contribution of Palestine, Persia
and other peripheral areas of the Greek world, and referring the entire concep-
tion of this cultural enterprise to the activity of one or more writing circles in
9th-century Constantinople.226 However, other scholars maintain that at least
the core group of these manuscripts must share a common background, that
should be confidently traced back to Late Antiquity.227
The existence of non-professional writing circles, often connected with
single personalities or groups of intellectuals, has emerged as a vital factor in
9th- and 10th-century book production, even ifdue to the collective nature
of the enteprisesthese men were less ready to leave a trace of their activ-
ity in the form of colophons or subscriptions.228 One wonders if this may be
the origin of other outstanding products of Byzantine scribes of the late 9th
century, namely Vat. Urb. Gr. 111 of Isocrates (deriving recta via from a late
antique edition of the orator)229 and Par. Gr. 2934 of Demosthenes (famously
preserving the text of the decrees and the martyriai reported in the trials, and
probably coming from a different late antique prototype than the three other
independent witnesses, themselves slightly later in date).230
In Greek-speaking Southern Italy, between the 9th and the 10th century we
do not encounter any known scholar or scholarly activity stricto sensu,231 nor
is the number and quality of Classical manuscripts at all remarkable. What
we understand, however, is that Greek lexicography and grammar, along
226 Cavallo [2005]; Ronconi [2008], with a special analysis of Par. Gr. 1962; Ronconi [2013], a
radical thesis against the very existence of a philosophical collection.
227 See, after Irigoin [1980] 200204, Rashed [2002], Marcotte [2007], and the forthcoming
proceedings of the conference La collection philosophique face lhistoire organised by
F. Ronconi and D. Bianconi in Paris, June 2013.
228 Orsini [2005].
229 Pinto [2003] 3840; Fassino [2013] 2832.
230 See the essays in Grukova-Bannert [2014] (particularly E. Gamillscheg, S. Martinelli
Tempesta, B. Mondrain, and J. Grukova).
231 Mazzucchi [2010b] has argued that some form of philological activity must have been
performed in Southern Italy on Dio Cassius Vat. Gr. 1288 and on Homers Ilias picta
Ambrosiana (Ambr. F 205 inf.).
342 Pontani
with technical disciplines such as rhetoric and medicine, did find readers in
Calabria and Sicily, possibly as a mirror of the peculiar interests displayed by
the late antique schools of Egypt, Syria and Palestine,232 but also following to
some extent a chain of indigenous transmission.233 To narrow down the focus
to grammar and lexicography, the circulation of the so-called lexicon of Cyril
in Southern Italy is testified very early (see e.g. ms. Vallic. E II); however, it
seems unlikely that ms. Marc. Gr. 622 of Hesychius, written in the 15th century,
should have an Otrantine originas once believedand thus derive from
an earlier Italian copy. Even if scholars are today more skeptical about the
Italian provenance of the two manuscripts of the Etymologicum Genuinum,234
and even if we do not know the exact origin of the so-called Etymologicum
Casulanum (12th13th century),235 the persistence of an important tradition
of lexicographical studies is confirmed by such a complex manuscript as Vat.
Barb. Gr. 70, to be assigned to the Terra dOtranto of the late 10th century.236
The archetype of the entire extant tradition of the so-called Etymologicum
Gudianum, a widely read and copied lexicon that draws on the same sources
of the Genuinum (notably the epimerisms to Homer and the Psalms, the lexica
of Orus and Orion, synonymic lexica and lexeis to Byzantine canons etc.),237
ms. Barb. Gr. 70, was the object of a remarkable philological work of diortho-
sis, addition and implementation (e.g. further Homeric scholia and notes by
Choeroboscus; the lexicon called Synonymicum Barberinum), that must have
taken place in the Terra dOtranto towards the end of the 10th century:238 it has
been plausibly argued that this activity, directed by the hand known as d, took
place in the context of school teaching.239
2.5 Arethas
The leading Byzantine scholar between the last quarter of the 9th and the early
decades of the 10th century is again an ecclesiastical, namely Arethas of Patras,
archbishop of Caesarea in Cappadocia (ca. 850post 932): neither a teacher
nor a literary critic, he contributed to the dissemination of ancient Greek
240 Wilson [19962] 120130; Cavallo [2002a] 139141; Lemerle [1971] 210239, admittedly very
generous; Cufalo [2007] xxix note 73 (with discussion of the individual manuscripts).
241 Aletta [2004].
242 Luzzatto [2010].
243 Ed. Share [1994].
244 Russo [2012] esp. 111 on the textual transmission of the scholia.
245 Quattrocelli [2008].
246 Sonny [1896] 83130. See most recently Panzeri [2011] 8890.
247 Manfredini [1975].
248 Wilson [19962] 129; Russo [2012] 83 note 66 with earlier bibliography.
249 Bethe [1900] vvi.
250 epist. 44 Westerink: the correct interpretation of the passage
(not entirely fallen apart,
nor depriving readers of its utility) has been restored by Ceporina [2011] against e.g.
Cortassa [1997b]; see also Ronconi [2003] 2023.
344 Pontani
It should not come as a surprise that Arethas attention was also drawn
towards Christian texts, e.g. ecclesiastical law (Vallic. F. 10, with scholia),251 the
apologete Fathers such as Justin, Athenagoras, Eusebius and Clement (Par. Gr.
451, completed in 914; Arethas paid 20 nomismata to the scribe, Baanes, and
6 for the parchment, and the codex is now a fundamental witness for all the
texts it carries), and theological treatises (Mosq. GMI 231, written by Stylianos
in 932, with scholia).
As for the scholia penned by Arethas in his books, some of them faithfully
reproduce ancient or late antique prototypes, others (such as the memorable
dialogues with Lucian, or the criticisms directed against Plato and Julian)
stem from his own pen and ideas, though the distinction is not always easy
to draw. Recent analysis of the scholia to Lucianby far the largest corpus
of Arethan scholia to one and the same authorhas shown how much
Hellenic doctrine Arethas mastered and loved to display: on the cottabus,
on the jussive infinitive, on the history of Delos, on the blindness of Homer, on
the Attic use of incidental , on hyperbata in Hermogenes, on the aitia
of the Attic Thesmophoria, on Pythagoras golden thigh, and above all on all
sorts of lexicographical issues.252 In finding modern equivalents for Lucians
words, in elucidating antiquarian or mythographical issues, in looking for the
authors stylistic and narrative characters even beyond the charges of impi-
ety and sacrilege, Arethas establishes a dialogue with the ancient sophist that
involves in a productive manner not only his erudition but also the principles
of his own faith and ethics. Let us read a note on the Apology, in which the
archbishop learnedly attacks Lucians inconsistence with respect to his earlier
treatise on the salaried posts in great houses (de mercede conductis):253
. ,
, , ,
,
, ,
,
.
[ 4647]
.
to complain of old age: this creates the image of a simple character. But,
o ribald, those whom you have cruelly attacked will resort to this same
excuse, and will show that you are a censorious and querulous charlatan,
a mosquitoas they sayimitating an eagle: perhaps you resorted to
this verbiage in the attempt to imitate Socrates, when he refutes Prodicus
of Ceos and the sophists around him. But it is enough to counter you
with Homers lines: Were you like this when you assembled oarsmen and
sailed over the sea etc.
For all his admiration for a paradigm of pure Attic prose, Arethas was notori-
ously fond of an obscure and difficult style, a choice he defended by invoking
the example of Thucydides, Herodotus and even Gregory of Nazianzus: the
following passage is interesting for his views on ancient and modern Greek
style:254
, ,
,
[cf. Greg. Naz. or. 4.92] ,
.
,
.
.
To tell the truth, they [scil. the Fathers] too (as one can gather from the
words of the divinely writing man) seek to adapt to their own works the
language of Thucydides and the learning of Herodotus, two authors who
prove to be the best among the Hellenes through their density of style
and thought, and cause serious trouble to their attentive readers down
to the present day. For you cannot grasp automatically what they mean,
unless you follow ineffable twists and complain greatly about your help-
lessness. Thus, not even the divine Fathers hold the conciseness, the acer-
bity and the elevated solemnity of style as disreputable.
255 Wilson [19962] 136140, with further bibliography; Irigoin [1980] 192193.
256 Sch. Luc. p. 154.15 Rabe ...: Kavrus-Hoffmann
[2010] 5556, with previous bibliography.
257 Luzzatto [1993]; Pernigotti [2001].
258 Luzzatto [2000].
259 See now Ronconi [2012c], showing that this codex is not the copy of a late antique proto-
type, but rather the work of a 10th-century compiler who collected and annotated differ-
ent Aristotelian texts.
260 On Laur. 59.9 see L. M. Ciolfi, in Grukova-Bannert [2014] 23962. Par. Gr. 1741 is our
fundamental witness of the Poetics, grouped together with Dionysius of Halicarnassus,
Menander Rhetor and other uncommon authors of literary criticism: it was clearly com-
missioned by someone favouring an alternative approach to rhetoric than the current
one, represented by Hermogenes and his commentators: Harlfinger-Reinsch [1970];
Conley [1990]; Fryde [2000] 3132; Orsini [2005] 310313.
261 Diller [1975] 4253.
262 Kavrus-Hoffmann [2010] 6566. Netz-Noel-Tchernetska-Wilson [2011].
263 On Coisl. 345 as in fact belonging to the late 10th century, see now Valente [2008]. See
however Ucciardello [2012] 9194.
Scholarship in the Byzantine Empire ( 5291453 ) 347
witness of Polybius first pentad (Vat. Gr. 124, ca. 947),264 an independent codex
of Plato (Marc. Gr. IV.1, siglum T), an Organon (Marc. Gr. 201), the basic extant
collection of Hippocratic writings (Marc. Gr. 269),265 as well as an important
sylloge of rhetorical handbooks (Vat. Urb. Gr. 130).266
This elementary and very incomplete list (which could easily be extended
to the later decades of the century with such illustrious manuscripts as Neap.
Gr. 4* of Diodorus Siculus,267 Vind. phil. Gr. 67 of Stobaeus, and Vat. Gr. 738
of Sextus Empiricus),268 serves only as an exemplification of how important
a contribution this century made to our knowledge of Greek prose authors:
even when these manuscripts, most of which bear clear traces of their
Constantinopolitan origin, are not our unique or our earliest witnesses for the
authors involved, they generally stand out for their philological accuracy and
completeness.
The 10th century is also the age of the first, massive appearance of poeti-
cal manuscripts in Byzantium.269 Perhaps the best-known exemplar is the
famous Venetus A of Homer (Marc. Gr. 454), the only extant codex to display
in its margins the critical signs and (large excerpts from) the commentaries
deriving from the textual criticism of the great Alexandrian scholars: a remark-
able manufact reproducing its archetype with impressive skill, ms. Venetus A
is perhaps one of the most pivotal codices in the entire history of Classical
philology.270 No less impressive is Laur. 32.9, a landmark in our knowledge of
both Greek tragedy (it is our earliest witness for both Aeschylus and Sophocles)
and Hellenistic epic (it carries Apollonius Rhodius Argonautica), and of their
ancient exegesis.271 The philological analysis of both these codices shows that
they most probably derive from prototypes of the 9th century, which means
that the discovery of Greek poetry definitely antedated the age of Arethas.
2.7 Schools
The intellectual life of 10th-century Constantinople did not consist only of
ambitious scholarly enterprises. The epistolary of an anonymous professor
of grammar and rhetoric, who lived and taught in Constantinople around
920930, represents an invaluable and unique document revealing the
dynamics and mechanisms of education in this age.281 Beside showing that
private teaching was fairly common in the capital, and that the pupils mostly
belonged to the educated elite of high state officers,282 this collection of let-
ters also yields historical elements concerning the competition and coopera-
tion among teachers, the practices of transcription, copying and borrowing of
manuscripts, the methods for teaching elementary and advanced grammar,
the fees pupils were required to pay and the help expected from ecclesiastical
authorities, the educational attitude and the special role of teaching assistants
assigned to advanced students.
However, what interests us more directly in the present context is the phil-
ological practice to which the Anonymous professor refers in one of his let-
ters, addressed to a patriarch, perhaps Nicholas I Mysticus (in office 901907
and 912925).283 Evidently aware of the distinction between the scholars
( and ) and the mere scribes and calligraphers (
279 Cameron [1993] 300307; on the parts of the ms. see Orsini [2005] 302305.
280 James [2012] esp. 144157.
281 The letters, preserved in ms. Lond. Addit. 36749, have been edited by Markopoulos [2000].
282 Lemerle [1971] 246257; Speck [1974a] 2935.
283 This is epist. 88 Markopoulos: see Browning [1954], Markopoulos [1982], and particularly
Cortassa [2001].
350 Pontani
2.8 Collections
As a matter of fact, the role of Constantine VII (912959) in the organisa-
tion of culture must have been quite remarkable: we know that he sought to
revitalise higher instruction by combining in his universitywhether this
was an entirely public institution or an episodic creature of the emperors
mecenatism287both praxis and theoria, and by appointing the best teach-
ers available.288 Amongst them was the aforementioned Alexander of Nicaea,
the author of catenae to the Old and New Testament and the owner of ms. Vat.
Gr. 90 of Lucian, on which he left scholia partly related to those of Arethas.289
But Alexander was also the bibliophile to whom Nicetas Magistros wrote in
937/938 complaining that he could not find commentaries on some well-
known orations of Demosthenes (the False Embassy, the Crown, the Against
Androtion etc.).290
That a well-developed interest in books and book-collecting should surface
in schools and literary milieux of this age, is understandable: the emperor him-
self, partly continuing the tradition of his father Leo VI the Wise, devoted
strong efforts to arts and letters, not only writing ambitious comprehensive
syntheses on the etiquette at the imperial court (On the Cerimonial at the Court
of Constantinople), on government (On the Administration of the Empire), on
the geography of the empire (About the Themes), but also guiding the compila-
tion of encyclopedic syllogae devoted to different areas of human knowledge,
from medicine to veterinary studies (ms. Berol. Phill. 1538, prepared for the
imperial library), from zoology (the ancestor of Par. Suppl. Gr. 495 and Athos
Dion. 180) to agriculture (a copy in Laur. 59.32 of the Geoponica),291 from mili-
tary technique (Laur. 55.4 of the Tactica) to human history.292
Rather than envisaging the faithful textual transmission of single works,
Constantine aimed at collecting in his library as many books as he could,
appointing teachers and scholars capable of working on these books, and
then digesting the useful knowledge gathered from them in suitable ency-
clopedias. By far the most ambitious of these collections embraced excerpts
from Greek historians of all ages (with a predilection, as far as we can tell, for
287 Lemerle [1971] 263266 vs. Speck [1974a] 2228 (stressing the role of the corporation of
teachers).
288 Theophanes Continuatus 6.14, p. 446.122 Bekker: see Agapitos [1998] 175176 and
Lemerle [1971] 264265.
289 Markopoulos [2004], no. xvii; Maas [1973] 46872.
290 epist. 9, Westerink [1973] 7779.
291 See most recently Amato [2006].
292 See the overview by Lemerle [1971] 288297.
352 Pontani
early Byzantine authors), arranged according to their topic: out of the 53 origi-
nal sections, only the book On embassies and partly the books On virtues, On
ambushes and On gnomic statements survive,293 but their bulk is such that one
wonders if the whole enterpriseconsisting of dozens of similar chapters
was ever brought to a conclusion, and at any rate if it was ever read or copied
by anyone, beyond being preserved in the imperial library.294
This incredible compilation was clearly intended not as a historiographical
achievement in its own right,295 but rather as an encyclopedic work, which
paid the price of de-contextualisation to the advantage of readability and
of a thoughtful selection, especiallythough by no means exclusively
orientated on the moral aspect.296 Its leading idea was to gather bits and pieces
from different historians from Herodotus to Georgius Monachus (no more
than 26 appear in the extant sections), and to select their most useful parts in
a spirit that has recently been compared with Justinians rationale in putting
together the Digest.297
,
,
,
,
,
,
,
,
. ,
,
293 Of the latter two we have the original copies realised for the imperial library, mss. Turon.
C 980 and Vat. Gr. 73: see Irigoin [1959] and [1977]. See the editions by De Boor Roos
Bttner-Wobst Boissevain [19031910]. A new edition of On embassies is in preparation
by P. Carolla (winter 2014).
294 Cohen-Skalli [2012] and [2013]; Nmeth [2010] and [2013]. Pittia [2002].
295 Flusin [2002]. Magdalino [2011].
296 Lemerle [1971] 280288. See however Nmeth [2010].
297 Nmeth [2010] and Cohen-Skalli [2013].
Scholarship in the Byzantine Empire ( 5291453 ) 353
, ,
,
.
The scale of this enterprise did not dissuade the compilerswhatever the
organisation of their work, itself a debated issue298from quoting long pas-
sages from the various historians verbatim, with a fidelity that is all the more
welcome as several of the original sources (from Nicholas of Damascus to
John of Antioch down to several books of Diodorus Siculus) went lost soon
after the production of this anthology. In its ambition to transmit the use-
ful parts of a massive cultural heritage for the benefit of future generations,
the scope of Constantines work as stated in the proem attains the status of
historians of the imperial and early Byzantine age, mostly subsumed from
Constantine Porphyrogenitus brand-new encyclopedia.305
Perhaps unsurprisingly, the longest articles are devoted to such important
figures as Homer, Jesus, Origen and Dionysius the Areopagite. As a matter of
fact, the main goal of this work is to produce a historical dictionary that might
revive the (best of the) glorious Hellenic and Roman past in contemporary
Byzantium: lexicographical and historiographical choices can be explained
precisely in the light of this veneration of the remote past.306 The peculiar
attention devoted to Roman history chimes in well with what one would
expect from a Byzantine work, but also derives largely from the peculiarity
and scope of the intermediate sources to which Suidas is indebted.307 If the
interest of the modern scholar is stirred primarily by the wealth of information
about ancient authors and intellectuals, going backsometimes in a desper-
ately confusing wayto Hesychius of Miletus, and is attracted by the numer-
ous fragments of otherwise lost works,308 it should be borne in mind that this
work represents an outstanding unicum in the panorama of scholarly literature
of the 10th century, for it does not limit its scope to that of a mere etymologi-
cal lexicon, but combines different sources and different areas of interest in
order to provide the Byzantine reader with a wealth of otherwise widely scat-
tered knowledge, to introduce the reader into a world of the past that might
open up a dialogue with the present. A recent analysis has even attempted to
demonstrate that the compiler(s) of this lexicon was/were particularly gifted
in conjectural criticism.309
syntactical issues were sometimes tackled),316 and they hardly survived the
general political and cultural decline in the decade following the disastrous
military defeat at Mantzikert in 1071, which marked a decisive turning-point in
the history of the Byzantine empire. There is no reason to assume that these
chairs were part of a more wide-ranging imperial university encompassing
a full-fledged curriculum, but it is clear that Monomachus initiative was an
attempt to gather under the same roof the intellectual elite of his time, whose
members, from John Mauropous to the mesazon Constantine Leichoudes to
Psellus himself,317 had since the age of Basil II been running private schools
in various areas of the capital (Chalkoprateia, Sphorakiou, Forty Martyrs,
Diakonissa, Orphanotropheion etc.):318 in those days men did not devote
themselves to letters for profit but cultivated learning on their own, whereas
most scholars do not follow this path in matters of education, since they con-
sider money as the prime reason for occupying themselves with learning.319
A case in point might be John Xiphilinus, the nephew of his namesake the
patriarch: his epitome of Cassius Dios books 3680 (these books are today
mostly no longer preserved) was inaugurated at the request of emperor
Michael Ducas (10711078) in order to give an account of the pre-history of
the Roman empire, and therefore focused particularly on Augustus accession
to imperial power given that our own life and political system to a very large
extent depend on those times.320
This elite of imperial teachers, scholars, scribes and amateurs, which
included civil servants, learned monks, high-ranking ecclesiasticals etc., was
deeply rooted in the intellectual and cultural atmosphere of the capital, not
only as far as instruction was concerned, but also with respect to their peri-
odical meetings, readings and recitals known as theatra. The theatra became
particularly frequent in later decades under the Comnenian emperors, and
focused primarily on rhetoric and learning, sealing the mutual links of cooper-
ation and exchange between their members: what we gather from the speeches
and the reciprocal encomia of these people is the picture of a network of
educated friends who conceived literature and art both as a vehicle of educa-
tion and as a form of performance and high-brow entertainment.321
This context is all the more important for two reasons: first of all, the pre-
dominance of the capital has inevitably induced scholars to neglect isolated
amateurs living in more remote regions of the empire, such as the Cappadocian
protospatharios Eustathius Boilas, whose will (1059) stands out because it men-
tions, along with a generous list of liturgical and ecclesiastical books, one copy
of Leucippe and Clitophon, one of Aesop, and perhaps one of Artemidorus
Onirocriticon.322 Secondly, the rhetorical turn of theatra and of learned com-
munication in the 11th century (declamations, encomia etc.) explains to a cer-
tain extent the flourishing activity of distinguished rhetoricians such as John
Sikelos, who wrote a lengthy and erudite commentary on Hermogenes On
types of style (De ideis), and later of John Doxapatres, the author of homilies
on Aphthonius which draw significantly on ancient and less ancient literature
from Dionysius of Halicarnassus down to John Geometres.323
321 Marciniak [2007]; Cavallo [2007a] 7378; Agapitos [1998] 177181; Kazhdan-Epstein [1985]
120158.
322 Vryonis [1957]; Lemerle [1977] 1363.
323 Rabe [1931] liliii and cxiii; Kustas [1972]. On Sikelos see Wilson [19962] 150. On Doxapatres
see Rabe [1907] and Hock-ONeill [2002] 234237.
324 Karpozilos [1982]. Wilson [19962] 151153.
325 Hrandner [2012] 63; Reitzenstein [1901], showing his debt to a lost Latin grammatical
source of the Augustan period.
326 Epigr. 43.45 Lagarde: /
.
Scholarship in the Byzantine Empire ( 5291453 ) 359
as for myself, I pay less attention to the copies than to the truth of the
matter, since I have no other way to detect in them the tradition pre-
served by time in a genuine and uncorrupted state.
to all such problems they [scil. scholars] put forth one solutionthat
copies are not infallible nor blameless in their text, an explanation that I
have also introduced with good reason in the discussion of this theologi-
cal inquiry.
It is also interesting to see that Mauropous, while dealing with the reading
Absalom instead of Joab in Gregory of Nazianzus (Oration 21.15), identifies
the category of Verschlimmbesserung as what the mass calls a mistake from
correction ( : epist. 18.9798;
perhaps a reminiscence of Origens terminology? or else a terminological bor-
rowing from the ethical/religious sphere?).
A pupil of Mauropous became the leading intellectual of his age: Michael
Psellus (10181092/93) would frequently boast of his expertise in the most
diverse fields of knowledge, such as astronomy and medicine, geography and
mythology, law and architecture, music and rhetoric, with a special emphasis
on the entire range of Greek literary history.331 Claiming to be well acquainted
with foreign cultures (Egyptian treatises and Chaldaean oracles, works on
magic and alchemy etc.), Psellus was a polygraph who constantly tried to pres-
ent himself as a philosopher (a physician, a theoretic philosopher, a theolo-
gian): as a matter of fact, he became hypatos ton philosophon in Monomachus
newly restored academy,332 and in his own autobiography, after describing his
early success as a pupil, he showcased his swift promotion to the rank of the
most learned and versatile professor of the capital (he taught amongst others
the Georgian scholar Petritsi):333
,
, [Hdt. 2.135.1],
.
, , ,
,
, ,
,
[Theophr. hist. plant. 1.6.9]. ,
[Athen. 1.47e]
.
334 Psellus, Chronicle 6.3544; epistle to M. Cerularius 2a3a; Agapitos [1998] 180183;
Criscuolo [1990].
335 Vassis [1991] 1632.
336 Ed. Boissonade [1851] 343371. See Cesaretti [1991] 90123.
337 See encom. matr. 6ac.
338 Cesaretti [1991] 6089. Wilson [19962] 161163. Roilos [2005] 121124. Angelidi [2005].
362 Pontani
,
, ,
,
.
,
,
.
Every Hellenic myth is really just myth, and just as the things that do not
exist and never had a chance to exist can never exist, so too the empty
mythology of the Hellenes can never acquire firm substance, nor will
their scattered wisdom ever become concrete. However, we should prac-
tice speech not only on the firm ground and on paths accessible to the
mission of rhetoric and philosophical discourse, but we should give dis-
cursive substance even to non-existent entities, so that we may obtain a
sweet drink not only from the drinkable waters but also from the exter-
nal, bitter ones.
339 Guglielmino [1974]; Cesaretti [1991] 2943; Wilson [19962] 149150. Psellos funeral oration
for Nicetas is still to be read in Sathas [1876] 8796.
340 Alleg. de Iove nato, p. 220.19 Sathas (see Cesaretti [1991] 81, and Roilos [2005] 122).
341 Papaioannou [2013] 2950.
Scholarship in the Byzantine Empire ( 5291453 ) 363
rhetorica, stems precisely from Psellos age):342 his focus on style influences
his judgment concerning Christian authors, e.g. the praise he bestows on
Symeon Metaphrastes for having recast the hagiographies in a neater style,
or on Gregory of Nazianzus as the finest paradigm of Greek prose (in various
genres, from panegyric to philosophical writing), or on John Chrysostom for
his clarity as opposed to Thucydides.343 Psellos most illuminating essays in
this field are the comparationes between Euripides and the 7th-century poet
George of Pisidia and between the novelists Heliodorus and Achilles Tatius.
While the conclusion of the former comparison is lost (but Euripides metrical
versatility and ethical shortcomings are both taken into account),344 the latter
ends in favour of Heliodorus, thanks to the quality of the complicated plot, the
richness of Attic diction, and the moral qualities of the female protagonist345
(a Byzantine penchant for Heliodorus is easy to prove: suffice it to think of the
allegorical interpretation by Philip-Philagathus of Cerami: see below 3.8).
A passage on Euripides will exemplify Psellos approach to ancient authors:346
[] [ ,
]
[. ] [] ,
[ ], [ ] [],
,
[], [ ]
. [ ]
[ ]
[] .
[Hec. 218ff.]
,
.
those of others. Euripides, on the other hand, took fewer pains over these
matters but devoted more effort than he to musical composition, i.e.
that in the words, and its use and to these three fairest of arts, viz. music,
rhythmic and metric, bringing as it were shawms, citharas and lyres into
conjunction with his own plots. When he has to use barbarian accents, he
imitates their speech in such a way that the same man is considered best
in using Greek and most precise in committing solecisms (the element in
barbarian speech which is contrary to Attic is solecistic). There are times
when he deviates from propriety and comes more under the sway of the
power of his own eloquence than of the strict demands of the poetry. For
instance, when he has brought Hecuba onstage as antagonist to Odysseus,
a man of noble birth and oratorical skill, he raises her up against him and
gives her the prize of honor; he has Odysseus declaim not without charm,
but he makes him inferior to a captive woman! [transl. A. R. Dyck]
Indeed, Psellus was particularly keen on ancient Greek dramaa rare occur-
rence in pre-Palaeologan Byzantium347, and there is a chance that he may
have composed a short treatise on tragedy, partly derived from Aristotelian
doctrine and partly original in its approach to metre, poetic diction, scenogra-
phy (ekkyklema etc.), musical styles, dance and actorial conventions.348
Even if Byzantine philosophy does not fall within the scope of the present
survey, it must be stressed that Psellus commitment to the recovery and the
study of Platonic and Neoplatonic thought (Porphyry, Iamblichus, Proclus, but
also the Hermetic corpus) proved propaedeutic not only to the development of
his own theoretical work (e.g. the important treatise On miscellaneous wisdom
[De omnifaria doctrina], largely indebted to Plutarch and Proclus), but above
all to a surge in the transmission and study of these authors in Byzantium.349
Psellus letter to the patriarch Xiphilinus is paradigmatic in this respect as
an attempt to rescue Plato and pagan philosophy in the sense indicated by
the Church Fathers and by Gregory of Nazianzus in particular.350 Much as
in the case of Homeric allegorism, Psellus keeps promoting pagan wisdom for
the exclusive purpose of gaining useful teachings for the present age and a cor-
rect understanding of nature, without any conflict with the Christian faith.351
This is why Psellus humanism has been often seen as a compromise that
gave no impulse to a first-hand scholarly activity on texts or to a true philo-
sophical revival, but fostered a new inquiry into the intellectual and literary
genealogy of well-known authors (Gorgias behind Hermogenes, Plato behind
Proclus etc.).352
Even so, Psellus Platonism earned him the sentence of having to make a
public profession of orthodox faith, and a brief relegation in a monastery.353
Much more severely, Psellus pupil John Italus, a native of Southern Italy who
succeeded him in the chair of philosophy in 1055, was put to trial in 1076 and
confined to a monastery six years later for his overt adherence to Platonism,
and for his attempt to deal on the same level with pagan philosophy and with
Christian theology.354 The text of the anathema against Italus and his works
(which, despite the damnatio, partly survive down to our own day) is a remark-
able caveat against those who study the Greek disciplines and not only for the
sake of educational training but also follow these vain doctrines and believe in
them as having certainty, so that they initiate others into these doctrines, some
by stealth, others openly, and teach them without hesitation.355 Although
Italus learning extended to grammar and poetry,356 his symbolic importance
in the cultural history of 11th-century Byzantium is connected with the vicis-
situdes and the meaning of his trial:357 the risk of a Neo-platonic haeresy
was to pop up frequently in the 12th century, not only at the imperial court,358
but also in the wider context of culture, from the inaugural lectures in the
359 Michael of Anchialos when promoted to the chair of hypatos ton philosophon in 1165/67:
see Browning [1977], iv.
360 Agapitos [1998] 187191; Podskalsky [2003].
361 Magdalino [1991] 11. See also Magdalino [2012] 19.
362 On the latter see Rhoby [2009] and Jeffreys-Jeffreys [1994].
363 Mullett [1983]; Jeffreys [2011]; Cavallo [2002b] 429431; Kazhdan-Epstein [1985] 121133.
Scholarship in the Byzantine Empire ( 5291453 ) 367
364 Browning [1981]; Katsaros [2003]; Flusin [2008] 390391; Fuchs [1926] 29 and 47; Luzzatto
[2000] 5354.
365 See the impressive list by Browning [19621963], completed by Constantinides [1982] 51;
Agapitos [1998] 190191.
366 See the pessimistic view by Speck [1974a] 6480, and Magdalino [1993] 325330.
367 Lemerle [1971] 8588 and 9596.
368 Martin [1956] 229230. Wilson [19962] 180181.
369 Ed. Koster [1922] 101113.
370 Ed. Boissonade II [1830] 340393; see Tovar [1969].
371 Hunger [1978] 20 and Hrandner [2012] 6466; Darrouzs [1960]; Roosen [1999] and
Antonopoulou [2003], with a fresh account of biographical and bibliographical data;
Nicetas poem in hymnic form on the pagan epithets of the Olympian gods is a remark-
able tour de force: Kaldellis [2009a] 1415; Browning [1963] 1417.
368 Pontani
. <>
.
clearly he did not call him brother, for some copies have peon [of the
peoi] as the fag-end of the line: in some of the exemplars we have found
poleon kudistate peon [most illustrious of my many peoi], and peos is a
generic noun for relative.
Many other scholars taught in the patriarchal academy during the first half
of the century. But a no less crucial feature of the Comnenian age consisted
in the imperial courts special interest in the ancient Hellenic heritage. Far
from the asphyctic walls of schools and academies, in a delicate historical
juncture where they needed to counterbalance the territorial contraction
of their Empire,374 the emperors and their circles rediscovered or reshaped
part of their Greek identity precisely around the Hellenic past and classical
exempla, in a double-tongued discursive synthesis that was not in conflict
with the Christian and Roman pedigree but supplemented it with the flavour
and the very substance of a glorious civilisation: the pagan gods were back
in fashion, political and historical speculation cautiously started to take into
account ancient Athens behind the old and the new Rome, and some of the
ancient works dealing with pagan history and mythology became essential
reading for every educated man.375 This is very evident in the fresh approach
to Homer and Homeric studies as a repertoire for historians, as a source of
inspiration for poets and writers, and as a working ground for philologists
(from Theodore Prodromus to Eustathius of Thessalonica, from the novelists
to Constantine Manasses to Nicetas Choniates).376 But perhaps even more
impressive is the creation, by the hand of an anonymous scholar deeply famil-
iar with ancient literature, of a refined cento of 5th-century Attic drama known
as Christus patiens, and dealing in very sophisticated and allusive (chiefly
Euripidean) terms with the Passion of Christ (the text used to be attributed,
wrongly, to Gregory of Nazianzus).377
To be sure, a suspicious attitude against profane culture, and specifically
against the teaching of ancient Greek grammar, continued to make itself felt,
e.g. when the rhetor George Tornices, metropolitan of Ephesus in the 1150s,
gave voice to the worries of Alexius I, and wrote that the wise emperors and
educators did show a high consideration for culture and studies, but:378
, ,
,
,
,
, .
men and treacherous for women and girls, whose eyes and ears, in their
view, should remain virgin.
.
, , ,
,
379 See Basilikopoulou-Ioannidou [19711972] 169199, and a thorough discussion of the reli-
ability of this terminology in Magdalino [1993] 382412.
380 Browning [1995] 23.
Scholarship in the Byzantine Empire ( 5291453 ) 371
,
.
, ,
,
.
.
381 Robins [1993] 127148; Browning [1975a] 9; Browning [1976]; Gallavotti [1983]; Festa [1931];
Polemis [1995]; Efthymiadis [2005] 266271; Ciccolella [2008] 113118; a synthetic over-
view in Hunger [1978] 2429.
382 Vassis [19931994] 1012, ll. 46:
, =
,
For he came to Byzantium bringing us good
news (but every village and every town are already full of Nicholas grace).
383 Miller [2003] 1213 (ll. 815).
372 Pontani
390 Only partly edited: see Sell [1968]; Berger [1972]; Lasserre-Livadaras [19761992]; Baldi
[2013]. Its interest lies inter alia in the fact that it faithfully reproduces the glosses of the
Genuinum, thus enabling their reconstruction when the two mss. of that lexicon are dam-
aged or missing.
391 Ed. Gaisford [1848]; letters - are also available in the synoptic edition of Lasserre-
Livadaras [19761992]. See Reitzenstein [1897] 241248. A much smaller compilation is
the so-called Etymologicum Parvum, ed. Pintaudi [1973].
392 Browning [19621963] 1920.
393 Kaldellis [2009a] 1617.
394 Ed. Montana [1995].
395 It shows the same sytem of parts of speech, and the same overlap with philosophical
categories: see Robins [1993] 163172 and Donnet [1967a].
396 Wilson [19962] 184187.
397 Giannouli [2007] and Demetrakopoulos [1979].
398 Gregory of Corinth, Exegesis on the iambic canon 17.4 and 2.56 Montana respectively.
374 Pontani
But the work that constitutes Gregory of Corinths best claim to renown
today is his lengthy treatise on the Greek dialects,399 whose degree of origi-
nality has, however, often been called into question. The author himself men-
tions among his sources Trypho and John Philoponus; modern research has
confirmed these debts, and has also detected debts to the scholia to Dionysius
Thraxall in all, little seems to proceed directly from Gregorys own autono-
mous learning. In the treatise, we find a series of (sometimes precious) quo-
tations from ancient authors digested according to their alleged provenance:
the preface mentions Aristophanes, Thucydides and Demosthenes for Attic,
Hippocrates for Ionic, Archytas (!) and Theocritus for Doric, Alcaeus for Aeolic
(though of course e.g. Euripides, Herodotus and Pindar often appear in con-
nection with their respective dialects, and Homer for all of them). Then we
read a definition also indebted to late antique models:400
,
[ Schaefer] . ,
, .
, , .
, . ,
. , ,
. .
Gregorys work on the whole lacks a coherent structure, and it shows a num-
ber of mistakes, oddities and inconsistencies, but it deserves consideration as
an ambitious attempt to systematise such a difficult issue for the benefit of
399 Ed. Schaefer [1811]; see Bolognesi [1953] and Consani [1991] 5968.
400 Pp. 912 Schaefer. For the reading instead of Schaefers conjecture see
Consani [1991] 66.
Scholarship in the Byzantine Empire ( 5291453 ) 375
401 Including one unduly attributed to the 13th-century scholar Manuel Moschopoulos: see
Hunger [1978] 3132 with references, and Cengarle [1971].
402 Gouma Peterson [2000].
403 Wilson [19962] 182183; Kaldellis [2009a] 3639; Browning [1962]. Conley [1990] 3840.
Fryde [2000] 5458. Arabatzis [2006] 1736.
404 Some of his works ended up in the Okeanos, namely ms. Laur. 85.1: Cacouros [2000]; Fryde
[2000] 193196. For Michaels interest in Aristotelian works on biology and zoology see
Hellmann in this volume.
376 Pontani
,
,
,
, ,
,
,
,
,
, , , ,
.
to introduce the myth in our own rightful court, one piously interprets
its sense in relation to our forefather: for the first created man deceived
the demiurge, violating his order, and was then exiled from Eden. Since
then, he has worn this thicker flesh as a leathern coat, and has been
allotted this toilsome life, being condemned to live in the valley of tears,
to eat bread with his sweat, to succumb to abominable maledictions and
to the worst of penalties, deathuntil Heracles, my Jesus, comes to set
him free.
409 The best edition is Rizzo [1971]. See also Wilson [19962] 180.
410 The treatises are edited by Kindstrand [1979] and Hinck [1873] 5788. On Isaacs ekdosis
see Pontani [2006a], with the editio princeps of selected scholia.
411 His interest in prosopopoea surfaces e.g. in his note to Il. 8.183, where we read an uncom-
mon reference to Lucian: ,
(cf. Catapl. 27).
378 Pontani
biography and the contents of Iliad and Odyssey must be reconstructed and
understood within their historical context and ideological horizon, not in the
arbitrary vacuum of a philosophical or theological stance.418 In this respect,
Tzetzes displays a very philological mind indeed.
Tzetzes is also the author of several corpora of scholia now lost or unpub-
lished, e.g. to Sophocles,419 to Oppian,420 to Porphyrys Eisagoge.421 But he is
best known for the impressive commentaries on Hesiods Works and Days,422
on Aristophanes plays,423 and on Lycophrons Alexandrathe latter is attrib-
uted in the manuscripts to his brother Isaac, but was in fact certainly com-
posed under Johns direct scrutiny.424 All three works, preserved intact down
to our day, address pivotal texts in the late antique and Byzantine curriculum
(the fortune of Lycophron may sound unexpected to us, but it proceeds from
the status of the Alexandra as a repository of exquisite mythographical learn-
ing, and as the only narrative poem in dramatic form),425 and they all interact
with the respective corpora of scholia vetera. However, Tzetzes attitude towards
the heritage of ancient exegesis is rarely a passive one: the Hesiod commentary
offers from the very beginning a proud statement of its novelty and original-
ity, and is replete with critical remarks against the predecessor Proclus, who
is reproached for his obscurity and manifold inadequacy.426 The remarkable
blend of literary, allegorical, moral, etymological and mythographical learning
in Tzetzes writings on epic is well exemplified by the following passage on the
Muses occurring in Works and Days 1 (29.1330.1 Gaisford = ll. 10933 Cardin):
,
.
,
, ,
, .
,
,
. , , ,
, ,
,
,
. ,
, , ,
,
.
, ,
, .
But Muse is the knowledge obtained through education, not the innate
and untaught wisdom. Mousa is so called because it is sought by every-
one or by the many, or because it inquires and investigates on many
things. The Muse is called collectively Muses in the plural: they are nine,
according to myth, namely thrice three, i.e. many times many, for number
three is the beginning of multitude. And even if the general knowledge
is uniform and simple, it extends itself endlessly through the partial dis-
coveries and inventions: the doctrines of men are faint and very partial,
not like the divine one which is unitary and entire. And those doctrines of
ours, the Muses, are born in Pieria, the dwelling of Zeus (i.e. of the intel-
lect), namely the place in our head whence flow the fertile, abundant
[piona, pimele] and wittiest thoughts, through a nine-day intercourse of
Zeus with Mnemosyne, i.e. through the intellects repetition and memo-
risation of what it has read. Being born in this true Pieria (not on the
mountain), the doctrines dance on the Helicon, celebrating their father
Zeus: this means that, being written down in books, they whirl [helisson-
tai] and circulate everywhere as if in a dance, announcing the intellect
that has given them life. And thus I call Pieria the dwelling of the intel-
lect, and Helicon the books through which the doctrines (i.e. the Muses)
circulate, and the literary works.
Greece (esp. lyric poetry and drama), a controversial history of the Pisistratean
recension of the Homeric poems and of the translation of the Septuagint,434
and a detailed reconstruction of the library and the scholars in Hellenistic
Alexandria (the authors interest in chronology emerges more fully in his bulky
Chiliadesa sort of universal chronicle grown out of distinct scholia to his
own lettersbut also in other parts of his work).435 A didactic treatise in iam-
bic trimeters on ancient Greek tragedy, largely devoted to its origins, structure,
and metre, was compiled by Tzetzes on the basis of scholia to Euripides and
Dionysius Thrax, the Anonymus Parisinus and other more remote sources.436
His inquiries into metre resulted in a short companion to the principal cola,437
but his brother Isaac surpassed him greatly in this respect: in his versified
treatise on Pindars odes438 Isaac paraphrased the metrical scholia but also
displayed a certain familiarity with complex lyric systems and responsion,439
in an age when Byzantine doctrine on metre still widely relied on the rules
put forth by Hephaestions handbook, variously excerpted, commented, and
amplified in the several, mostly anonymous short treatises devoted to the
subject.440 Following in Isaacs footsteps, his later contemporary Trichas pro-
duced an influential Synopsis of the nine metres, based on the lines of a hymn
to the Virgin.441
We will never know what Tzetzes book of reasonings ( )
looked like, if it ever existed: in a passage of his commentary on the Frogs,442
he tells us it contained a critical discussion of 52 (!) Euripidean dramas and 119
books of various authors, while yet other books bear my scattered reasonings
on other wise authors, dealing with flaws, contradictions and various types
of lapsus to be found in each of them. The latter statement has found some
partial corroboration in the notes detected on two important manuscripts
of ancient Greek historians: Tzetzes autograph annotations on ms. Laur.
70.3 of Herodotus address issues of prosody, grammar, and accentuation,443
while the fifty marginalia he penned in the margins of a late 9th-century codex
of Thucydides (Heid. Pal. Gr. 252) display a greater ambition, and number
among the most surprising, if unsystematic, philological achievements of the
Comnenian age.444
In addition to the acute observation of some palaeographical peculiarities
of the scribe (the position of final sigma, the old-fashioned system of accen-
tuation and punctuation), Tzetzes offers here some more wide-ranging pro-
tests against the world of learning around him (we have mentioned his violent
attacks on the new fashion of schedography and its popularity in the patriar-
chal academy: see also his note to Aristophanes Frogs 1160a, p. 139 Koster), or
against the obscurity (asapheia) or the syntactical soloecism of the author he
is commenting on (notes on Thuc. 5.17.2 and 18.15). More specifically, when
defending the reading against in Thuc. 1.123.2, he adds
an extremely violent invective against those philologists (offspring of pigs, of
illiterate barbarians) who correct sound texts without any guiding criterion,
e.g. eliminating the psilosis from Ps.-Herodotus Life of Homer ( ,
instead of the respective forms with ; he later pleads in favour of
the manuscripts in the speech of the Thebans in 3.61.1although he
wonders why this is the only dialectal feature left in the entire speech):445
,
, ,
,
.
in with their opinions, / and argues that one should write according to
the technes norm / both in poetry and in prose, / not polluting in any
respect the principles of the techne.
Together with many other minor philological suggestions, some of which sup-
ported by references to his knowledge of literature (e.g. Euripides Alcmeon
quoted on Thuc. 2.102.5), or to unconventional etymologies (Italia from Latin
vitulus, or Gela from river Gelas: the latter shows that he had access to a fuller
copy of Stephanus of Byzantiums Ethnika), Tzetzes overtly addresses issues
of manuscript transmission when praising the antiquity of manuscripts as
a criterion for their textual reliability. This is the same idea lurking behind
his lengthy metrical addition to the scholium on Aristophanes Wealth 137
(p. 41.1228 Massa Positano), against the who
blemish recent codices:
,
,
,
...
.
because the man who had urged me to write / did not find for me an old
book, / or at least two or three of the recent ones, / so that I may cor-
rect one line from one, one line from the other, / I then found two recent
books ... / and I redressed the ship of speech / until the wave of awkward-
ness remained tiny...
3.7 Eustathius
Tzetzes aforementioned invective in the Thucydides notes was directed,
amongst others, against the professors of the patriarchal school (the Tholos of
St. Sophia) and of the imperial academy (the so-called Senate of the Philo
sophers, close to the Portico of Achilles). It is possible that Tzetzes included
among his targets also the most distinguished figure among such professors,
although we have no clear information about his personal and intellectual
relationship with him: I am alluding to the other great Classical scholar of the
Comnenian age, perhaps the most learned man of the Byzantine millennium,
namely archbishop Eustathius of Thessalonica (ca. 11151195/96).
386 Pontani
Eustathius spent the first part of his life in the capital as a private teacher,
then from 1168 onwards as a professor of rhetoric (maistor ton rhetoron) in the
patriarchal academy and as the organiser of an important reading circle; in
1174 (or 1177) he was appointed archbishop of Thessalonica, the city whose
surrender to the Latin invasion in 1185 he described in a detailed historical
monograph.446 His letters and his opera minora, while dealing chiefly with
moral, ecclesiastical or theological issues, show a vast amount of Classical
learning, which he regularly manages to prevent from clashing with Christian
ethics and ideology. In this respect, Eustathius appears as one of the most
paradigmatic examples of Byzantine Christian humanism, all the more so
as he proves conversant with an incredible number of ancient authors and
texts. His pupil Michael Choniates, in a heartfelt eulogy of Eustathius extra
ordinary moral and intellectual qualities, affirmed that he subordinated
Greek philosophy to the divine Christian wisdom like a serious servant to a
noble landlady.447
Eustathius outstanding place in the history of scholarship is ensured,
once again, by his commentaries: there is scanty evidence about his notes to
Oppian,448 Aristophanes,449 or the Greek Anthology,450 and we will overlook
here his (as yet largely unpublished) exegesis on John the Damascenes Iambic
canon for Pentecost.451 The most important documents of his scholarly work
are definitely his commentaries on Pindar, on Dionysius the Periegete and on
the Homeric poems.
Of the Pindar commentary, probably the earliest in date, only the proem
is extant, and it is fair to wonder if he ever wrote down in full the entire
exegesis, or if it remained in the state of an unsystematic set of notes from
his lectures. Be that as it may, even the proem452 offers some penetrating
insights on Pindars poetic style (his dialect, the compound adjectives, the
hyperbata, the profusion of metaphors and allegories, the ekphraseis, the
frequent use of digressions and gnomai, obscurity as a Stilprinzip), as well
they [scil. the epinicians] are especially popular because they are more
human, concise and otherwise relatively less obscure.
, ,
, ,
my heartfelt aim was to stroll through the Iliad and provide what is use-
ful to the reader, not to the learned man (for it is unlikely that he should
ignore any of these things), but to the young man who is learning at this
moment, or perhaps has already learnt but needs some reminders.
456 On this term, see Kambylis [1991a] 1516. Kolovou [2012] 151153.
457 Martini [1907], Pontani [2000], and above all the revealing codicological analysis by
Cullhed [2012].
458 Commentary on the Iliad 2.2123 (p. 3.58 van der Valk). See Browning [1992a] 141142.
Scholarship in the Byzantine Empire ( 5291453 ) 389
Not especially conversant with textual criticism (Maas idea that he might
have restored the text of Athenaeus when producing its epitome has been
proved to be wrong),465 careful but sometimes enthusiastic with allegorical
interpretation (especially when applied to traditional myths rather than to a
poets isolated fanciful imagination),466 Eustathius strikes the modern reader
for his immense devotion to study, and his manifold interest in ancient life and
literature. Rhetorical, allegorical and grammatical explanations occur freely
here and there, and individual passages are sometimes first summarised and
briefly commented on from an overall perspective and then taken up line by
line. This procedure responds to the different needs inherent in Eustathius
commentary: on the one hand the task of locating and explaining passages of
a complex poetical text, on the other the preparation of this text as a poten-
tial source of quotation and imitation for the pupils. Hence the importance
of Hermogenian terminology:467 Eustathius rhetorical comments on Homer,
as recent research has shown, are not purely erudite notes on a dead author,
but they rather presuppose a widespread mimetic practice on the part of
Byzantine writers, in which both the stylistic features and the mythological
content of the Homeric text are paramount to the refinement of oratorical or
historiographical prose.468
The use of allegory in Eustathius is not primarily defensive for confessional
purposes, and never aims at Christianising the poem (as in Psellus or to a certain
extent in Galenus): rather, while refraining from a purely contextual or rhetori-
cal rendering of the myths involved, Eustathius sticks to the general credibility
of Homers plot,469 but adds interpretive dimensions to a text whose multiple
hermeneutic facets the pupils were invited to discover. This is why, even in the
wider allegorical framework presenting, for instance, Penelope as philosophy
(the homeland of the spiritual journey of Odysseus/man), in several cases vari-
ous diverging types of allegoresis can be juxtaposed without choosing between
them (Calypso as matter, bodily life, or astronomy; Proteus as a dancer, the idea
of friendship, primal matter, or number ten).470
Eustathius guiding principle is in fact the utility of Classical works for the
education of the young. The utility () of the poem does not reside
in its alleged hidden Christian message, but more deeply in a moral reading,
465 Van der Valk I [1971] liilvi; Erbse [1950] 7592; A. Pontani [1995a] 341342.
466 Cesaretti [1991] 207274.
467 Lindberg [1977].
468 Nnlist [2012c]; Cullhed [2014] 3843 and 4954.
469 Cesaretti [1991] 207274, to be read with Cullhed [2014] 4449.
470 Pontani [2011a] and [2013].
Scholarship in the Byzantine Empire ( 5291453 ) 391
,
(
),
,
,
,
, ,
,
,
, ,
, ,
,
.472
places lend nobility to Homers poetry; mythical stories, not only those
told by the poet in a peculiar manner, but sometimes even more widely,
as they emerge from the accounts of other writers; finally myths, some of
which untempered, incurable and considered only at face value, others,
on the other hand, equipped with an allegorical or anagogic treatment.
The importance of Eustathius teaching and cultural activity must have been
indeed remarkable in the latter part of the Comnenian age: two learned broth-
ers acquired proficiency in the realm of letters precisely thanks to their daily
synanagnoseis (common readings) with the future archbishop. I am refer-
ring to Michael and Nicetas Choniates, the former himself an archbishop of
Athens since 1182, the latter a writer of history and theology. Neither was a clas-
sical scholar, but Nicetas historical prose is perhaps the most sophisticated
example of the adoption of Classical quotations and models in Byzantine
prose, as recent studies have shown, with regard, for instance, to the depiction
of emperor Andronicus Comnenus in Odyssean terms,473 or even on a more
down-to-earth lexical level.474
Nicetas brother, Michael Choniates (11381222), overtly describes in his
epist. 102 Kolovou his own participation in a reading circle of the capital, and
recalls Eustathius contributions of doctrine and liveliness to these meetings:475
Michaels special feeling for antiquity is revealed by a long elegy on the ruins
of Athens, perhaps one of the most important pieces on this subject in the
whole of extant Byzantine literature.476 His entire oeuvre is permeated by a
sense of Christian humanism coloured by Stoic accents and by a strong fidelity
to Atticist style, as well as by a constant dialogue between his Greek, Roman
and Christian backgrounds. Michaels proximity to antiquity is shown inter alia
by the fact that he is the last known man to quote at first hand Callimachus
Hecale and Aitia, yielding not only invaluable elements for the reconstruction of
the epyllion, but also one of the most remarkable instances of stylistic and rhe-
torical appropriation of a Hellenistic author in Byzantium. As we learn from
his letters, Michael owned a large personal library, which included Euclid,
Thucydides and other Classical authors: it might well have suffered losses after
the fall of Athens to the Latins in 1205, but it certainly subsisted after that date
during Michaels exile on the island of Keosindeed most of the Callimachean
quotations in his works belong to the second decade of the 13th century.477
477 Pontani [2011b] 115117, with further bibliography; Wilson [19962] 204206.
478 Cavallo [2000].
479 Magdalino [1993] 323325; Bianconi [2010a] 7779 and n. 6.
480 Cavallo [2000] 231233 lists several manuscripts, arranged according to their paleographi-
cal facies.
481 Mazzucchi [2003] and [2004].
394 Pontani
, . ,
. .
Note that this Theocritus was corrected against the book of Calabrus the
paedagogue, the pupil of Psellus. The Periegete, on the other hand, was
corrected by comparing the text with Stephanitzes the Periegete [scil.
Stephanus of Byzantium], and Pindar from Trichas.
One of the hands found in Ambr. C 222 inf. appears together with other
scribes in a series of 21 manuscripts produced by the so-called scriptorium
of Ioannikios.484 The only known scribe of this group, Ioannikios, must have
been the leading figure of a team specialised in manuscripts of philosophical
and medical content (Aristotle, Galen, Aetius, Paul of Aegina), while showing
an interest for literary authors as well, e.g. Sophocles and Euripides (the impor-
tant Laur. 31.10, with scholia), Homers Iliad (Vat. Gr. 1319) and Apollodorus
(Par. Gr. 2722, ff. 1632). Earlier research had connected this activity with
Southern Italy, because five of the codices identified so far bear the mark of
Burgundio of Pisa, an Italian scholar and translator who spent his life between
Constantinople, Messina and his hometown (more on him below 3.9); how-
ever, palaeographical and codicological analysis has proved inconclusive, and
the exact collocation of this very interesting enterprise still remains sub iudice;
an interesting case has been made for Constantinople, in connection with the
local interest in Aristotelian philosophy in the early 12th century.485
3.9 Italy
Ioannikios manuscripts are linked to Burgundio of Pisa, one of the learned
Westerners who spent part of their careers in Constantinople, in an age when
Venice, Genoa and other cities intensified their commercial links with the
Byzantine Empire.486 Burgundio translated the Exposition of the Orthodox
Faith by John of Damascus as well as other patristic treatises, but also works of
Galen and Nemesius, and Aristotles Meteorologica. Other outstanding mem-
bers of this heteroclite group of Italians are James of Venice, to whom the West
owes the first systematic activity of translation of Aristotle from the Greek
rather than from the Arabic,487 and Moses of Bergamo, a book collector and
translator of Greek ecclesiastical and grammatical works (including a treatise
on the oblique case of nouns like ), whose hand has recently been
identified as the annotator and interlinear glossator of Theognis in the glorious
ms. Par. Suppl. Gr. 388.488 Burgundio, James and Moses took part as interpret-
ers and experts in the famous theological dispute of 1136 between patriarch
Nicetas and Anselm of Havelberg in Constantinople.489 Other scholars and
translators, particularly keen on ecclesiastical issues, were to follow in later
decades, among them Leo Tuscus and Hugo Etherianus.490
But, as we have seen above ( 2.6), first-hand Greek doctrine did circulate on
the Italian soil: whether or not the Ioannikios manuscripts belong to Southern
Italy, a brief note must be devoted here to Greek written culture in Sicily and
Apulia between the 11th and the 12th century. First of all, once more we find
an absolute predominance of liturgical and theological manuscripts; in the
face of this, scholars have uttered conflicting judgments about the range and
the extent of the circulation of Classical texts in the area. The optimistic view
insists on the continuity of some textual traditions preserved in Magna Graecia
since antiquity;491 a more painstaking and more cautious analysis stresses the
persistent incompatibility between monastic culture and pagan literature
(even in the most important monasteries such as the Patir at Rossano or San
Salvatore at Messina), and the paucity of manuscripts of Classical authors that
can be safely ascribed to Southern Italy before the early 13th century (other
492 Luc [2012b] and [1993] respectively. See also Luc [1990]; Jacob [2002].
493 See Luc [1993] 2930 and especially 6388, with a painstaking examination of the cul-
tural milieu and the manuscripts (though he is wrong on Philip-Philagathus, see below).
Some form of philological attention was devoted to Christian texts (see e.g. Luc [1989]),
but the Norman domination did not bring about a flourishing of schools, academies or
cultural activities in Southern Italy.
494 Luc [1993] 8586 and [2007a] 80; Dorandi [2002b].
495 Martinelli Tempesta [forthcoming]; Carlini [20022003]; Dorandi [2002b] 3 and Berschin
[1980] 292295, with earlier bibliography. For an updated survey on Eugenius verse, and
on the cultural atmosphere of Norman Sicily, see Torre [2007] and [2008] 6389; Luc
[1993].
496 E.g. in the Salernitan school: see Irigoin [2006a].
Scholarship in the Byzantine Empire ( 5291453 ) 397
mid-12th century. This man stands out for his acquaintance with Classical texts
(from Homer to Plato, from Menander to Lucian to late Greek rhetors), which
emerges both from the learned quotations, the philosophical depth and the
erudite style of his homilies,497 and from the most surprising of his writings,
namely a complex allegorical introduction to Heliodorus novel, preserved in
the Otrantine ms. Marc. Gr. 410, copied shortly after Philips times.498 This alle-
gory in Neoplatonic fashion,499 written upon the request of some students, is
staged as an educational dialogue between Philip and his students, intended to
show that the Aethiopica, albeit clad in a pagan atmosphere, are actually con-
cerned with the fight between good and evil and the contemplation ()
of God, whereby the characters Theagenes and Chariclea represent models of
sophrosyne, and Kalasiris (etymologically, ) the hierophant
who drags mankind towards the good.
By the late 12th century, Sicily and Calabria lost ground as opposed to the
rapid ascent of southern Apulia, which became a more solid hearth for
the transmission of the Greek written heritage. A paramount role was played
in this context by the hegoumenos Nicholas-Nectarius (ca. 1155/601235) at
St. Nicholas of Casole, in the region of Otranto.500 Active as a diplomat and an
interpreter, he travelled extensively in Italy and to Byzantium, wrote epigrams
and theological treatises, annotated manuscripts (most notably Par. Gr. 3 of
the Old Testament), and devoted efforts to the creation of a rich library, also
seeking to promote dialogue with the Roman Catholic Church, and to forge
a Salentine circle of Greek-speaking poets and intellectuals. One of Nicholas
pupils, Palaganus of Otranto, is the concepteur of two outstanding manu-
scripts, namely the codex unicus of Aristaenetus Letters (Vind. phil. Gr. 310)
and the oldest preserved Odyssea cum scholiis (Heid. Pal. Gr. 45, anno 1201).
The latter, the product of a team-work of eight scribes, carries an autograph
epigram by Palaganus (the son of the powerful komes Pelegrinus) and a colo-
phon written by the hand of Nicholas-Nectarius, and it displays on its margins
chiefly glossographical and mythographical material, partly drawn from the
497 Torre [2008]; Bianchi [2011] 169 (contributions by A. Corcella, C. Torre, M. Dulus,
G. Zaccagni); Cupane [1978].
498 An edition with a learned introduction in Bianchi [2006] 147. An earlier attribution of
this text to an Alexandrian philosopher of the 5th century is untenable.
499 Roilos [2005], 130133.
500 See on him Hoeck-Loenertz [1956]; Schiano [2011]; Jacob [1980] and [2008]; von
Falkenhausen [2007] 5560.
398 Pontani
4.1 Nicaea
Nicetas Choniates account of the dramatic siege and fall of Constantinople to
the Latins in 1204 (books 1719), weaves together Biblical tones and Classical
reminiscences, and makes intense appeals to Gods justice and to ancient
Nemesis, as well as incorporating hints of the disintegration of Christian
icons and the fusion of pagan statues in the Hippodrome.502 Having inspired
Edward Gibbons image of the massive destruction of the books and the sym-
bols of learning in the aftermath of the fall,503 Nicetas pages on the Fourth
Crusade serve as an excellent introduction to the long period (12041261) of
the Latin Empire of Constantinople, when the Byzantine court was exiled
in the provincial city of Nicaea, present-day Iznik.504 The need for a more
immaterial and at the same time more substantial foundation of a national
identity, as well as the growing and increasingly conflictual relations with the
Latin West, promptedin the wake of a wider trend of which more above in
3.3a further attachment to Hellenic identity, and particularly to the heri-
tage of ancient Greek language, art and culture.505
Naturally, intellectual life in Nicaea could not immediately be revived. First
of all, there was a predictable shortage of books: though no estimate can be
made, a number of texts must have gone lost in 1204,506 and the manuscript
production that can be safely assigned to the Nicaean period is extremely
501 Arnesano-Sciarra [2010] 433440; Pontani [2005b] 218225; Jacob [1988]; Irigoin [1969] 51
([1980] 248249).
502 See the new commentary by A. Pontani [2014].
503 Decline and Fall, ch. 60: To expose the arms of a people of scribes and scholars, they
affected to display a pen, an inkhorn, and a sheet of paper, without discerning that the
instruments of science and valour were alike feeble and useless in the hands of the mod-
ern Greeks.
504 Angold [1974]; Giarenis [2008].
505 Browning [1983] 124; Angold [1974] 2933; Magdalino [1991]; Flusin [2006].
506 Wilson [1980] 285.
Scholarship in the Byzantine Empire ( 5291453 ) 399
scanty, especially as far as the profane authors are concernedand in this case,
grammars, lexica and rhetorical handbooks once again prevailed, with very
little room left for literary texts.507 But the most serious problem concerned
intellectuals and scholars: some of them, starting from Nicetas Choniates and
Nicholas Mesarites, found refuge in Nicaea upon fleeing the capital, but a new
generation had to be raised and trained in order to replace them, and this was
the long-term project of two important emperors such as John Vatatzes (1222
1254) and Theodore II Lascaris (12541258).
The latter must be regarded not only as a man of rhetorical and intellectual
standing (he wrote short texts on rhetoric and annotated a copy of Aristotles
Physics and On Heavens),508 and a keen admirer of the grandeur of ancient
Greece (his epist. 80 Festa is a tribute to the ruins of Pergamon),509 but also
in the wake of his predecessor, who had founded libraries on every art and
subject in provincial towns510as a collector of books which he then left at
the disposal of interested readers or students,511 and the promoter of a school
of grammar and rhetoric in the newly restored premises of St. Tryphon in
Nicaea. The teachers in this school (we are told in Theodores epist. 217 Festa,
which also includes details about the curriculum) were a certain Andronikos
Phrankopoulos and a somewhat better known Michael Kakos Senacherim,
who in addition to attaining the grade of protasekretis in the imperial admin-
istration also devoted efforts to the exegesis of Homer, as can be proved by a
handful of scholia to the Iliad and Odyssey attributed to him in manuscripts.512
In the early 1240s, Theodore Lascaris had been the pupil of Nicephorus
Blemmydes (11971272), a theologian and an ascetic, but above all the most
important teacher and scholar of philosophy in the entire Nicaean age.513
We have considerable knowledge about his training and career from his
4.2 Southern Italy between the 13th and the 14th Century
Recent studies have stressed the role of Southern Italy in Moerbekes edu-
cation and training;525 as a matter of fact, of all the marginal areas of the
Greek-speaking world, throughout the 13th century it is Southern Italy, and
particularly Apulia, that earns pride of place. The Salentine pupils of Nicholas-
Nectarius (on whom see above 3.9) were particularly active, both as writers
and as book-collectors: one need just think of John Grassus, a literatus and an
amateur poet, who sent an Odyssey to his friend George Bardanes the met-
ropolitan of Kerkyra (the receiver was naturally intrigued by the books on
Scheria, the Homeric equivalent of his island), and later annotated a copy of
Diodorus Siculus, which had been brought from Constantinople to Otranto by
a man named Nicholas (Par. Gr. 1665).526
Despite the slow decline of the Casole monastery, book production in the
entire Salento grew conspicuously thanks to the efforts of low-brow ecclesias-
tical figures or isolated scholars,527 not only in Otranto but also in minor cen-
tres such as Gallipoli, Aradeo, Zollino, often connected to libraries or scholastic
institutions.528 Of 177 Otrantine manuscripts of the 13th and early 14th cen-
tury listed in the latest census, almost 50% are of profane character, and many
of them are written in a peculiar style, conventionally known as Baroque
minuscule.529
What is surprising is that we do not find only servile aids to students of med-
icine, law or rhetoric (though of course etymologica, lexica etc. are frequent),
or schedographic and grammatical collections incorporating philosophical
and technical definitions,530 but also ambitious copies of Homer equipped
with bulky and prestigious corpora of scholia, such as ms. Lond. Harl. 5674 (our
most important witness of the scholia to the Odyssey), Ang. Gr. 122 (a pivotal
text for the so-called h-scholia to the Iliad), Vind. phil. Gr. 49 (the only wit-
ness of Demos Homeric allegories), Vind. phil. Gr. 56 (an Odyssey probably
deriving from a copy of Nicholas-Nectarius, copied in the year 1300), and Oxon.
New College 298 (an Iliad with many exegetical and allegorical materials partly
stemming from Tzetzes and the Constantinopolitan milieu).531 An Otrantine
provenance and a date in the 13th century has also been assumed for other cru-
cial copies of ancient Greek poets, e.g. (to mention but a few) mss. Par. Gr. 2773
and Vat. Gr. 2383 of Hesiods Works and Days,532 ms. Scor. R-I-18 of Lycophron
(of 1255, with Tzetzes commentary), ms. Laur. Conv. Soppr. 152 of Sophocles (of
1282),533 ms. Vat. Gr. 1135 of Euripides;534 not to mention miscellanies of philo-
sophical (Laur. 71.35 of Porphyry and Ammonius, anno 12901291), rhetorical
(Par. Gr. 2970)535 or medical content (Marc. Gr. 273 of Dioscorides).
In fact, despite this remarkable production of books, no real indigenous
philological activity or advanced scholarship can be assumed for Southern
Italy either before or after the reigns of Frederick II Hohenstaufen and
Manfredboth emperors, for that matter, proved attentive to Greek, the
527 Arnesano [2008] 1315; Jacob [1980] and [1987]; Luc [2012a] 590593.
528 On libraries see Wilson [1980] 295299; on schools see Arnesano-Sciarra [2010] 440454;
Efthymiadis [2005] 274275.
529 Arnesano [2008] 73122.
530 Frstel [20022003].
531 Sciarra [2005b]; Cavallo [1989]; Pontani [2005b] 203241.
532 Arnesano [2005] 143145; Turyn [1964] 7173.
533 Arnesano [1999].
534 Irigoin [1982].
535 Arnesano [2011].
Scholarship in the Byzantine Empire ( 5291453 ) 403
536 Irigoin [2006a] 132136; McCabe [2007] 239244 with earlier bibliography.
537 Weiss [1950]; Cavallo [1980] 233235.
538 Berschin [1980] 314317; Ciccolella [2008] 9297; Dorandi [2013b].
539 The expression is by Theodore Metochites (Mineva [19941995] 324, concerning Nicaea:
). A concrete example might be found in the most
important manuscript of the Greek novelists (Laur. Conv. soppr. 627), written in the 1260s
or 70s, but carrying also some letters of Theodore II Lascaris: Wilson [19962] 225; Bianchi
[2002] 183184.
404 Pontani
and population; indeed the very quarrel around the union of the Churches had
profound and unfortunate consequences on the Byzantine intelligentsija and
on the teaching system in Constantinople.540
Despite this political shakiness, the reigns of Michael VIII and Andronicus II
were a golden age for scholarship (and classical scholarship in particular), and
a period to which we owe an immense debt for our knowledge of Greek pagan
literature, both in matters of quantity and of quality. In quantitative terms,
the comparison between the sheer number of manuscripts of any classical
author dating before 1204 and after 1261, helps to understand the extent of the
change that came about;541 this process went hand in hand with the recovery
of many books previously scattered far from the capital,542 and with the rise
of a new public for books and book-collecting, also thanks to the increasing
popularity of bombycine and then particularly of (cheaper) Italian paper.543
In qualitative terms, the scholars of the Palaeologan age, from Planudes to
Moschopoulos, from Pachymeres to Triclinius, produced outstanding editions
and commentaries of tragedy and comedy, of Pindar, of hexametric poetry,
of Plutarch, Ptolemy and Strabo, of Plato and Aristotle and Proclus...: for all
these texts, and many others, the Palaeologan age can be said to represent a
vital turning-point in the history of their tradition.544
This is by and large what is commonly understood under the label
Palaeologan renaissance (or revival), a phenomenon that originated in the
reawakening of learning in and around the imperial and patriarchal milieux,
and involving not only the capital, but also spreading to Thessalonica (the
second city of the empire), and extending to more peripheral centres such as
Ephesus, Trebizond, Cyprus, Crete, and later the Peloponnese. Once more, it
should be stressed that we are not dealing with a sudden revolution, but rather
with the intensification of a contact with texts (and particularly ancient texts)
that had never entirely disappeared from the horizon of Byzantine learned
elites (those that met regularly within active scholarly circles and literary
practicing the typically Byzantine art of transcribing excerpts from the texts
of his interest.555 Chosen on account of their linguistic features, proverbial
nature or antiquarian importance, these excerpts (drawn from such diverse
authors as Homer and Sophocles, Philo and Synesius) reveal a series of inter-
esting variant readings, which must be ascribed either to his conjectural skill
or to the use of better manuscripts than the ones extant today556corrections
and annotations by Gregory of Cyprus have been spotted in the margins of
manuscripts of Demosthenes (Par. Gr. 2998), Plato (Scor. y.I.13), Proclus (Marc.
Gr. 194)557 and Aelius Aristides (Par. Gr. 2953, a very influential hyparchetype
in the tradition of this author, probably deriving from Acropolites copy).558 An
outstanding teacher and a prolific excerptor, George/Gregory of Cyprus tried
his skill at rhetorical declamations, progymnasmata and paraphrases (e.g. of
Aesops fables),559 and he became one of the most distinguished paremiog-
raphers of the Byzantine age, setting up an epitome of the old collection of
Diogenianus.560 His intellectual physiognomy, however, will become clearer
only once his theological oeuvre is taken into account, including some inter-
esting linguistic remarks e.g. on the values of the prepositions and in
Gregory of Nyssa, related to the widespread 12th-century debates on the pro-
cession of the Holy Spirit.561
Acropolites other pupil, John Pediasimus Pothos (ca. 12501310/14), prob-
ably trained first in Thessalonica and later in the capital, became a deacon and
was first appointed hypatos ton philosophon in the 1270s, was later promoted to
the rank of chartophylax of Ochrid (where he also taught), and finally in 1284
megas sakellarios again in Thessalonica.562 Not a particularly original scholar,
, he
was poor and absolutely fond of books: his hands were relatively skilful at writing, and
since he could not buy his favourite works with money, he bought them through his own
toil and he became a copyist of more numerous books than any other lover of culture ever
copied.
555 Kotzabassi [2010].
556 Prez Martn [1996].
557 Menchelli [2010].
558 Constantinides [1982] 145146 and 153; Prez Martn [forthcoming].
559 Kotzabassi [1993]. On the teaching of rhetoric in the early Palaeologan age, see
Constantinides [1982] 153155.
560 Prez Martn [1996] 313319.
561 Antirrhetica 59: see Larchet [2012] 240; on the wider issue see Bucossi [2009].
562 Constantinides [1982] 116125; Mergiali [1996] 2123; Wilson [19962] 242; Bianconi [2005a]
6072, who also refers to a catalogue of books (in ms. Vat. Gr. 64) that may have something
to do with Pediasimos.
Scholarship in the Byzantine Empire ( 5291453 ) 407
Parmenides (ms. Par. Gr. 1810) represents a true edition of the text, full of con-
jectures, corrections of philosophical terms and ideas, and equipped with a
long supplement which he designed suo Marte.573 Pachymeres was fully con-
vinced, against the ideas of patriarch Athanasius I, that philosophy is impor-
tant for man,574
,
,
, .
for its aspect is similar to Hermes moly, which is difficult to extract for
the many and thus provokes their hatred, so that the sweet appears as
bitter, because of the negligence of the simple men, those whom Circe
destroyed by means of a drink.
The exact place of Pachymeres in the patriarchal school (the sources credit
him with the titles of dikaiophylax and megas didaskalos) remains unclear, as
does the distinction and hierarchy between the offices of didaskalos ton didas-
kalon and katholikos didaskalos. However, the deep links between the patri-
archal school and the imperial milieu are demonstrated by the fact that the
first teacher of logic and rhetoric (appointed in 1265 at the Church of the Holy
Apostles, where generations of professors had taught) was no other than the
imperial secretary Manuel-Maximus Holobolus (ca. 12451310/14), a monk,
poet and scholar whose adventurous public life earned him imprisonment,
mutilation and rehabilitation (as a fierce opposer of the union with the Latin
Church, he was persecuted by his former patron Michael VIII).575 Holobolus,
who became in his youth a rhetor ton rhetoron and later an oikoumenikos didas-
kalos, had experienced the difficult cultural situation of Nicaea, and after the
recovery he successfully prompted Michael VIII to implement the teaching
of grammar, poetry and rhetoric. He spent his career teaching and devoting
his time to the study and exegesis of Aristotle (he made an overarching para-
phrase of the Stagirites works, as well as commentaries on Physics and Prior
573 Steel-Mac [2006]. Ed. Westerink et al. [1989]. Fryde [2000] 206208 on Pachymeres
Platonic studies.
574 This part of the proem is edited by Golitsis [2009] 213, who also detects the reference not
to Homer but to his most important exegete: Eustathius, Commentary on the Iliad 10.1719.
575 Constantinides [1982] 5259. Prez Martn [1995] 414417.
Scholarship in the Byzantine Empire ( 5291453 ) 409
, ...
, ,
.
582 Constantinides [1982] 6687; Wilson [19962] 230241; Fryde [2000] 226267; Mergiali
[1996] 3442; Wendel [1950] is still very useful.
583 Fuchs [1926] 5962; Constantinides [1982] 6871; see his epist. 23 Leone on the rough
selection of his pupils, amongst whom Manuel Moschopoulos, George Lacapenus and
the Zarides brothers.
584 Fryde [2000] 157158; Constantinides [1982] 72 and n. 33; Martin [1956] 295299.
585 Prez Martn [2006]; Wilson [19962] 232233.
586 For I want to collate it with my own (epist. 33, p. 66.15 Leone
).
587 epist. 67 (to Mouzalon), p. 99.2429 Leone.
Scholarship in the Byzantine Empire ( 5291453 ) 411
588 Mittenhuber [2009]; Pontani [2010a]; Burri [2013]. For other instances of the direct deriva-
tion of Palaeologan manuscripts from late antique prototypes see Fryde [2000] 153155.
589 See Fryde [2000] 237241 (with further bibliography) and several essays collected in Diller
[1983]; specifically on the text of Pausanias (whose archetype was probably prepared by
Planudes) see Diller [1980] 489491.
590 E.g. Laur. 59.30: Fryde [2000] 248253.
591 E.g. Vat. Gr. 191 carrying technical and scientific authors: see Bianconi [2004b] 324333.
592 Quattrocelli [2009]. Gaul [2011] 174181.
593 The former was the antigraphon of Ficinos Laur. 85.9: see Bianconi [2008b].
594 dAcunto [1995].
412 Pontani
, ,
, .
This passage is very obscure because the text of the old copies, worn away
in many places, does not yield a continuous and consistent sense: I have
seen an old book with many blank spaces, left by the scribe for he was
clearly unable to find the missing words, but hoped perhaps to find them
elsewhere. Here, on the other hand, in place of what was missing one
finds continuous writing, because there was no hope of finding the miss-
ing parts.
595 epist. 106, p. 169.1819 Leone , , (as you know, I like him
very much).
596 Rollo [2008b], with further bibliography. See Stadter [1973] (and A. Pontani [1995b]
9293) on the manuscripts fate in Quattrocento Italy.
597 Bianconi [2011b].
598 Devreesse [1954] 9091. Wilson [19962] 236.
599 See Constantinides [1982] 136 on this and other instances of scholars facing problems
with the parchment supply.
Scholarship in the Byzantine Empire ( 5291453 ) 413
useful venues for taking more sporadic notes on selected works (exemplaires
de travail).600 The former category embraces some other famous codices of
prose (e.g. the aforementioned Vat. Gr. 191) and above all illustrious verse codi-
ces, which it is now time to examine.
Planudes achievements in the domain of poetry are most impressive,
although he did not deal in depth with either lyric or dramatic poetry. As a
matter of fact, scholia to Hesiod, Pindar and some scenic poets might stem
from his pen,601 and his milieu probably produced selections of teaching
textspartly equipped with scholiasuch as Vat. Gr. 915 (Pindar, Lycophron,
Homer, Theognis etc.).602 But Planudes major feats of scholarship lie else-
where: ms. Laur. 32.16, the product of six scribes working under his supervision
as early as 128083,603 is a voluminous anthology of hexametric poetry ranging
from Nonnus Dionysiaca (for which work it is our codex unicus) to Theocritus
and Apollonius Rhodius (for both texts it is an independent witness of the
utmost importance),604 from Hesiods Theogony (of which it offers our earliest
preserved complete copy), down to Oppian, Moschus, Nicander, Tryphiodorus
and some autobiographical poems of Gregory of Nazianzus.
Perhaps Planudes most remarkable achievement is his outstanding collec-
tion of Greek epigrams, digested by subject in 7 books605 in Marc. Gr. 481, writ-
ten in 1299 or 1301.606 This codex embraces both Nonnus Paraphrase of the
Gospel of St. John and a rich selection of Cephalas sylloge (see above 2.6),
including 388 epigrams lacking in the Heidelberg manuscript and now mak-
ing up the Appendix Planudea (or book 16) in modern editions of the Greek
Anthology. Planudes enthusiasm for this genre, as testified by his scribal care
and conjectures, as well as by other copies of the same collection (he slightly
later supervised the realisation of ms. Lond. Addit. 16409, a fair copy of the
Marcianus), did not prevent him from mutilating or bowdlerising the more
, ,
.
,
.
We should note that the reading of Hellenic literature has always been
an object of longing and delight for lovers of learning, and particularly
the reading of the poems of Homer, because of the grace and variety of
the language. That is why the present metrical paraphrase has been writ-
ten in heroic metre, to give pleasure to lovers of learning and literature.
[transl. R. Browning]
One genre where Planudes was indeed sometimes obliged to resort to censor-
ship were his translations of Ovids amatory poems. In fact, relying on a solid
though not impeccable knowledge of Latin610 acquired perhaps through his
familiarity with the Westerners living next door to the Akataleptos monastery,
and refreshed during his diplomatic mission to Venice in 12961297, Planudes
was the first Byzantine to devote a systematic effort to the Hellenisation of
Roman masterpieces, from Christian (Augustines On Trinity and Boethius
Consolation of Philosophy) to philosophical works (Ciceros Dream of Scipio
with Macrobius Commentary), from Ovidian poetry (including the Heroides
607 On censorship in general throughout the Byzantine millennium see Wilson [19962] 1218;
A. Pontani [1995a] 322327.
608 Karla [2006]; Valerio [2011].
609 Browning [1995] 21; Fryde [2000] 10; De Stefani [2002] 44.
610 See Bianconi [2004a] 554564 on various categories of his mistakes in the translations,
such as for perperamwrongly spelled per peramin Aug. de trini-
tate 13.5.8.
Scholarship in the Byzantine Empire ( 5291453 ) 415
621 Constantinides [1982] 103108; Mergiali [1996] 4952; Fryde [2000] 295298; Wilson
[19962] 244247; Gaul [2008] 169171 argues for an earlier date of death, shortly after his
teacher Planudes.
622 Constantinides [1982] 141; Browning [1960] 13. Among Nicephorus books was the Plato
mentioned above 4.4 and the Odyssey Caes. Malat. D.XXVII.2 (Pontani [2005b] 297300).
623 Ed. Titze [1822] 1743. See Hunger [1978] 14; P. Ippolito [1981]; Constantinides [1982] 105
106; Mergiali [1996] 5052; Fryde [2000] 219221.
624 Webb [1994]; Keaney [1971] 303313; Gaul [2011] 305307. Moschopoulos Erotemata
and enjoyed great popularity throughout the Byzantine period and the
Renaissance. The epimerisms to prose authors created by an otherwise unknown
Staphidakes are discussed by Gaul [2008] 191194.
625 See Lindstam [1925], and Gaul [2011] 181183 for the success of Philostratus in Palaeologan
Byzantium.
626 Webb [1997]; Webb [1994] 8591.
627 Cengarle [1970] and [1971]; Hunger [1978] 32.
Scholarship in the Byzantine Empire ( 5291453 ) 417
]
, ,
, ,
, ,
, ,
. ,
.
hid the fire] Zeus had kept the fire hidden until then: Prometheus, the good
son of Iapetos, stole it from Zeus the decision-maker in a hollow reed, for
the sake of the mortals, escaping the notice of Zeus the thunder-bearer
(terpikeraunos), whose namethey sayderives from the verb trepein
(to turn) through metathesis of the rho, not from the verb terpesthai
(to be delighted): this means that Zeus puts to rout the enemies with
his thunder rather than delighting himself with it. Then he adds a sort of
ethopoeia, imagining which words Zeus would speak to Prometheus in
anger due to the theft of fire.
evident success throughout the Palaeologan age, and was often copied up to
the 1330s.634 Moschopulos inclination to anthologies and chrestomathies in
general also led him to the creation of the so-called Anthologie des quatre
(excerpts from Philostratus Images, Marcus Aurelius, Aelians Natural History,
and the so-called Sylloge Vaticana of the epigrams), which was designed as a
textbook for the intermediate level between the primary learning of grammar
and the exegesis of more advanced poetical texts. This collection, which still
awaits closer study (as do other products of the Palaeologan anthologising
fashion),635 marked the first instance of prose texts being integrated in a fixed
school curriculum.636
While the real extent of his philological contributions to the establishment
of the texts of Attic drama is still hotly debated in the present day (especially in
the case of Sophocles, an author he edited around 1290), it can be affirmed that
Moschopoulos does not stand out as a first-rate textual critic, although he was
definitely well acquainted with the iambic metre, and certainly inserted metri-
cal conjectures in his recensions. Some of his good readings, however, may in
fact derive from deliberate editorial choices rather than from the inspection
of better sources. Be that as it may, Moschopoulos editions certainly enjoyed
a great success among Byzantine schoolmen and scholars, and were by far the
most widespread ones before Triclinius.
Moschopoulos owed his success also to his activity as a grammarian and
schedographer. In fact, the fashion of elementary grammatical exercices,
epimerisms, and schedography was so common in Palaeologan Byzantium
that a man like George Lacapenus,637 a teacher of grammar and rhetoric living
in the capital, even applied this method to a wide selection of Libanius letters,
and later to his own letters exchanged with Michael Gabras, with Planudes
pupil John Zarides and with other learned friends over a considerable span
of years (12971315): these epimerisms were later arranged in alphabetical
order.638
644 Constantinides [1982] 108109; Fryde [2000] 208210; Mergiali [1996] 8587; Wilson
[19962] 243244.
645 See on this Gielen [2011] and above all the excellent overview in Gielen [2013].
646 Searby-Sjrs [2011]; Pontani [2010b] 2336; Constantinides [1982] 125126.
647 References in Reinsch [1974] 1722.
648 Mergiali [1996] 99102; Reinsch [1974]; Browning [1992a].
649 epist. 20, p. 115.1130 Reinsch (Nov. 1326), with Reinsch [1974] 1116.
650 Browning [1992a].
651 Ed. Matranga [1850] II.52024; see Reinsch [1974] 6675.
Scholarship in the Byzantine Empire ( 5291453 ) 421
Sirens as two different forms of pleasure; Scylla and Charybdis as moral and
physical sin etc.).652 These works, albeit no outstanding feats of scholarship, all
attest to a lively enthusiasm for ancient poetry, substantiated by a remarkable
erudition and familiarity with old exegetical and allegorical works; however,
this fondness is never devoid of a sense of guilt for the pleasure of reading a
poetry that he qualifies as , sordid, rotten verse.653
Finally, another learned teacher and politician who devoted his time to lit-
erate instruction was John Glykys, a former pupil of George of Cyprus, and an
imperial officer who later became patriarch of Constantinople in 13151319.654
Glykys is the author of a treatise entirely devoted to syntax,655 which abandons
the ambition of a general overview and opts, instead, to provide solutions to
particular problems (zetemata), above all the nominal cases, the uses of the
participle and the issue of solecism and barbarismthe latter a very popular
theme in Byzantine grammars but also in smaller independent, mostly anony-
mous notes or treatises to be found scattered in Byzantine manuscripts.656
, , ,
... , ,
, , ,
658 Gaul [2011] 215219; Bianconi [2005a] 1931 and 5160; Katsaros [1997]; Laourdas [1960].
659 Bianconi [2005a] 7290; Fryde [2000] 299301; Mergiali [1996] 5455; Wilson [19962] 247
249; Gaul [2011], esp. 220240.
660 Ritschl [1832].
661 Gaul [2007] 296326.
662 Gaul [2008] 184190.
663 Lenz [1963].
664 Gaul [2011] 121163.
665 An analytical catalogue of the manuscripts and the modern editions of Magistros scho-
lia to the scenic poets (mostly, but not always, coinciding with that of Triclinius and
Moschopoulos) is provided by Gaul [2011] 387401.
Scholarship in the Byzantine Empire ( 5291453 ) 423
vided the triads of Aeschylus, Euripides, Aristophanes and Sophocles (the lat-
ter with the addition of the Antigone) with important introductory texts, such
as biographies and summaries, and with commentaries largely indebted to
earlier scholia (the same is true for Pindars Olympian Odes and, less certainly,
Pythian 14).666 In most of these cases Magistros did not constitute genuinely
new texts of the poets involved: rather, he added his own exegesis and some
emendations to the vulgate text of Planudes and Moschopoulos,667 at times
displaying great tolerance for variae lectiones of the same passage and their
multiple meanings.668 This partially explains why it is sometimes very hard to
disentangle in the manuscripts the exact attribution to the different Byzantine
scholars, and to sort out the contributions of each one to the constitution of
the text.669 However, as far as Magistros notes are concerned, the obvious mor-
phological paraphernalia and a rather uncertain command of metre do not
obscure the attention he dedicated to mythological and stylistic features, as
well as his pride in distinguishing his own approach from that of his colleagues.
The analysis of the scholia to the Oedipus rex (the only ones well edited in
modern timesin addition, naturally, to the Aristophanes scholia)670 shows
a much more wide-ranging and ambitious approach than Moschopoulos.671 A
good example is provided by the ambitious note on Aristophanes Wealth, post
626 :672
,
,
.
,
, ,
.
Note here that, although it would be necessary to insert a choral song until
they arrive at the temple of Asclepius and can see Ploutos once again, the
poet immediately represents Karion as he brings the old men the good
news of Ploutos recovery. He does so not without reason, both follow-
ing the habits of New Comedy, in which the parabasis had disappeared
(as we mentioned above), and wishing to show that Ploutos recovered
his eyesight very quickly.
The Thessalonican scholar Demetrius Triclinius, who may have been a pupil
of Magistros and must have had some contact with the school of Planudes,
was the most famous philological genius of the early 14th century,673 and
probablywith a breadth of vision well above the average of the literary
coteriesthe only one to conceive of Classical studies not exclusively as sub-
servient to rhetorical or stylistic aims.674 The acme of his activity must be
placed in 13051320, yet despite the great renown associated with his learning
his biography remains very imperfectly known, and no information is available
with regard to his profession (where was he trained? did he become a monk?
did he ever teach in a school?). What can be reconstructed is an unceasing
commitment to the study of Classical texts (poetry in particular), practiced
in the wake of his fellow countryman Magistros, probably in the same erudite
circles of Thessalonica. More generally, the number and quality of the scribes
involved in the copying of manuscripts, as well as the constant dialogue or
overlap between the philological activities of the various leading scholars of
this age, point to the existence of a cercle dcriture having immediate connec-
tions with the capital.675
As opposed to his other colleagues and predecessors, Triclinius devoted
most of his own efforts to textual criticism, above all concerning dramatic
and lyric texts, which he emended and assessed by relying on the collation
of several, often old and forgotten manuscripts, and on a firm knowledge of
metre, ranging from the more obvious iambic, trochaic, dactylic and anapaes-
tic sequences, to the more elaborate responsions of lyric strophae and antis-
trophae. This resulted not only in a special ability to make sense in metrical
terms of some of the most difficult lyrical sequences in Greek tragic choral
673 Bianconi [2005a] 91118; Fryde [2000] 268292; Wilson [19962] 249255; Mergiali [1996]
5557.
674 Gaul [2008] 163.
675 Bianconi [2005a] 9296.
Scholarship in the Byzantine Empire ( 5291453 ) 425
676 Basta Donzelli [1994] on Laur. 32.2; De Faveri [2002], to be read with Magnani [2004];
Tessier [1999].
677 Lamagna [1996]; Wilson [19962] 252; Smith [19811982] and [1992].
678 Tessier [1999] 4449; Gnther [1998] 61166.
679 Gnther [1998] 167185; Irigoin [1952] 331364; Bianconi [2005a] 105.
680 Koster [1957]; Wilson [1962] and [19962] 251253.
681 Tessier [2005] xviixix; Aubreton [1949] 2941; Bianconi [2005a] 100104, a fundamental
overview of the mss., adding Triclinius notes to the Ajax in ms. Par. Gr. 2884.
682 Smith [1975] 1 and 3440; Turyn [1943] 100116; Dawe [1964] 5964; Fryde [2000] 270.
426 Pontani
683 Ms. Laur. 32.2 (on which see Zuntz [1965] 128134; Browning [1960] 15) also included
Sophocles, some tragedies of Aeschylus, Hesiods Works and Days and the Theocritean
section of Par. Gr. 2722, partly written by Planudes.
684 Turyn [1957] 222258 and 2352 (with Bianconi [2005a] 119122: the scribe of the
Angelicanus is the same of Par. Suppl. Gr. 463 of Aristophanes); Zuntz [1965] 136140;
Magnani [2000] 2951; De Faveri [2002] for the edition of the metrical scholia.
685 Derenzini [1979].
686 Gallavotti [1982]; Bianconi [2004b] 343344; Bianconi [2005a] 39 and 99, with further
bibliography.
687 Turyn [1957] 250252.
688 Derenzini [1984]; Bianconi [2005a] 124126. This is of course relevant for our knowledge
of the relationship between Planudes and Triclinius: see Wilson [1978].
Scholarship in the Byzantine Empire ( 5291453 ) 427
Vat. Gr. 83, and then in a series of newly produced codices, he was helped by
the scribe Nicholas Triclines, who often emerges as his collaborator, and dis-
plays a lively copying activity, for instance in ms. Laur. 70.6 (Herodotus) and in
the margins of Ang. Gr. 83.689
The scholarship of Triclinius is not easy to reconstruct in detail, consisting
as it does in a vast proliferation of manuscripts, often produced by a network
of well-trained scribes.690 His work was aided by unceasing progress in the
knowledge of ancient metre, but also prompted by a divine and secret inspira-
tion ( ) that urged him to textual emendation and
conjecture,691 though his sound method led him to declare regularly the exact
provenance of the collected or concocted scholia (e.g. by prefixing a cross to
Moschopoulos notes, a capital letter to Magistros, and the word to
his own).692 Less frequently, he also showed an interest in other exegetical
approaches, including allegory (an astronomical reading of Iliad 4.14 recalls
the Neoplatonic tradition stretching from Porphyry to Michael Psellus).693
Despite the respect it commands, Triclinius activity has received contrast-
ing assessments.694 What is certain is that: a) it was not merely the fruit of
an isolated genius, but it involved an erudite circle keen on editions of poetry
and prose (from rhetoric to historiography, from grammar to philosophy
and science);695 b) it became of paramount importance both in the field of
Byzantine teaching696 and as the touchstone for the tradition of many Classical
texts in Italian humanism and later.697 This partly justifies the disparaging atti-
tude by which Triclinius, in the prolegomena to his Aeschylus edition, justifies
his interest for lyric metres of Greek drama:
, ,
[Orph. fr. 1a =
101 Bernab] ... ,
, ,
689 Bianconi [2005a] 106107 and 124141; on Libanius particularly Bianconi [2005c].
690 Bianconi [2005a] 102182.
691 Smith [1975] 257. Tessier [2005] x.
692 Smith [1975] 36.
693 The note has been transmitted by Angelo Poliziano: Maer [1954].
694 A negative tone e.g. in Diggle [1991] 99101; Zuntz [1965] 194197.
695 Bianconi [2005a] 178182.
696 Bianconi [2010b] 494498.
697 Fryde [2000] 289290.
428 Pontani
, .
For not even those who invented these excellent things published them
for such (ignorant) people, but for the intelligent ones, taking little care of
the dull people, as someone once said I sing to the intelligent: close the
doors you laymen.... Therefore, in order not to give the impression that
I damage the knowledgeable, I offered them what has come to my mind,
and they will realise if it is correct or not, because what I have excogitated
on the choral songs and the other genres contained in the plays, working
a lot through a divine and secret inspiration on these texts and on their
metres, all this has been expounded in the plays.698
703 Fryde [2000] 322336; Mergiali [1996] 6067; Wilson [19962] 256264; evenko [1962]
and [1975]. On Metochites and Chora see Underwood [19661975]; Teteriatnikov [1996].
704 Misc. 93.3.1 Agapitos .
705 Misc. 93.1.1 Agapitos
, , .
706 Agapitos (et al.) [1996].
707 Hunger [1952]; Beck [1952] 75; Wilson [19962] 264; Bazzani [2006]; Bianconi [2008a].
708 Featherstone [2012].
709 Misc. 1.2, in Hult [2002] , , ,
.
710 Wilson [19962] 258259.
430 Pontani
,
,
The Chora library and scriptorium718 thus ended up being a repository of illus-
trious manuscripts (e.g. the glorious Clarkianus of Plato, but also Vat. Gr. 130
of Diodorus Siculus),719 and a sort of scholarly circle, a reading-group where
the activity of preserving, copying and annotating could become a collec-
tive enterprise, as is shown by the fact that often several scribes (not neces-
sarily professional scribes) were engaged in the production, restoration, or
philological study of one single book. Recent studies have made considerable
headway towards the identification of some of these copyists: for instance
the so-called Metochitesschreiber, who worked in close connection with the
master720 and produced, besides copies of Metochites own works, such land-
marks as the codex Crippsianus (Burney 95) of the minor Attic orators, or the
Aelius Aristides Vat. Urb. Gr. 123, or the monumental Aristotle Par. Coisl. 157,721
has been identified with Michael Clostomalles, an imperial notary who wrote
many official documents throughout the first decades of the 14th century.722
But the most important figure in this context is Metochites pupil and cul-
tural heir Nicephorus Gregoras (ca. 12961361), a teacher, an intellectual and
a multifaceted scholar, capable of writing a fundamental historiographi-
cal work on his own times, of annotating and editing Classical authors, but
also of calculating the exact length of the astronomical year centuries before
the Gregorian reform.723 No less convinced than his master Metochites
to the edition of whose writings he devoted painstaking effortsthat not
only for us Christians do all good things belong to God, but also for the most
learned of the ancient Hellenic writers,724 his passion for science led him to
717 Carm. 1.115357 Treu (transl. Featherstone [2010], adapted). See also Browning [1960] 13.
718 Bianconi [2003] 541543; Bianconi [2005b]; Prez Martin [1997a]; Frstel [2011].
719 Mazzucchi [1994].
720 Wilson [19962] 229.
721 Prato [1994] 123131.
722 Lamberz [2006] 4448.
723 Blachakos [2008]; Mergiali [1996] 6378, with special attention to his astronomical work;
Wilson [19962] 266269; Fryde [2000] 357373; Guilland [1926]; evenko [1962]; Fuchs
[1926] 6265 on his teaching activity.
724 Epist. 4.16164 Leone
... . The whole letter is remarkable for its philo-
logical speculation on Aristides and Homer.
432 Pontani
,
,
, .
, ,
,
.
,
,
, ,
.
,
. ,
.
, .
Palamas took this sentence from here, but he did not believe it carried a
correct meaning as it stood: corrupting and partly altering its wording,
he transformed radically the whole sense. He omitted altogether one syl-
lable from the holy writ, he changed another syllable and substituted a
third one, for in his own, illegal copy instead of cooperation (synergia)
he wrote act (energeia), an etymologically related word, by no means
uncommon for his deceitful games. The original text of the old book went
the indivisible and unmixable substance, invoking the cooperation of
the other substance, and it meant the indivisible and umixable sub-
stance of Christ (namely his divinity), invoking the cooperation of the
other substance (namely of humanity) performed the miracles. But this
shameless man, stealing away one article (ten) and concealing it in the
abyss of silence, impiously and incorrectly assigned the other article to
the two cola, as you (and whoever else wishes) can see. Thus, deleting the
following word, he brought an impious confusion into the holy sentence,
as is customary for him: but now, since we have cleaned the holy sentence
of the cursed blasphemy, let us restore the pious meaning.
room left for the study of Hellenic literature and culture suddenly became very
narrow.739
This is not to say that hesychasts were necessarily more hostile to pagan
learning than their earlier counterparts: according to Gregory Palamas (1296
1359), the coryphaeus and leader of this spiritual movement,
,
,
just as we can gain some utility even from snakes, provided we kill, dis-
sect, prepare and use them reasonably against their very bites, so their
(scil. the pagans) doctrine is inasmuch useful to us as we may use it in
our fight against them.740
This implies that Hellenic wisdom could still circulate in schools, but had
to be considered as purely subservient to the right religious instruction,741
and by no means the basis for a philosophical development comparable to
the Western scholasticism that had grown out of Aristotles works.742 Hence
the slow but inexorable marginalisation of the non-aligned scholars, starting
from Nicephorus Gregoras himself, who was satirised by patriarch Philotheus
Coccinus (13001379) as a sort of Sophoclean Ajax failing to realise that the
pagan poets and philosophers were actually dead,743 down to those schol-
ars who eventually converted to Catholicism, first and foremost Demetrius
Cydones.
There were partial exceptions: Theodore Meliteniotes ( 1393), didaskalos
ton didaskalon in 13601388 and later archdeacon of the Patriarchate, wrote
a commentary on the Gospels but also Three Books of Astronomy which
attempted, decades after Metochites, to reconcile the Hellenic tradition going
back to Ptolemy and Theon with the new results of Arab science.744 It is inter-
esting to note that Meliteniotes was the owner and annotator of an important
Iliad (now ms. Genav. Gr. 44):745 this seals a link between astronomical and
Homeric studies which occurs with other contemporary figures as well, e.g.
George Chrysococcas the Elder (the scribe of Vat. Pal. Gr. 7 of the Odyssey, and
a proficient student of Persian astronomy at Trebizond)746 and Constantine
Lucites (a pupil of Theodore Hyrtacenus, and the owner of the Iliad Ambr. I
58 sup.).747 More broadly, although the scholarly interest in science during
the later Palaeologan era had deep roots in the polymathy of some outstanding
scholars of the Planudean age,748 it was also a very important phenomenon
as far as the transmission and emendation of prose texts in manuscripts is
concerned.749 Another example is provided by the astronomer Joseph
Bryennios, who had been appointed didaskalos ton didaskalon or didaskalos tes
ekklesias since 1390: an expert in Latin language, grammar, and exact sciences,
he spent many years of his life in diplomatic missions to Crete and Cyprus, but
was above all a faithful theologian of the Patriarchate, as well as a book collec-
tor and an enthusiastic reader of Marcus Aurelius Meditations.750
But on the whole the second half of the 14th century was dominated by
other concerns, above all by the civil and external wars, the dramatic effects
of the plague (13481362), and the increasing need for a serious interchange of
ideas with the Latin West, connected with the tightening of commercial links,
the progressive sense of danger that was seizing the empire (after the unsuc-
cessful Ottoman siege of Constantinople in 1394, emperor Manuel II addressed
a petition to Western rulers), and alsoon the cultural levelthe spreading
of Planudes translations of the masterpieces of Latin literature.
A key moment in this process was the attempt to transplant on Byzantine
soil the scholastic philosophy of Thomas Aquinas: this challenge was pursued
by Demetrius Cydones (1324/251397/98),751 an intellectual and politician who
served as mesazon to emperors John VI Cantacuzenus and John V Palaeologus,
and became a tutor and later a correspondent of the learned Manuel II, an
emperor curiously capable of showing enthusiasm for a manuscript of Plato
or Suidas.752 The author inter alia of a very large epistolary (451 letters) in pure
745 The scholia are edited by Nicole [1891]; on the margins of the Genavensis we also find
Moschopoulos notes: see Browning [1960] 1415.
746 Fryde [2000] 350; Lampsides [1938]; Pontani [2005b] 329330.
747 Constantinides [1982] 9395 and 142.
748 Fryde [2000] 341343.
749 Mondrain [2012].
750 Rees [2000]; Mergiali [1996] 156160.
751 Ryder [2010], esp. 537; Mergiali [1996] 113130; Fryde [2000] 381386.
752 Mergiali [1996] 123124, with references.
Scholarship in the Byzantine Empire ( 5291453 ) 437
,
,
I had barely one book from which I had to translate, so that it was not
easy to discover a corruption in the text, or to correct it, since there
was no other copy to collate with that one text in order to seek the true
reading.756
The hesychast movement caused a split in the cultural elite of the late 14th
century: the anti-Palamites included Cydones pupil Manuel Calecas (ca. 1360
1410), who in 1390 opened a school of enkyklios paideia in Constantinople, and
wrote a grammatical work for his teaching activity,757 although the compara-
tively small success of this school was due partly to the difficult political and
economical moment (he had to flee the capital in 1396 upon converting to the
Roman faith), and partly to Calecas own problematic character.758 On the side
of the admirers of Gregory Palamas one finds a pupil of Thomas Magistros,
based in Thessalonica during the third quarter of the 14th century, namely
the aforementioned Philotheus Coccinus, later patriarch of Constantinople.759
In addition to writing long essaysas we have seenagainst Gregoras and
pagan wisdom, Coccinus wrote a panegyric for Gregory Palamas phrased in a
pure Hellenic style and language, and insisted that Classical learning should be
used only as a rhetorical frame and as a help in reading the Scriptures, not for
its own sake, much less in the search for truth.760
Perhaps the most fervent supporter of Palamas, and the most implacable
opponent of Gregoras, the Calabrian monk named Barlaam (1348), spent most
of his life as a scholar and a teacher at the Soter monastery (a centre de copie of
ecclesiastical as well as of Classical authors),761 before being appointed bishop
of Gerace in 1340.762 Barlaams theological works need not detain our attention
here, even if they reveal a close acquaintance with Plato and Aristotle; what is
more interesting is the emblematic role of Barlaam as one of the last important
figures of Greek culture in Southern Italy during its slow but inexorable decline
under the Anjou and Aragonese domination.763 Several decades later, the trav-
eller Athanasios Chalkeopoulos was to describe the sad state of decay of both
libraries and linguistic knowledge in Southern Italy.764
Barlaams successor to the see of Gerace in 1348 was another man of
letters, later appointed to the diocese of Thebes (13661383), called Simon
Atumanus:765 as well as being a close friend of Cydones, and the first transla-
tor of Plutarch in the West (On controlling anger; he may also be the translator
into Greek of the Hebrew Old Testament in ms. Marc. Gr. 7),766 he owned and
partly annotated several important manuscripts that were mentioned earlier,
such as the Planudean Plato Vind. phil. Gr. 21, the Salentine Odyssey Vind. phil.
Gr. 56 and the main witness of the alphabetical series of Euripides tragedies,
Laur. 32.2.
758 On Calecas see Mergiali [1996] 134135 and 163164; Loenertz [1947].
759 Bianconi [2005a] 227237.
760 Life of St. Sabas 5, in Tsames [1985] 168169.
761 Bianconi [2005a] 166167 and [2004b] 350351.
762 Fyrigos [2001]; Bianconi [2008a] 456458.
763 See Fyrigos [1997], andmore specifically on manuscriptsLuc [2006]; Perria [1999];
Cavallo [1982] 581591.
764 Laurent-Guillou [1960]; Luc [2007b] 4546.
765 See on him Fedalto [1968]; Wilson [19962] 268269; Mergiali [1996] 140141; Pertusi [1960].
766 Weiss [1977] 207210; Mercati [1916].
Scholarship in the Byzantine Empire ( 5291453 ) 439
Perhaps the most iconic link between the late Byzantine age and the roots
of Italian humanism is represented by a pupil of Barlaam named Leontius
Pilatus. While it is unclear whether he was a Calabrian or a Greek, and whether
he attendend Barlaams classes in Italy or in Thessalonica,767 Leontius is uni-
versally known today as the first translator of Homer into Latin since the age
of the Roman empire. This task, which he performed during his stay in
Florence in the early 1360s, was entrusted to him by no lesser intellectuals than
Francesco Petrarca and Giovanni Boccaccio, who were eventually very unsat-
isfied with the literary quality of the translation, as well as with the poor out-
come of his teaching of Greek in Florence.768
If the substantial failure of this attempt delayed the project of restoring
Greek as a pivotal part of humanistic curriculum, Leontius Greek culture in
and on itself has been assessed more positively in recent times: his contact
with Aristotelian and scholastic philosophy,769 his restoration of the Greek
passages in the illustrious Florentine codex of the Pandects,770 his partial trans-
lation (in two subsequent redactions) of Euripides Hecuba (on Ioannikios
ms. Laur. 31.10),771 as well as the painstaking annotations he scattered in the
margins of his Homeric versions (partly indebted to glossaries, mythogra-
phers, and to the ancient and Byzantine exegesis to Homer, Lycophron and
other authors),772 show that Leontius must be considered as a scholar conver-
sant with pagan wisdom, and a worthy source of Petrarchs notes in his own
Latin copy of the Iliad and of Giovanni Boccaccio in his Genealogiae deorum
gentilium.773 Let us consider for example the following cluster of passages:
767 Cortesi [1995] 458461; Rollo [2007] 721 (with earlier bibliography) revises the tradi-
tional idea that he should be regarded as a Calabrian, and believes that at least some of
his education must have taken place in Crete.
768 Cortesi [1995] 458461; Pertusi [1964].
769 Harlfinger-Rashed [20022003].
770 Di Benedetto [1969]. On the ms. see Baldi [2010].
771 Rollo [2007].
772 We still have the autograph manuscripts (Marc. Gr. IX.2 and 29): see the analysis by
Pertusi [1964] and Pontani [20022003].
773 Pertusi [1964] 295380.
440 Pontani
Petrarch in his note on Odyssey 1.344: Hellada dicta est Grecia. Greci
Hellines ab Hellino rege ut vult Leon. Ego partem Grecie puto circa
Athenas, etc.
Boccaccio, Genealogies of the Pagan Gods 4.48 Ellanum dicit
Theodontius filium fuisse Deucalionis et Pyrre (cf. schol. D in Il. 12.117 et
alibi), quem ait Barlaam, patre mortuo, adeo nomen suum et imperium
ampliasse, ut fere omnis Grecia, que in Egeum mare versa est, a nomine
suo Ellada nominata sit, et Ellades Greci.774
774 See Pertusi [1964] 299300; Pontani [20022003] 307; Rollo [2007] 101 and 77.
775 Thorn-Wickert [2006]; Wilson [1992] 812.
776 Weiss [1977] 227254; Mergiali [1996] 137139; Hankins [2002].
777 A. Pontani [1999]; Rollo [2002b]; Thorn-Wickert [2006] 150165.
778 Rollo [2002a] 7980.
779 Pontani [2005b] 209 and 239242.
Scholarship in the Byzantine Empire ( 5291453 ) 441
the texts he wished to translate (e.g. in ms. Vat. Gr. 226 of Platos Republic and
Vat. Gr. 191 of Ptolemys Geography).780 In this context, it is interesting to note
that Chrysoloras theory of translation, based on a certain degree of liberty
though refraining from modifying the Graeca proprietas, proved essential for
generations of humanists who attended to the painstaking task of Latinising
the bulk of ancient Greek literature.781
In addition to Chrysoloras charism, spelled out by his pupil Leonardo
Bruni, who regarded him as the restorer of Greek letters in Italy after 700
years of silence,782 the long-term influence of his teaching consisted above
all in the extraordinary tool he fashioned for learning the language, namely
a grammar in the traditional question-and-answer form (known already from
Moschopoulos), butfor the first timedesigned for the needs of a Latin-
speaking public.783 Chrysoloras Erotemata, thanks to its multiple and yet
coherent structure, enjoyed such widespread success both in the West and
in Constantinople and prompted so many abridgments and compendia by
the hand of the authors pupils and successors (from Guarino da Verona to
Constantine Lascaris), that, in the lack of an autograph, the reconstruction
of its original form remains problematic today:784 a new critical edition now
offers many answers to the open philological issues, and will no doubt be the
starting-point for future research in this delicate field.785
In a famous letter to emperor Manuel II, Chrysoloras observed in
astonishment, in the wake of a similar statement by Cydones,786 that
, ,
,
780 Gentile [2002] and [1992]. The notes to Lucian in ms. Vat. Urb. Gr. 121 belong to a pupil of
Chrysoloras: Berti [1987].
781 See Cencio de Rusticis report of Chrysoloras views in Cortesi [1995] 470471 and Wilson
[1992] 11.
782 Hankins [2007] 32223.
783 Wilson [1992] 812; Pertusi [1962].
784 Ciccolella [2010]; Nuti [2012].
785 Rollo [2012].
786 Apology p. 366.9596 Mercati: the Latins break much sweat in order to walk in the laby-
rinths of Aristotle and Plato, for which our people never showed interest (
,
).
442 Pontani
Sunt qui tres, sunt qui quattuor, sunt qui quinque, sunt qui novem esse con-
tendant. Omissis reliquis sequamur hos extremos qui novem fuisse dicunt.
De ipsis igitur summatim intelligendum est musas notiones quasdam et
intelligentias esse, quae humanis studiis et industria varias actiones et
opera excogitaverunt, sic dictas quia omnia inquirant vel quia ab omnibus
inquirantur, cum ingenita sit hominibus sciendi cupiditas. enim
graece indagare dicitur; igitur indagatrices dicantur.
Some say they are three in number, others four, yet others five or nine.
Let us follow straight away the latter, who say they were nine. One has
to understand about them that the Muses represent the notions and
skills which through human application and hability produced the vari-
ous works and actions. They are thus named because they investigate
everything or are investigated by everyone, for men naturally long for
knowledge. Mosthai means to investigate in Greek; mousai thus means
investigators.
That the translation of ancient Greek works could have a great impact on
contemporary mores is overtly declared by Guarinos Venetian pupil Francesco
Barbaro (13901454), who owned and annotated books of poetry and prose,
from Homer to Plato, from Plutarch to Lucian. Precisely Plutarchs Lives rep-
resent the favourite work of humanist translators: thus Barbaro himself trans-
lated Aristides and Cato the Elder, while his fellow-countryman Leonardo
Giustinian translated the Lives of Cimon and Lucullustwo men whose deeds
and ethos, in Barbaros view, provided an excellent model and paradigm for the
best energies of the Venetian youth.793
In 1423, Guarinos other pupil Vittorino da Feltre (13781446) was sum-
moned by the Gonzaga family to establish a school near Mantua: the Ca
Zoiosa, one of the milestones in the chain of Western paedagogical thought,
focused on humanities, sciences and gymnastics,794 and gave pride of place
to Greek literature in the original language, with special emphasis on poetry
(Homer, Nicander, Oppian, Dionysius the Periegete, Apollonius Rhodius).795
The school also gave shelter and an ubi consistam to Greek scribes working
for Vittorino, such as Petros Kretikos and Girard of Patras, and occasionally
hosted outstanding teachers such as George of Trebizond and Theodore Gaza.
Although this did not result in the creation of an advanced scholarly circle
793 Wilson [1992] 2325; F. Barbaro, epist. 2 Griggio (to Palla Strozzi).
794 Garin [1958]; Mller [1984]; Cortesi [2010].
795 Cortesi [2000].
444 Pontani
,
.
, ,
, ,
.
.
book entertains the utmost similarity with ours, not however with the
one at the beginning of the book (for this belongs to another man named
Melissenos), but rather with the small, rounded characters to be found in
the middle of the codex, and even somewhat before that point. But the
title First book of Aristotles Physics at the beginning of this manuscript,
is also in Theodores hand.
, ,
, ,
,
, ,
, .
One of Plethos indisputable merits lies in the creation of a school where most
of the brightest intellectuals of his time were either trained or hosted for some
time: these include Bessarion (who stayed in Mystra between 1431 and 1436,
after some philosophical studies in the capital with John Chortasmenus and
George Chrysococcas), but also lesser intellectuals or learned scribes such
as Nicholas Secundinus, Matthew Camariotes, Constantine Lascaris, John
Eugenicus, John Doceianus, Charitonymus Hermonymus, Demetrios Trivolis,
Michael Apostolis, Demetrius Raul Cabaces, Demetrius Chalcondylas. Some
of them would remain active in the Peloponnese over the following decades,
others would move to Constantinople, yet others would flee to Italy and
become pioneers of Greek humanism in various European courts or cities.
Chalcondylas (14241511), who arrived to Italy in 1449 and was to become the
editor princeps of Homers Greek text in 1488.827
Another scholar who came to Italy for the Council and later travelled back
and forth before fleeing Constantinople for good just before the Fall of 1453,
was the philosopher John Argyropoulos (ca. 14101487).828 A professor of rhet-
oric and philosophy under the patronage of emperor John VIII until 1441, and
later at the Xenon of the Kral after obtaining his degree at the University of
Padua, Argyropoulos was perhaps the most influential translator and scholar
of Aristotle in the Italian Quattrocento, but he also read and copied Plato and
Plotinus. His most serious colleague and rival in the Byzantine capital was
George Scholarius, the future patriarch Gennadius:829 although a high officer
at the court of John VIII, Scholarius organized and ran a school in his house
throughout the period from 1430 through 1448, and at the same time wrote both
theological works (despite being a Palamist, he taught Western Scholasticism
and even translated into Greek Thomas Aquinas On being and essence) and
philosophical commentaries on Aristotle and Porphyry, as well as a handbook
of grammar not phrased in the current question-and-answer format.
In a letter to his beloved pupils, Scholarius complained that830
,
, ,
,
,
,
, .
,
, ,
.
Scholarius words tackle a vital issue, namely the fact that the process of
translatio studiorum revolved around the creation of libraries.831 The size
of public or institutional collections of Greek manuscripts in Mantua, Rome
and Florence, grew dramatically during the second quarter of the century;
amongst private libraries mention should be made at least of those of Giovanni
Aurispa, of Palla Strozzi, and primarily that of Bessarion, which was enlarged
through acquisitions from other scholars collections (e.g. Aurispas) as well
as through book-hunting campaigns in Southern Italy (e.g. in Otranto), and
finally bequeathed to the Republic of Venice (1468), where it built the core of
the newly founded Library of San Marco.832 Bessarions rationale for the gath-
ering of his library is well explained in his letter to Michael Apostolis, written
soon after 1453:833
,
, ,
, ,
, , ,
,
, ,
as long as the common and single hearth of the Greeks [scil. Constan
tinople] remained standing, I did not concern myself [scil. with gather-
ing manuscripts], because I knew that they were to be found there. But
when, alas, it fell, I conceived a great desire to acquire all these works, not
so much for myself, as I possess enough for my own use, but for the sake
of the Greeks who are left now, and who may have better fortune in the
future (for many things may happen in the course of the years), so that
they may have a place where to find their own entire culture (the culture
preserved down to our own day) intact and safe, and multiply it, and may
thus avoid the risk, which would ensue from the loss of what is extant on
top of all the many excellent works of the divine ancient authors that are
lost since time immemorial, of remaining forever dumb and under all
aspects identical to barbarians and slaves.
The development of libraries is of course part and parcel with the develop-
ment of scholarship. In an age when Greek literature in the East seemed to
be on the brink of entirely perishing,834 this massive transfer of books from
the Levant to Italy created the indispensable premise for the rise of a new
philological activity performed on Greek texts by Western scholars. Bessarion
himself (14031472),835 becoming in 1440 a cardinal of the Roman Church and
almost a pope, insisted on gathering a remarkable cenacle of scribes and schol-
ars (Gaza, Callistus, Giannozzo Manetti, Giorgio Valla etc.) at his house near
the old Forum Augusti: it cannot be fortuitous that, for instance, the first, brief
Western mention of ms. Venetus A of the Iliad occurs precisely in a writing of
a member of this circle, Martino Filetico.836
A translator of Xenophon and Aristotle, and a philosopher deeply engaged
in the Platonic/Aristotelian controversy, Bessarion did not stand out as a first-
rate philologist, but he showed a special familiarity with the Classical heritage
both in the marginalia he left on his books (as when he tried to restore a cor-
rupt text of Aristotles On Heavens in Marc. Gr. 491 by resorting to a medieval
Latin translation, or when he detected the hand of Eustathius of Thessalonica
in ms. Marc. Gr. 460 of his commentary to the Odyssey) and in the preface
and notes added to his translation of Demosthenes First Olynthiac, where the
patriotic words of the great Athenian orator were interpreted as an exhortation
to the Westerners to take action in favour of Byzantium against the Turkish
threatwith the pagan, Hellenic past still somewhat opposed to the Christian
present.837
The patronage of pope Nicholas V (14481455) marked an unprecedented
Aufschwung in the history of Greek learning in Rome, for it gave a common
834 This is the exclamation of Lauro Quirini in his letter to pope Nicholas V soon after the
Fall (Pertusi [1977] 227.9799): nomen Graecorum deletum; ultra centum et viginti milia
librorum volumina, ut a reverendissimo cardinali Rutheno accepi, devastata. Ergo et lin-
gua et litteratura Graecorum tanto tempore, tanto labore, tanta industria inventa, aucta,
perfecta peribit, heu peribit!.
835 Bianca [1999]; Wilson [1992] 5767.
836 Pincelli [2000] esp. 8587.
837 Lamers [2013] 91124; Wilson [1992] 6264; Bisaha [2011] 8081.
Scholarship in the Byzantine Empire ( 5291453 ) 453
network and background to the many brilliant humanists who were living in
the city, amongst them Bessarion, George of Trebizond, Theodore Gaza, and
the Italian Lorenzo Valla (14071457).838 The latter, in particular, was not only
a brilliant translator of Homer, Herodotus, Thucydides, and Demosthenes
(his devotion to ancient historians was prompted by the commissions of
Nicholas V),839 but proved capable of exploiting his philological genius (already
at work in his renowned confutation of the Donation of Constantine) and his
knowledge of Greek both in dealing with ancient inscriptions and in promoting
textual analysis in a field which had so far remained almost devoid of system-
atic treatments.840 Vallas Collation of the New Testament with the Greek Truth
was a Lebenswerk that went through at least two redactions in the 1440s and
early 1450s:841 the second version eventually persuaded Erasmus of Rotterdam
to publish it in 1505. This displayed to the wider public Vallas unprecedented
ability to collate several manuscripts of the same text, to detect or conjecture
scribal errors of different kinds behind textual corruptions, to invoke the role
of the indirect tradition, to spot interpolations, to blame Jerome for his mis-
takes, in a word to adopt even for the Holy Writ a sophisticated philological
method with a systematic thrust virtually unparalleled in Byzantine quarters,
and surpassing the tradition of the 13th-century Latin correctoria of the Bible.
One example will suffice:842
Matth. 4.6 Quoniam angelis suis mandavit de te] Graece est mandabit,
quam culpam librarii arbitror, aut alicuius temerarii correctoris tam
hic, quam in psalmo: nam e Psalmo [scil. 90.11] hic locus est sumptus
, ....4.10 Et illi soli servies]
, hoc Graeco verbo non utuntur nostri, sed nomine quod est
latria, volentes hunc actum soli Deo deberi: is etsi frequenter Deo exhibe-
tur, non tamen semper. Nam quantum ego sentio, magis ad homines per-
tinebat; ideoque principes Graecae linguae, quorum autoritate nitimur
in verborum significationibus, ita usurpant, ut apud Xenophontem lib.
3 [3.1.36] in Cyripaedia:
, hoc autem est: equidem Cyre ego vel anima
redimerem nequando ista serviret. Haec viri sunt verba, tum uxor eadem
loquitur alio verbo, ut appareat inter ipsa verba latriam et duliam nihil
interesse [3.1.41]: , . Et
certe cum priores fuerint autores gentiles quam fideles, sive Graeci sive
Latini, nimirum multo plus obtinent autoritatis, quippe cum eos omnis
posteritas, tam fidelium quam infidelium, habeat autores, et eatenus
recte loquatur, quatenus ab illorum usu non discrepat. Nam consulto qui-
dem et de industria velle ab illis dissentire, nisi vehemens causa cogerit,
insania est, inscientem vero hoc facere, inscitia; quanquam sint qui
negent theologiam inservire praeceptis artis grammaticae. At ego dico
illlam debere servire etiam cuiuslibet linguae usum, qua loquitur, nedum
literatae.
Matthew 4.6 Quoniam angelis suis mandavit de te] in Greek we have man-
dabit, which I regard as a mistake of the scribe or of some bold correc-
tor both here and in the Psalm (for this passage is taken from the Psalm
he ordered to his angels about you that...)...4.10 Et illi soli servies]
latreuseis: the Latins do not use this Greek verb, but the noun latria,
insisting that this act pertains only to God; now, even though the word
appears often in relation to God, this is not always the case. As far as I can
judge, it used to pertain to men, which is why the princes of the Greek
language, on whose authority we rely in matters of semantics, use it in
this sense, as in Xenophons Cyropaedia book 3: O Cyrushe saidI
would pay with my life so that she might never be a slave. These are the
mans words, then the woman says the same with another term, so that
it might become apparent that there is no distinction between the words
latreia and douleia: that he may pay with his life so that I might never be
a slave. And of course pagan authors came earlier than Christian ones,
both in Greek and in Latin: which is why they have a greater authority,
because later generations (both pagans and Christians) regarded them as
models, and used language correctly only insofar as they did not depart
from their usage. For anyone who wishes to part with their example on
purpose and overtly, without being forced by a strong necessity, is a fool,
and to do so by mistake is a sign of ignorance; some, however, deny that
theology should follow the rules of grammar. But I argue that theology
should follow the usage of any language in which it speaks, all the more
so in the written form.
Scholarship in the Byzantine Empire ( 5291453 ) 455
By May 29th, 1453 most Greek scribes and scholars were active in Western
Europe, at least some Italian erudites were capable of performing philologi-
cal activity on ancient Greek texts, and the translatio of Greek scholarship
to the West was almost complete. In the following year, Angelo Poliziano
was born.843
chapter 1
Eleanor Dickey
1 Introduction
2 Papyri
3 Extant Lexica
3.1 Early Lexica
3.2 Atticist Lexica
3.3 Pollux
3.4 Ammonius
3.5 Author-Specific Lexica
3.6 Hesychius
3.7 Etymologica
3.8 Suda
3.9 Other Lexica
4 Extant Grammatical Works
4.1 Dionysius Thrax
4.2 Apollonius Dyscolus
4.3 Herodian
4.4 Theodosius
4.5 Philoponus
4.6 Choeroboscus
4.7 Gregory of Corinth
4.8 Other Grammarians
5 Extant Commentaries
5.1 Early Commentaries
5.2 Galen and Medical Commentaries
5.3 Philosophical Commentaries
5.4 Mathematical Commentaries
5.5 Ancient Work on Literature
5.6 Epimerismi Homerici
5.7 Eustathius
5.8 Other Byzantine Works
1 Introduction
Our knowledge of ancient scholarship comes from four types of source. Some
material has been recovered on papyrus; this method of learning about ancient
scholarship is in some ways preferable to all others, bypassing the medieval
manuscript tradition and all the errors and distortions it introduces. But the
amount of material that survives on papyrus is tiny compared to that preserved
in other ways, it is usually very fragmentary, and (despite what one might
expect) it is not always uncorrupt. Moreover most scholarly papyri belong to
the school tradition; material from the best ancient scholarly work is far less
common in them than basic lexical information. Thus although papyri are an
important resource where we have them, they do not significantly reduce our
reliance on manuscript sources.
Another type of source is works that have survived through their own man-
uscript tradition; these are occasionally intact but more often shortened and/
or reworked versions of the original texts. Such works are greatly valued and
are not uncommon for certain types of material: lexica, Roman-period gram-
matical treatises, works on style, and commentaries on religious, philosophi-
cal, and scientific works stand a reasonable chance of survival either intact or
in substantial epitomes (abbreviated versions). Other types of scholarly mate-
rial, however, are almost never preserved in this fashion: works of the great
Alexandrian scholars, commentaries on and discussions of literary works, and
early grammatical treatises fall into this category.
The sources of our knowledge of ancient scholarship 461
For works that are not preserved intact the next best source is fragments
quoted in the text of other works. Such fragments are often tantalizingly short
and stripped of context, but sometimes they can be substantial. Fragments
are usually quoted with attribution to a specific author, and often with the
title of the work in which they appeared as well. It is sometimes possible to
reconstruct part of a lost work by putting together the preserved fragments,
but when using such reconstructions a reader needs to be aware of the extent
to which the text at hand is the creation of a modern editor. A wide range of
material can appear in fragments, but grammatical treatises and commentar-
ies on works for which other commentaries survive intact are the most com-
mon fragmentary material.
The fourth source for ancient scholarship is scholia, that is material that has
survived in the margins of literary works transmitted via the manuscript tradi-
tion. Scholia are extremely important to the student of ancient scholarship,
as they are our main source of knowledge of Alexandrian scholarship and of
ancient commentaries on literary works. But they are also highly problematic,
as they are severely abbreviated notes, often without attribution to a source,
and very prone to corruption. Scholia tend to be published in editions that col-
lect the marginalia to a given work and arrange them in the order of the work
on which they comment. Many scholia are unpublished, for marginalia have
little value when they were copied from works that have also survived inde-
pendently, and marginalia of the Byzantine period are often thought to have
little value in any case (though this opinion is now declining in popularity). The
published scholia are normally those that provide our only witness to scholar-
ship that is, or could be, of the Hellenistic, Roman, or late antique periods.
2 Papyri
long (covering Philippics 9, 10, 11, and 13) and thus far longer than the vast
majority of surviving scholarly papyri. Didymus is explicitly named as the
author of the commentary, and the papyrus dates to the early second cen-
tury AD, making it relatively close to Didymus own time (late first century BC
and early first century AD), but the work appears nevertheless to have under-
gone some abbreviation and alteration in the interval; it may even be a set
of excerpts from Didymus commentary rather than the complete work.1 This
find is in many ways very exciting, but in others somewhat disappointing, as
the papyrus does not contain quite what modern scholars had expected that
an original commentary of Didymus would have. Pearson and Stephens [1983]
give a good text, Gibson [2002] a translation and commentary, and Harding
[2006] all three.
Another important fragment of commentary explicitly attributed to a
famous scholar is the papyrus of Aristarchus commentary on Herodotus. The
attribution to Aristarchus gives this fragment great importance, but it is in
many ways less useful to us than the Didymus papyrus. It is short, containing
only one legible column (the end of the commentary on book I), and consid-
erably later than Aristarchus himself, probably from the third century AD. In
terms of content it is even more disappointing than the Didymus papyrus and
therefore is normally considered to be an abridgement or set of extracts rather
than a full version of the original commentary. Paap [1948] gives a text.
Most scholarly papyri,2 however, do not carry attributions to named schol-
ars; no doubt many originally had authors names attached, but they no longer
survive. Some of these papyri are nevertheless of considerable significance.
A relatively large group of papyri preserve hypotheses, brief summaries of
literary works. In modern times hypotheses of dramatic texts are widely known,
as they are often printed in editions of the texts to which they relate. This privi-
leging of hypotheses over other types of ancient scholarship (which are much
less likely to be printed in a modern edition of a literary text) occurs in part
because some dramatic hypotheses may go back to the work of Aristophanes of
Byzantium and preserve crucial information. But the convention also extends
to hypotheses that have no scholarly information and simply summarize the
plot, a type much commoner than the scholarly hypothesis. Such plot-summary
hypotheses, which often have a mythographic slant, were common in antiquity
as well as the Byzantine period, and many of them have been found on papyri.
In antiquity, moreover, such hypotheses were written not only for dramatic
texts but for a wide variety of different genres: there were even hypotheses
for the individual books of longer works such as the Homeric poems. These
hypotheses are often found in groups in papyri, without the original texts
they summarize; the most famous example of this phenomenon is the strings
of hypotheses of Euripidean tragedies known as the Tales from Euripides
(see below 8.4). Such summaries are obviously of considerable value to us
where the original texts have been lost, but they are also of great interest for
the window on ancient readers interests with which they provide us. A collec-
tion of these papyri with discussion is provided by Rossum-Steenbeek [1998].
Some papyri dating from the first/second to the fifth century AD contain
remains of the Mythographus Homericus. This term, invented in modern
times, is used to refer to the author of a lost work, probably composed in
the first century AD, that related the full versions of myths alluded to in the
Homeric poems. The work could be called a mythological commentary, for
it was arranged in the order in which the allusions occurred in the poems. It
tended to give only one particular version of each myth, attributed to a specific
source; a number of the attributions can be shown to be genuine, and it seems
that the compiler was using important and now lost scholarly commentar-
ies, probably Alexandrian. Although most of this compilers work is lost in its
original form, a number of papyrus fragments have survived, and much mate-
rial from the commentary was incorporated into the medieval D scholia (see
below 8.1). For a text and discussion see Rossum-Steenbeek [1998] 278309.
Of other papyrus material the largest group is the Homerica, a diverse body
that in addition to the Mythographus includes commentaries, glossaries, and
paraphrases. Some of this material is incorporated into editions of scholia by
Erbse [19691988], van Thiel [1992], and Pontani [2007]; guides to the rest can
be found e.g. in M. L. West [2001a] 130136 and Lundon [1999c]. Another impor-
tant group is the Aristophanes papyri, which are often the focus of debates on
the process by which ancient commentaries were turned into medieval scho-
lia; these are collected with translation and discussion by Trojahn [2002] and
Montana [2012a]. A number of papyrus commentaries on philosophical works
survive, including a long fragment of commentary on Platos Theaetetus (nor-
mally dated to the second century AD, but perhaps as early as the late first
century BC) and a fragment of commentary on Aristotles Topica from the first
century AD; these have been collected in CPF [1995].
There is also a substantial body of grammatical papyri, containing doctrine
that is often anonymous but usually of considerable antiquity. Many of these
papyri have been collected and discussed by Wouters [1979]. Papyrus lexica
form another common genre; most of them appear to be focused on the works
of a particular author or group of authors, but more general lexica are not
unknown. The lexica have no collected edition; Naoumides [1969] offers a list
of papyrus lexica with discussion of their characteristics, but more have been
464 Dickey
3 Extant Lexica
The lexicon was a popular form for recording knowledge in antiquity; lexica
contained a broader range of information about each entry than a modern
dictionary and in some respects resembled our encyclopedias. At the same
time the older lexica tended to be narrowly focused, concerning topics like
names of fishes or Homeric words rather than being general collections like
a modern dictionary (cf. Tosi in this volume). The tradition of Greek scholarly
lexica dates back at least to Aristophanes of Byzantium and is one in which
surviving material is particularly likely to be old; a large number of ancient and
Byzantine lexica survive, intact or abbreviated, and fragments of many oth-
ers can also be found. These lexica are the source for our knowledge of many
elements of Greek vocabulary and for much of our information on lost works
of literature, and much still remains to be learned from them. They must how-
ever be used with care, as they are usually poorly transmitted and often inad-
equately edited.
except for a few sets of entries taken over from alphabetizing sources). Many
entries consist of a non-Attic word, usually but not always from the koin (e.g.
two), an injunction against using it, and the appropriate Attic replace-
ment (e.g. two), while others give the proper Attic syntax of the lemma
(e.g. happen must be accompanied by being when it means
happen to be) or the difference between easily confused words (e.g. a
girl is female, but a youth is male). Phrynichus sources include
several lost works of ancient scholarship, and his work is valuable both for
preserving such fragments and for the light it sheds on the way the Atticists
worked and on the type of mistakes that Greek speakers trying to write clas-
sical Attic were likely to make in the second century. For texts see de Borries
[1911] and Fischer [1974].
Phrynichus opposite on the Atticist spectrum was another second-century
lexicographer now known as the Antiatticist or Antiatticistaa misleading
name since this scholar (whose real name has been lost) was clearly an Atticist
in principle but differed from Phrynichus in having a different canon. The
Antiatticist admitted a larger group of authors into his canon and apparently
held that the use of a word by any Attic author made it acceptable as Attic,
even if a more recherch alternative existed. Until recently it was believed
that the Antiatticist was a contemporary of Phrynichus who wrote in response
to the first book of Phrynichus Ecloga and against whom the second book of
the Ecloga was then directed, but now some hold that Phrynichus attacked
the Antiatticist throughout the Ecloga, and others that Phrynichus used the
Antiatticists work rather than attacking it, suggesting that the Antiatticist may
have been a predecessor rather than a contemporary.
The lexicon seems to have originally consisted of a list of Attic words, with
definitions and references to the words occurrences in classical texts; many of
the words listed were ones whose claim to be considered properly Attic had
been disputed by the stricter Atticists, and the Antiatticist seems to have made
a point of showing that those words were indeed attested, often by quoting the
relevant passage. Unfortunately the work survives only in the form of a drasti-
cally reduced epitome from which most of the quotations have been excised,
leaving only tantalizing references to lost works. Enough remains, however,
that the work is useful for information on lost literary works, historical details
about classical Athens, and fragments of Hellenistic scholarship, as well as for
understanding the controversies of the Second Sophistic period (cf. Matthaios
in this volume). There is a text in Bekker [18141821] vol. I, 75116.
A work in iambics from around 200 AD entitled
On Attic Controversy about Words is attributed to the glos-
sographer Philemon; this Philemon is not to be confused with a much earlier
468 Dickey
3.3 Pollux
Julius Pollux (or Polydeukes) of Naucratis, a rhetorician of the latter part of the
second century AD, was the author of the Onomasticon, a wide-ranging lexicon
in ten books. The work now survives only in the form of an epitome that has
suffered interpolation as well as abridgement, but it is still of considerable bulk
and primarily Polluxs own work. It is based on works of classical literature
and Alexandrian scholarship, including many no longer extant; among these
sources are Aristophanes of Byzantium and Eratosthenes.
The Onomasticon is organized not in alphabetical order like other surviving
ancient lexica, but by topic; in this it preserves a very early method of organiza-
tion that originally predominated in Greek scholarship and was only gradually
replaced by alphabetical ordering (cf. Matthaios and Tosi in this volume). Some
entries are very brief, but others are complex and detailed, offering much more
than a simple definition. Perhaps the most famous section is Polluxs discus-
sion of the classical theater and its paraphernalia, including a description of
76 different types of mask for different characters in tragedies, comedies, and
satyr plays, which is an invaluable source of information on the ancient stage.
Much other historical information can also be found in the Onomasticon, as
can fragments of lost works, better readings of extant works, and definitions
(including some earliest attestations) of obscure words.
The standard edition of the Onomasticon is that of Bethe [19001937], and
some useful discussion is provided by Tosi [1988] 87113; see now also Bearzot
et al. [2007] and Matthaios [2013a].
The sources of our knowledge of ancient scholarship 469
3.4 Ammonius
A lexicon entitled On Similar and Different
Words or De adfinium vocabulorum differentia On the Differences between
Related Words is preserved in late manuscripts under the name of Ammonius,
but it is generally agreed not to have been composed by any of the known
bearers of that name. The work is closely related to a number of other lex-
ica that survive only as epitomes, of which the most significant are the
On the Differentiation of Words attributed to an unidenti-
fied Ptolemaeus and the On Different Meanings or
De diversis verborum significationibus On the Different Meanings of Words of
Herennius (or Erennius) Philo. It is thought that the ancestor of all these works
was probably a lexicon composed by Herennius Philo in the second century
AD, which was severely epitomized both with and without its authors name
and preserved (probably still in a reduced form, but one of substantial size)
with the substitution of Ammonius name.
The lexicon consists primarily of pairs of words that are similar or identical
in some way, with an explanation of the difference between them. It is often
called a lexicon of synonyms, and in the majority of cases the paired words are
in fact synonyms (e.g. city and city, or well and well),
but in other cases they are homonyms, similar or identical in form but different
in meaning (e.g. there and thither, or populace and
fat). Some are similar in both form and meaning, and occasionally an entry
consists of a single word followed by a list of synonyms. The sources include
classical literature, Alexandrian scholarship, and scholarship of the early
Roman period, most now lost; sometimes literary quotations are included to
exemplify the meaning or usage of a particular word. While the vast majority
of the entries contain information that is correct by the standards of classical
usage, and some of them preserve really valuable scholarly information, there
are also a few mistakes and a certain amount of banality. For text and discus-
sion see Nickau [1966] and [2000].
than the average of ancient scholarship. Unusually for a work of this period,
Harpocrations glossary follows complete alphabetical order (i.e. words are not
merely grouped together by their first letters, or by their first two or three let-
ters, but fully alphabetized as in a modern dictionary); there is however some
debate about whether this feature can be traced to Harpocration himself or
was added at a later stage of transmission.
Harpocrations work survives, in a contaminated and somewhat abridged
form, in a number of late manuscripts; this version is known as the full version
in contrast to our other main witness to the text, an epitome dating probably to
the early ninth century. There is also an early papyrus fragment of the glossary,
from the second or third century AD, as well as extracts from Harpocration
preserved in Photius and in scholia to the orators. All texts of Harpocration
are unsatisfactory; one edition, and references to the others, can be found in
Keaney [1991], on which see Otranto [1993].
From the same century comes a Hippocratic glossary by Galen, which was
based heavily on earlier glossaries; unlike Galens Hippocratic commentaries
it is largely scholarly rather than scientific in orientation, and the preface con-
tains much useful information on the work of earlier scholars. Galens glossary
also employs complete alphabetical order, though this feature may not be due
to Galen himself. A poor text of the glossary can be found in Khn [18211833]
vol. XIX.
Of uncertain date is a lexicon to Plato attributed to an otherwise unknown
Timaeus the Sophist, which survives in a single manuscript. The work has
clearly suffered significant additions and subtractions at later periods, leading
to the inclusion of many non-Platonic words and to non-Platonic definitions
of words that do occur in Plato. It is nevertheless important as the sole surviv-
ing witness to a genre: two other Platonic lexica, by Boethus and Clement, are
known only from insubstantial fragments. Timaeus seems to have used earlier
commentaries on Plato that are now lost, and his lexicon also appears to be
one of the sources of our extant scholia. Bonelli [2007] and Valente [2012] pro-
vide editions.
An anonymous glossary of Herodotean words known as the Word-list
survives in two versions; version A is arranged in the order of the words appear-
ance in Herodotus text and version B in alphabetical order. Sometimes the title
Word-list of Herodotus is reserved for version A, while version
B is called the Lexicon of Herodotean Words.
Version A is older; its date is unknown, but it was clearly written to accompany
an unaccented version of the text (i.e. before c. 900 AD). It seems to be based
(at least in part) on a commentary, for it sometimes offers definitions intended
to clarify the interpretation, in a specific context, of common words easily con-
The sources of our knowledge of ancient scholarship 471
3.6 Hesychius
Hesychius of Alexandria composed in the fifth or sixth century AD a lexi-
con of obscure words that survives in an abridged and interpolated version
(cf. Matthaios in this volume). Hesychius based his work on a lost lexicon com-
posed by Diogenianus in the second century AD, which he claims to have sup-
plemented from the works of Aristarchus, Heliodorus, Apion, and Herodian.
Such claims are now difficult to verify or refute, but the work clearly contains
material from lost sources much earlier than Hesychius himself. The lexicon
consists of a list of poetic and dialectal words, phrases, and short proverbs. The
words are often in inflected forms (as they appeared in the original texts from
which Hesychius predecessors extracted them) rather than the dictionary
forms used today, and they are alphabetized usually by the first three letters.
Hesychius lexicon is useful for several reasons. It is the only source for a
large number of rare words that occur nowhere else in extant literature (par-
ticularly dialect forms). It also preserves, and provides information on, many
words that would be omitted from a modern dictionary for being proper names
(thus for example it is one of our main sources for the names of Attic clans).
In some cases Hesychius entries can be used as independent witnesses to the
texts of extant authors and can supply correct readings of words corrupted in
the transmission of those texts. Hesychius also tells us what ancient scholars
thought his obscure words meant; this information can be useful both as a
guide to the actual meanings of the words and as a source of insight into the
ways that ancient scholars understood and interpreted literature.
The lexicon in its current form is substantially different from the one
Hesychius wrote. Not only was his work severely abridged in transmission (a
process that eliminated, among other things, most of Hesychius indications
of his sources for the various words), but it has been heavily interpolated as
well. About a third of the entries are Biblical glosses from Cyrillus lexicon, and
material from other sources has also been added to Hesychius original core.
The interpolations must have occurred rather early, for material from Cyrillus
was already in Hesychiuss work by the eighth century.
A further complication is the state of the text. Only one manuscript of
Hesychius survives, and it is late (fifteenth century), damaged, and seriously
472 Dickey
corrupt. The best edition is that of Latte [19531966], P. A. Hansen [2005], and
Hansen-Cunningham [2009].
3.7 Etymologica
A number of enormous, anonymous Byzantine etymological lexica have sur-
vived more or less intact and preserve much valuable ancient scholarship.
Though traditionally referred to as etymologica because of the attention they
pay to etymology, they contain much other information as well. The oldest and
most important of these is the Etymologicum genuinum True Etymological
Dictionary, which was compiled in the ninth century, though our only wit-
nesses to it are two tenth-century manuscripts of unusually poor quality. From
the original version of this work, with various excisions and additions, are
descended almost all the other etymologica, of which the most important are
the Etymologicum magnum Great Etymological Dictionary from the twelfth
century, the Etymologicum Gudianum Gudian Etymological Dictionary from
the eleventh century, and the Etymologicum Symeonis Etymological Dictionary
of Symeon from the twelfth century. The Etymologicum (Florentinum) parvum
Little (Florentine) Etymological Dictionary, for which we have only entries
from the first half of the alphabet, is somewhat older but much less useful
because of its small scale and lack of quotations.
The sources of the etymologica vary but generally date to the second cen-
tury AD and later; major sources include Herodian, Orus, Orion, Theognostus,
Choeroboscus, scholia, and the Epimerismi Homerici. But since these works
were themselves usually based on earlier scholarship, the etymologica are
indirect witnesses to a considerable amount of Hellenistic scholarly work, as
well as preserving numerous fragments of classical literature otherwise lost.
Texts of these works are incomplete, scattered, and often unsatisfactory; for
a complete list see Dickey [2007] 9192.
3.8 Suda
The Suda is a huge dictionary / encyclopedia compiled in the late tenth cen-
tury. From the twelfth until the mid-twentieth century the work was referred
to as Suidae lexicon lexicon of Suidas, but now it is generally thought that the
Suda in manuscripts is the works title, not the authors name, and in
consequence the work is usually called the Suda and considered to be anony-
mous. The Suda may have been compiled by a group of scholars, but author-
ship by an individual cannot be ruled out.
The work consists of c. 30,000 entries of varying types; some lemmata are
followed by short definitions as in a lexicon, and others by detailed articles
resembling those in a modern encyclopedia. They are arranged in a form of
alphabetical order adapted to Byzantine Greek pronunciation (i.e. vowels not
The sources of our knowledge of ancient scholarship 473
4 Grammatical Works
The date of the origin of the Greek grammatical tradition is hotly debated
(cf. Montana and Matthaios in this volume), and as with the lexica, the very
earliest works have not survived. But we have surviving works from three
ancient grammarians, Dionysius Thrax, Apollonius Dyscolus, and Herodian,
as well as a number from the late antique and Byzantine periods; portions of
numerous other grammatical works survive as fragments and will be discussed
in section 6 (for typology of grammatical treatises and the definitions of techne
grammatike see Valente and Swiggers-Wouters in this volume).
BC. Others maintain that Aristarchus and his followers already possessed an
advanced grammatical system and that the date of the therefore makes
little difference to our view of the evolution of grammar.
The itself is relatively straightforward; it consists of a concise expla-
nation of the divisions of grammar and definitions of the main grammatical
terminology. Because of its extreme brevity, it accumulated a large body of
explanatory commentary (this material is all traditionally known as scholia,
but it includes continuous commentaries as well as marginal scholia), which
are in many ways more interesting and informative than the text itself, though
clearly later. The is also traditionally accompanied by four supplements,
which are probably old but later than the text itself: or De
Prosodiis On Prosody, On the (Grammatical) Art or Definitio Artis
Definition of the (Grammatical) Art,
or De Pedibus et de Metro Heroico On (Metrical) Feet and the Heroic Meter, and
a paradigm of the declension of beat derived from the Canons
of Theodosius. Some of these supplements are the subjects of additional com-
mentaries. Both scholia and supplements contain valuable information about
other ancient grammatical writings, particularly the lost works of Apollonius
Dyscolus, and cover a wide variety of topics.
Good texts of the can be found in Lallot [19982], with translation and
excellent discussion, and Uhlig [1883]. The supplements and scholia are best
consulted in Uhlig [1883] 103132 and Hilgard [1901].
considers. Although his works are primarily important for their portrayal of
Apollonius own ideas, they are also useful as sources of information on the
lost writings of earlier scholars, since they include numerous references to
Zenodotus, Aristarchus, and others. Apollonius seems to have been particu-
larly indebted to Trypho, though (perhaps because the latter was a scholarly
grandchild of Aristarchus) Aristarchus direct and indirect influence is also
considerable. Apollonius style is notoriously opaque and elliptical, and his
terminology is idiosyncratic; indeed since antiquity one of the explanations
offered for his nickname troublesome has been a reference to the
sufferings he inflicted on his readers.
Good texts of the Syntax can be found in Uhlig [1910] and Lallot [1997] with
translation and discussion; the minor works can all be found in Schneider
[18781902], and individually with translation and discussion in Dalimier
[2001] for Conjunctions and Brandenburg [2005] for Pronouns. Glossaries and
other aids to understanding Apollonius difficult Greek can be found in most
of these works and in Schneider [1910], and there are additional useful discus-
sions in Blank [1982] and [1993].
4.3 Herodian
Aelius Herodianus (2nd cent. AD), son of Apollonius Dyscolus, is responsi-
ble for most of our knowledge of ancient accentuation (cf. Matthaios in this
volume). His main work, the On Prosody in General,
is said originally to have been an enormous work giving the rules for attach-
ing accents and breathings to Greek words, with explanations based on their
terminations, number of syllables, gender, and other qualities; it now survives
in several epitomes (most notably those of Arcadius and Joannes Philoponus
of Alexandria) and is one of the major extant grammatical works despite being
considerably reduced in size. The only one of Herodians works to survive
intact, however, is the On the Anomalous Word, a trea-
tise on anomalous words.
Many of Herodians rules were meant to apply to classical and Homeric words,
i.e. to words from six centuries and more before his own time. Alexandrian
scholars from Aristophanes of Byzantium (c. 257c. 180 BC) onward worked on
accentuation, and Herodian built on a tradition going back to these scholars.
Even the Alexandrians were too distanced from classical and Homeric Greek to
possess any native-speaker knowledge of those dialects; yet their pronounce-
ments can sometimes be proven right by modern techniques of comparative
philology, to which they did not have access. Many modern scholars believe
that the Alexandrians drew on a living tradition of accentuation going back to
478 Dickey
the classical period and perhaps beyond, but there is some debate as to the form
and extent of that tradition. For discussion see Probert [2006] 2545.
Numerous works of Herodian survive as fragments (see below 7.2). There
are also a number of surviving works doubtfully or spuriously attributed to
Herodian, most of which were composed considerably after his day. There are
no good texts of Herodians genuine works; for explanation of the options avail-
able and their pitfalls see Dickey [2014] and [2007] 7577, and Dyck [1993a].
4.4 Theodosius
Theodosius of Alexandria, who lived probably in the 4th and 5th centu-
ries AD, was the author of the Canons, a set of rules and paradigms
for declensions and conjugations (cf. Matthaios in this volume). This long
and detailed work was a teaching tool intended to supplement the of
(ps-) Dionysius Thrax (see above 4.1) and appears to be the ancestor of the
fourth supplement to that work. It gives all theoretically possible forms of the
words it illustrates (most famously in an ultra-complete paradigm of
beat), thus producing a large number of forms unattested in actual usage. Partly
as a result of this inclusiveness, the are not highly respected today, but
for many centuries they exerted an important influence on Greek textbooks.
Two lengthy commentaries on the survive; that of Choeroboscus
(8th9th cent.) is intact, and that of Joannes Charax (6th8th cent.) is pre-
served in an excerpted version by Sophronius (9th cent.). These commentaries,
particularly that of Choeroboscus, are now considered more important than
the themselves.
Theodosius is also credited with short treatises entitled
- On the Declension of Barytone Words Ending in - and
- On the Declension of Oxytone Words Ending in
-, and he may be responsible for the On Prosody supple-
ment to (ps-) Dionysius Thraxs . Spurious works include a long
On Grammar and shorter works entitled On
Dialects and On the Accent.
The best text of the and its commentaries is that of Hilgard [1889
1894], which also provides a good introduction to the works; for further infor-
mation see Dickey [2007] 8384 and Kaster [1988] 366367.
4.5 Philoponus
The sixth-century philosopher Ioannes Philoponus of Alexandria (cf. Matthaios
in this volume), who is known primarily for his heretical Christian theology and
his commentaries on Aristotle, is also credited with several grammatical works,
three of which survive. One, the Rules for Accentuation,
The sources of our knowledge of ancient scholarship 479
4.6 Choeroboscus
George Choeroboscus, who lived in the eighth and ninth centuries AD, was a
Byzantine teacher and author of a number of grammatical works (cf. Pontani
in this volume). Choeroboscus works were not intended as contributions to
the advancement of grammatical theory; they are clearly part of his teaching
materials and were often intended for fairly elementary students. Their signifi-
cance lies in three areas: the light they shed on grammatical teaching in the
9th century, the influence they exerted on later scholars (including Eustathius
and the compiler of the Etymologicum Genuinum), and their extensive use
of earlier grammatical treatises (Choeroboscus is for example responsible
for much of the preservation of Herodians fragmentary
On the Declension of Nouns).
The longest and most important of Choeroboscus works is a gigantic com-
mentary on the of Theodosius, evidently composed as a teaching
tool, which survives both intact and drastically excerpted in a short collec-
tion of extracts on accents entitled On Accents. Choeroboscus also
480 Dickey
5 Extant Commentaries
The vast majority of intact commentaries (and other exegetical works focus-
ing on particular authors or texts, which are grouped here with commentaries
for organizational convenience) discuss religious, philosophical and scientific
texts and concern their philosophical or scientific aspects. They are usually
omitted in discussions of ancient scholarship because the questions with
which they are concerned are not normally considered to be within the bounds
of scholarship, and because they were often composed at a late period with
little or no reference to Alexandrian or other early work. But it is important to
482 Dickey
be aware of their existence, as such commentaries are in some ways our best
evidence for what was going on in the world of late antique scholarship. We
have only recently come to appreciate the extent to which scientific and even
Biblical scholarship influenced the development of work on, and the transmis-
sion of work on, pagan literary texts (cf. McNamee [2007] 7992).
six and On the nature of the child only fragments survive in Greek (though more
exists in Latin). For editions and further bibliography see Ihm [2002a].
these authors survives, almost all from the late antique and Byzantine periods
but sometimes incorporating earlier work.
The earliest commentaries come from the third century AD. The philoso-
pher Porphyry has left us two works: a commentary on Ptolemys Harmonica
(text in Dring [1932] updated by Alexanderson [1969]) and an introduction
and explanation of Ptolemys or , which con-
cerned astrology (text in Boer-Weinstock [1940]). From Iamblichus we have
a commentary on the Introductio arithmetica Introduction to Arithmetic of
Nicomachus of Gerasa; Pistelli [1894] provides a text. There is also a portion
of an elementary commentary on Ptolemys Handy Tables
surviving from the early third century; text and translation in Jones [1990].
From the fourth century we have more material. Of a commentary by
Pappus on the first six books of Ptolemys Almagest we now have the portion
on books five and six; text in Rome [19311943] vol. I. Pappus commentary on
Euclids Elements has fared less well: two books on book ten of the Elements
survive, but only in Arabic translation. They include a philosophical introduc-
tion to book ten as well as detailed mathematical discussion. There is a good
edition with translation in Junge-Thomson [1930].
The works of Theon of Alexandria (cf. Matthaios in this volume), who also
lived in the fourth century, slightly later than Pappus, are better preserved.
Theons commentary on books one to thirteen of Ptolemys Almagest survives
apart from the section on book 11 and portions of the section on book 5. The
commentary on book three provides a rare glimpse of ancient scholarship
produced by a woman, for it was based on a text edited by Theons daugh-
ter Hypatia, who was an important Neoplatonist teacher until lynched by
Christian monks; she also wrote her own commentaries, which unfortunately
do not survive. For a text of the commentary on the first four books of the
Almagest see Rome [19311943] vols. IIIII; for the rest one must resort to
Grynaeus-Camerarius [1538]. On Hypatia see Dzielska [1995].
Theon also composed two works on Ptolemys Handy
Tables. Both are self-standing treatises rather than commentaries in the strict
sense of the word. The Great Commentary originally comprised five books,
of which the first four are still extant, and the Little Commentary, which has
survived intact, is in one book. For text and translation see Tihon [1978], [1991],
and [1999] and Mogenet-Tihon [1985]. We also have an introduction to Euclids
Optica attributed to Theon; text and translation in Heiberg [1882] 139145.
Proclus, a Neoplatonist of the fifth century AD, has left us an intact four-
book commentary on the first book of Euclids Elements. The commentary is
based on a number of earlier works, including Eudemus of Rhodes lost History
of Geometry (c. 330 BC), lost works of Porphyry (3rd century AD), and commen-
The sources of our knowledge of ancient scholarship 487
taries on Euclid from the Roman period. The commentary is oriented toward
the curriculum of the Neoplatonist school and has philosophical and histori-
cal as well as mathematical value; it is frequently cited by modern scholars in
discussions of philosophy, mathematics, Euclid, and its lost sources. Friedlein
[1873] provides a text and Morrow [19922] a translation. Also attributed to
Proclus, but probably incorrectly, is a paraphrase/commentary on Ptolemys
Astrological Matters (text in Allatius [1635]).
Eutocius of Ascalon, who lived in the fifth and sixth centuries, has left us
three commentaries on works of Archimedes. They are important mathemati-
cal works in their own right and significant for our understanding of Greek
mathematics and its history. For text and translation see Mugler [1972] and
Netz [2004]. We also have a commentary by Eutocius on the surviving half
of the Conica of Apollonius of Perga (c. 200 BC). Though not as famous as
Eutocius commentary on Archimedes, this work has some philosophical and
mathematical value. For text and translation see Heiberg [18911893] vol. II.
Other surviving work includes an introduction to Euclids Data by Marinus
of Neapolis, a pupil of Proclus who lived in the fifth and sixth centuries (text in
Oikonomides [1977]). There is a long anonymous commentary from sometime
in the late antique period on Ptolemys that has no modern
edition (text in Wolf [1559]). Later commentaries on most of the mathematical
writers also exist. For further information on many of the works mentioned
here see Knorr [1989] and Mansfeld [1998].
From the Roman period there are also a variety of surviving works of literary
criticism. The literary works of Dionysius of Halicarnassus, from the late first
century BC, include critiques of Plato, Thucydides, and the major Attic orators,
a treatise On Literary Composition concerning word order and euphony in
both poetry and prose, and letters on various questions of literary history and
stylistic criticism. The treatise On the Sublime attributed to Longinus is prob-
ably from the first century AD and covers both poetry and prose from a wide
variety of genres and periods. Treatises surviving from the second century AD
and later are less important, but the rhetorical textbooks of Hermogenes of
Tarsus (cf. Matthaios in this volume), who was admired by the emperor Marcus
Aurelius, are particularly notable. Of the large corpus attributed to Hermogenes
only two works are clearly genuine, but the spurious works are also useful, as
are commentaries on the corpus by the fifth-century Neoplatonist Syrianus.
For more information on these works see Russell [19952].
We also have some surviving exegetical work on individual texts, particu-
larly the Iliad and Odyssey. The earliest of these is the
Homeric Problems (also known as Allegoriae Homericae Homeric Allegories)
attributed to Heraclitus and written in the first century AD (cf. Matthaios in this
volume); this work offers allegorical interpretations and defenses of Homers
treatments of the gods. Heraclitus sources included Apollodorus and Crates
of Mallos, and there is some debate about whether his work can be considered
particularly Stoic in orientation. Buffire [1962] provides a text and translation
and Bernard [1990] a discussion; cf. also Konstan-Russell [2005] and Pontani
[2005a].
Plutarch has left us several works on particular authors (cf. Matthaios in
this volume): Comparison of
Aristophanes and Menander, epitome (Mor. 853a854d) and
On the Malice of Herodotus (Mor. 854e874c), in addition to his
works on Plato mentioned above. A substantial essay entitled
On the Life and Poetry of Homer is attributed to him
but probably dates to the second or third century AD; the first part contains a
short biography of Homer, and the second part discusses interpretation. There
is a text in Kindstrand [1990] and discussion in Hillgruber [19941999].
The third-century Neoplatonic philosopher Porphyry has left us two works
on Homer (cf. Matthaios in this volume). One is an extended allegory on Odyssey
13.102112, the cave of the nymphs; this piece is crucial for understanding the
Neoplatonic interpretation of Homer. For discussion of its various editions see
Alt [1998] 466. Porphyry also composed a treatise entitled
Homeric Questions, which is believed to be based in part on Aristotles six-
book Homeric Questions (now lost except for a few frag-
ments). Porphyrys work is exegetical in nature and consists not of a linear
The sources of our knowledge of ancient scholarship 489
5.7 Eustathius
Eustathius, archbishop of Thessalonica, wrote a number of commentaries on
ancient authors in the twelfth century AD (cf. Pontani in this volume). The most
important of these are his massive works on Homer, but we also possess several
others. Eustathius based his commentaries on an impressive range of ancient
sources, many of which are now lost to us in their original form. He consulted
different manuscripts of the texts with which he worked and recorded variant
490 Dickey
readings, thus preserving for us the readings of manuscripts that have since
disappeared. He also made extensive use of scholia, lexica, and other scholarly
works, some of which no longer exist. In addition, he used works of ancient
literature other than the ones upon which he commented and thus sometimes
preserves fragments of those texts and variants otherwise lost.
The longest and most important of Eustathius works is his commentary on
the Iliad. This was written for students and educated general readers, rather
than for scholars, and is designed to be read with or without the text of the
Iliad. The author provided it with a marginal index, which appears to be an
invention of his own. The main source is the Homeric scholia (both those we
possess and others now lost), but many other works are also used. The com-
mentary on the Odyssey is similar but much shorter and less important. For
the Iliad commentary, of which we are fortunate enough to possess Eustathius
own autograph manuscript, the best text and discussion is that of van der Valk
[19711987]; for the Odyssey commentary see Stallbaum [18251826].
Eustathius other surviving commentary concerns a second-century didactic
poem by Dionysius Periegeta that describes the world. Though the poem itself
is not highly regarded today, the commentary (which is far longer than the
poem) is important for its preservation of portions of Strabo and of Stephanus
of Byzantium that do not survive elsewhere. Mller [1861] gives a text that has
been corrected by Ludwich [18841885] II.553597. Eustathius also wrote a
commentary on Pindar, of which we now have only the introduction. This is
useful primarily for quotations from odes that have since disappeared. For text
and discussion see Kambylis [1991a] and [1991b].
6.1 Hephaestion
Hephaestion of Alexandria, who lived in the second century AD, was the
author of the most important ancient metrical treatise and is now our main
source for ancient metrical theory, analysis, and terminology (cf. Matthaios
in this volume). His treatise originally comprised 48 books, but after repeated
epitomizing, much of it conducted by the author himself, we now have an epit-
ome in one book, known as the Encheiridion Handbook. There are also some
fragments of disputed authorship that could be excerpts from fuller versions
of the work, entitled On the Poem, On Poems,
and On Diacritical Marks.
The most important of these survivals is the Encheiridion, which discusses
and explains different metrical structures, illustrating them with extensive
quotations from ancient poetry. The two fragments on poems, the contents
of which overlap to a great extent, concern the analysis of poetic texts by met-
rical structure, and the discusses the use of the coronis, diple,
asteriskos, and other diacritic marks in different types of meter. Though not
designed as an introduction to the field, the Encheiridion soon became a text-
book because of its straightforward, systematic presentation and was used as
such for much of the Byzantine period. In consequence it accumulated an
extensive body of scholia and commentary, including a detailed and informa-
tive commentary by Choeroboscus (early 9th century). A reworking in verse by
John Tzetzes is also extant. Hephaestion continued to be the basis of metrical
theory until the 19th century, and while modern work on meter has tended to
move away from Hephaestions theories, his terminology is still standard in
the field.
Recently Hephaestion has been used chiefly in work on ancient metrical
theory, for which Hephaestions own work is crucial and the ancient commen-
tary on it is also valuable. The collection is however also very important as
a source of fragments of lost poetry, and for our understanding of Byzantine
classical scholarship. The standard text of all Hephaestions surviving work,
Choeroboscus commentary, and the scholia is that of Consbruch [1906];
Ophuijsen [1987] gives a translation and commentary.
6.2 Photius
Photius, patriarch of Constantinople in the ninth century, was the most impor-
tant of the Byzantine scholars (cf. Pontani in this volume). His influence was
responsible for the preservation of many ancient texts that would otherwise
have been lost, and his own work drew on, and thus preserves fragments from,
many other works that subsequently disappeared. For his Lexicon see above
492 Dickey
7 Fragments
The works discussed so far are all exceptional in that they survive as indepen-
dent entities; the vast majority of ancient scholars work is now either lost
or preserved only in fragmentary condition. The fragments may come from
extant works of ancient scholarship (thus Hipparchus commentary on Aratus
preserves fragments of Attalus commentary on the same subject, and the ety-
mologica preserve many fragments of Aristarchus), from scholia (see below
8), or from extant non-scholarly literature. Strictly speaking only the third
category really counts as a source of our knowledge about ancient scholarship
The sources of our knowledge of ancient scholarship 493
distinct from other sources discussed in this chapter, but for convenience a
number of important fragmentary authors will be grouped here even if some
or all their fragments come from sources discussed elsewhere in this chapter.
When working with fragmentary authors it is often convenient to use a col-
lected edition of the fragments, where they can be easily located and compared
with one another and where an editor has usually provided some information
about their probable original context. For this reason I refer to such collected
editions frequently in this section. However, many collections of fragments
are woefully out of date, relying on grossly inadequate editions of the texts
in which the fragments occur, and some of them were inadequate to begin
with owing to reliance on fanciful hypotheses about the nature of the original
work. So when using a collection of fragments it is important to pay attention
to the editions of source texts that were available to its compiler: when the
source text has received a better edition since the publication of the collection
it is essential to consult that edition as well as the collection. For example, if
one were to use Lentzs edition of Herodians fragmentary
On the Prosody of the Odyssey without consulting Pontanis edition
of the scholia to the Odyssey [2007], one would be relying on a text a century
and a half out of date and wholly unreliable.
Accentuation, and the rest come from a wide variety of other works of both
authors; predictably, the main source of the fragments is the Homer scholia.
Haas [1977] provides a text and discussion.
Philoxenus of Alexandria, a grammarian who worked in Rome in the first
century BC, wrote a variety of works that now exist only in fragments (cf.
Montana in this volume). His main work, On
Monosyllabic Verbs, was etymological (probably in the synchronic rather than
the historical sense) and concerned with deriving the Greek vocabulary from
a core of monosyllabic verbs (as opposed to the Stoic view that the base words
were nouns). The surviving fragments therefore come principally from Orion
and the etymologica, though scholia are also a major source. Theodoridis
[1976] gives a good text; see also Dyck [1982c], Koniaris [1980], Lallot [1991b],
and Pagani [forthcoming].
From Orus, who lived in the fifth century AD, we have a fragmentary trea-
tise on ethnics entitled On Ethnics or
How Ethnics Should be Spoken; the fragments come from Stephanus and the
Etymologicum genuinum, and Reitzenstein [1897] provides a text and discus-
sion; cf. also Billerbeck [2011].
There are numerous other fragmentary grammarians, most notably
Lesbonax (see Blank [1988]), Comanus of Naucratis (see Dyck [1988b]), Epa
phroditus (see Braswell-Billerbeck [2008]), Agathocles, Hellanicus, Ptolemaeus
Epithetes, Theophilus, Anaxagoras, and Xenon (see F. Montanari [1988]). There
is good information on late antique and Byzantine grammarians and their edi-
tions in Hunger [1978] II 383.
has been done in a scattered range of works. See especially Matthaios [1999],
Schironi [2004], Ludwich [18841885], and van Thiel [2014].
Other early scholars that can usefully be consulted in collections of frag-
ments include Crates of Mallos (Broggiato [2001] and [2013]), the Homerist
Heliodorus (Dyck [1993b]), and Theon (Guhl [1969]). There are elderly collec-
tions of fragments of Zenodotus, Aristonicus, Didymus, and Nicanor, but it is
not advisable to use them.
Larger commentary fragments tend to come from the same type of com-
mentary that is most often preserved intact, those on philosophical or sci-
entific works. Attalus of Rhodes produced an astronomical commentary on
Aratus Phaenomena in the early second century BC, and considerable frag-
ments of it survive via Hipparchus commentary on the same work (text in
Maass [1898] 124). Extensive fragments of a commentary to books one
through nine of Euclids Elements by Heron of Alexandria (1st century AD)
are preserved in Proclus commentary on the Elements and in a tenth-century
commentary by Anaritius (Al-Nayrizi), which was originally written in Arabic
and translated into Latin. (For editions and translations see Mansfeld [1998]
26 n. 90.) Anaritius commentary also preserves fragments of a commentary by
Simplicius (6th century) on book one of the Elements.
The third-century Neoplatonist Porphyry has left, in addition to surviving
works on Homer, Aristotle, and Ptolemy, fragments of commentaries on sev-
eral of Platos dialogs (texts in A. Smith [1993] and Sodano [1964]). From the
same century we have a substantial volume of fragments of commentaries on
the dialogs by Iamblichus (Dillon [1973]).
8 Scholia
The term scholia is a complex one. Its primary usage today, and the only one
nomally employed by those working on literature, is for explanatory notes
written in the margins of medieval manuscripts. In some contexts, however
(chiefly works on medical and philosophical writers, works written in modern
Greek, and some older work), scholia is simply a synonym for commentary
and says nothing about the way the material referred to has been transmit-
ted. Scholars differ about whether to call marginal notes in papyri scholia, as
such notes tend to be different in character and origin from those in medieval
manuscripts. Here I use the term only for marginalia in medieval manuscripts;
annotations in papyri are discussed above (see 2).
Scholia are of crucial importance to the study of ancient scholarship, as
they provide most of the information we have about Alexandrian scholarship
498 Dickey
8.1 Homer
The scholia to the Homeric poems are the largest and most important group of
Greek scholia, containing a significant amount of Alexandrian material. They
are traditionally divided into three different groups (excluding the Byzantine
scholia, which form several groups of their own but are often disregarded as
they do not transmit a significant amount of ancient scholarship); the divi-
sion is based on the history of the Iliad scholia, which is much clearer than
that of the Odyssey scholia, but it is generally considered to apply to both sets
of scholia.
The three groups are known as the D scholia, the A scholia, and the bT scho-
lia, classified according to the manuscripts in which they are found. As some
scholia are found in several different types of manuscript and therefore fulfill
the criteria for more than one group, the classification is hierarchical: identifi-
cation as a D scholion takes precedence over either of the others, and identifi-
cation as an A scholion takes precedence over identification as a bT scholion.
The D scholia are unfortunately named after Didymus, with whom they are
now known to have no connection; they are also known as scholia minora or
scholia vulgata, and from the latter name they can be called V scholia. They are
the largest group of Homeric scholia, and our earliest manuscript evidence for
them is older than that for the other types of scholia, for the chief witnesses to
the D scholia are manuscripts Z and Q, which date to the ninth and eleventh
centuries respectively. D scholia are also found in a wide range of other manu-
scripts; many are very short and can appear as interlinear glosses, but others
are more substantial.
The D scholia represent the ancient vulgate tradition of interpretation, the
explanations familiar to laymen and used in elementary instruction rather
500 Dickey
scholia are clearer than is the case with most scholia, for at the end of almost
every book the scribe added a subscription indicating their source:
,
Written beside (the text) are Aristonicus Signs and Didymus On the
Aristarchean edition, and also some extracts from Herodians Iliadic prosody
and from Nicanors On punctuation. The principal basis of the A scholia is
therefore the four works cited in this subscription (all of which are now lost
except insofar as they are preserved in the scholia), but it is unlikely that the
scribe who wrote it was actually copying from the works themselves. Rather
his source, or more likely his sources source, was a compilation of these four
works (and some other material) probably made around the fourth century
AD and known today as the Viermnnerkommentar four mens commentary
or VMK.
All four elements of the Viermnnerkommentar represent Alexandrian
scholarship to a significant extent. Aristonicus treatise on signs, com-
posed in the Augustan period, was a compilation of excerpts from one of
Aristarchus commentaries and from other works, focusing on critical signs.
Didymus work, probably also from the Augustan period but later than that
of Aristonicus (which Didymus probably used), was a compilation based pri-
marily on Aristarchus commentaries, though his focus was on textual variants.
Herodians treatise on Homeric accentuation, from the late second century
AD, also drew heavily on Aristarchus commentaries, and Nicanors work on
punctuation, from the first half of the second century AD, was based on ear-
lier works including those of the Alexandrians. The A scholia are thus a major
source of information about the opinions of Aristarchus and, to a lesser extent,
other Alexandrian scholars; they contain more than a thousand explicit refer-
ences to Aristarchus. They are of crucial importance for our knowledge of the
text of Homer, the goals and methods of Alexandrian scholarship, and ancient
systems of accentuation, punctuation, etc. (cf. Montana in this volume).
The A scholia also contain material that probably does not derive from the
Viermnnerkommentar. This information is more interpretive in nature and is
related to material found in the bT scholia; A scholia of this type are also called
exegetical scholia and as such are grouped with the bT scholia.
The bT scholia are so called because they are found in manuscript T
(11th century) and in the descendants of the lost manuscript b (6th century).
They contain some Alexandrian material (much of it attributable to Didymus)
but seem to come more immediately from a commentary of the late antique
period (known as c), of which b produced a popular and T a more scholarly
version. These scholia are also known as the exegetical scholia, because they
502 Dickey
are concerned primarily with exegesis rather than textual criticism. They
include extensive extracts from the Homeric Questions of
Porphyry and the Homeric Problems of Heraclitus (see
5.5 and Matthaios in this volume). Until recently the bT scholia were thought
to be much less valuable than the A scholia (whose worth has been recognized
since the eighteenth century), because of the limited extent to which they can
aid in establishing the text of the Homeric poems. In the past few decades,
however, an increasing interest in ancient literary criticism has brought these
scholia into new prominence, and they are currently at the center of modern
work on ancient Homeric scholarship.
The scholia to the Odyssey are much fewer and less well preserved than
those to the Iliad. This distinction goes back to antiquity, when the Iliad was
considered the superior work and so was read and copied much more often
than the Odyssey. Nevertheless it is clear that the Alexandrians produced texts
and commentaries on both poems, and that ancient scholars discussed the
interpretation of the Odyssey as well as that of the Iliad. Thus equivalents of
all three groups of Iliad scholia can be found for the Odyssey scholia: there
are Alexandrian text-critical scholia, exegetical scholia of the bT type, and
D scholia (often called V scholia in this context). However, because there is no
equivalent of Venetus A among the Odyssey manuscripts the different types
are not so easily separable by manuscript source.
In addition to the uses of the Homer scholia already mentioned, they are
important for the understanding of post-Homeric literature. Much of this liter-
ature, both Greek and Latin, was based to some extent on the Homeric poems,
but not on the Homeric poems as we read them: rather on the Homer of the
scholiasts. Authors such as Apollonius Rhodius and Virgil drew on and alluded
to Homer based on the readings and interpretations current in their own time,
and therefore the scholia provide us with information crucial for understand-
ing their poems.
Most of the A and bT scholia to the Iliad are best consulted in the superb
edition of Erbse [19691988]. This edition is highly selective and tries to rep-
resent an early stage of the A and bT traditions, a feature that makes the most
famous scholia readily available and easy to consult but also results in the
omission of many scholia from different traditions, some of which are impor-
tant. The omitted material includes all the D scholia, the bT scholia derived
from Porphyry and Heraclitus, and some other material that cannot easily be
assigned to any of the three main groups of scholia, not to mention all the
Byzantine scholia. The seven volumes of Erbses edition thus represent only a
small fraction of all the preserved scholia, and since many scholia appearing
in codex A are omitted from the edition because they belong to the D family,
The sources of our knowledge of ancient scholarship 503
while others appearing in manuscripts of the b family are ignored because they
come from Porphyry or Heraclitus, the edition is not even a complete collec-
tion of the scholia appearing in the manuscripts included. The D scholia can
be found in van Thiels edition [2000], and the Porphyry and Heraclitus scholia
can be found in Sodano [1970] and Buffire [1962] respectively; for Porphyrius
cf. MacPhail [2011]. These last two groups have been translated (Buffire [1962]
and Schlunk [1993]), but the rest of the Iliad scholia, like nearly all extant scho-
lia, have never been translated into any language.
A comprehensive edition of the Odyssey scholia is in progress (Pontani
[2007]), but so far it covers only scholia to the first four books of the poem.
For the rest of the poem one can use Ernst [2006] for the D/V scholia and the
inadequate work of W. Dindorf [1855] for the others. Of the enormous litera-
ture on these scholia some important works are those of Erbse [1960], van der
Valk [19631964], Montanari [1979], Schmidt [1976], Meijering [1987], Schmit-
Neuerburg [1999], Schlunk [1974], and Rengakos [1993] and [1994].
8.2 Hesiod
The scholia to Hesiod are voluminous, useful, and of impressive antiquity.
Ancient scholarship on Hesiod began early, for lost interpretive works appear
to date at least as early as Aristotle, and the first critical text was produced
by Zenodotus. Zenodotus, Apollonius Rhodius, Aristophanes of Byzantium,
Aristarchus, Crates, Aristonicus, and Didymus all left textual or interpretive
comments on Hesiod that are still preserved under their names, though they
did not all write full commentaries on the poems (cf. Montana in this volume).
The oldest portion of our surviving scholia comprises the remains of a com-
posite commentary of uncertain authorship (Choeroboscus and Dionysius
of Corinth have both been suggested, but the author could be completely
unknown). This commentary was a compilation of earlier writings, including
both grammatical and critical notes from Alexandrian and other scholars and
paraphrases from school texts; an important source seems to be the commen-
taries of Seleucus (1st century AD). In general, the material seems mostly to
come from before AD 100.
In addition to the direct transmission of this commentary as scholia attached
to the text of Hesiod, there is an indirect transmission via several etymological
works, particularly the Etymologicum Genuinum. The authors of these etymo-
logica quoted extensively from the scholia to Hesiod, and the scholia to which
they had access were better preserved than those in the manuscripts we pos-
sess, as well as being unmixed with any later commentaries.
In the fifth century AD the Neoplatonist Proclus wrote a philosophical com-
mentary on the Works and Days. Proclus made extensive use of the earlier
504 Dickey
composite commentary, of which he had a fuller version than that now pre-
served in the scholia, and he also drew heavily on a commentary by Plutarch
on the Works and Days. Plutarchs commentary is now lost in its original form,
but Proclus survives largely intact in the scholia and preserves significant
portions of Plutarchs work. In our manuscript scholia to the Works and Days
Proclus commentary has been mixed with the scholia derived from the earlier
composite commentary, but a few manuscripts mark the notes from Proclus
commentary with special symbols, so they are relatively easy to separate. There
is also a substantial amount of Byzantine commentary on Hesiod.
The most important editions of Hesiod scholia are Di Gregorios [1975] for
the Theogony and Pertusis [1955] for the Works and Days. For an overview of
discussion see M. L. West [1978] 6375, with bibliography p. 91, and Montanari
[2009a].
8.3 Pindar
The voluminous scholia to Pindar offer abundant ancient material remark-
ably unmixed with later additions and are useful for a number of different
purposes. Because of the extent to which these purposes diverge, discussions
and even editions of Pindar scholia often cover only one type of material. The
main divisions are between metrical and non-metrical and between old and
Byzantine scholia.
There is a large body of old metrical scholia, compiled probably in the fifth
century AD and based on a metrical analysis of the Odes written in the sec-
ond century AD. This analysis incorporated a commentary by Didymus that
transmitted the work of Alexandrian scholars and was based on the text and
metrical divisions established by Aristophanes of Byzantium (cf. Montana in
this volume); its medieval transmission was in part separate from that of the
text of the Odes and their non-metrical scholia. Scholars now generally agree
that Aristophanes colometry and the Alexandrian metrical analysis do not go
back to Pindar himself and that in consequence the metrical scholia are of lit-
tle use for understanding Pindars own metrical intentions. They are however
very important for our understanding of ancient metrical theory, since their
detailed, line-by-line analysis (with continuous texts often resembling a trea-
tise rather than traditional scholia) offers one of the few surviving examples
of the practical application of the theories preserved in Hephaestions manual
(see 6.1 above).
The exegetical scholia to Pindar are more numerous than the metrical scho-
lia and have an equally impressive pedigree, since they preserve the remains
of commentaries by Aristarchus and several of his successors, incorporated
into a comprehensive work by Didymus and then epitomized in the second
The sources of our knowledge of ancient scholarship 505
century AD. Like the old metrical scholia, they are virtually free of late interpo-
lations, so that almost any piece of information found in them can be assumed
to come from the Alexandrians (though not necessarily without abridgement
and alteration). These scholia attempt to explain the difficulties of the odes
and offer an interpretation of the poets meaning. In doing so they invoke
historical, biographical, and mythological data, some of which appear to derive
from accurate transmission of information going back to Pindars own time,
though parts seem to be simply Alexandrian conjecture based on the poems
themselves. The proportions in which these two types of material occur, and
therefore the extent to which one can rely on information provided by the
scholia but not otherwise verifiable, are the subject of debate. It is however
clear that the interpretations found in the scholia were widely accepted in
antiquity, for they are reflected in later poetry influenced by Pindar, such as
that of Theocritus, Callimachus, and Horace.
The scholia to Pindar are frequently cited by modern scholars, most often in
discussions of Pindaric interpretation, for which they remain crucial, but also
for historical and mythological information that can be used for other pur-
poses; they are of course also very useful for work on ancient metrical theory
and on the evolution of scholia. Their value for establishing the text of Pindar
is high, as they sometimes preserve the correct reading for passages that have
been corrupted in all extant manuscripts of the text.
The best text of the Pindar scholia is that of Drachmann [19031927], but for
the metrical scholia it is better to use Tessier [1989]; Arrighetti et al. [1991] pro-
vide a concordance. Discussions include those of Irigoin [1958a], Deas [1931],
and Lefkowitz [1991].
8.4 Tragedy
Scholia to Euripides, Sophocles, and Aeschylus generally seem to have similar
origins and history, but as work on Euripides is better preserved than that on
the other two tragedians extrapolations need to be made from Euripides scho-
lia to those on Sophocles and especially Aeschylus.
Only nine of the nineteen surviving plays of Euripides have preserved scho-
lia: a large amount of annotation exists for the Byzantine triad (the texts usu-
ally read in the later Byzantine period) of Orestes, Hecuba, and Phoenissae,
and less extensive but still substantial notes survive on the Medea, Hippolytus,
Alcestis, Andromache, Rhesus, and Troades. For most plays the scholia are eas-
ily divisible into old and Byzantine scholia, though in the case of Rhesus and
Troades the two types are more difficult to separate.
The oldest Euripides scholia go back to the work of Aristophanes of
Byzantium, who established the Alexandrian text and colometry of Euripides
506 Dickey
pect that only some material from Dicaearchus epitomes survived as part
of a collection compiled in the first century BC or AD; cf. now the edition by
Meccariello [2014].
Scholia to Sophocles are less numerous than those to Euripides but
more evenly distributed among the different surviving plays. The old scholia
are based on a composite commentary by Didymus (drawing on Alexandrian
sources), along with material from the Roman-period scholars Pius,
Sal(l)ustius, Herodian, Diogenianus, and others. For reasons that are not quite
clear, the Oedipus at Colonus has the most useful and informative old scholia.
The most important manuscript of Sophocles, the tenth-century L, has only
old scholia and is our primary source for the ancient material. However, some
other manuscripts also contain old scholia, which they sometimes report more
fully than does L, and the Suda and the Etymologicum Genuinum contain rem-
nants of more old scholia in a fuller form than that found in L. The old scholia
are frequently used for historical, textual, lexical, and interpretive informa-
tion. There is also a large body of Byzantine scholia, attached primarily to the
Byzantine Triad of Ajax, Electra, and Oedipus Rex.
The hypotheses to Sophocles plays show many similarities to those of
Euripides. As in the case of Euripides, multiple hypotheses to individual plays
have been preserved via the manuscript tradition, and it is clear that several
different types of hypothesis existed already in antiquity, with the oldest being
based on the introductions written by Aristophanes of Byzantium. Papyri of
non-Aristophanic hypotheses without the plays themselves exist, indicating a
phenomenon like that of the Tales from Euripides, but because these papyri
are fewer and differ in some important respects from the Tales from Euripides
papyri, the nature and purpose of these hypotheses is less well understood
than that of their Euripidean equivalents.
The old scholia to Aeschylus are found primarily in the tenth-century man-
uscript M. They contain material from the Hellenistic and Roman periods,
including some that is almost certainly Alexandrian; it is sometimes argued
that these scholia derive from a commentary by Didymus, but this theory
remains unproven. Also of considerable antiquity, but more altered in trans-
mission, is the material in the A or F scholia, which derive from a commentary
ascribed (probably falsely) to John Tzetzes. These scholia are sometimes nearly
valueless, but at other times they provide ancient material omitted or abridged
in M; it is clear that their author was using a manuscript with ancient scholia
very similar to those in M but without some of Ms errors and omissions. The
F scholia are much longer and more numerous than the other classes of scho-
lia but exist only for the Byzantine triad (Prometheus, Persae, and Septem, the
plays normally read in the later Byzantine period). In addition, there are many
Byzantine scholia to Aeschylus.
508 Dickey
Texts of the scholia to the tragedians are scattered and in some areas incom-
plete. The most important editions are Schwartzs text [18871891] of the old
scholia to Euripides; texts by Christodoulou [1977], De Marco [1952], Longo
[1971], Papageorgius [1888], Xenis [2010a] and [2010b] of scholia to Sophocles;
and texts by O. L. Smith [1976b] and [1982] and Herington [1972] of scholia to
Aeschylus. On editions of scholia not covered by these texts see Dickey [2007]
3138. Hypotheses are normally printed with editions of the plays rather
than with the scholia. The hypotheseis on papyri are now edited in CLGP I 1.1.
Important discussions include Gnther [1995] on Euripides and O. L. Smith
[1975] on Aeschylus.
8.5 Aristophanes
The scholia to Aristophanes are among the most important sets of scholia, in
part because they provide background information without which many of the
jokes and allusions in the comedies would be incomprehensible. They are rela-
tively well preserved, and most of them can be found in a sound and reliable
modern edition, making them easier to use than many scholia.
The old scholia to Aristophanes are derived from a variety of sources going
back to the beginning of Alexandrian scholarship. Callimachus, Eratosthenes,
and Lycophron (a contemporary of Zenodotus) all worked on Aristophanes to
some extent, and the first continuous commentary on his plays was produced
by Euphronius, the teacher of Aristophanes of Byzantium. Aristophanes of
Byzantium himself produced an edition of the plays, providing an introduc-
tion to each (the extant verse hypotheses of the plays are thought to be distant
descendants of these introductions) and may also have written a commen-
tary; Callistratus and Aristarchus probably wrote commentaries on the plays,
and Timachidas of Rhodes wrote one on the Frogs (for Alexandrian works on
Aristophanes plays see Montana in this volume).
The work of these and other scholars was combined into a single commen-
tary by Didymus in the late first century BC or early first century AD, and some-
time in the first two centuries AD Symmachus compiled another commentary,
using Didymus as his main source but also consulting other works. At a later
date Symmachus commentary or one of its descendants, along with some
other material, was copied into the generous margins of a book of the plays of
Aristophanes and formed the archetype of our extant scholia.
Perhaps the most important of the additional sources of our scholia is the
metrical commentary on Aristophanes written by Heliodorus around AD 100.
This commentary is often studied apart from the other scholia, for it is crucial
for our understanding of ancient metrical theory but of limited use in under-
standing Aristophanes. Heliodorus work has been preserved to varying extents
The sources of our knowledge of ancient scholarship 509
for the different plays; one can reconstruct from the scholia nearly all of it for
the Peace, as well as substantial sections of it for the Acharnians and Knights
and some fragments for the Clouds and Wasps, but little else.
In addition to the direct tradition of the scholia, which is well attested in
several manuscripts, there is an indirect tradition via the Suda, whose writer
had access to the same body of material when it was more complete and there-
fore often preserves scholia that did not survive in the direct tradition. There
are also a number of papyri and ancient parchment fragments with commen-
taries or scholia on Aristophanes; on the whole, those of the fourth century
and later seem to reflect a body of material very similar to the ancestor of
our scholia (though in some places more complete), while the earlier ones,
which are much rarer, apparently belong to different traditions. There are also
Byzantine scholia to Aristophanes, especially on the triad of plays made up of
the Plutus, Clouds, and Frogs.
The best edition of the scholia is a multivolume work edited first by Koster
and later by Holwerda [Koster-Holwerda 1960], which includes both old and
Byzantine scholia, usually in separate volumes. Rutherford [18961905] pro-
vides translation and commentary for many scholia (cf. also Chantry [2009]),
and White [1912] 384421 extracts the Heliodorus fragments from the scholia
and groups them together with a good discussion. For further discussion see
Montana [2006a].
a ppealing series of illustrations was provided. Most of the new material came
from a work known as the Catasterismi of Eratosthenes, which appears to be
the late epitome of a lost astronomical treatise probably written (by the third-
century BC scholar and mathematician Eratosthenes) as an elementary and
literary astronomy manual designed to complete and explain Aratus. The edi-
tor of apparently took extracts from this original work and re-arranged them
in the order of Aratus poem to enhance the appeal of his new edition.
The edition proved wildly popular and soon replaced the scholarly edi-
tion entirely in the West; in the Byzantine world both editions existed side by
side, resulting in extensive cross-fertilization of the explanatory material. As
a result, while some surviving manuscripts (most notably M) contain scholia
largely derived from the earlier edition and others (notably S and Q) contain
substantial amounts of explanatory material from the edition, manuscripts
of the earlier edition generally show at least some influence from . Much of
the commentary has however been lost in Greek; the Anonymus II con-
sists primarily of a Latin translation of the edition made in the seventh or
eighth century and known as the Aratus Latinus. There is also a third-century
Latin translation of preserved as scholia to the first-century Latin translation
of Aratus attributed to Germanicus Caesar. For texts and discussions see Jean
Martin [1974] and [1998] and Maass [1892] and [1898].
There is a considerable body of scholia to the Alexandra of Lycophron, in
fact much larger than the poem itself. It is divided into two groups: old scholia,
which go back to Theon, and Tzetzes scholia. The scholia are rich in mytho-
graphical information and also useful as evidence in the debate as to whether
the author of the Alexandra can be identified with the Lycophron who was
a tragedian of the third century BC or whether the poem was composed by
another Lycophron in the second century BC. Scheer [1908] vol. II provides a
text and discussion. There is a new edition by Leone [2002].
The Theriaca and Alexipharmaca of Nicander both have large bodies of
scholia, with sources including Theon and Plutarch. They cover a wide vari-
ety of topics; while much of this material is late, some of it preserves valuable
ancient commentary. The scholia are used particularly for the information they
provide on the history of the poems and Nicanders other writings. Crugnola
[1971] and Geymonat [1974] provide texts.
The scholia to the Batrachomyomachia are mostly Byzantine and have
attracted little attention in recent years; for text and discussion see Ludwich
[1896] 117135 and 198318.
chapter 2
Definitions of Grammar
Alfons Wouters and Pierre Swiggers
1 Introduction
2 Eratosthenes of Cyrenes Definition of Grammar
3 Dionysius Thrax Definition of Grammar
4 From Dionysius Thrax to Sextus Empiricus: Ptolemy the Peripatetic,
Asclepiades of Myrlea, Chaeris and Demetrius Chlorus
5 Sextus Empiricus: The Deconstruction of (the Definitions of ) Grammar
6 Dionysius Thrax Definition of Grammar and Its Afterlife
7 Conclusion
1 Introduction
1 See for more details Swiggers-Wouters [1995a], [1995b], and [2004], and Montana in this
volume.
Ptolemy III Euergetes (cf. Fraser [1970]; Geus [2002]; Matthaios [2008] 556569
and Montana in this volume). This definition, which would reach back to the
3rd century BC, has been transmitted, indirectly, in the scholia (Sch. Dion. T.
160.1012) to Dionysius Thrax grammar manual. The definition, also briefly
explicated by the scholiast, runs as follows:
. The immediately subsequent explication concerns the last word
() of the definition: , by letters
(Eratosthenes was) intending written texts / writings.
Taken on itself the definition looks straightforward, given its limpid struc-
ture: it consists of the definiendum followed by the copula introducing the
definiens. However, the definition raises various problems of different nature.
Matthaios [2011a], in a foundation-laying study, analysed Eratosthenes defini-
tion from a threefold point of view: historical, theoretical and ideological. We
will focus here on the linguistic-historiographical aspects, but two preliminary
remarks are in order:
The passage raises many textual, and even more exegetical problems, not
only for the identification of the exact source responsible for the change
5 , but also for the identification of the subsequent
defenders of the views involved. A text-external complication comes from the
fact that in Suetonius De grammaticis 10.4 Eratosthenes is said to have claimed
the title philologus for himself:
3 As it appears from Sch. Dion. T. 3.2426; 7. 2425 (cf. infra) and 448.6, must be
a mistake for . The grammarian Antidorus is otherwise unknown, but lived very
probably in the first half of the 3rd cent. BC. Cf. Pfeiffer [1968] 157158.
4 See Pfeiffer [1968] 158: The literary work of Praxiphanes (the Peripatetic scholar of the 4th
3rd cent. BC) is here regarded as foreshadowing the work of the Alexandrian .
5 The Alexandrian poet and philologist Philetas of Cos (4th3rd cent. BC) appears to be the
first who adopted this title. Cf. Strabo XIV 657: .
518 wouters and Swiggers
6 See also Dihle [1998] 89: Eratosthenes Anspruch auf die Bezeichnung knnte [...]
bedeuten, dass er es ablehnte, auf Dichterexegese festgelegt und eingeschrnkt zu werden,
und dass er auf die Vielseitigkeit seiner wissenschaftlichen Ttigkeit verweisen wollte.
7 See the references in Matthaios [2011a] 64, n. 34.
8 Cf. Matthaios [2011a] 65, who cites a fragment of Philicus of Corcyra (dated to 275/4 BC).
definitions of grammar 519
The genesis of the name for this activityand the institutional achieve-
ment of establishing such a field of scholarly studycan be traced back to the
early stages of Alexandrian philological activity. Our sources on this, viz. Sextus
Empiricus (cf. infra) and the scholia on Dionysius Thrax, seem to have derived
their information from Asclepiades of Myrleas work (2nd
1st cent. BC; cf. Pagani [2007a] 3134). It was probably through Asclepiades
report on the evolution of the discipline and of its name, that in the scholia on
Dionysius Thrax we read that it was Antidorus who gave the name
to the discipline consisting in (having / producing / showing) knowledge
about written texts:
(Sch. Dion. T. 7.2425).
In Antidorus view grammar was thus defined in relation to a (further
unspecified) knowledge / knowing of writings. Although we cannot be sure
whether this was indeed Antidorus authentic wording, the account of the
scholia conforms to what we may reasonably expect: an initially broad and
open conception of a field constituting itself, and primarily defined by its
material object (viz. texts). Eratosthenes was to provide a further specification
of the discipline with his (more) explicit definition, which we will now com-
ment upon.
Eratosthenes definition, as transmitted by the scholia on Dionysius, con-
tains three components calling for an analysis: the substantive , the quali-
fying adjective and the adverbial syntagm . As we have
seen, the scholia explicate as referring to writings (),
an equation which is attributed to Eratosthenes himself (although we have no
further textual evidence for such an equation).
The two other components require a close analysis. The term , which is
used as the definiens (at least if we read the passage as a definition of the form
x is (a) y of such and such nature), should be interpreted as designating the
genus under which the falls.This interpretation also holds when
we read the passage as the () is the that .... The term
has a wide range of meanings, which can be reduced to two major semantic
fields: (a) experience, habit, customized behavior, (b) skill, faculty, capacity.
It is especially the latter semantic field that occurs in the syntactic combina-
tion + followed by the complement of the preposition. The matter has,
however, to be considered in the context of ancient theory of knowledge and
philosophy of science. The term has a long-standing history in philosophi-
cal discussions concerning the nature and the various types of knowing. The
following chronological sequence may serve as an illustration (for a more
detailed analysis, see Matthaios [2011a] 7076):
520 wouters and Swiggers
(3b) Zeno of Citium (335263 BC) seems to have formulated the core defi-
nition of , as a track-making capacity / experience, following a way
and a method (the latter explicative part may be a later addition):
, ,
(Sch. Dion. T. 118. 1416 = SVF I 72).
(4) The term , used to define or used in relation with , can still
be found in the context of epistemologically based discussions on the status of
grammar, as we find them in the works of Sextus Empiricus (cf. infra) and in the
scholia on Dionysius Thrax. The scholia provide us with the following defini-
tions, the first of which may be more or less contemporary of Sextus times (or
simply a rewording of Sextus definition), the latter two being distinctly later:
(4a)
(Sch. Dion. T. 118. 1012).
(4b)
,
(ibid. 3. 1113).
(4c)
,
.
(ibid. 164.58).
While the definitions under (4) postdate the one attributed to Eratosthenes,
it is clear from the preceding list that there was, specifically since Aristotle,9
a tradition of defining cognitive aptitudes and habits in terms of , and
that the latter term occurred in relation with an appraisal of the status of
. Eratosthenes use of the term was thus a deliberate epistemo-
logical choice, and a clear indication of the recognition of (or of
the ) as belonging to the realm of sciences. Eratosthenes seems
to have understood as that property (or skilled experience) of the
one who possesses it, which deals with writings. Now, this property / skilled
experience / habitual capacity is said to be . This qualifying adjective
is not functioning as a differentia specifica here, and for this reason we cannot
take the phrase attributed to Eratosthenes as a definition that would be fully
in line with the ordering of concepts in a Porphyrean hierarchical structure.
The term is not a subspecifying term, but rather a laudatory epithet,
highlighting the value or importance of the in question. The problem then
is to properly understand what Eratosthenes may have meant with .
Since later definitions of grammar (cf. infra) do not contain the word ,
9 As noted by Matthaios [2011a] 75, in his De partibus animalium, Aristotle distinguishes
two types of : one that exists in the knowledge () of an object, the other being
acquired through education.
522 wouters and Swiggers
and are more explicit on the material object of the grammarians knowledge or
capacity, we may infer that the evolutionary line in the definitions of grammar
was one of narrowing down the scope of the grammarians field of knowledge /
experience / capacity. Since Eratosthenes, following shortly after Antidorus,
stands at the beginning of a process in which the general knowledge and
know-how of the was first narrowed down to skillful knowledge
and handling of writings, and was later reduced to literary writings (and to
what is mostly found in the writings of poets and prose-writers; cf. infra), we
may read into Eratosthenes phrase a global and holistic valuation of the
of the grammarian: the latter is creditedin an absolute10 waywith a full-
fledged skill / property, with respect to the object of his study, viz. writings. The
object itself is left in its full generality: , without further determina-
tion or specification.
In his definition of (or of ), Eratosthenes thus
propounded a very large view of the field of grammar, invested with encom-
passing cognitive aims. As Matthaios [2011a] 79 judiciously points out,
Eratosthenes connected the potential of the philological discipline with a
demand for universal knowledge [...]. His grammar definition goes along
with the universality of knowledge that Eratosthenes claimed for himself with
the title . The phrase merely serves to specialise the field
of universal knowledge, which grammar covers.
10 Matthaios [2011a] 78 speaks of the intensification of the term with the attribute
.
11 Dionysius Thrax Techn has been transmitted to us in medieval and Renaissance manu-
scripts dated between the 10th and 18th century. The text was translated, somewhere in
the fifth century, into Syriac and Armenian. The manuscript tradition of the Syriac and
Armenian translations extends between the 7th and 9th century (for the Syriac manu-
scripts), and between the 14th and the 17th century (for the Armenian manuscripts).
Starting with the oldest manuscript testimonies the transmitted text of the Techn is
accompanied by four supplements (cf. G.G. I 1, 105132): (1) on diacritic signs for accents,
quantity, aspiration; (2) on the definition of techn; (3) on metrical feet; (4) a paradigm list
of verb forms for tupt to beat/hit (see Dickey in this volume).
definitions of grammar 523
(a) the main gist of Dionysius Techn reflects the state of grammatical
knowledge attained in the first centuries BC
(b) whatever the relationships that can be established between Dionysius
Techn and the grammatical contents of later Greek and Latin manuals, the
initial sections of the Technwhich include the definition of grammar
should be considered authentic, i.e. dating back to the 2nd1st century BC.
It is at the beginning of the Techn, in the first section, that we find a defini-
tion of grammar. The Greek text reads as follows:
.
,
,
, ,
, ,
(G.G. I 1, 56.3).
This passage contains (a) the definition of grammar, properly speaking; (b) a
division of grammar into six parts, which are successively enumerated. The
definition is philologically oriented, and refers to the didactic practice of trans-
mitting philological expertise and the relevant theoretical and terminological
foundations. The components of the art of grammar are defined with refer-
ence to the classroom-situation of text reading and explanation. In the descrip-
tion of the parts of grammar almost nothing is said about language structure,
12 See the most recent surveys of the discussion by Matthaios [2009a], Callipo [2011] 2834,
and Pagani [2010b] 393409; [2011] 3037.
524 wouters and Swiggers
except for the subdivision dealing with analogy; an important role is attributed
to the reading (aloud) of texts, and to the appreciation of poems.
The definition of grammar given in the Techn reflects the activity of the
or teacher of literature which included the following parts
(cf. Sch. Dion. T. 453. 1923): the reading of the text, the indication of the poeti-
cal figures of expression, the explanation of problematic words and historical
references, the analysis of the origin of lexical items, the indication of (certain)
grammatical regularities and finally a judgement of the literary work.
These components are said to constitute the empeira of the grammarian. As
noted above, the notion empeira refers to a particular skill, based on acquain-
tance or familiarity (in this case, without theoretical knowledge). Through its
occurrence in the title of the treatise techn has a twofold meaning: it refers to
(the) art (of grammar), but it also designates a treatise in which the principles
of the art in question are expounded.13 The sixth part of the grammarians
empeira, viz. the judgement on poems, is called the finest part of all those
[contained] in the techn. Although one might be tempted to consider this
a slight inconsistency in the (original?) text, one could also view this passage
as reflecting an attempt on Dionysius part to overcome the tension between
the concepts empeira and techn in the definition of grammar, by interpreting
techn as the systematization (also for didactic reasons) of the practical exper-
tise (i.e. empeira).14
In Dionysius Thrax view, this empeira concerns
. Traditionally the passage is understood
as the expressions used for the most part among poets and prose-writers: the
15 See for example the following translations: Steinthal [18901891] II, 174: Grammatik ist
die Kunde der bei Dichtern und Prosaikern durchschnittlich vorkommenden Redefor-
men; Kemp [1987] 169: grammar is the practical study of the normal usage of poets and
prose writers; Lallot [1989] 41: La grammaire est la connaissance empirique de ce qui
se dit couramment chez les potes et les prosateurs; Robins [1996] 10: is
the of the general usage of poets and prose writers; Bcares Botas [2002] 35:
La gramtica es el conocimiento de lo dicho sobre todo por poetas y prosistas; Callipo
[2011] 57: La grammatica la conoscenza empirica delle cose dette per lo pi da poeti e
prosatori.
16 Whereas some modern commentators have made a strong case of the difference between
the two adverbial phrases and , others (e.g. Lallot [1989] =
[19982] 69 and Sluiter [2001] 310 n. 46, with some reservation) are inclined to consider the
two formulations as equivalent. In fact, the scholiasts on the Techn tend to treat the two
formulations as equivalent. The long and complex transmission history of the Techn,
and of commentaries on the text, may also have led to the non-distinctive use of the two
formulations.
17 A strong additional argument for relating in the text of the Techn to
, is offered by the definition of the ars grammat-
ica given by Varro (fr. 234 Funaioli), viz. ars grammatica [...] scientia est <eorum> quae
a poetis historicis oratoribusque dicuntur ex parte maiore (grammar is the knowledge
526 wouters and Swiggers
a rgument is that the translation (as given by Lallot) ce qui se dit couramment
chez les potes would be contradicted by the third of the parts of grammar
(cf. infra) that are listed immediately after the definition, viz. the
or llucidation des mots rares (Lallot [1989] = [19982] 69). But this
argument does not seem valid. Lallot [1989] = [19982] 7778 has convincingly
demonstrated that for the ancient grammarians the term indicates
those words that were unusual in the grammarians own language, i.e. obsolete
words or words from the various Greek dialects, which over centuries were
used in poetry and prose. The term therefore should not be taken to
imply that the words in question were unusual in the language of the authors
explained by the grammarian.
In short, as we have already argued elsewhere in more detail (cf. Swiggers
Wouters [1994] 532534), Patillons counterarguments to the traditional inter-
pretation and translation of Dionysius definition are not acceptable. It is fur-
thermore unclear what other texts than those of poets and prose-writers would
constitute an object of study for the grammarian. However, the debate clearly
is not finished. Ventrella [2004] 110 prefers for the Techn the text reading
(cf. the passage in Sextus) instead of and considers
it a qualification (che venga pi volte ripetuta, spesso esercitata) of ,
which explains his translation: la grammatica pratica ripetuta / esercitata di
ci che si dice presso poeti e prosatori.
We will now briefly comment upon the division of grammar as specified in
Dionysius Thrax definition.
of what is said by poets, historians and orators for the most part). This definition is gener-
ally considered to be a literal translation of Dionysius definition, although prose-writers
have been divided into historians and orators. There can be no doubt that in Varros text
too the qualification ex parte maiore, which is the counterpart of Dionysius
, has to be understood as a restriction of the empirical domain of grammar.
18 According to Di Benedetto [2000] 396, on the contrary, the has nothing to do
with the reading of the text, but refers only to the determination of the correct accent of
a word.
definitions of grammar 527
this extension, well established by the time of the grammarian Herodian in the
second century AD, may have been the achievement of Dionysius Thrax.
(b) The second part of grammar is the exegesis of the or
poetical phrases which may obscure the meaning of the text (cf. Schenkeveld
[1991] 153156). In case of a hyperbaton for example, the grammarian will indi-
cate the normal word order (Sch. Dion. T. 455. 2430); he will also explain
metaphors in the text (ibid. 457. 2229), allegories (ibid. 456. 814) etc.
(c) The third part of grammar deals with , i.e. foreign or obsolete
words, and with , a term covering the domain of realia (persons, geo-
graphical data, myths etc.), as commonly assumed by most ancient scholiasts
(cf. Sch. Dion. T. 303. 4: Histora is the narration of ancient facts). As to the
term which qualifies [rendering], according to the scho-
liasts (Sch. Dion. T. 14. 19; 169. 15; 567. 41) it has the meaning here of
[ready], an interpretation which inspired the translations of Kemp ([1987]
172: ready explanation) and Lallot ([1989] 41 = [19982] 43: la prompte lucida-
tion). To this, however, Patillon [1990] 694 objected: La promptitude na rien
faire ici; il sagit de lexplication claire. The term can indeed also
mean obvious, readily accessible, easy (see LSJ 1541), and within a didactic
context a readily understandable rendering19 probably makes more sense
than a prompt explanation.
(d) The inclusion of etymologia (in its ancient conception and practice cf.
Lallot [1991a] and Sluiter in this volume) within the grammatical expertise of
the philologist should be seen in the light of establishing, by means of etymo-
logical reasoning, the correct meaning and/or form of rare and unusual words
occurring in older poetry. But etymologia also served as a tool for establishing
the correct pronunciation of words (cf. Sch. Dion. T. 454. 2129).
(e) The fifth part of grammatik is what we, from our modern point of view,
would consider the properly structural one. The key term here is analogy, a
term which in a grammatical context refers to a patterning recurrent through-
out series within the language system. The concept seems to have been already
extensively used by Dionysius teacher Aristarchus (cf. Ax [1982] and Matthaios
[1999] 400 sqq.), and maybe also by Aristophanes of Byzantium before him
(cf. Callanan [1987] 107122). It is probable that already at a rather early stage
lists of rules of nominal and verbal inflection (kannes; cf. Sch. Dion. T. 309.9)
had been set up; it is also likely that, in order to determine or explain unknown
or uncertain inflectional forms, grammarians and grammar teachers referred
to words displaying a similar inflectional pattern (cf. Lallot [1989] = [19982] 80).
This assumption is supported by ancient scholiasts like Sch. Dion. T. 15. 1423:
19 Callipo [2011] 57 translates: la spiegazione accessibile delle parole rare e dei contenuti.
528 wouters and Swiggers
20 See also Morocho Gayo [1999] 356. Morgan [1995] 8889 on the other hand believes that
the judgement could bear on the educational or moral value of the literary works. For
definitions of grammar 529
(I) He first mentions Ptolemy the Peripatetics objection against the definiens
in Dionysius Thrax definition of grammar, without quoting Ptolemys
own definition (provided he gave one); anyhow, it is clear that Ptolemy favored
a definition of grammar as a . On this point, he is immediately corrected
by Sextus:
.
.
(
,
, ),
,
,
[...] ( Math. I 6061).
Irvine [1994] 45 the krsis embraced all the aspects mentioned: Criticism involved textual
criticism and judgment on the authenticity of works as well as literary, esthetic, and ideo-
logical criticism.
530 wouters and Swiggers
, ,
, , <>
,
< > ,
, .
,
.
.
, [...] ( Math. I 7374).
< >
, ,
.
,
[...]
, , ,
. ,
. (Math. I 7780)
In the first book of his On Grammar Chaeris says that complete grammar
is a skill which diagnoses from expertise <and research> the things said
and thought by the Greeks as accurately as possible, except those things
which come under other kinds of expertise. This last he did not add idly,
since of the things said and thought by the Greeks some come under vari-
ous kinds of expertise and others do not, and he does not think that
grammar is the expertise or skill of those which come under other kinds
of expertise [...]. For it is immediately obvious that he freed grammar
from the sorites puzzle and separated it from alien precepts, those of
music and mathematics, inasmuch as they are not relevant. However, he
by no means saved it from being non-existent, but rather he actually
helped to establish that it was just so (transl. Blank [1998] 1718).
.23 .
(Math. I 84).
24 See also Dalimier [2006], according to whom Sextus was attacking a type of (school)
grammar that was no longer in use at his time.
25 Siebenborn [1976] 105, n. 1 dates him immediately after Dionysius Thrax (c. 17090 BC).
definitions of grammar 533
(3) In the third place, we should consider the doctrinal content of Sextus
documentation.
The definitions commented upon by Sextus Empiricus are those of Dionysius
and subsequent scholars. Dionysius definition has already been analysed
(cf. supra) and we can now take a closer look at those of his successors.26 From
Sextus text it does not appear whether Ptolemy the Peripatetic proposed a
definition of his own: if he did, it seems that his definition of did
not contain the term as a definiens, but rather the term . We can-
not exclude the possibility that Ptolemy defined as a knowledge
dealing with writings (), or something similar (e.g., writings in poetry
and prose), but it is also possible that he merely replaced in Dionysius
definition with . Anyhow, the crucial element in Ptolemys deviation from
Dionysius was the insistence on the nature of grammar / grammatical knowl-
edge as being an art, a practical skill, acquired by study and training, the lat-
ter combined aspect being the distinctive feature with respect to
(observation), on the one hand, and (joint exercise) on the
other. Clearly, Ptolemy also wanted to define the status of grammar with refer-
ence to Platos and, especially, Aristotles theory of knowledge and philosophy
of science.
The two remaining authors in what we have labeled the third phase
(cf. supra), viz. Chaeris and Demetrius, clearly follow in the steps of Dionysius,
while correcting his definition. Interestingly, Chaeris takes up again (delib-
erately?) the term which we encountered in Eratosthenes definition of
grammar, although he links to : as a matter of fact, he defines gram-
mar as a skill / aptitude / condition () derived from art / expertise ()
or making use of . The syntagm (which Barwick, followed by
Blank [1998] 18, n. 23, restores as , on the basis of Sch.
Dion. T. 118.11) can be read as indicating an absolute origin or as indicating an
26 Blank [1994] 155157 and [1998], Lallot [1995b] 7881, and Prencipe [2002] 5659.
definitions of grammar 535
and grammatikos (cf. supra), distinguished between their respective roles, and
who placed the grammarian on an inferior level: for Crates, the grammarian
dealt with linguistic forms (i.e. their phonetic, prosodic, and morphological
aspects), whereas the critic was required to possess an overall knowledge. Of
course, philological activity, when viewed in its most extensive conception,
had to go beyond the competence of the grammarian. Sextus (I, 248249) also
reports that a pupil of Crates, Tauriscus, proposed a tripartition of the field
of kritik, dividing it into a rational (logikon), empirical (tribikon) and histori-
cal (historikon) part. In this tripartition30 the subject matters dealt with by
the grammarian are redistributed over the three parts of the kritik, and are
thus not assigned to a single subdivision. As pointed out by Blank [2000] 406,
Tauriscus division of kritik parallels ancient debates concerning levels of
knowledge and types of approaches in medical practice:
30 The logikon part covers phonetics, orthography, morphology, and syntax; the tribikon part
the analysis of dialects and styles/registers; the historikon part deals with glosses, rare
words and realia.
538 wouters and Swiggers
Dionysius
Asclepiades
The division of grammar into parts was to become a prominent topic in the
scholia on Dionysius Thrax (cf. infra).
It is clear that Asclepiades in some sense aspired to be master above
master, improving upon a text and an author commonly used in philologi-
cal circles: his emendation of Dionysius definition of grammar is based on a
very subtle argument, making a distinction between grammar and the gram-
marian. The latter distinction could hardly have been acceptable to authors
like Eratosthenes (and maybe also Dionysius), but it allows Asclepiades to
show his acquaintance with Platos theory of knowledge (cf. Grg. 464c465a
and Phlb. 55e56d), through the distinction between firm or fixed ()
, i.e. the non-mathematical kinds of expertise, and the conjectural
() , which are subject to chance, such as navigation and
medicine. Asclepiades assigns grammar (not the short-lived activity of the
grammarian) to the domain of firm arts, just like music and philosophy. He
then qualifies grammar, just like Demetrius, as an , but a knowledge
embracing all the things said in poets and writers: while the grammarian, in
view of his short-livedness, will be acquainted only with most of the speech
of the poets and writers, grammar is the knowledge of all that (
[sc. ] < >
, ,
) (Math. I 73). From this argumentation two facts follow, which
appear in Asclepiades definition:
definitions of grammar 539
As will be clear from the preceding analyses, our main source for the ancient
Greek definitions of grammar is Sextus Empiricus. As a matter of fact, in his
Against the grammarians Sextus deals explicitly with the status of grammar as
a discipline. His examination is especially interesting for two reasons:
(a) on the one hand, Sextus wants to document his critical examination
through a scrutiny of definitions of grammar available to him;
(b) on the other hand, since his aim is to assess the status of grammar, he
specifically pays attention to the way in which the authors define the nature
and function of grammar.
At the beginning of the third section of his treatise Sextus explicates his
approach, with reference to Epicurus methodological principle:
, ,
(Math. I 57).
31 Sextus term prolepsis is generally translated as preconception, but this term should not
be taken in its present-day acceptation: it refers to a basic, first-hand idea that we have
of a particular (type of) object. It is therefore not a preconceived idea, but an empirically
based apprehension.
540 wouters and Swiggers
The key terms here are (a) , and (b) / . The term
refers to a rational, principle-based study (which can be either an art
or a science) of an object (or, rather, a domain of objects). The two qualify-
ing adjectives and more precisely define the nature of the
aimed at: it should be systematic / coherent () and real(istic)
().
Now, Sextus examination will consist in showing that the definitions of
grammar given by grammarians are incoherent and that they entail that their
object cannot be real. Let us have a look at how Sextus proceeds.
,
; ,
,
.
(Math. I 83).
definitions of grammar 541
(Math. I 90).
Of all the definitions of grammar that are attested before Sextus Empiricus
only one seems to have enjoyed lasting continuity. Whatever may be the rea-
sons for the eclipse of other definitions, the one proposed by Dionysius was
perpetuated in the training of philologists and in school room practice.
542 wouters and Swiggers
It is likely that in its original form Dionysius Thrax Techn covered the six
parts of grammar specified at the beginning of the work; subsequently, the
work may have undergone a condensation, with a focus on the technical part
of grammar, a part for which specialized terminology as well as basic distinc-
tions were already in place in the 2nd/1st c. BC (cf. Matthaios [1999]). We lack
precise information about the evolution of the organization and doctrinal
refinement of Dionysius Techn, but one cannot dispute the fact (a) that the
work was already an important reference for Asclepiades and, later, Sextus
Empiricus, and (b) that in the course of time, the work developed into a man-
ual for the instruction of grammar, as is clear from the papyrological tradition
(cf. SwiggersWouters [1995a] 9697) and from the Byzantine scholia on the
Techn.
In the papyrological documentation (published by Wouters [1979]) we find
confirmation of the process of doctrinal uniformizationallowing for some
fluctuation and variation (cf. SwiggersWouters [1995a]), but the papyri
contain little information on the definition of grammar and on the status of
grammar as a discipline. Only in P.S.I. 1.18 (5th c. AD; Wouters [1979] n. 5),
which contains the opening paragraph of the Techn and Supplement 3 (cf.
supra, note 11) on metrical feet, do we find the almost literal reproduction of
the definition and division of grammar as found in the transmitted text of the
Techn. In the definition of grammar given in P.S.I. 1.18, ll. 914 there is only a
slight variation with respect to that of the Techn as transmitted by the medi-
eval manuscripts (cf. supra, 3):
(Wouters [1979] 122 and 124), whereas
the medieval manuscripts read . Together with P. Hal 55A (5th
c. AD; Wouters [1979] n. 4)a fragmentarily preserved parchment text, with
parts of 12 and 20 of Dionysius manualP.S.I. 1.18 should be considered
the first direct copy of the Techn.
As to definitions of grammar, the Byzantine commentaries precisely focus
on the one given by Dionysius, and only offer scant information on other defi-
nitions, which seem to have been marginalized by the focal interest on the
Techn. A divergent definition32 mentioned in the scholia is the one attributed
to Tyrannio of Amisos (2nd/1st c. BC), a pupil of Dionysius Thrax: his definition
32 Another definition of a first century BC author is preserved only in a Latin source (and
in a Latin translation), viz. Aristo of Alexandrias definition quoted by Marius Victorinus:
Grammatica est scientia poetas et historicos intellegere, formam loquendi ad rationem et
consuetudinem dirigens (G.L. VI, 4.78). This definition reflects the view of grammar as
dealing with poetry and prose writings, but in addition it seems to apply a larger forma-
tive, ethical role of the grammarian (ad rationem et consuetudinem dirigens).
definitions of grammar 543
7 Conclusion
The above survey and analysis of definitions of grammar in ancient Greece has
shown:
33 For a very useful tabulation of the contents of the scholia dealing with the division of
grammar (Sch. Dion. T. 10.8ff, 12.3ff, 13.7ff, 115.8ff, 168.19ff, 170.17ff, 452.34ff, 471.8ff), see Blank
[2000] 412413.
34 See Usener [1892], who attributes the quadripartition of grammar to Tyrannio.
544 wouters and Swiggers
(b) that the definitions given of grammar testify to an explicit concern with
assessing the epistemological status of grammar, in its relation to the various
degrees of knowledge; as such, the definitions of grammar transmitted to us
should be contextualized in the larger frame of ancient philosophy of science;
(c) that all the extant definitions reflect a conception, and a practice, of
grammar as a (literary) text-based discipline, with only a sporadic integration
of the study of common speech;
(d) that, through the tradition of philosophical training and of grammar
teaching in schools, the definition of grammar as the (propaedeutic) disci-
pline preparing students for dealing with literary texts seems to have imposed
itself in ancient Greek culture and in Byzantine times. This evolution cannot
be separated from a factual development: the creation of didactic tools for
the teaching of grammarmanuals, paradigm tables, lists of exercises, a
process in which grammar became an object, and a primordial condition of
education in general and thus a cornerstone of cultural capital. In the course
of this process grammar became, next to an object of study, an instrument of
knowledge.
chapter 3
1 Introduction
2 Typology
2.1 Broad-Band Types
2.1.1 Annotations
2.1.2 Commentaries
2.1.3 Monographs
2.2 Special-Purpose Types
2.2.1 Catalogs and Lists
2.2.2 Scholarly Introductions to Literary Works
2.2.3 Dictionaries: Glossaries, Lexica, Thematic Dictionaries
3 Typological Synopsis
4 Outlook: Philology, Philological Writings, Systems Theory
4.1 External and Internal Differentiation of Philological Writings
4.2 Further System Characteristics
4.3 System, Discipline, Profession, and the Scholar
1 Introduction
2 Cf., e.g., Pfeiffer [1968] 3 and 134; Bhler [1977] 44; Wilson [1997a] 87; Dickey [2007] 3.
3 Cf. Montanari [2011a] 15 with a general definition of philological-scholarly works as text on
a text or text about a text.
4 For primary texts in a related context (and for the terms origin in Michel Foucaults Lordre
du discours, 1970), see Dubischar [2010] 4142.
5 The philological works of the disciplines truly foundational figures (Zenodotus, Callimachus,
Aristophanes of Byzantium, or Aristarchus) are all lost. The survey by Wilson [1997a] 9495
of what little has survived intact lists only five titles.
6 Cf. Pfeiffer [1968] on beginnings to the end of the Hellenistic period; Phlmann [1994] on
beginnings to the time of Septimius Severus. Other concise historical overviews include
Montanari [1993b] 259281; Wilson [1997a]; [1997b]; Kaster [1997]; Dickey [2007] 317.
7 Scholia, Commentaries, and Lexica on Specific Literary Works in Dickey [2007] 1871.
typology of philological writings 547
Hypomnemata (commentaries)
Alphabetical lexica
Anthologies
Mythographical historiai
Hypotheseis, with further distinctions between reworkings and summaries
on the one hand and real hypotheseis, which come in two subtypes, on the
other
The so called Scholia minora, with further distinctions between the Scholia
minora proper and two variants, namely, paraphrases and lexica
11 Cf. Asper [2007] 46 on the difference between mere Textklassenbenennungen and more
developed Gattungssysteme.
12 Gansel [2011] 1314 emphatically suggests that text-linguistic studies move from what-
questions (Was-Fragen) to how-questions (Wie-Fragen). Examples of comparable
approaches in neighboring ancient intellectual or literary areas include Ax [2005] on
Roman grammar; Asper [2007] on ancient scientific literature texts; Dubischar [2010]
on mostly popularizing auxiliary texts. Ancient philological writings with their own, spe-
cific functions remain yet to be more fully explored.
13 This caveat, commonly made by scholars who propose new classifications, has perhaps
never been expressed more directly than by Genette [1997a] 1: At the time of writing
(13 October 1981), I am inclined to recognize five types of transtextual relationships.
typology of philological writings 549
d ivisions and boundaries where there are really gradual transitions.14 In addi-
tion, classifications are always contingent on the choice of the distinguishing
criterion. Any one classification or system will highlight certain differences
and establish certain relations among the studied objects but eo ipso eclipse
others. Different distinguishing criteria produce different results.15 Therefore,
what will be presented here, in spite of its at times systematized appearance,
makes no claim to being definitive. If the following typology of philological
writings serves as a preliminary base for further discussion and exploration
of this overall not yet sufficiently studied topic, it will have served its purpose.
2 Typology
The following typology will observe both the various formal structures and the
functions of philological writings. This essentially text-pragmatic approach
acknowledges that philological writings, in this respect like other scientific or
technical texts, are essentially functional writings. They are each composed
for a specific purpose, and their function largely determines their formal
structure.16
The most fundamental distinction for the following typology of philologi-
cal writings is that between broad-band and special-purpose types of writings.
discussed and taught orally. Typologically, this would constitute the zero-level, so
to speak, at which no textualization of philological efforts takes place. Our testi-
monies for such oral philological instruction and discussion are naturally scarce.
But even in the absence of any reliable information, we would have to assume
the existence of oral venues of philological activity wherever philology was prac-
ticed.17 In fact, however, there is some evidence for oral ( ) teaching
and discussion at Alexandrias Museion.18 Timon of Phlius satirizing image of
the Museion as a cage of squabbling birds points to a strong oral (and of course
contentious and competitive) element in the discourse cultivated at this famous
institution.19 More specifically, Pfeiffer plausibly assumes that Zenodotus inter-
pretations were transmitted orally because the written commentary had not
yet been invented.20 In addition, Suda (s.v. ) tells us that
in his early years Aristophanes of Byzantium heard () Zenodotus and
Callimachus.21 And Aristophanes himself, as Pfeiffer again assumes, may have
talked more fully to his pupils about lexical issues that would have come up as
he was compiling material for the various sections of his .22
2.1.1 Annotations
With brief annotations, placed in the interlinear or marginal spaces
directly on a given copy of the primary text, we enter the realm of writing.
Annotations can address issues of textual constitution as well basic aspects of
understanding. Before the advent of the codex, marginal or interlinear annota-
tions had to be short. They were thus most suitable for commenting on specific
points concerning individual lines or even individual words in a line of the
primary text. But since annotations are written, they share the characteristic
advantage of all writing: the addresser and the addressee no longer have to be
in the same place at the same time because the acts of writing and reading,
that is, of sending and receiving, can be, and usually are, temporally and/or
spatially separated.23 Marginal or interlinear annotations thus give the philo-
17 On the predominance of oral teaching and training in other technical areas see Meiner
[1999] 139141; Meiner [2003]; also van der Eijk [1997] 96.
18 For an overview of the first Alexandrian scholarship see also Montana in this volume.
19 Cf. Lloyd-Jones Parsons [1983] 372373 (fr. 786); Pfeiffer [1968] 9798, but also Cameron
[1995] 3132; Long [1978] 74 et passim.
20 Cf. Pfeiffer [1968] 108 with n. 1.
21 Cf. Slater [1986] 1 (T 1, 13); Pfeiffer [1968] 172.
22 Cf. Pfeiffer [1968] 202.
23 For analyses, informed by and contributing to general communications theory, of the
transition from oral to written communication and some of its consequences see Ehlich
[1983]; Luhmann [1997] 249290.
typology of philological writings 551
logical scholar the opportunity to add his own comments to an existing pri-
mary text, by fixating them permanently onto the very same scroll that already
contains the primary text. The range of potential reception, that is, of the num-
ber of people who can be reached at present or in the future, is thus at once
greatly expanded.
From this point on, the original Texttrger (text carrier) holds two kinds
of texts: the primary text that elicited the scholars annotations as well as the
added philological paratext.24 Even so, paratextual annotations are a rather
elementary, if not rudimentary, form of writing. They are still far cry from
exploring more fully the formal, dispositional, intellectual, and stylistic fea-
tures or strategies that characterize more developed technical or scientific
written discourse.25 Instead, the annotations contents are, by necessity, not
only directly tied to specific points of the primary text; what is more, they also
lack physical autonomy since they are written on an already existing text copy.
Annotations are nevertheless philological-scholarly tools of great practical
value andas the first manifestation of serious philological efforts in writing
even of historical importance. Two main types can be distinguished: signs and
explicit notes. The first philological sign (), famously invented already
by the first head of the Alexandrian Library, Zenodotus of Ephesus, in his edi-
tion of Homer, is the , a short horizontal dash (), which Zenodotus
used to mark spurious lines.26 Considering the obvious practicality and effi-
ciency of the obelos, its invention may have been only a small and obvious
step to take for Zenodotus. But it was a giant leap for the history of philology.
Inaugurating a method of non-destructive criticism,27 Zenodotus marked
the lines that he believed to be not genuine, but he did not delete them. For
the first time, an editor had provided the serious reader and scholar with an
opportunity of appraising his critical judgment.28
It was not until several generations later that Aristophanes of Byzantium,
fourth head of the Library,29 seems to have added other critical signs to
the obelos, such as the (asterisk: *), to mark lines that are dupli-
cated from another place, as well as the (sigma: ) and the
24 Cf. Genette [1997b] in particular 319343 on notes and 337339 on allographic notes.
See also Moennighoff [2008]; on related topics also Barney [1991] and Grafton [1999].
25 Cf. Asper [2007] 2735.
26 See Montana in this volume.
27 Jacob [1999] 13.
28 Cf. Pfeiffer [1968] 115; also, e.g., Phlmann [1994] 2728.
29 See Montana in this volume.
552 dubischar
(inverted sigma: ) for two consecutive and interchangeable lines of the same
content.30 The number of philological signs and in some cases their mean-
ings were modified and further developed by Aristarchus of Samothrace, sixth
head of the Library. While there is uncertainty about some details, it seems
that he used both critical and exegetical marginal signs in his editions of the
Homeric poems. Critical signs were the obelos () as used by Zenodotus and
Aristophanes; the (dotted diple, ) to point to a verse in
which Aristarchus text differs from that of Zenodotus;31 the asterisk (*) for
lines that are wrongly repeated elsewhere in the text; the obelos added to the
asterisk (*) where the repeated line is out of place; the ( ) indicating
suspected spuriousness; the antisigma () to indicate lines in disturbed order,
with the stigme ( ) denoting the line that should immediately follow the line
marked with the antisigma. The famous diple (>), widely applicable and fre-
quently used,32 was of a different nature. It marked lines whose language or
content was perhaps also exegetically noteworthy (not only with regard to tex-
tual criticism) and pointed to a corresponding explanation in a commentary
(see below 2.1.2).33
Aristarchus semeia became the standard philological signs for centuries to
follow, also adopted early on by scholars in Rome,34 even if a certain diver-
sity and flexibility in the signs uses must be accounted for.35 Some papyrus
fragments in fact contain un-Aristarchan signs whose use was fairly consis-
tent nevertheless. For instance, the so-called ancora, an anchor-shaped diag-
onal upward or downward pointer often marks places where text had been
omitted or draws attention to text-critical restoration in the top or bottom
margin; and there were other lunate signs whose roughly moon-shaped
form and meaning are related to the antisigma.36 The meanings of other
signs surviving on papyrus are less clear or vary more strongly, as those of the
30 Cf. Nauck [1848a] 1518; Pfeiffer [1968] 178, also Slater [1986] 210.
31 See also Ludwich [18841885] II 5864.
32 Gudeman [1922b] 1918.
33 Examples of Homeric lines to which Aristarchus applied his symbols in Ludwich [1884
1885] I 22; cf. also McNamee [1992] 28 (Table 1); Pfeiffer [1968] 218.
34 Cf. Phlmann [1994] 47.
35 Cf. Ludwich [18841885] I 2021 and elsewhere; Gudeman [1922b] 19161917; McNamee
[1992] 11. Cf. Hephaestions dictum in (On Signs):
(The signs for the poets are used differently for different
poets, p. 73 ed. Consbruch [1904]).
36 Cf. McNamee [1992] 115.
typology of philological writings 553
diple () in non-Homeric texts, the simple diagonal stroke, the dotted obelos, or
the letter (chi).37
Other kinds of signs facilitate a texts reading or understanding. Even
though they are generally thought of as less philological than especially the
critical , they too deserve to be mentioned. They include diacritical signs
(accents and breathings), metrical or colometrical signs, lectional signs (indi-
cating, for instance, speaker change or word separation), and various kinds of
punctuation.38 Some papyri also bear marginal signs that give practical rather
than critical or exegetical aid, as they indicate how certain texts were used, for
instance, again the letter (chi) or the monogram chi-rho, likely symbols for
the words (passage) or (useful).39
Philological signs are one kind of paratextual annotation. Marginal or inter-
linear notes, written not as symbols but as explicit words, are the second type.
To be sure, these are not the scholia, that is, not the Late Antique or Byzantine
excerpts from earlier, usually Alexandrian scholarly writings (commentaries,
but also monographs or lexica) that the scholiast adds in the margins or inter-
linear spaces of a papyrus or later, more typically and with more space, in the
wider margins of a page of a codex. Instead, the marginal or interlinear notes
to be discussed here are direct manifestations of productive philological work
and are in this regard not all that different from philological signs. They, too,
can be associated with the ancient practice of producing a critical edition,
an obviously important process about which, however, we have less reliable
information than we would like.40 The uncertainties result from the scantiness
of our evidence as well as from the fact that text-editing is itself a complex pro-
cess that means partly different things in different times and contexts.41 The
meanings of two key terms, and ,42 however, seem to be rela-
tively stable. The former comprises the steps of manuscript collation and text
emendation (with or without conjectures); the latter refers to the releasing of
a text which could then be read or copied by others.
37 Cf. McNamee [1992] 1521; also McNamee [1992] 2948 (Tables 2 and 3); Montanari [2011b]
11 and elsewhere.
38 For papyrus evidence of the paragraphus (horizontal stroke), double dot, various kinds of
punctuation, lectional signs, accents, hyphens, breathings, coronis, and others see Turner
[1971] 1018.
39 Cf. McNamee [1992] 2022.
40 Cf. van Groningen [1963]; Irigoin [1994]; Montanari [2011b]. Galens uvre is a storehouse
of relevant information that still awaits full exploitation; cf. Hanson [1998].
41 Cf. Most [1998] x.
42 See Montana and Montanari in this volume.
554 dubischar
2.1.2 Commentaries
The commentary ()47 is typologically situated on the next higher
level of textual autonomy. But there are functional similarities to the manifes-
tations of philology mentioned thus far. Like oral explanations and marginal
or interlinear annotations, philological commentaries can aim at elucidat-
ing a primary text with regards to textual criticism and exegesis. The new
and defining feature of the commentary is that it is a physically independent,
self-standing48 text. Written on a separate Texttrger, the commentary still
has a primary text as its object but is no longer physically attached to it.
53 Cf. Luppe [2002] 5758; Gibson [2002] 1415; Lundon [2011a] 162163 and 171.
54 Cf. Pfeiffer [1968] 218; McNamee [1992] 11; Montanari [1993b] 242.
55 For Aristarchus as inventor of the commentary cf. Montana in this volume.
typology of philological writings 557
56 The occasional affinity between commentary and lexicon, pointed out by Gibson [2002]
1718, 20, and 172174, has the same reason. With some typological irony, the production
of scholia recreates the paratextual problem of annotations that are tied materialiter to
a specific primary-text copy, now a codex. In response, modern scholia editions again
separate these annotations from their particular copies of the primary text.
57 Cf. Pfeiffer [1968] 160161; Dickey [2007] 29.
58 Pfeiffer [1968] 2123.
59 Cf. Dickey [2007] 1871 and in this volume for an impressive and instructive parade of
Greek primary authors who became subjects of ancient commentaries.
60 Pfeiffer [1968] 208. Cf., e.g., Luppe [1978]; [2002] 5863; Montanari [1993b] 243281;
Lundon [2011a]. For ancient commentaries on Demosthenes (on papyrus) see Gibson
[2002]; Harding [2006]; Montana and Dickey in this volume.
558 dubischar
68 Cf. also Sluiter [1999] 188 in the context of commentaries on Hermogenes and Aphthonius.
69 Cf. Montanari [1984] 130132; Haslam [1990]; Montanari [1995c]; Dickey [2007] 26 and in
this volume; Rossum-Steenbeek [1998] 85118.
70 Cf. Holwerda [1964]; [1967]; Dickey [2007] 29.
71 Cf. Dyck [19831995]; Dickey [2007] 2728; Pontani in this volume.
72 The distinctions made here and in the following paragraphs have similarities but are not
congruent with the tension between charity and criticism, discussed by Sluiter [2000a]
187 and 189190 as one of four dialectics characteristic of commentaries.
73 Paraphrastic commentaries may thus be the fullest embodiments of some commentators
tendency to focus on content more than on form; cf. Sluiter [2000a] 190.
74 On the inherently close ties between written commentaries and teaching see Sluiter
[1999]; [2000a] 190192.
75 Cf. Ihm [2002a] 6465; Dickey [2007] 44.
560 dubischar
76 For Themistius see Dickey [2007] 4950; for (Ps.-)Proclus see Dickey [2007] 68; for
Munatius see Dickey [2007] 6365; for Eutecnius see Dickey [2007] 6566 and 70.
77 Cf. Vallance [1999] 223228.
78 Cf. Dickey [2007] 4850; also Vallance [1999] 228242 and 242244 on non-submissive
medical and mathematical commentaries by Galen and Proclus.
79 Cf. also Sluiter [2000a] 189190; Asper [2007] 3542. More generally, Lloyd [1996] 2046;
[1987] 101108.
80 Cf. Fowler [1999] 427, but see also 430; Shuttleworth Krauss [2002] 2; Montanari [2011a] 16.
81 Cf. Manitius [1894].
typology of philological writings 561
2.1.3 Monographs
Possessing the greatest degree of textual autonomy, monographs ()
conclude this sequence of broad-band types of philological writings.88 Like the
hypomnema, the monograph is a self-standing text. Unlike the hypomnema,
however, it can be independent from any particular primary text not only
physically but also in terms of its contents. While commentaries are devoted to
specific primary texts that they set out to explain, monographseven philo-
logical monographsare freer in their choice of subject and contents. If com-
mentaries were said to have a liberating effect on philological efforts in writing,
monographs provide an even greater freedom as they are convenient vessels
for any topics discursive, extensive, and fully in-depth treatment.89
Scholarship on ancient syggrammata, whether philological or other, is still
in its early stages. The largest surviving distinct corpora of ancient Fachtexte
(technical literature in the broadest sense) are those of Hippocrates, Aristotle,
Galen, and, to a certain extent, Plutarch. Each of them has its own character
and typological profile, so that the results achieved about any one of them
cannot easily be generalized or applied to the study of monographs in another
thematic field, such as philological scholarship.90 In their surveys of the
disciplineeach valuable and still essential for certain aspects of ancient
scholarshipPfeiffer, Montanari, and Dickey do not systematically differenti-
ate further between subtypes of philological monographs. And while Markus
Aspers typological study of Greek scientific texts (Wissenschaftstexte in a nar-
rower sense, excluding philological scholarship) is rich in important insights,
the classification established there does not fit the realm of philological
syggrammata well.91
The following observations will thus propose some more finely-tuned typo-
logical distinctions within the field of philological monographs. Based on
87 Cf. Jones [1999] 167168 on the different intended audiences for Theons Little Commentary
and the Great Commentary; Dickey [2007] 6768.
88 Historically, of course, monographs existed long before hypomnemata. The sequence pre-
sented here is typological, not chronological.
89 Cf. Shuttleworth Kraus [2002] 23; Asper [2007] 213 and 57 with n. 1.
90 See also van der Eijk [1997] 8991.
91 This text-typological incompatibility points to traits that set philology apart from other
intellectual-scientific disciplines; see below, Outlook: Philology, Philological Writings,
Systems Theory.
typology of philological writings 563
92 Cf. Ihm [2002b] 316317, with more examples. The situation is complicated even further
by the fact that Galen also wrote unmistakable passage-by-passage commentaries on
some of the Hippocratic primary texts in question; see Ihm [2002a] 8889. On Galen as
commentator see Flemming [2008].
93 Cf. Harding [2006] 1320.
94 Cf. Lundon [2011a] 163166.
95 Cf. Novokhatko in this volume.
564 dubischar
96 Cf. Asper [2007] 7175 with n. 102; see also Gudeman [1927c].
97 Cf. MacPhail [2011].
98 Cf. Wehrli [1953] 5154.
typology of philological writings 565
135 Pfeiffer [1968] 143; specimens also in Slater [1976] e.g. 241. See also above 2.1, and
Montana in this volume, on Timon of Phlius bird-cage fragment.
136 Cf. Bowersock [1969] 5975 and 89100; von Staden [1997b] 3337. Fichtner [1985] 170 lists
titles of nine polemical works by (or attributed to) Galen, and even his non-polemical
writings contain frequent attacks against poorly trained or poorly performing colleagues.
137 Pfeiffer [1949] 351352 (fr. 460); cf. Pfeiffer [1968] 95 with n. 4, 125 n. 1, 135136.
138 Cf. Pfeiffer [1968] 141, 146147.
139 For the following titles see Pfeiffer [1968] 213.
140 Cf. also Pfeiffer [1968] 91.
141 Cf. also Pfeiffer [1968] 289; Dyck [1988b] 221265.
570 dubischar
aradox seems to be that the special-purpose writings meet very basic philo-
p
logical needs. They arise in many situations, in connection with many texts,
and they concern, first, identification and classification of texts, second, over-
view and contextualization, and, third, semantic understanding. To be sure,
such issues can also be addressed, along with many others, in the broad-band
types of philological writings. Special-purpose genres, however, are devoted
solely to one select aspect, to which they are perfectly tailored. This single-
purpose orientation gives each of these text types its unique character, func-
tion, and form.
In the realm of ancient philology, the most ambitious and most influential
philological catalog is Callimachus Tablets or Tables, the famous
that comprised more than 120 books (i.e., papyrus scrolls).153 They served as
an annotated library catalog representing and classifying the Alexandrian
librarys astounding holdings of Greek literature. Only the first generation of
philologists in Alexandria worked without the aid of the Pinakes.154 But in the
longer run, it is clear that to amass hundreds of thousands of rolls in the library
would have been of little use without a sensible classification that enabled the
prospective reader to find the books he needed.155 Once again, in theory the
invention of a new type of philological writing, in this case the systematized
annotated library catalog, appears as a necessary and quite obvious practical
step to be taken; however, in light of this undertakings huge dimensions at
Alexandria, it undoubtedly qualifies as yet another giant leap for ancient liter-
ary scholarship.
In producing his Pinakes, Callimachus seems to have adhered to certain prin-
ciples of method and systematization: Authors and their works were grouped
according to genres of poetry (epic, lyric, tragic, comic) and prose writing
(rhetoric, laws, philosophy, historiography, medicine, miscellaneous). Within
each group, individual authors were presented in alphabetical order, which
was likely also the case, wherever feasible, for titles of works by one author.
The opening words of a work also seem to have been part of the standard Pinax
entry because they were a means of identifying texts that had no explicit titles
or multiple texts that had the same title. To add to the Pinakes philological
practicality and usefulness, short author biographies were included as addi-
tional basic contextual information.156
Considering the very basic nature of the need for identification, classifica-
tion, and orientation, it is not surprising that we also know of several other
ancient book catalogs or lists, even ifthis, too, is no surprisenone of them
matched the monumental Callimachean Pinakes in scope or in philological
rigor. Callimachus himself also wrote two special lists. One of them contained
the names of dramatic poets () in chronological order and from
the beginning (
).157 he nature of the other special Pinax cannot be determined
153 Cf. Pfeiffer [1949] frr. 429453; cf. Pfeiffer [1968] 127133; Regenbogen [1950] 14191424;
Blum [1977] 223243; Montana, this volume.
154 Cf. Pfeiffer [1968] 128.
155 Pfeiffer [1968] 133.
156 Cf. Blum [1977] 1112 on Biobibliographien.
157 Cf. Pfeiffer [1949] 349350 (frr. 4546); Blum [1977] 198207.
typology of philological writings 573
with certainty, but it may have been a bibliographical list as well.158 Alexandrias
rival institution, the library of Pergamon, had its own Pinakes, known as the
.159 Furthermore, fragments of book lists of various kinds
and relatively simple organization have survived on papyri. They reflect the
holdings of libraries or of private collections of differing sizes, and their char-
acter ranges from the expected lists of prominent and well-known classical
authors to lists that reflect highly specialized or professional interests and
collections.160 Finally, genre-specific or discipline-specific Werkverzeichnisse
(e.g., of orators or of medical writings) have even come down to us through
direct transmission,161 of which Diogenes Laertius extensive lists of philoso-
phers works are the richest surviving specimen. Also preserved, on account
of their obvious practicality, are Galens two annotated auto-bibliographies,
De libris propriis and De ordine librorum suorum.162
158 Cf. Blum [1977] 208223; but also Pfeiffer [1968] 132.
159 Cf. Pfeiffer [1968] 133 and 236; Regenbogen [1950] 1423; Blum [1977] 246.
160 Cf. Otranto [2000]; Houston [2009] 234247.
161 Cf. Regenbogen [1950] 14261438; Blum [1977] 246299.
162 Khn [1830] XIX 848 and 4961; Singer [1997] 322 and 2329; Boudon-Millot [2000].
163 Important are Holwerda [1976] 178198; Bud [1977] 2933; Meijering [1987] 107133.
164 Cf. Mossmann [2010] 249 with notes 10 through 13.
574 dubischar
came from another genre: 5th-century Athenian drama, that is, tragedy and to
a lesser extent comedy.
Aristophanes learned introductions were intended to be a necessary help
for the scholarly reader,171 of these plays. Their function is to providein con-
cise and systematic form and based on thorough eruditioncontextualizing
and other orienting information concerning a given play. No scholarly hypoth-
esis by Aristophanes has survived intact, but several preserved introductions
clearly originate from Aristophanic hypotheseis,172 whose organization has
thus been reconstructed, of course with some uncertainty, as consisting of the
following elements:173
(1) A concise summary of the plays plot, in one or two sentences, often
beginning with the main characters name.
(2) The so-called , i.e., brief information whether the same mytho-
logical subject was treated by the other two famous tragedians:
...[name of the playwright] ...[title of the tragedy], pos-
sibly followed by ......or instead, when appropriate,
.
(3) Brief information about the place of the action and the identities of the
chorus and the prolog speaker: ( ) ()
(or or )...[name of the place], () ...[iden-
tity of the chorus], ...[name or identity of the prolog
speaker].
(4) The ,174 i.e., an enumeration of the plays main events:
, followed by a string of nouns; e.g., Soph. Ant. (hypoth. I):
, ,
.
(5) The , including the year of the original performance:
...[name of the eponymous archon] ;175 the result of that
years tragic contest: ...[name of the winning poet], ...,
...; the title and the number of the play in the Alexandrian library
or the remark .
(6) A judgment concerning the quality of the play as a whole, as first-rate or
only second-rate: , or:... ), as well as
of individual parts of the play.
edies and comedies in classical Athens are remarkably stable; in a ddition, that
very institutional framework also required that new plays be produced at a
continuously high rate. This leads to oppressively large corpora of plays that
are thematically and structurally relatively uniformthe individual poets
theatrical and poetical innovations notwithstandingand whose differences
and individual characteristics become visible only after closer inspection. The
third problem is the plays lengths: Dramatic textseach play usually taking
up one papyrus scrollare too long to allow for quick and convenient over-
view and orientation. Thus, to summarize the difficulties: the textual corpora
of the great Athenian playwrights are, in and for themselves, far from user-
friendly. Each consists of an unwieldily large number of decontextualized dra-
matic texts (papyrus scrolls) that to the hasty beholder look very much alike
and whose individual lengths make quick identification and grasping of other
essential information impossible.
Aristophanes hypotheseis are a straightforward means by which this extraor-
dinarily difficult situation for these extraordinarily valuable texts is improved.
The urgency of the outlined problems and, correspondingly, the importance
of the service rendered by scholarly hypotheseis are confirmed by the fact that
Aristophanes seems to have produced his hypotheseis right along with his first
philological edition of dramatic texts. Aristophanes introductions are in fact
seen as the most substantial remains of Aristophanes editions of tragedies
and in a lesser degree of the comedies.178 This means that the competent
edition and reception of Athenian drama is practically impossible without
immediately supplementing it with additional orienting and contextualizing
information for each play. Aristophanes introductions do precisely that.179
Hypotheseis, now also including simple summaries, have emerged in many
fields of Greek literature. However, a coherent comprehensive typology has
not yet been achieved. At this point, no more is possible than to present some
typologies that cover different sectors of Greek literature: first, Greek tragedy,
second, Greek drama including comedies, third, Greek poetical works in gen-
eral, and finally, other Greek literature.
The most fundamental classification of tragic hypotheseis was formulated
around the middle of the last century by Zuntz, who distinguished the follow-
ing three types:
Zuntz important classification has generally been adopted,185 but it was also
partly refined some twenty years later by Bud. Basing his analyses on both
external and internal textual criteria, Bud suggests that there is a second type
of learned hypothesis, which does not go back to Aristophanes of Byzantium
and which he calls saga-hypothesis.186 This result has been accepted by
Rossum-Steenbeek and, earlier, Holwerda (while Buds further conclusion,
that the saga-hypotheseis resemble Dicaearchus hypotheseis, is generally
rejected).187 Rossum-Steenbeek herself, on the other hand, replaces Zuntz
188 Cf. Rossum-Steenbeek [1998] 12 with n. 4; Mossman [2010] 251; Bing [2011] 205 with n. 19.
189 Cf. Bud [1977] 2984.
190 Cf. Mllendorff [2010].
191 Cf. Mllendorff [2010] 275277.
580 dubischar
There are, third, also ancient hypotheseis to non-dramatic texts, an area care-
fully charted by Rossum-Steenbeek, whose results are of general relevance
even thoughand partly becauseshe focuses on fragments of hypotheseis
preserved on papyri. Rossum-Steenbeeks classification, which of course also
includes dramatic hyptotheseis, is the following:
also prose writers such as the Attic orators Demosthenes and Isocrates.197 Not
to be confused with these introductory or quasi-introductory hypotheseis are
the self-standing summaries and abridgments that were widely produced
and used in antiquity from the Hellenistic period onward. Usually referred to,
individually, as an (or, in Latin, epitoma),198 these texts differ funda-
mentally from the hypotheseis discussed here in that they are long enough to
have been written and transmitted as independent texts. As such self-standing
abridgments and summaries, however, they already lie outside of what can still
be conceived of as philological writings and will therefore not be investigated
further at this point.199
197 References in Dickey [2007] 6365, 52, and 55, respectively. For the somewhat special case
of Libanius hypotheseis to Demosthenes, see Gibson [1999]; [2003].
198 Other similarly used terms are or breviarium; cf. Opelt [1962] 944946.
199 On epitomai see Opelt [1962]; Dubischar [2010]; Mlke [2010].
200 See Tosi in this volume.
201 Cf. Tosi [1994b] 172174; also Pfeiffer [1968] 197; Gibson [2002] 20.
582 dubischar
202 Cf. Pfeiffer [1968] 115. On the term for rare, difficult, or obsolete words, especially
in Homeric poetry, see Pfeiffer [1968] 7879; cf. also Montana in this volume.
203 On alphabetization see Tosi [1994b] 151155, and in this volume.
204 Cf. Dickey [2007] 5354 and 4546.
205 Pfeiffer [1968] 115 for this reason calls it a model for the future.
206 Cf. again Dickey [2007] 4546 and 5354.
207 Cf. Degani [1988] 11691170; Matthaios [2010a] 166 with n. 3. See also the similar later pub-
lication Degani [1995] here 505507.
208 Pfeiffer [1968] 36.
209 There is an analogy to the first scholarly hypotheseis produced by Aristophanes of Byzan-
tium in connection with his first edition of dramatic texts (see above, 2.2.2). Zenodotus
and Aristarchus are responding to different but in both cases fundamental philological
needs that arise immediately from their primary texts.
typology of philological writings 583
it had emerged in the archaic period, was continually used and produced for
more than one thousand years, into late antiquity and beyond.210
While Greek glossographys initialand never ceasingconcern was
Homeric language (cf. Ps.-Apion), beginning in Hellenistic times glossaries
were also devoted to other individual authors such as Herodotus (cf. the men-
tioned and ), Plato (cf. Timaeus
the Sophist), or Hippocrates (cf. Erotian or Galen)211 as well as to other liter-
ary genres such as tragedy, comedy, or oratory.212 Thus, glossaries have passed
many tests of time and have proved to be functionally optimized text types that
provide indispensable semantic aid regarding specific primary texts, authors,
or text corpora. The ancient glossaries efficiency is owed also to the fact that
their entries were limited to rare and difficult words but did not include, as is
common practice today for the sake of completeness, basic everyday words
that no one would have trouble understanding.213
In the Hellenistic period, the focus of scholarly dictionaries also expanded
beyond mainly glossographical interest. To be sure, to a considerable extent
Hellenistic grammarians and scholars still collected their lexicographical
material from the primary texts themselves.214 But the purposes for which
words were collected by scholars became more manifold. Correspondingly,
two new types of dictionaries emerged. One of them can be called the -
type, after Aristophanes of Byzantiums famous (roughly meaning:
nteresting Words).215 Lexeis reflect the emergence of the study of words as a
component of grammar understood as an autonomous discipline that no lon-
ger needs to serve the ends of exegetical philological Textpflege. This type of
dictionary is dominated by a linguistic (including stylistic) interest in words,
word meanings, and word usage. Their various specific points of focus lead to
210 Cf. Degani [1988] 11701172, 11741175, 11771179, 1185, 11871188; Gibson [2002] 19 notes that
certain lexicographical branches underwent no evolution. In other words, their stable
function resulted in a stable form.
211 In addition to Degani [1988], cf. Dickey [2007] 2425 on Homer, 5354 on Herodotus, 47
on Plato, 45 on Hippocrates; also Dickey [2010] 1315, and in this volume.
212 Cf. Degani [1988] 11741175 on tragedy, 1175 on comedy, 11781179 on orators; on
Demosthenes see Gibson [2002] 1819, 157171, and 190199.
213 Cf. Dickey [2010] 2123.
214 Cf. Matthaios [2010a] 169170 et passim; but see also Pfeiffer [1968] 197.
215 Cf. Pfeiffer [1968] 197203; Dickey [2007] 93. Against the common opinion, however,
Slater believes that what most take to be individual thematic sections of the were
in fact independent lexicographical works and that the never existed as one big
work; cf. Slater [1976] 237 n. 11; [1986].
584 dubischar
216 E.g., Aristophanes of Byzantium, Ammonius/Herrenius Philo, Orus, Photius (cf. Dickey
[2007] 9294, 9495, 99, 101102 respectively, and in this volume).
217 E.g., Pamphilus (cf. Matthaios [2010a] 175 with n. 24).
218 Cf. Degani [1995] 519521; [1988] 11791181. Examples are Phrynichus, the Antiatticista,
Moeris, Philemon, Aelius Dionysius and Pausanias, Orus and Orion (Dickey [2007]
96100, and in this volume). However, behind the lexicographical work already con-
ducted by Alexandrian poetae docti on Homer and other early poets there is a double
purpose of both semantic-exegetical help and aid for their own literary production; cf.
Tosi [1994b] 145 and Matthaios [2010a] 168 with n. 10. On the transmission and appropria-
tion of atticistic lexicographical material in the late antique and Byzantine periods see
Matthaios [2010a] 186197.
219 See also Matthaios [2010a] 167 n. 4.
220 E.g. Pollux and Stephanus of Byzantium (cf. Dickey [2007] 99 and 101 respectively, and in
this volume).
221 Cf. again cf. Matthaios [2010a] 169170 et passim, with Pfeiffer [1968] 197.
typology of philological writings 585
3 Typological Synopsis
222 For the distinction between glossai and lexeis see Degani [1988] 1169; [1995] 505506 and
508; also Pfeiffer [1968] 198; Montanari [1993b] 250251. For Tosi [1994b] 144, however, the
main conceptual distinction is between lessicografia (which includes, however, what is
commonly, and here, referred to as glossography) and onomastica.
223 Cf. Dickey [2010] 1921.
224 Cf. Dickey [2010] 20.
586 dubischar
Broad-Band Types
Special-purpose types
The view of the large field of philological writings in its typological diversity
invites some farther-reaching observations and reflections of more theoreti-
cal nature. It was already noticeable that philological writings do not come in
a random or arbitrary variety of text types. Instead, an underlying nexus ties
them together, which will now be studied more closely. How are philological
writings different from other kinds of writing? Why have philological writings
assumed their particular forms? How are the various types of philological writ-
ings related to one another? Which dynamics influence the creation of philo-
logical writings and their types? These and other questions will be addressed
in this outlook. For this purpose, it will be helpful to view the field of philo-
logical writings through the lens of systems theory. This means that philology
and philological writings will be studied essentially as manifestations and pro-
cesses of communication.225
225 The relationship between a system and communication cannot be explicated here in
detail. In short: systems operate; this is how they exist. The operation that constitutes
social systems (soziale Systeme) is communication. See Luhmann [1984] e.g., 6667, 79,
588 dubischar
and elsewhere; Stichweh [1994] 62 and elsewhere; Gansel [2008a]; [2008b]; Dubischar
[2010] 6063.
226 Cf. Stichweh [1994] 6468 on publications as the most basic elements of scientific
communication.
227 Cf. Stichweh [1994] 15.
228 Cf. Stichweh [1994] 18: Disziplinen bilden sich um Gegenstandsbereiche und Problem-
stellungen herum.
229 Cf. Stichweh [1994] 21.
230 On Textpflege, see above, 1. The Aristotelian notion of a telos in this context is not an
erratic import; cf., e.g., Stichweh [1994] 67 on the telos of modern science. Disciples of
pure systems theory may prefer the term Sinn (as a systems ultimate meaning or pur-
pose); cf. Luhmann [1984] 93147.
typology of philological writings 589
236 On text types as routinely sought solutions for recurring communicative needs see Asper
[2007] 20, 24, and elsewhere.
237 Cf. Asper [2007] 12, who recognizes the close link between textual conventions of scien-
tific writings and the character of the particular field of knowledge which they cover.
238 Cf. Hanson [1998].
239 Cf. Erler [1993].
240 Cf. Niehoff [2011].
typology of philological writings 591
241 Even Pfeiffers account, despite its unmistakable humanist-idealist coloring, invites such
an approach. Throughout his thoughtful and richly documented narrative, Pfeiffer shows
592 dubischar
that in many of its phases the development of ancient, in particular Alexandrian scholar-
ship follows a certain internal logic that causally links later scholarly projects and writings
to previous ones. Some terms used programmatically by Pfeiffer are strikingly compat-
ible with systems theory: e.g., the origination of scholarship as a separate intellectual
discipline (cf. external differentiation in systems theory), the notions that scholarship
arose (cf. emergence in systems theory) and became one selfconscious discipline (cf.
self-referentiality in systems theory)to name examples just from the books opening
page; cf. Pfeiffer [1968] 3.
242 Cf. Stichweh [1994] 44.
243 Cf. Pfeiffer [1968] 9293, 9599.
244 See Montana in this volume.
245 Stichweh [1994] 42 and 44. This is an example of the general pattern in the evolution of
social systems that Luhmann [1984] 219 characterizes as a quasi-hydraulic repression and
redistribution (but never elimination!) of problem pressure in social systems through the
evolution of media of communication.
246 Montanari [1998d] 1819 reconstructs this kind of causal chain even within an individual
scholars philological career (Aristarchus). The same principle, however, also shapes the
development of the discipline at large.
typology of philological writings 593
emerge to meet this demand are mainly the composite commentaries and
various types of meta-philological writings.
However, infinite growth is impossible for any system. In regard to scien-
tific-scholarly communication, the reason is obvious. The reservoir of poten-
tial problems in any given field or discipline is not inexhaustible. Therefore,
the characteristically high frequency of innovations (Innovationshufigkeit)
during a disciplines formative period will at some point begin to decline, and
gradually a period of saturation (Sttigungsphase) sets in.249 The rates at
which additional internal differentiations or even just connecting communi-
cation (Anschlusskommunikation) take place goes down. The discipline can,
however, even then maintain its communicative momentum if its concepts
are successfully transferred from the original areas to new and thus far unex-
plored fields of study (Konzepttransfer).250 Here of course the same trajec-
tory toward at some point exhausting the reservoir of potential problems is
still inherent in the system.
Realistically, however, concept transfer to new areas within a discipline
is not perpetually possible. If the disciplines pool of worthwhile questions,
which alone secures the continuation of scientific communication, has been
substantially drained, the systems prior differentiations will, after a period of
stagnation, eventually be reversed. In other words, the discipline will decline
and perhaps disappear, especially since professional scientific or scholarly
disciplines tend to depend on extracting resources from their environment,
which may at this point decide to reduce the allocation of resources necessary
to sustain the systems capability of producing connecting communication
(Anschlusskommunikation).251 Its prior growth, accompanied by increasing
internal differentiations of communication and, thus, of types of writings, will
then be reversed gradually or drastically.
These processes too are observable during different phases in the history of
Alexandrian philological scholarship. Suffice it to recall that many philologi-
cal practices, methods, and their corresponding types of writings were initially
developed for the study of the Homeric texts. Their use, however, was subse-
quently expanded also to other text corpora, such as the lyrical poets, drama-
tists, prose writers, and the Hellenistic poets themselves. Overall, the formative
period between Zenodotus and Aristarchus was one of high innovation fre-
quency. It produced not only a myriad of individual philological contributions
on many primary authors but, even more significant in the present context, a
wide array of types of philological writings. Didymus, on the other hand, the
author of innumerable works of largely compilatory nature, representsafter
a preceding period of crisis owed considerably to external factorsa period
of renewal but also of saturation, if not over-saturation. Shortly thereafter,
Alexandrian philology experiences its final decline. Or in systemic terms: the
many text-typological and thematic differentiations that have taken place in
the system during its formative periods are now largely reversed. Later, scho-
liography will take this process of reversal even further by largely abandon-
ing all typological differentiations and turning the radically fragmented and
selected primary materialtaken, for instance, from hypomnemata, syggram-
mata, or glossaiinto a relatively uniform mass of scholia.
and philological problems as well as entire types of writings are said simply
to emerge?
While no social system consisting of communication can ever emerge
without the presence of and contributions by individuals,255 for a more
nuanced insight, the concept of system rationality (Systemrationalitt)
will be helpful.256 In functional systems, Systemrationalitt reflects the sys-
tems expectations at any given point. And the system expects, to phrase it
abstractly, that true to the systems telos those operations will be chosen that
secure or at least enable the most and most effective connecting communica-
tion (Anschlusskommunikation).257 To phrase it colloquially, the system favors
what makes the most sense, given the nature of the system and the specific
situation at hand.
This concept may cast a new light on the famous Alexandrian inventions of
types of philological writings. The quotation marks here and elsewhere already
indicate that the notion of an invention should not to be taken too literally
in this context. It is true that critical semeia (Zenodotus and Aristarchus),
philological glossai (Zenodotus), Pinakes (Callimachus), learned hypotheseis
(Aristophanes), lemmatized hypomnemata (Aristarchus), to name only the
most prominent types, are specialized and highly efficient functional types of
writings.258 But it was remarked repeatedly that their inventions, in the situ-
ations in which they occurred, were to a certain extent obvious steps to take.
We can now say that these typological developments reflect the disciplines
Systemrationalitt. The pull of the systems inherent rationality made the
inventions of these functional types of writings just about inevitable. Within
the exceptionally favorable framework that existed for some generations in
Ptolemaic Alexandriathis is the real miracle, so to speak, more so than the
specific successive stages in the development of philological scholarship once
the discipline had been set on its path259 these types of writings, arguably,
255 In pure systems theory, however, people are not considered part of the system, only com-
munication is; even the concept of a person and his or her actions undergoes consider-
able revision; cf. Luhmann [1984] 155 and elsewhere.
256 Cf. Gansel [2011] 4951, 8694 with other scenarios that show Systemrationalitt at work.
257 Cf. Luhmann [1984] 62, 122, and elsewhere; also, e.g., Gtje [2008] 205209. Continued
connecting communication is necessary for the existence of any social system. Where
there is no Anschlusskommunikation, the system immediately ceases to exist.
258 In this sense, even the writings characterized above as broad-band are specialized. In
the larger systemic context, their distinct specialization is precisely that they (and only
they) can serve a broad range of purposes.
259 The emergence and subsequent further development of philology in Alexandria con-
firms Luhmanns theory that communication is, on the one hand, fundamentally unlikely
598 dubischar
simply had to emerge, that is, had to be invented at some point, even more or
less in the very chronological order in which they in fact did emerge. In addi-
tion, we should keep in mind that these types of writings were not created
ex nihilo. Rather, they represent professionalized, systematized, and perfected
variants of practices that had been or could be cultivated in less systematic
and sophisticated fashion in pre- or sub-philological contexts.260 Therefore,
to put it bluntly, it does not require breathtaking ingenuity to develop the idea
of using marginal signs or interlinear annotations for philological purposes, or
to produce scholarly dictionaries, a catalog for the Librarys holdings, or intro-
ductions or commentaries to works of literature. The room for real individual
achievement on part of the formative figures of Alexandrian philology, as typo-
logical inventors, seems thus to become even smaller.
However, there is still something arguably breathtaking about the devel-
opment of types of philological writings in Hellenistic Alexandria, and the
distinguished literary scholars fame for having established them is deserved.
By introducing important new types of writings, Zenodotus, Callimachus,
Aristophanes, Aristarchus, and others were effective agents if not embodi-
ments of philologys Systemrationalitt. At crucial points in the disciplines
development they seem to have recognized what needed to be done, they had
the capacity (intellectual, organizational, infrastructural) to do it, and they did
it so well (systematically, thoroughly, comprehensively) that their solutions
lasted, even serving as tools and models for others.261 If this interpretation
is correct, the Alexandrian typological inventors may be praised for having
exercised a quasi-Hegelian freedom (Freiheit ist Einsicht in Notwendigkeit) in
recognizing and internalizing philologys necessities as they arose in certain
situations.
Moreover, the fact that in Hellenistic Alexandria philology rose for the
first time to the level of a distinct and self-conscious professional discipline
makes the famous scholars of Alexandria (in addition to their other accom-
plishments) truly formative protoi heuretai (first inventors) of the main types
* My gratitude to Klaus Alpers, Ian Cunningham, Marco Ercoles, Barbara Fero, and Stephanos
Matthaios for their helpful remarks on first drafts of this paper; I am also thankful to Ian
Cunningham for kindly revising my English.
1 For the sources of knowledge see Dickey in this volume. The following typology should be
mostly considered as a theoretical abstraction depending on the surviving work-titles and
fragments. A rigorous and exhaustive analysis of ancient grammatical theories and their
relation with the remainder of ancient treatises goes beyond the purposes of a typological
approach, which I have intentionally adopted. About the definitions and tasks of grammar as
well as its partitions (viz. the tripartite system of Asclepiades and the quadripartite one), see
e.g. Pagani [2011] 20f. nn. 1621 with rich bibliographical references. For the sake of brevity,
I will mainly refer to the LGGA-articles (www.aristarchus.unige.it/lgga/) on the single gram-
marians, where detailed and updated bibliography can be easily collected.
2 One must keep in mind that in ancient Greek grammar has never been an autonomous
discipline, being linked to rhetorical, philosophical and scholarly studies (see e.g. Blank
[2000] 400).
3 I follow here Ax [1982] 97 (= Id. [2000] 128f.) and Matthaios [1999] 15f. with bibl. (see
Id. [2007] 13f.), with required adaptations due to the typological approach (see n. 1). For
the Latin world see the useful model sketched by Ax [2005] (in part. 259), which will here be
taken into due account. Obviously, there are many overlaps within the groups and a rigorous
distinction between them according to the modern conception of grammar is impossible
(see e.g. Barwick [1922] 227 n. 2: Auch die ars grammatica [scil. ], soweit
sie die Flexionslehre behandelte, kann hier [scil. among works /de latinitate]
genannt werden).
4 Fehling [1956] 247 n. 1 suggested that the -type could have been created on the basis of
the (see below, 2), particularly on the introductory part dealing with
the doctrine of parts of speech and of inflection (see also Pinborg [1975] 112). For the reflec-
tion on the parts of speech in ancient Greek scholarship see also Swiggers-Wouters in this
volume (section III.2).
5 On and its criteria, see Pagani in this volume with an exhaustive discussion and
further bibliographical references.
6 See e.g. Fehling [1956] 258ff.; Pinborg [1975] 112.
602 Valente
studies.7 It must be remembered that grammar in its full and complete sense
was understood to be knowledge of literary compositions, accompanied only
in some cases by knowledge of what is said and thought in Greek according to
the common usage (Pagani [2011] 17).
Early definitions of grammar and its purposes were at first formulated in
Hellenistic philological works. The theoretical framework of grammar was
namely philology, as it is attested by the first definition of grammar as a com-
plete mastering in written literary works ( )8 formu-
lated by Eratosthenes.9 The philologically oriented perspective of Alexandrian
grammar,10 defined by Dionysius Thraxs as the maximally
extensive experience of what is said by poets and prose writers,11 would be
later questioned, among others, by Asclepiades, who stressed that grammar is
a techne of thing said by poets and prose authors.12 As Di Benedetto [2007]
417ff. rightly points out, Asclepiades definition suggests the evolution and
renewal of grammar during the 1st c. BC.
7 See Di Benedetto [2007] 417f.; Pinborg [1975] 110ff.; Cribiore [2001] 210; Matthaios [2009a].
See also Ax [2006c] 250: Der ars-Typ ist eine didaktisch motivierte Darstellung der
Sprachkonstituenten, die der Identifikation der Sprachelemente bei der Lektre und
weniger normativen Richtigstellungen dient. Influences of Stoic treatises (see
below, 1.2) dealing with the smallest elements of speech, , , parts of speech, as
well as and must be acknowledged (see e.g. Barwick [1922] 91f., 229f.).
8 See Pagani [2011] 17f. with n. 3; Matthaios [2011a]. For the translation of see L. Pagani,
PAWAG (http://www.aristarchus.unige.it/pawag/) s.v.; see also Swiggers-Wouters in this
volume (section II.2).
9 It could have been located in his : however, very little can be inferred about the
structure and the contents of this work in two books (see Geus [2002] 52 n. 38, 291 nn. 11f.,
304f., and Matthaios [2011a] 62f. with rich bibliography). See also Schenkeveld [1994] 263.
Pfeiffer [1968] 162 cautiously suggested its possible influence on Asclepiades :
see below.
10 Alexandrian scholars used to call themselves (on the Cratetean polemics
against the term in favour of see Sext. Emp. (Math. 1.79, 248) with Blank [1998]
140f., 259; Eratosthenes was the first to call himself (Suet. Gram. 10): see Pfeiffer
[1968] 158f., 238; Schenkeveld [1994] 265; Blank [2000] 404f.
11 Sext. Emp. Math. 1.57:
., transl.
Di Benedetto [2000] 395 (= Id. [2007] 522). See Di Benedetto [2007] 392ff.; Schenkeveld
[1994]263f.
12 Sext. Emp. Math. 1.74, transl. Pagani [2011] 19. On the polemics against Dionysius and on
the following definitions of grammar formulated by Ptolemy the Peripatetic, Chares/
Chairis, Demetrius Chlorus, and Tyrannion, see Blank [1998] 124146 and Pagani [2011] 19f.
Typology of Grammatical Treatises 603
13 Sext. Emp. Math. 1.250f. (~ Dion. T. GG 1.1.5.26.3), transl. Di Benedetto [2000] 395 (= Id.
[2007] 523). See Blank [2000] 410f., 413 tab. 58.4. See also Quint. Inst. 1.8.1321 and Ax [2011]
384404 with bibl.
14 See Matthaios [1999] 22 with nn. 46f. with bibl.
15 See Di Benedetto [2007] 419; Blank [1998] XLVf.; L. Pagani [2009a]; Montana in this
volume.
16 The and the might have been part of a complete work
in eleven books at least: see Usener [1913] 309 n. 125; Pfeiffer [1968] 158, 162 n. 8.
17 See Schenkeveld [1994] 264f.; Blank [2000] 412 tab. 58.1, especially on the basis of Sext.
Emp. Math. 1.9196, 252f.; see also above n. 7; Swiggers-Wouters (section II.2) in this
volume. On the later quadripartite system, counting a among its , Blank
[2000] 412 tab. 58.3; Pagani [2011] 21 with nn. 1921 for bibl.
18 On this topic cf. Montana in this volume.
19 See e.g. Asclep. Myrl. ap. Sext. Emp. Math. 1.44 with Blank [1998] 113 and n. 64. This bipar-
tition (elementary grammar vs. higher grammar aiming at the interpretation of classical
texts) is also attested by Phil. Congr. 148150 (3.102f. C.-W.): Blank [2000] 402.
604 Valente
20 Cribiore [2001] 210. She also points out that a limited systematisation of grammati-
cal tools already existed in the work of the Alexandrian scholars, where it served their
philological pursuits, and it is conceivable that some aspects of embryonic normative
grammar were already part of school practice in Hellenistic times. Only later, however,
did grammar arrive at a distinct methodology and a systematic articulation of a definite
body of knowledge; see also ibid. n. 121: the absence of grammatical exercises and text
in Hellenistic times is highly significant and cannot be explained, in my opinion, solely
on the basis of the limited number of literary and semi-literary papyri preserved from
this period. On a possible existence of some ars-type in the Roman world already in the
1st c. BC, see Ax [2011] 1315 (his hypothesis cannot however be automatically transposed
to the Greek world because of the lack of any certain proof).
21 School exercises are also good witnesses to this ongoing practice: Cribiore [1996] 38ff.
22 See Wouters [1979] 43; Schenkeveld [1994] 267f.
23 See Wouters [1979] 44f.
24 Grammar, and grammatical teaching, was not a process of uniform transmission of a
set of codified doctrines. Taken together with the other -papyri, our text [P.Berol.
9917] shows that grammarians proceeded through adoption, adaptation and transforma-
tion, and through retouching definitions, technical terms, and lists of examples. This
approach par bricolage explains the differences in formulation, in exemplification, and
in organisation of the textual testimonies for the -genre (Wouters-Swiggers [2011]
329). Differences concern, for instance, the number and sequence of the parts of speech,
their definitions and accidents (, viz. consequential attributes): see Wouters
[1979] 38f., 60 n. 49.
25 The attribution of this booklet to Dionysius Thrax has been strongly questioned since
antiquity: see Di Benedetto [2000] 397399; Id. [2007] 381ff.; Matthaios [2009a] (with
further bibl.); Pagani [2010b]; [2011] with further bibl.; Montana in this volume. For the
papyri carrying the text of the , not without modifications and adaptations, see
P.Hal. 55A (Wouters [1979] 109119, no. 4, 5th c. AD). The incipit of the is also pre-
served by PSI 1.18 under the title (5th c. AD: Wouters [1979] 120124, n0. 5),
after the treatise (~ GG 1.1.119.69), one of the ancient supplements to the
in the Mediaeval mss. (suppl. III: Uhlig [1883] LIIf.), which the scholiasts to the still
found in that order (see sch. [Vat.] Dion. T. GG 1.3.128.28ff. with Di Benedetto [2007] 459f.).
26 Works with the title have been also written by Lupercus (3rd c. AD: see
Ucciardello [2008] with bibl.) and by Astyages (see Ucciardello [2006a]), but no fragment
Typology of Grammatical Treatises 605
[1979] 141155, no. 8) dealing with and (here there are coincidences
with other grammatical doctrines such as Apollonius). See also P.Kln 4.177 (4th c. AD) on
vowels and consonants (see Cribiore [1996] 267, no. 371).
37 Wouters [1979] 196 with bibl.; see also Cribiore [1996] 266; Di Benedetto [2007] 459f.
38 On the false ascription, see Di Benedetto [2007] 405410; Wouters [1979] 9092 with rich bibl.
39 See Cribiore [2001] 212 with n. 127; Swiggers-Wouters [2003a].
40 See Wouters [1979] 6192, no. 2, in part. 81f. (see also 40 with nn. 37f. with bibl.); Cribiore
[2001] 212.
41 See below, 2.3.
42 See e.g. Sext. Emp. Math. 99130 with Blank [1998] 153170.
43 See e.g. Pohlenz [1939]; above, n. 7.
44 See Di Benedetto [2007] 416, 479; Schenkeveld [1994] 272f.
Typology of Grammatical Treatises 607
56 See Glck [1967] 33ff.; Blank [1998] 189191. See also Apoll. Dysc. 491.13: see Lehrs [1848]
424; Dyck [1983] 3 n. 3.
57 Dyck [1983] 3; see above, n. 47.
58 See Glck [1967] 31f.; Blank [1998] 189: the best known example are Priscians Partitions of
the first twelve verses of the Aeneid (Partitiones duodecim versuum Aeneidos principalium,
GL 3.457515).
59 See above, n. 31.
60 See Lehrs [1848] 426; Glck [1967] 32f.; Dyck [1983] 4f. Typical cases are the epimerisms to
Homer (originally in the form of scholia-epimerisms, which will be later alphabetized in a
lexicographic structure: see Dyck [19831995]). On the Byzantine practice see also Robins
[1993] 125148.
61 Fragments are collected by Lentz [1867] XVIIXXXIII and supplemented by Dyck [1981]
and [1993a] 793. On Herodians work and its relationship with the pseudoepigraph
Epimerisms (Boissonade [1819]), see Dyck [1981] and [1993a] 792f. with bibl.
62 See Cribiore [2001] 197.
63 See Brandenburg [2005] with the review by Schmidhauser [2007].
64 See Blank [2000] 414.
Typology of Grammatical Treatises 609
form).65 The deals first with the name of a part of speech (), then
its definition (), its (kinds), syntax and (viz. assigning words
to the discussed part of speech); the handles forms and pros-
ody of those words, their dialectal forms and affections ().66 Nonetheless,
each part of speech (noun, verb, participle, article, pronoun, preposition,
adverb, conjunction) was treated separately by various grammarians from the
1st c. BC onwards: the most influential works in this field of research were those
by Trypho, Apollonius Dyscolus and Herodian.67
Treatises on nouns were written by Apollonius Dyscolus (
or )68 and Herodian ( or or ).69
Subsections or separate monographs were also devoted to single topics, such
as Philoxenus (On comparatives and superlatives, frr. 331353
Theodoridis),70 Tryphos (frr. 8393 von Velsen),
in one book and in one book (in Suda
1115 Adler), Habros (frr. 1118 Berndt) and (frr. 9f.
Berndt), Apollonius Dyscolus ,71 Herodians (GG
3.2.849f.)72 and (GG 3.2.897903).73
Among the monographs on verbs, Apollonius Dyscolus or
74 was one of the most important works, receiving a commentary by
Zenobius (Schoemann [1881] 8f., fr. 5 ap. Etym. Gen. 124 Lasserre-Livadaras)75
and being extensively used by Herodian and later grammarians (such as
Priscian and Choeroboscus).76 The same general approach was seemingly
113 See Lentz [1867] LXXIIIXCVI; Hunger [1967a] 29; Dyck [1993a] 783786; Dickey [2007] 75.
114 See Ucciardello [2008].
115 See Blow-Jacobsen [1982] and Parsons [1982] respectively.
116 Wouters [1979] 188197, no. 14; Cribiore [1996] 266, no. 368.
117 On the origins of metrical studies and on ancient metrical doctrines, see now Ercoles
[2014] with bibl. The discipline was fully investigated by Aristoxenus of Tarentum in the
Elementa Harmonica: see Rocconi [20082].
118 See Ippolito [2006b].
119 Intermediate steps are represented by the epitomes in eleven and three books respec-
tively (Choerob. Prol. Heph. p. 181.1116 Consbruch).
120 See Dickey [2007] 104f. Suda 659 Adler adds a generic among the works
of the grammarian.
121 On Heliodorus see Rocconi [2004] with bibl.
122 See Ippolito [2006c].
123 See Ucciardello [2006a].
124 A survey in Reiter [2012] 155 n. 1.
125 See Reiter [2012].
126 See Tsantsanoglou [20102].
614 Valente
(pp. 73.1176.16 C.) testifies to the usage of the critical signs introduced by the
Alexandrian scholars.136
2 Monographs on Hellenismos
136 could have been the title of a work by Diogenianus (2nd c. AD):
see above n. 49.
137 See above, n. 5.
138 See e.g. Fehling [1956], and Ax [2005] 250 on Latin grammar: Der Typ De latinitate [...]
ist eine Zusammenstellung sprachlicher Zweifelsflle, die mit Hilfe der Sprachnormen
Analogie, Sprachgebrauch und literarischer Tradition einer Klrung zugefhrt werden
sollen. Sie ordnet das Material meist systematisch nach Wortarten mit dem Schwerpunkt
auf den Irregularien des Nomen.
139 See e.g. Ax [2011] 148ff. with bibl. Some Byzantine treatises on solecism and barbarism
survive, which made use of older grammatical sources: see e.g. Nauck [1867] 283312 and
Pontani [2011c] 102f. with further bibl.
140 See also Strab. 14.2.28 with Radts commentary: see Fehling [1956] 222.
141 See Ax [2005] 248f. with further references.
142 See Kleist [1865] 13 n. 15; Theodoridis [1976] 10.
143 See Reitzenstein [1897] 382; Theodoridis [1976] 10.
144 To be possibly identified with his (Suda 1115 Adler): see
Ippolito [2008].
145 See Razzetti [2003e].
146 See Razzetti [2002b].
147 His seven books or probably were a lexicon
according to Suda 29 Adler ( ): see Reitzenstein [1897] 383ff.
616 Valente
148 See Fehling [1956] 259, who groupes under this category treatises on and analogy;
Schenkeveld [1994] 283292. Some of the following works could be inserted within the
grammatical works on single parts of speech as well (see above, 1.1 and n. 4).
149 See Lentz [1867] CVIII; Dyck [1993a] 789. P.Flor. inv. 3005 = P.PisaLit. 26 (4th c. AD: Carlini
[1978]; Wouters [1979] 216224, no. 17) preserves fragment of an epitome of the work.
150 The origins of such grammatical investigations date back to the Sophistic movement: see
e.g. Pfeiffer [1968] 38 (on Protagoras).
151 See Schneider [1910] 5868.
152 See Ucciardello [2008].
153 Suda 691 Adler reports that in these books : see Ucciardello
[2008] for a survey of critical remarks and suggested corrections to the text, as well as a
general interpretation.
154 See Cohn [1884a] 7: ea verba tractavit, quorum formatio vel declinatio a vulgari usu rece-
dere et ab analogiae ratione abhorrere videbatur.
155 Orus wrote a (Suda 201 Adler); the title for
this work of Herodian is also attested (GG 3.2.907, see Lentz [1867] CXVI). Models for this
work could have been Seleucus and
(frr. 71f. M.): see Razzetti [2002b]; Ucciardello [2006c].
156 Lentz [1867] CXVIICXXII. See also Dyck [1993a] 790f.; Sluiter [2011].
Typology of Grammatical Treatises 617
were later included in scholarly activity. The codification of such texts was
brought about during the 4th/5th c. AD by Theodosius of Alexandria169 in his
,170 containing rules and
tables on the declensions of nouns (GG 4.1.4.136.13), on the position of the
accent in nominal declension (36.1442.8), and on the conjugation of verbs,
with the paradigms of the verbs and (4399). This work also pro-
vided a suitable supplement for the .171
Some treatises on analogy dealt specifically with some parts of speech,
such as Tryphos (fr. 20 von Velsen)with
some possible overlap in content and/or inspiration with Philoxenus
172, his (fr. 21 von Velsen)173
and his in one book (fr. 81 von Velsen).174
As regards dialectal issues, it is not always easy to tell, on the basis of the
preserved titles and fragments, whether some works had a lexicographic
or rather a grammatical structure: this is the case of Demetrius Ixions
(fr. 40 Staesche).175 According to the two surviving frag-
ments, the by Aristocles contained some prosodic remarks
P.Berol. inv. 22141 (7th c. AD: Cribiore [1996] 268, no. 377). See Cribiore [2001] 214: some of
the verbs used in exercises, such as didaskein, graphein, and typteinto teach, to write,
and to thrashseem to allude to a school practice. For declension tables see P.Berol.
sine num. (2nd/3rd c. AD: Cribiore [1996] 272, no. 384), PSI inv. 479 (5th/6th c. AD: Cribiore
[1996] 267, no. 372: declensions of , and of the adjective ), PSI inv.
2052 (5th/6th c. AD: Cribiore [1996] 267, no. 374: declensions of and of ).
169 See Robins [1993] 111123; Cribiore [1996] 52f.; Pagani [2006b]; Dickey [2007] 83f.
170 On the term and its derivatives, see Norden [1905] 508528.
171 Cribiore [2001] 214. Theodosius work would enjoy an enormous fortune in the Byzantine
age (as it was transmitted together with the as its supplement: Uhlig [1883] XLVII)
and in the Renaissance, being also the model for modern Greek grammars: see Pagani
[2006b] with further bibl. On the presence of the athematic declension see Cribiore [2001]
214f. On the later and no longer extant by the grammarian Astyages
(Suda 4259Adler), see Ucciardello [2006a].
172 See von Velsen [1853] 22.
173 De argumento nihil traditur; non ambigendum tamen, quin Trypho in iis quoque sub-
stantivis, quae fere anomala vocitantur, analogiam quandam valere demonstravit (von
Velsen [1853] 23). Ippolito [2008] suggests that these two works could be subsections of
the same.
174 See above, 1.2.
175 The title of the work by Demetrius Pyktes has also been questioned (see
Pagani [2007b] with bibl.), and the surviving fragment refers to the etymology of /
. The content of the by the grammarian Astyages is obscure as well:
see Ucciardello [2006a].
Typology of Grammatical Treatises 619
176 Hdn. GG 3.2.18.16 (An. Ox. 3.298.25299.1 Cramer, Lehrs [1848] 370.26f.; cf. Hdn. GG
3.1526.1f.) on the Attic termination - of non-oxytone nouns in -; Etym. Gud. 353.39f. (s.v.
) = Etym. Magn. 545.8 on the Attic in . See Corradi [2007] with bibl.
177 See Kowalski [1928] 167; Pagani [2005c].
178 See Kleist [1865] 13f., 52ff.; Theodoridis [1976] 1012. The attribution of a
to Herennius Philo, still asserted by Palmieri [1988] 4446, is wrong
(see Theodoridis in apparatus to frr. 311 and 321, and above, n. 83).
179 Reitzenstein [1901] 87: Die Fragmente lassen uns erkennen, da Philoxenos wirklich ver-
sucht, mittelst seiner Abhandlungstheorie das Latein lediglich als Abart des Griechischen
zu erweisen. Im einzelnen ist der Einflu der Stoa unbestreitbar. However, the treatment
of Latin as a dialectal variety of (or a derivation from) Greek seems to have been suggested
by (Claudius?) Didymus in his (frr. 447450 Funaioli): see
now Braswell [2013] 9092.
180 See Haas [1977] 98, 176f. See also Pagani [2009c] with bibl.
181 See Cohn [1894b] 2805.
182 See Schndeider [1910] 138f.
183 It can by no means be established whether this treatment of Aeolic belonged to a more
extensive work (Wouters [1979] 294).
620 Valente
184 See Valente in this volume (section III.2) with further details and bibl., and Id. [2014a].
185 See Dyck [1993a] with bibl.; Alpers [2004] 1ff. with bibl.; Dickey [2007] 7577 with bibl.
186 See Lallot in this volume.
187 Synt. 51.152.5 (with Blank [1982] 15), where orthography is also mentioned. The same
parallelism with orthography also appears in 7.614, see Blank [1982] 9, 18, [1998] 195, and
Valente in this volume (section III.2).
188 See Blank [1993] 711f.; Lallot [1997].
Typology of Grammatical Treatises 621
189 See also Blank [1982] 9f.; [1993] 721727; [2000] 411415 with rich bibl.; Schenkeveld [1994]
293.
190 See Blank [1982] 27f.; [1993] 724f.; [2000] 415; Schenkeveld [1994] 293298.
191 On the absence of any systematic treatment of syntax in the , see above, 1.1.
192 See Schenkeveld [1994] 298.
193 See Dion. Hal. Comp. 4 (3.22.823.2); Pinborg [1975] 102; Schenkeveld [1994] 273; Blank
[2000] 403.
194 Trypho started to develop syntactical remarks pursuant to a more grammaticalised frame-
work, without elaborating any comprehensive theory on syntax: see Matthaios [2003].
195 Remarks on Attic syntax can also be discovered in Atticist lexica, but the sources of such
remarks are not easily traceable: see Valente [2014b] with bibl.
chapter 5
1 Structural Distinction
1.1 Horizontal Onomastic Structure
1.2 Vertical Lexicographic Structure
2 Scope Distinction
2.1 General Lexica
2.2 Lexica Concerning a Single Author
2.3 Lexica Regarding One Genre
2.4 Lexica on Particular Themes
3 Content Distinction of General Lexica
3.1 Lexica with Various Contents
3.2 Grammatical Lexica
3.3 Lexica Whose Content is More Properly Morphological and
Orthographic
3.4 Etymological Lexica
3.5 Synonymic-Differential Lexica
4 Distinction among Lexica Concerning a Single Author
4.1 Real and Proper Lexica, Whose Entries are Ordered
Alphabetically
4.2 Glossaries, Whose Entries Follow the Order in Which
They are Found in the Text
5 Distinction among Lexica on Particular Themes
5.1 Geographic Lexica
5.2 Biographical Lexica
1 Structural Distinction
Herond. 7,5761, where one finds a long list of shoe names, similar to the list of
Poll. Onom. 7,94). Callimachus must have been particularly
important (however, it is not possible to concur with Schoenemanns assertion
that the poet from Cyrene was the first to adopt the onomastic system) and
the by Aristophanes of Byzantium was undoubtedly also important. In
Callimachus work numerous lists of animals, objects and phenomena with
frequent ethnic-dialectal indications were present. Pfeiffer [1968] 135, build-
ing on the conjectures of several scholars, puts forward some suggestionsin
the light of Suda 227 A.concerning the possible titles of some sections.
However, the only certain one (fr. 406 Pf.), attested by Athenaeus (7,329a) and
Eustathius (1936,14), is the title of the section about fish (see also Montana in
this volume).
Even in Aristophanes, while the introductory section adopted an embryonic
form of lexicon, the later sectionsorganized by semantic fieldswere ono-
mastic (the titles are , and
). In fact, Wendel [1939a] 508 regarded them as closely related
to Callimachus because the clarification of the geographic
area in which a given term is used often appears here (but Aristophanes was
thereby simply responding to a generalized interest of the whole of Hellenistic
culture).
Other authors dealt with technical language: Eratosthenes compiled ono-
mastics entitled and , which dealt with the terms
of artisans work and with those regarding domestic utensils, in a strong relation
to studies on comedy. These works became one of the main sources of Pollux,
especially in the tenth book. Apollonius wrote a whilst Nicander of
Colophon and Philemon of Aixone (3rd c. AD) collected the names of objects
useful in daily life, and Suetonius and Telephus from Pergamon (2nd c. AD) the
names of items of clothing.
Particularly relevant, in this context, were the medical works: Amerias com-
piled a , a collection of the names of medicinal herbs, Xenophon
of Cos, Apollonius of Memphis and, in the Traian age, Rufus of Efesus and
Soranus composed lexica collecting the names of body parts, while Soranus
also wrote the . Similarly, in the culinary field it is worth recall-
ing Artemidorus (1st c. BC), which probably included
explanations for terms and references to passages of comedy. Another figure
engaged in a vast range of activity was Tryphon (1st c. BC), who, among other
works, also compiled the (concerning the names of edible plants), a
and collected the names of musical instruments.
Other works concentrated on the terms peculiar to a local parlance. Among
the most notable, mention should be made of the by the above
Typology Of Lexicographical Works 625
These lexica are also classifiable according to how strictly the criterion of
alphabetical order is followed. In this respect, particularly lucid and useful is
the study by Esposito [2009] 259265, who also highlights the possibility
concerning above all the late lexica of an encyclopedic typethat bodies of
glosses deriving from various sources were maintained compact, thereby inter-
fering with the alphabetical order. She also hypothesizes that, at least origi-
nally, the dimensions of the lexicon would be responsible for the order (few
glosses could be distributed alphabetically only according to the first letter):
exemplary in this regard is the case of Orion (cf. Wendel [1939b]). Esposito
[2009] also presents (p. 263) a table of papyrus lexica, classified according to
whether the alphabetical order:
1. is limited to the first letter (P.Yale II 136 [2nd c. AD], P.Oxy. XV 1804 [3rd
c.AD], Bodl.Ms.Gr.Class. f. 100 [P] fr. 1 [4th5th c. AD], P.Oxy. XV 1803 [6th
c. AD], P.Ness. II 8 [7th c. AD])
2. concerns the first two letters (this is the most numerous group and
includes the most ancient findings: P.Hib. II 175 [3rd c. BC], P.Berol. inv.
Typology Of Lexicographical Works 627
9965 [3rd2nd c. BC], P.Heid. I 200 [3rd2nd c. BC], P.Heid. inv. 3069v.
[2nd c. AD], P.Oxy. XVII 2087 [2nd c. AD], P.Oxy XV 1801 [2nd3rd c. AD],
P.Sorb I 7 [2nd3rd c. AD], P.Mnch. II 22 [3rd c. AD], P.Oxy III 416r. [3rd
c. AD])
3. concerns the first three letters (P.Ryl. III 532 [2nd3rd c. AD], MPER N.S.
XV 142 [6th7th c. AD])
4. is rigorous (P.Oxy. XV 1802 + LXXI 4812 [2nd3rd c. AD, on which see
Schironi [2009a], Esposito [2011], P.Oxy XLVII 3329 [4th c. AD], PSI VIII
892 [4th c. AD]).
1. on different columns;
2. divided by an empty space;
3. separated by a dot on the top;
4. no marker of separation.
2 Scope Distinction
Another distinction concerns the scope of the lexicon. In this respect, four
types of lexica are found:
regard, see Alpers [1981] 121123) and Philostratus of Tyre. P.Oxy. XV 1804 (3rd
c. AD) preserves a fragment of a lexicon of this type whose author is Diodorus.
The lexicon by Harpocration, who lived in the 2nd c. AD, has come down to
us in two compilations, one quite extensive and perhaps close to the original,
the other being Byzantine and highly epitomized. Harpocrations lexicon is of
crucial importance because it was the source for various Byzantine lexica.
This type of lexicon is undoubtedly the most common and best docu-
mented, also because it aptly responded to the needs of Byzantine encyclo-
pedism. Clear examples are the Suda (on its plural sources, see Adler [1928]
XVIXXII), the Pseudo-Zonaras (whose value was underestimated for too long,
until its revaluation by Alpers [1972]) and the Lexicon Vindobonense by Andreas
Lopadiotes (with regard to the latter work, see in particular Guida [1982]).
Alpers [2004]) or the Laur. 59,16, of which Adler [1928] XVI noted the rela-
tionship with the Suda, edited by Massa Positano-Arco Magri [1966].
Atticist lexicon. This typology was also employed in the Byzantine age: between
the 13th c. and the beginning of the 14th we have the works of Moschopulus
and Thomas Magister (on the persistence of Atticism in Byzantium, see the
enlightening pages of Wilson [1983] 48).
4.1 Real and Proper Lexica, Whose Entries are Ordered Alphabetically
Here we should include the lexicon by Apollonius the Sophist on Homer and
those on Plato compiled by Timaeus and the Pseudo-Didymus (effectively only
in its first part: the second part is completely chaotic), but there are also papy-
rus findings that belong to this category, such as P.Oxy. XXIV 2393, reproducing
fragments of a lexicon on Alcman.
4.2 Glossaries, Whose Entries Follow the Order in Which They are Found
in the Text
Obviously, such instruments require a consultation strictly linked to the cor-
responding text (see Dubischar in this volume): appropriately, Montanari
636 Tosi
Within lexica on particular themes, attention should focus above all on two
types:
Edited by
LEIDEN | BOSTON
Cover illustration: Venetus A: Marcianus Graecus Z. 454 (= 822), folio 12, recto. Center for Hellenic Studies,
Harvard University.
Brills companion to ancient greek scholarship / Edited by Franco Montanari, Stephanos Matthaios, and
Antonios Rengakos.
pages cm. (Brills companions in classical studies)
Includes bibliographical references and index.
ISBN 978-90-04-24594-5 (hardback set : alk. paper) ISBN 978-90-04-28190-5 (hardback, v. 1 : alk.
paper) ISBN 978-90-04-28191-2 (hardback, v. 2 : alk. paper) ISBN 978-90-04-28192-9 (e-book)
1. Classical philologyHistoryTo 1500. 2. Classical languagesGrammar, Historical. I. Montanari,
Franco, editor. II. Matthaios, Stephanos, editor. III. Rengakos, Antonios, editor.
PA53.B74 2015
480.09dc23
2014032536
This publication has been typeset in the multilingual Brill typeface. With over 5,100 characters covering
Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities.
For more information, please see www.brill.com/brill-typeface.
issn 1872-3357
isbn 978-90-04-24594-5 (hardback, set)
isbn 978-90-04-28190-5 (hardback, vol. 1)
isbn 978-90-04-28191-2 (hardback, vol. 2)
isbn 978-90-04-28192-9 (e-book)
Volume 1
Prefaceix
List of Contributorsxvi
Part 1
History
2 Hellenistic Scholarship60
Fausto Montana
Part 2
Disciplinary Profiles
2 Definitions of Grammar515
Alfons Wouters and Pierre Swiggers
Volume 2
Part 3
Between Theory and Practice
Section 3.1
Scholarship
Section 3.2
Grammar
3 Syntax850
Jean Lallot
contents vii
6 Orthography949
Stefano Valente
Section 3.3
Philological and Linguistic Observations and Theories in
Interdisciplinary Context
3 Mythography1057
Claudio Meliad
Bibliography1267
General Index1427
Passages Index1463
part 3
Between Theory and Practice
section 3.1
Scholarship
chapter 1
The Hellenistic age has rightly been seen as a civilization based on books,
that is to say, a society in which the spread of written copies of poetic-
literary works gradually intensified and became customary. Possession of
books and personal reading became considerably more significant than in the
past, even though use of written books had already begun to play an increasing
role in the preceding two centuries.1 As stated by R. Pfeiffer: It is obvious that
we have reached the age that we called hesitatingly a bookish one; the
book is one of the characteristic signs of the new, the Hellenistic, world. The
whole literary past, the heritage of centuries, was in danger of slipping away in
spite of the learned labours of Aristotles pupils; the imaginative enthusiasm
of the generation living towards the end of the fourth and the beginning of the
third century did everything to keep it alive. The first task was to collect and
to store the literary treasurers in order to save them for ever.2 The idea that
scholars should be concerned with preserving the magnificent culture and
education (paideia) of previous centuries was certainly not restricted to the
material aspect of book production and the collection of exemplars. The decisive
1 For a survey of the history of classical scholarship see Pfeiffer [1968], Montanari [1993b],
Montanari [1994a], Matthaios-Montanari-Rengakos [2011], Montana [2012c], Montanari
[ forthcoming], and Montana in this volume; LGGA is a specific lexicon of the figures of the
ancient scholars; Dickey [2007] provides an overview of the materials of ancient scholarship
(see also Dickey in this volume); for an outline of the ideas and concepts of literary criticism
present in these materials, see Meijering [1987], Nnlist [2009a] (with the rev. by L. Pagani
[2009b]).
2 Pfeiffer [1968] 102.
3 On the role of Aristotle and of the Peripatos, see Montanari [2012d] with the bibliography
(in particular Montanari [1994a], Montanari [2000a]); see also Montana, Hunter, and Nnlist
in this volume.
4 On the typology of philological writings, see Dubischar and Tosi in this volume.
5 Suidas, ( 74 Adler).
6 Tzetzes, Prolegomena de comoedia, Prooem. I 112, Prooem. II 14, 2239 Koster; Alexander
Aet. TrGF 1, 100 T 6 = T 7 Magnelli; Lycophron TrGF 1, 101 T 7; cf. Pfeiffer [1968] 101, 105106;
and Montana in this volume.
Ekdosis. A Product of the Ancient Scholarship 643
other manuscripts as well as by his own conjectures. Diorthosis can be the term
for either kind of correction. It is hard to imagine any other way.7
So the first diorthotes of Homer selected a copy he considered to be suit-
able and worked on it in various ways. On this important point Pfeiffer and
K. Nickau, who has produced fundamental studies on Zenodotus,8 are in
agreement. H. van Thiel likewise believes the Alexandrian ekdosis consisted of
the copy chosen by the grammarian from among those available, and provided
with a series of annotations.9 M. West suggests that the particular eccentricity
of Zenodotus text could not have been due merely to his judgment and opin-
ions, but must in part have reflected an eccentricity of the tradition on which
he based himself: he may have worked on a rhapsodic exemplar produced in
an Ionian context, which thus reflected a line of tradition differing from that
which subsequently became widely accepted and which was predominantly of
Attic origin. Thus it may have been an exemplar he had perhaps brought with
him to Alexandria from Ephesus, his native city. According to this hypothesis,
Zenodotus worked on a Homeric text characterized by idiosyncratic aspects:
consequently his Homeric text, resulting from the combination of the base-
text plus the annotations in the margins, was necessarily influenced by this
circumstance.10 Of course, this is no more than a mere hypothesis, which, how-
ever, is based on the same vision with regard to the manner of working of the
pioneer of Hellenistic philology: namely, choosing a copy and performing a
diorthosis, i.e. carrying out corrections on the copy in question, in order to pro-
duce his own ekdosis of Homer.
By pondering on these themes over the years, I have come to the conclusion
that the problem of the characteristics of the Alexandrian ekdosis can be prof-
itably addressed by starting from its concrete form in terms of its production
as a book, on the basis of the following presupposition: in order to understand
the nature of what we call a grammarians ekdosis of a text and what it con-
tained, it is crucial to examine the way in which it was materially constructed.
I have therefore tried to emphasize the importance of the relationship between
the bookshop artefact on one hand and the text as an object of philological
editing, with its various paratextual elements such as annotations and semeia
(critical signs), on the other.11 We must take into account and award suitable
prominence to what we know regarding the creation of new copies of texts,
in the scriptoria by professional scribes or also privately by individuals, along
with insights that can be gleaned from surviving examples. To look at the prob-
lem in this perspective, the papyri are an essential source of information that
cannot be disregarded; we will thus start from the papyri to search for data
helpful to illuminate these issues.
It is an accepted and well documented fact that new examples of literary
works were normally re-read and corrected through additional further com-
parison with the antigraph, at times even on the basis of a collation with
other copies. Numerous types of evidence for this can be adduced on the
basis of papyrus fragments of literary texts, and papyrologists are fully aware
of the phenomenon of corrections introduced in order to improve an exemplar
in the framework of book production. Naturally we are particularly interested
in the most ancient evidence, although we are hampered by the fact that the
papyri datable to the period between the last decades of the 4th and the 3rd
century BC (the era of Zenodotus) are very limited in number. This notwith-
standing, some small corrections of material errors can already be observed
in the two most ancient surviving literary papyri, the Persians of Timotheus
(PBerol. inv. 9875) and the renowned Derveni Papyrus,12 dated to the last
decades of the 4th century BC (recall the dates of Zenodotus: ca 325ca 260).
Such examples suggest that these corrections were not the result of a system-
atic revision, but were made by the scribe, perhaps in scribendo. Though not
classifiable as a highly striking phenomenon, these occasional corrections of
small errors certainly represent the most ancient and visible evidence of a con-
cern for a correct text, or better, of the intention to correct a text in which an
error could be perceived.13 A few decades later we already find some consider-
ably richer and more significant witnesses, which I will now briefly summarize.
One noteworthy witness is the Homeric roll P.Ilias 12, of which substan-
tial parts are preserved, pertaining to books XXI, XXII and XXIII of the Ilias,
dated between 280 and 240 BC, thus still in the Zenodotean era or shortly
11 Montanari [1998d], [2000b], [2002a], [2004], [2009b], [2009d], [2011b], and [forthcom-
ing], with extensive bibliography.
12 Turner Parsons [1987] 92; text in Kouremenos Parssoglou Tsantsanoglou [2006].
13 See Montanari [2009b] 146147; Montanari [2011b] 34.
Ekdosis. A Product of the Ancient Scholarship 645
14 P.Heid.Lit. 2 (inv. 12621266) + P.Hib. 1.22 (Bodl.Libr. inv. Ms.Gr.Class.b3(P)/2) + P.Grenf. 2.4
(Bodl.Libr. inv. Ms.Gr.Class.b.3(P)) = MP3 979: cf. S. West [1967] 136191; Sforza [2000].
15 S. West [1967] 136137.
16 It is well known that the witnesses (both direct and indirect) of the Homeric text which
date from the early Hellenistic age (roughly up to the 2nd c. BC) show the presence of
additional lines as compared to the numerus versuum fixed at a later stage, which cor-
responds to that of the modern editions: cf. further on and Haslam [1997] for an effective
overview.
17 The so-called vulgata can be defined as the Homeric text that prevailed in the trasmis-
sion, cf. shortly below.
18 S. West [1967] 137.
19 S. West [1967] 133, 137; cfr. Pfeiffer [1968] 218; Nickau [1977] 261; McNamee [1992] 9 nn.
56, 15 n. 31, Table 1 p. 28, Table 2 p. 38, Table 3 p. 43; Montanari [1998d] 16, and Montanari
[2012c] 2829, with additional bibliography.
646 Montanari
At the side of l. 23.157 there are traces of a sign rather like a diple:20 if it
were genuinely a diple, then one would have to raise the question of when the
signed was marked on the papyrus. That is to say, one would have to endeavor
to ascertain whether or to what extent it dates from a time later than the
base text, given that it is normally believed that the diple was introduced by
Aristarchus21 (born around 215 BC), and in effect a sch. of Aristonicus to this
line provides information on an Aristarchean observation (and draws a paral-
lel with 2.278).22 This papyrus is a witness that should certainly be the object
of an in-depth re-examination from all points of view, including from the per-
spective of paleography, above all to determine the time gap between the base
text and the subsequent interventions.
Of a slightly more recent date, but equally significant, is P.Odyssey 31, dated
to between 250 and 200 BC, which contains parts of books IX and X of the
Odyssey.23 This too has been examined by S. West: The text had undergone a
double process of correction and collation. The original scribe appears to have
had two MSS. at his disposal, and in several places he has cancelled readings cor-
rect in themselves in favour of readings which are no better and are sometimes
obviously worse. Presumably he had more faith in the MS. which he used in
order to correct than in that from which he originally copied it...Corrections
have also been inserted by a second hand, which can usually be distinguished
without difficulty from that of the original scribe. The readings inserted by m.
2 agree with the Vulgate, but in several places where the text diverges consid-
erably from the Vulgate there is no trace of a correction. There are no marked
affinities between this text and that of any of the Alexandrian critics. There
are plus-verses (they are quantitatively fewer as compared to the previous
exemplar) and marginal signs, most of which are probably of a stichometric
character (but the left-hand margin is often lost). The most important fact is
that the roll underwent a twofold process of collation and correction: the first
scribe would seem to have had two exemplars available, and he often corrected
his text on the basis of another manuscript, after which a second hand inserted
readings that are in agreement with the vulgata.
These are witnesses of great importance for the question we are examining
here: we have two Homeric exemplars from the mid-3rd century BC, there-
fore definitely and decidedly pre-Aristarchean, which in addition to various
kinds of often somewhat problematic critical signs and the expected plus-
20 S. West [1967] 138. The usual form of the diple is >.
21 Pfeiffer [1968] 218.
22 The observation concerns the use of the verb in the plural with a singular but collective
subject ( ), as is also found in Il. 2.278: cf. Matthaios [1999], p. 384.
23 P.Sorbonne inv. 2245 A = MP3 1081: cf. S. West [1967] 223224.
Ekdosis. A Product of the Ancient Scholarship 647
verses, also show rather clear evidence of collation with other copies and a
conspicuous number of interventions performed on the base-text at various
times and in various different ways.24 M. Haslam says: Our earliest Homeric
manuscripts, those of the 3rd cent. B.C., are characterized by their startling
degree of difference from the text that prevailed later, sometimes known as
the vulgate...We now have fragments of about forty Homer manuscripts
written c. 150 B.C. or earlier...Several of these early manuscripts give evi-
dence of having been collated with another exemplar25 (so wild is hardly the
word for them) and sometimes reveal that vulgate readings coexisted along-
side eccentric ones...Of the variants entered from a second exemplar in the
most extensive of the early Iliad papyri (P 12 [scil. P.Heid.Lit. 2 + P.Hib. 1.22 +
P.Grenf. 2.4]) four coincide with the vulgate...A similar picture is presented
by the most extensive of the early Odyssey texts, P 31 [scil. P.Sorbonne inv. 2245
A]... The Homer of readers in the 3rd and early 2nd century...was appre-
ciably more flaccid than the Homer of subsequent readers.26 This was the situ-
ation that Zenodotus and his earliest successors found themselves facing.27
From the 3rd century BC we have the Milan papyrus with epigrams by
Posidippus, P.Univ.Milan. 309, another important piece of evidence in view of
the quantity of corrections and annotations the text presents.28 The majority
of the corrections were made by the same scribe, clearly in scribendo (in
general amounting to one and never more than three letters and all aimed
at correcting minor slips in the drafting stage), but subsequently, after the
copyists corrections, two other hands intervened with further emendations
and the differences in approach should be recognised.29 The third person to
make changes to the text in col. XI recorded a variant on the reading of l. 30,
noting it in the upper margin: at col. XI 30 we can read [ and in the
upper margin it is written (the last three letters are not visible in the
photograph but can be seen in the original document).30 It is extremely likely
that this is a correction or a variant, probably for the of the text, an
inversion has been proposed, but it is not clear, owing also to the
fact that the rest of the verse has not been preserved.31
The papyrus fragments of the following centuries, and in particular of the
three centuries of our era (the era with regard to which the papyrus findings
are most abundant) provide us with rich and valuable documentation of exem-
plars with interventions of deletion, addition and corrections of all types. The
following significant examples will suffice for our purposes;32 they could be
easily increased with fragments from various periods.
POxy. 2161, of the 2nd century AD, contains Aeschylus Diktyoulkoi. The
scribe has occasionally corrected some of his own errors: for instance, in l. 831
he wrote , but then crossed this out with an oblique line through each let-
ter, writing supra lineam the correct reading ].
PBerol. inv. 9872 (BKT II), of the 2nd century AD, is a long papyrus roll
(75 columns plus various fragments) that contains a commentary on Platos
Theaetetus with a substantial number of corrections. The most recent editors of
the roll, G. Bastianini and D. N. Sedley, write: The volumen has been proof read
and corrected in many places: letters or words omitted in the original drafting
stage have been restored, superfluous letters or words have been cancelled,
letters judged to be mistakes have been replaced by those considered correct.
All these changes do not appear necessarily to presuppose a collation with an
exemplar different from that of the copy...The variety of ways the corrections
have been made may lead one to suspect that the roll had been corrected on
various occasions: the first hand (a diorthots in the scriptorium) added the
missing words, which are marked in the upper margin...or lower...or are
placed after the line directly in the intercolumnium...A later hand or hands,
appears to have gone through the whole text, cancelling with a line in ink all
the letters judged to be wrong.33 For example, at col. LXIII, l. 6 the scribe had
written , omitting some words. In the intercolumnium to the
left, the corrector has put the sign of an upwards-pointing ancora (something
similar to an arrow) and in the space between and has written ;
in the upper margin, one can read the words | (),
which were probably preceded by a downwards-pointing ancora now lost
in lacuna. The corrected text is therefore
|.
POxy. 2256, of the 2nd3rd centuries AD, contains hypotheseis of various
tragedies by Aeschylus. The fragmentary hypothesis of fr. 3 recalls the vic-
tory, with the trilogy of which the Danaids was a part, against Sophocles and
another author, probably Mesatos (l. 5). After the name of the latter and at the
beginning of the following l. 6, round brackets can be clearly seen, which are
generally used as a sign to indicate expunction in literary texts and non-literary
documents. It is clear here that the round brackets were placed in scribendo,
which can be explained solely by imagining that the scribe copied from an
exemplar where the expunctions were already present to indicate that the
plays placed between brackets had been mistakenly placed after the name of
Mesatos.34
The copy of the Gospel according to St. John contained in PBodmer 2 dates
to the 3rd century AD. The scribe has corrected the text in a variety of ways.
There are supra lineam additions (ll. 2 and 12) and words rewritten above parts
of the text cancelled with a sponge: in ll. 910 has been written over
a word that has been scrubbed out and which continued in the following line,
where the letters can be made out in the remaining space; the second
part of l. 10 has been rewritten; at the beginning of l. 11 is the remains of an
eliminated reading, subsequently punctuated with dots as well as small round
brackets supra lineam.
I turn now to a manuscript which, I believe, provides us with what can be
termed an anthology of the techniques and methods available for correcting
and improving a text: POxy. 2404 + PLaur. inv. III/278, a fragment of a papyrus
roll (late 2ndearly 3rd century AD) containing a part of 5153 (POxy. 2404) and
of 162163 (PLaur. III/278) of Aeschines oration Against Ctesiphon.35 It seems
quite evident that this copy of the oration of Aeschines has been collated with
a second exemplar and has been the object of detailed and systematic correc-
tion seeking to identify the textual structure by distinguishing cola and periods
and to correct copying errors for the benefit of the reader, and to emend the
text in places judged unsatisfactory, by means of various different methods
of deletion and by writing the alternative readings above or beside the wrong
interpretations.36 We can see that the work of proof reading was not limited
solely to correcting minor errors as discretely as possible in order to reduce
the possibly negative impact of emendations on the appearance of the text.
In fact, more evident corrections, albeit written with care and precision, have
been made, with the apparent aim of improving the text and enabling it to be
read according to the intention of the corrector or correctors. The methods of
corrections and cancellation used in this papyrus are: the use of dots above
a letter, an oblique (single or double) line through a letter in question and,
for longer sequences, a line above or through the letters to be deleted, or by a
combination of these methods. We have also seen the widespread practice of
simply writing the correct letters above those judged incorrect as way of indi-
cating a deletion, as it were, automatically without the need for other material
indications. Another form of correction is the addition of words between the
lines or in the margins. As regards punctuation, the scribe provided the text
only with paragraphoi, whilst copious punctuation was added (at least, so it is
thought) by a later hand.37 Most of these are dots, placed slightly higher than
the letters, which had already been written, making sure that the dots were not
above a letter but in the narrow space between the end of the preceding word
and the beginning of the next. A lower dot can also be seen at col. I, l. 17. The
system can be described as follows: the upper dot combined with the para-
graphos marks the end of a sentence; the upper dot on its own distinguishes
the cola of the sentence; the lower dot indicates a weaker pause.38 This consti-
tutes proof of a serious attempt to highlight the syntactic and rhetorical struc-
ture of the text, leading us to consider the role of punctuation in Alexandrian
philological exegesis (rather than the complex and idiosyncratic system cre-
ated by Nicanor, one can mention the simpler and more widely-used system of
the three stigmai of Dionysius Thrax).39
I wish to emphasize at this stage why I have drawn attention to these manu-
scripts and their characteristics, with a choice of significant examples, to which
others could easily be added.40 The point is not that they may be considered as
exemplars of a grammarians ekdosis: there is absolutely no evidence for such
a suggestion. Rather, in my view they are of value because they highlight the
importance of the techniques adopted in the workshop for book production
and the effect such craftsmanship had on the development of a philological
practice that sought to ameliorate and emend texts regarded as unsatisfac-
37 Turner [1957] 130: The second hand not only revised the text for errors but collated
its readings with an exemplar different from that from which it was copied; cf. Turner
[19802] 212; Colomo [2008] 1516, 24. On punctuation marks in papyri, see Turner [19802]
9293; Turner-Parsons [1987] 910.
38 Colomo [2008] 1516.
39 Dion. T., Ars Gram. 4; cf. Colomo [2008] 1522; Montana [2009b]. For the punctuation
system employed by Dionysius Thrax and Nicanor, see also Montana, Matthaios, Dickey,
and Valente (2.4) in this volume.
40 Other useful material can be found in S. West [1967] passim and Haslam [1997] 6369.
Ekdosis. A Product of the Ancient Scholarship 651
tory due to the (real or supposed?) errors they contained. The papyri provide
ample evidence of the different methods used to improve an exemplar of a
book, in other words to correct the (new) copy of a text. It was considered
appropriate to add, remove or modify letters or words that had been omitted
or written erroneously, or cancel what was regarded as erroneous and replace
it with what was judged to be correct by writing the correction above the line,
in the margins and in the intercolumnia, at times with specific markings to
indicate the position referred to. Sometimes the forms presumed to be cor-
rect were introduced in replacement of the previous words once these had
been materially eliminated, or at times by simply writing the correct letters or
words between the lines or in the interlinear space above the form judged to
be incorrect, as a way of indicating, as it were, an automatic deletion without
the need for other material indications. On occasion, a horizontal or oblique
line could be drawn through the letters or words to be deleted; another method
was to mark these letters or words by dots or lines above or below or enclosed
within round brackets, or even to erase them with a sponge.41 Thus there was
a veritable tool kit for diorthosis. Often the interventions were carried out by
the diorthotes of the scriptorium, whose task within the atelier was to re-read
and correct the text, if necessary also by comparing the copy with the model,
in other words through a practice of collation.
E. G. Turner and P. J. Parsons write: One of the questions the palaeographer
should ask about any literary manuscript is whether it has been adequately
compared against its antigraph (the exemplar from which it was copied), a
task which, in a publishing house, was the duty of the diorthotes, corrector,
or whether it has been collated with a second exemplar (a procedure often
carried out by private individuals to secure a reliable text)...But several of
our surviving papyrus manuscripts, and especially those which are beautifully
written, contain such serious unnoted errors that it is clear their proof-read-
ing was of a summary, superficial kind, if done at all...Those ancients them-
selves who set store by having a dependable copy (persons like Strabo and
Galen) were aware of this weakness and adopted a routine to counter it: they
themselves (or their secretaries) checked the copy to be used against another
exemplar. If, therefore, the text had been checked against its first exemplar,
and was later collated with a second, it may well bear the marks of this double
checking.42
Best practice in the book production consisted in a comparison between
copies and corrections of mistakes, carried out by a professional or occasional
41 Turner-Parsons [1987] 1516, with reference to examples in plates; see also Turner [19802]
93 and Pl. VIII; Bastianini [2001].
42 Turner-Parsons [1987] 1516; Turner [19802] 93 and Pl. VIII.
652 Montanari
diorthotes, who had adequate resources for deleting, adding, replacing and
marking various aspects and features of the text in order to improve it and
increase its reliability. Even a private copy could be subjected to the same kind
of treatment, with the use of the same tools and procedures, for personal rea-
sons springing from the cultural or research interests of the owner. Analogies
with what we understand by philological practice are evident and need to be
stressed, as the methods and techniques adopted in handicraft book produc-
tion honed the skills that would gradually be developed and applied by gram-
marians. Little by little, a procedure that probably did not appear particularly
strange or extravagant among those for whom use of books was an everyday
practice developed into an extraordinarily innovative principle: the diortho-
sis of the diorthotes of the scriptorium became the diorthosis of the philolo-
gist, diorthotes not of an individual copy of Homer but diorthotes of Homer.
Effectively, concerns and emendments of a specifically book production and
commercial nature became those of a critical and philological-grammatical
nature.43
The aim pursued by the corrector of a publishing house was to produce an
exemplar that would represent the best possible workmanship, a good copy
suitable for sale on the book market or to a client, perhaps intended to be the
personal copy of a scholar or an educated man, who did his own corrections
and annotations (we will note the case of Galen). In contrast, the grammar-
ians underlying objective in correcting the text of his personal copy was more
ambitious, because he sought to find the true and proper form of the work he
was dealing with. He worked on a copy with the aim and intention of achiev-
ing, as it were, the model exemplar, which would display what in his view was
the genuine form of the literary work in question. This conception led to the
possibility of indicating doubts or a textual aporia, a perspective that certainly
did not belong to the mental system and operational horizon of the craftsmen
of the scriptorium.
Thus with Zenodotus, drastic and univocal deletion (a typical action of the
craftsman in book production, meaning dont write these words in the new
copy) for the first time was accompanied by the sign of philological uncer-
tainty, namely the obelos, a simple horizontal stroke on the left of the line. This
marked a decisive intellectual change: attention began to focus on the work in
its own right rather than merely on perfecting the individual copy. It is vital not
to underestimate or downplay the invention of this critical sign, which had a
momentous impact because it could also be applied systematically to poems
of the great length and cultural importance of the Iliad and the Odyssey. By
means of his simple semeion Zenodotus was able to indicate his suspicion that
a given line might not be a genuine line of Homer, but that he was not suf-
ficiently sure of his judgment to be able to proceed with clear-cut and defini-
tive deletion of the element in question. Later, the progress of the discipline
gradually increased, with further development of the system of critical signs
and the markings of exegetic reflection and erudite comment. By the time of
Aristarchus the system of semeia had become complex and refined, but it had
all begun with Zenodotus small obelos and its radically new meaning for a
reader of his texts.44
I believe that the philological work of the Alexandrian grammarians, start-
ing from the first generation, represented something new in cultural history
and marked significant intellectual progress. The reality and significance of
this revolution becomes more evident and tangible if we succeed in grasping
a fundamental chain of circumstances: namely, the aspects and procedures of
book production had moulded a material and, in a sense, craft-oriented base
of tools and working procedures that were subsequently adopted and utilized
by grammarians for quite different purposes and in a different perspective.
Thus the tools and methods of book production became the tools and meth-
ods of scholarship by virtue of an innovative and decisive intellectual change,
which signalled a transition from the aims of pure craftsmanship, i.e. from
correcting an individual copy in the scriptorium and thereby creating a good
product, to an intellectual aim of a philological nature, namely producing the
exemplar that would contain what was held to be the correct text of the work.
Thus no longer would the copy be an exemplar of the work: rather, it would be
the text of the work in itself, and this implied a sharp difference between cor-
recting a (single) copy and editing a text itself.45
Let us recapitulate. Zenodotus worked on an exemplar of Homer that was
available to him and which he deliberately chose for the specific purpose of
producing his ekdosis. However, he had more than a few reservations about
it, concerning both the numerus versuum and also a certain quantity of read-
ings. He had doubts about the authenticity of some lines, and adopted a sign
indicating his suggestion that the line should be expunged, the obelos, which
he marked in besides the lines: this was athetein, the athetesis. But it has always
been more difficult to determine how he proceeded with lines he believed
should most certainly be deleted from the text as definitely spurious and really
to be rejected. Normally, such lines would have been present in his base-text
(as were those for which he proposed the athetesis by means of the obelos).
44 For the evolution of the system of semeia see also Montana in this volume.
45 Montanari [2011b].
654 Montanari
For such cases of deletion, in the scholia one finds the expressions ou gra-
phein (do not write: the most frequent), ouk einai (is not there), ou pheresthai
(is not handed down) and a few others.46 The task of reconstructing exactly
and concretely what was the difference between the operations indicated
by this terminology has always been problematic. A good idea of what was
meant can be gained by examining more specifically the practices used in liter-
ary papyri for different modes of deleting something that is present in a text.
Thus Zenodotus, as we have seen, marked some lines with the obelos on his
copy, but he also used one or other of the graphic methods mentioned above
for lines that were clearly intended to be deleted from the base-text. But the
grammarian could equally well jot down another similar or equivalent term
for line deletion;47 alternatively, he could rely purely on a deletion sign with-
out verbal annotations, in which case the terminology for line deletion may
have been noted down by those who followed his teachings at the Museum, or
it may have been created, modified or extended by the subsequent tradition
(this is conceivable above all for ouk einai and ou pheresthai).48 I would argue
this is the most plausible explanation for the different terminology used in the
scholia for Zenodotus text alterations of athetesis and line omission, including
the problem of the material difference in the copies between athetised and
deleted lines.49
A somewhat skeptical attitude towards this vision has recently been
expressed by A. Rengakos, who puts forward some objections. 1) The expres-
sions in the scholia in reference to the lines deleted by Zenodotus more clearly
describe the situation of lines genuinely not included and therefore absent;
2)it is strange that Zenodotus invented a critical sign for athetesis, namely the
obelos, whereas there is no trace of a sign of the same type for deletion, even
though it has a more radical impact on the text; 3) with regard to passages
where Zenodotus eliminates some lines but information concerning his inter-
ventions on the deleted lines is preserved, Rengakos believes there is nothing
to confute the idea that those lines were effectively absent in his text. Thus the
Zenodotean ekdosis may have been a veritable continuous text, with the obeloi
and the variants and without the deleted lines.50 My opinion is that the vari-
ability of expressions used in the scholia to indicate line deletion suggests that,
at least in part, these were devices developed (perhaps much) later within the
tradition to put into written form what were predominantly (with the possible
exception of ou graphein) the material modes of deletion used in book produc-
tion, as described above. These devices were the signs for line deletion: there
was no need to invent any ex novo, because they were well known and available
in the practice. But the real innovation was the athetesis, which did indeed
call for a special sign. I still find it difficult to imagine that Zenodotus began
his task of diorthosis by writing a continuous and definitive text: to do so, he
would have had to begin once he had already made all the decisions, with poor
opportunity for second thoughts and new corrections, and the definition of
diorthosis / diorthotes does not point in this direction.
However, it is not unlikely that the paratextual apparatus on the working
copy may have given rise to problems of comprehension and readability, espe-
cially with the accumulation of interventions over time, and in places where
the multiple interventions on the text became interlaced with one another.
The copy bearing the work of diorthosis resulted materially in the philologists
own ekdosis, and we can imagine this as a product of years of study that led
over time to a series of interventions on the same exemplar. Together with
critical semeia, explanatory annotations must have been present in the work-
ing copy starting from Zenodotus onwards, and are likely to have continued
to be used by grammarians in their editorial and exegetic work. I therefore
feel it is far more plausible to assume that the ekdosis became ekdotheisa, i.e.
published and therefore available for consultation by scholars, poets and
intellectuals, as soon as the grammarian himself, or someone working on his
behalf, had had a copy made that followed the indications in the base-text
on which diorthosis had been performed, so as to create an exemplar that
was a correct and fair copy of the work,51 but still bearing the name of the
grammarian who was the author of the copied diorthosis, with the marginal
50 Two cases that have given rise to particularly extensive discussion are Il. 2.111118 and Il.
2.156169: the scholia (Aristonicus and Didymus) say that Zenodotus deleted the lines for
a shortened version of the passage, but on the other hand they preserve text interven-
tions by Zenodotus himself on these very lines (in actual fact, the intervention on the first
passage strikes me as debatable, and calls for further clarification): cf. Rengakos [2012]
250252. I see no difficulty in thinking that with regard to the lines in question Zenodotus
copy had one of the well-known deletion signs and that a possible variant for one of the
deleted lines was indicated in the margin. See below and n. 85.
51 Perhaps ou graphein, which is the most frequent expression for line elimination, may go
back to Zenodotus himself and may have been an indication to whoever transcribed his
ekdosis that the element in question was not to be copied.
656 Montanari
annotations which would still be necessary once the text had been properly
prepared. In short, first there was a working copy belonging to the diorthotes,
with all his interventions and annotations, after which it was possible to pro-
ceed to reproducing it as a fair copy of his ekdosis. Thus it was a step-wise
production, which we should obviously imagine to have been done not only
for Zenodotus but also for all his successors. This can also explain the conser-
vation and transmission of the interventions and textual choices made by the
grammarians.52
The material form of the ekdosis after Zenodotus must have remained very
similar: a grammarian chose, according his own preferences, an exemplar
that he considered suitable as a basis of his work. But Zenodotus choice of
the base-text of Homer seemed highly debatable and was open to criticism,
which is why Aristophanes and Aristarchus chose exemplars with noticeably
different characteristics.53 Consequently, a line of tradition predominantly of
Attic origin gradually spread, partly by virtue of the base-text of working cop-
ies used by grammarians active in a later period than Zenodotus. The latter
base-text proved decisive above all as regards the numerus versuum, whereas
the readings suggested by individual grammarians generally did not become
standard in the vulgata. The plus-verses present in the Zenodotean text (as
well as in several pre-aristarchean copies) were not his own interpolations but
were instead typical of exemplars that were current in his day:54 they disap-
peared because the work of Aristarchus led to general recognition of a text
that had a very similar number of lines to our vulgata. The Aristarchean nume-
rus versuum, which became the standard of the vulgata, was essentially the
outcome of the particular working copy selected by Aristophanes and, finally,
by Aristarchus. It is significant that Aristophanes did not go so far as to carry
out the drastic act of line deletion: in other words, it is significant that he
abandoned the use of performing material cancellations on his own copy with
the graphic techniques mentioned above. The obelos became the prime tool
for expressing a cautious doubt concerning parts of the text; ou graphein disap-
peared, leaving only athetein.55 Aristarchus followed exactly the same proce-
dure. This explains why many of the lines Zenodotus had decided to eliminate
52 Helpful confirmation comes from a testimony by Galen, quoted a little further on.
53 Montanari [2002a] 123125; M. L. West [2001a] 36: Clearly Aristophanes and Aristarchus
were not dependent on Zenodotus text but followed another source or sources more
similar to the vulgate; cf. M. L. West [2002] 138.
54 Haslam [1997]; M. L. West [2001a] 40; cf. above.
55 Or else, if genuine deletions were still carried out, they were of such minor relevance that
all knowledge of them was lost: this is possible, but not demonstrated.
Ekdosis. A Product of the Ancient Scholarship 657
from the Homeric text once and for all but which were present in the copies
chosen by later grammarians were preserved in the numerus versuum that
became the generally accepted tradition after the Aristarchean age and thus
remained in our vulgata.56 The abandonment of the drastic practice of mate-
rial deletion highlights the increasing sense of caution that had developed in
the meantime, and accounts for the fact that many of the lines deleted by
Zenodotus were in effect no longer deleted57 and thus were not obliterated
from the tradition.
The work of Aristarchus marked the period in which Alexandrian philo-
logical production included the drafting of extensive hypomnemata. The great
continuous commentary, which followed the text step by step, notably facili-
tated and enriched the communication and preservation of the arguments and
motivations put forward by the grammarians, so that the material which has
come down to us from this tradition is much more substantial.58 Yet the ekdosis
as an annotated working copy by no means went out of use, as clearly testified
by the information on the Aristarchean edition(s). On the other hand, the pos-
sibility of dwelling at length in the hypomnema on arguments pertaining to
text criticism and exegesis constituted an important resource. In practice, the
need to write on the copy chosen as the base-text was no longer so strongly
felt, especially as regards philological-exegetic arguments. Previously, before
the rise of separate hypomnemata, there had been a greater need to write on
the actual text of the working exemplar, but with Aristarchus the particularly
elaborate system of critical signs placed next to the lines59 as well as the vari-
ants and the readings to be adopted must have been present in the margins
and interlinear spaces, while the philological-exegetic treatment was mostly
developed in the commentary, although marginal annotations continued to be
utilized whenever they were felt to be of practical use, e.g. for short notes and
textual proposals.
The number of Aristarchus ekdoseis of Homer and their philological and
chronological relation to the commentary or commentaries is still a subject of
dispute. I will not go over the entire background here, nor report the treatment
already given elsewhere: I will restrict myself to summarising the results, in
order to set them within the framework that is being delineated. On the one
hand we have the frequent unequivocal references to the plural for Aristarchus
editorial work on Homer: (scil. or ),
and similar. On the other, the titles of two works by the grammarian
Ammonius, the direct successor of Aristarchus:
and , the
former in apparent contradiction with the latter and with the scholiastic cita-
tions that indicate two editions. Nevertheless, the solutions proposed go as far
as to hypothesize a carried out by Aristarchus immedi-
ate pupils, probably by his successor Ammonius, who in any case was familiar
with it.60
As regards the hypomnemata, I think it is difficult to deny that Aristarchus
made two successive versions: a first commentary based on the ekdosis by
Aristophanes of Byzantium is explicitly cited in sch. Il. 2.133 a:
. In contrast to this stands the citation of
perfected () hypomnemata in sch. Il. 2.111 b: it is perfectly plausible
to assume that Aristarchus produced a second version of the hypomnemata in
which he took into account the progress achieved over time by his work as a
Homeric philologist.61
We have two Homeric passages on which Aristarchus is known to have
changed his mind in comparison to his first text choice, as reported in the
scholia to Il. 10.397399 and 19.365368. Such a situation has many parallels in
the scholiographic documentation, describing Aristarchean second thoughts
and changes of heart.62 However, in the case of these two Homeric passages,
it can confidently be stated (despite uncertainties about details) that later
philologists were searching for information on the reasons and circumstances
for his change of mind and on the text situation that had ensued, and since
they were far from certain, they consulted the cited works by Ammonius. In
the attempt to explain this situation and reconcile the apparent contradiction
between the two titles of the successor in the school, it has been suggested
that Aristarchus himself personally composed only one ekdosis, as sug-
gested by the first title, and that the mentioned in the
second one was actually composed later, after the masters death, in the circle
of his first and most senior pupils, possibly even by Ammonius or else by other
followers. However, I believe that a slightly different hypothesis can lead to a
better understanding of what really happened.
63 See Montanari [1998d], [2000b] and [2002a]: this solution of the problem ekdosis / ekdo-
seis of Aristarchus is approved by Slings [1999]; Nagy [2003] and [2009] 2137; Rengakos
[2012] 252; I am unsure whether it was perceived by M. L. West [2001a] 6167.
660 Montanari
64 Pfeiffer [1968] 217: Whether Didymus was able to work on copies of these originals
and of Aristarchus and of his monographs, the , is an
insoluble problem; cf. M. L. West [2001a] 6167.
65 Pfeiffer [1968] 110.
Ekdosis. A Product of the Ancient Scholarship 661
copies, has had a number of supporters, starting above all from the positions
of M. van der Valk,67 whose line of interpretation was also adopted (of course
with individually differentiated stances) by H. van Thiel68 and most recently
by M. L. West, to whom we will return later.69 This tendency leads recta via to
a (quite unfair) underestimation of the importance and the value of the work
performed by the Alexandrians. Arguments against it have been put forward
by M. Haslam, Martin Schmidt, G. Nagy, J.-F. Nardelli, A. Rengakos and myself.70
On the question of the Konjekturalkritik often and abundantly attributed
to the Alexandrians, Rengakos observes that it is a theory based on the false
presupposition that we have general criteria for distinguishing between con-
jectures and genuine variants when we are faced with the overall set of read-
ings contained in the erudite sources, whereas such criteria do not exist at all.
Furthermore, in the sources there is no explicit testimony referring to conjec-
tural interventions, and it is impossible to demonstrate that a given reading
is the fruit of a conjecture by the philologist to whom the textual choice is
attributed. Rengakos has very clearly recapitulated that, on the contrary, there
is actually a considerable amount of plausible evidence of the Alexandrian
philologists knowledge of variants deriving from a comparison among copies.71
In addition to the arguments already illustrated above, based on the papyri
and on the general practice of book production, Rengakos has dwelt on this
problem, presenting very precise and cogent arguments concerning the testi-
mony offered by the poets of early Hellenism, i.e. of the Zenodotean age, who
67 Sharp criticism of van der Valks ideas (van der Valk [1949] and [19631964]) has been
made in a number of papers: for ex. Rengakos [1993] 3848; Rengakos [2002a] 146148;
Haslam [1997] 70 n. 31: ...he does not concern himself with the transmission. In catego-
rizing readings he operates with an opposition between original, old readings and only
subjective conjectures...a schematization that is surely too simple to cope successfully
with the complex vicissitudes of the Homeric text.
68 H. van Thiel [1992] and [1997] (see also [1991], Einleit., and [1996], [20102], Einleit.) has
argued that the readings which the tradition attributes to the Alexandrian grammarians
were actually exegetic glosses or mere indirect references or reminiscences of parallel
passages, written in a Rand- und Interlinearapparat, which Didymus, Aristonicus and
others then wrongly interpreted as textual variants; I discuss this rather idiosyncratic
vision in Montanari [1998d] 46; van Thiel [1992] is discussed by Martin Schmidt [1997],
with a reply in van Thiel [1997]: see above and n. 9.
69 M. L. West [2001a], [2001b] and [2002]: discussion in Montanari [2002a], [2004], [2009b],
[2011b] and [forthcoming]; see further on.
70 Haslam [1997]; Martin Schmidt [1997]; Fhrer Schmidt [2001] 67; Nardelli [2001] (par-
tic. 5270, in direct opposition to Wests theories); Nagy [2000b], [2003], [2004], [2009];
Rengakos [2002a], [2002b], [2012].
71 Rengakos [2012].
Ekdosis. A Product of the Ancient Scholarship 663
72 Rengakos [2002a] 149; cf. Rengakos [1993], [2001], [2002b], [2012]; an interesting case per-
taining to Zenodotus is highlighted by Fantuzzi [2005].
73 Lately Rengakos [2012].
74 Pfeiffer [1968] 110114: the citation is on p. 114; the three examples adduced by Pfeiffer
concern Il. 1.5, Il. 1.225233 and Il. 16.432458, Il. 4.88. Pfeiffer normally attributed the col-
lation of copies to the great philologists who succeeded Zenodotus: cf. for example p. 173.
Pfeiffers arguments should have been awarded greater consideration.
75 Cf. Rengakos [2012] 244248, with bibliography. The above mentioned evidence of
Didymus in sch. Il. 9.222 b is rightly underlined by various scholars (Nagy, Janko, Rengakos
and myself) and cannot be dismissed out of hand, as does M. L. West ([2001a] 37 n. 19):
I will examine closely meaning and importance of this sch. in a paper forthcoming in
Lemmata. Beitrge zum Gedenken an Christos Theodoridis. On Aristarchus second
thoughts see Montanari [1998d] and [2000b].
664 Montanari
All this evidence indicates that when engaging in text criticism, the
Alexandrians starting with Zenodotus and reaching the most refined
method with Aristarchus based themselves not only on text-internal con-
jectural proposals but also on external and diplomatic resources, consisting
in choice among variants they found or noticed in a non-univocal tradition
composed of the copies they had available and were thus able to consult. It
would seem, therefore, that the burden of proof is on whoever seeks to strip
the Alexandrian grammarians of any knowledge of variants deriving from col-
lation of copies, attributing to them only arbitrary conjectures, rather than
the opposite: the fact is that we have, at the very least, convergent evidence in
favor of knowledge of variants and I would go so far as to say that we have
real proof.
An interesting testimony concerning these problems can be found in the
recently discovered De indolentia by Galen, an author of major importance in
the history of ancient philology, not only on account of his activity and his
thought, but also by virtue of the information Galens text provides. It has
begun to be studied and appreciated from this point of view, but certainly
much fruitful investigation remains to be done.76 The new text is preserved
in a copy which, overall, has many incorrect forms and results in consider-
able uncertainty of interpretation, also affecting the points of interest here,
but it is worth commenting on the material and singling out several pieces of
information.77 In the 192 fire of Rome, Galen lost, among other things, all the
books he possessed, and in this work (composed in epistolary form as an answer
to a Pergamene friend) he talks extensively about his activity as a scholar and
about his books. Those that had been lost included texts corrected in my own
hand ( 6); there were also rare books that were not available elsewhere, and
books which, while not rare, constituted unrepeatable exemplars due to the
particularly accurate and carefully written text, such as the Plato by Panaetius
and two Homers by Aristarchus, and others of this kind ( 13). There follows
a rather tortuous passage, which may perhaps contain a reference to copies
with marginal annotations and bearing the name of the person who had made
the marginal jottings.78 A little further on Galen relates that he had also lost
books he himself had worked on, in which he had corrected various errors in
order to compose an ekdosis of his own. The task he had set himself, he says,
76 Manetti Roselli [1994]; Manetti [2006], [2012a] and [2012b]; Roselli [2010] and [2012a];
see Manetti in this volume.
77 Editions: Boudon-Millot Jouanna [2010], Kotzia Sotiroudis [2010], Garofalo Lami
[2012].
78 Cf. Roselli [2010] and [2012a], Stramaglia [2011], Manetti [2012b].
Ekdosis. A Product of the Ancient Scholarship 665
79 Cf. Boudon-Millot Jouanna [2010] xxxiiixxxiv, Manetti [2012a], Roselli [2012a]; on lec-
tional signs, punctuation and accentuation cf. Pfeiffer [1968] 178179.
80 Cf. Manetti Roselli [1994] 16251633; Manetti [2012a]; Roselli [2012a]; see Manetti in this
volume.
81 Roselli [2012a] 6467.
666 Montanari
Over the past few years, the arguments put forward mainly by myself and
by A. Rengakos82 have prompted a debate above all with the positions of M.
West, the most radical advocate of the theory that the Alexandrian philolo-
gists from Zenodotus to Aristarchus known as authors of ekdoseis of Homer
did not carry out any collation of copies: In fact, the first scholars known to
have cited manuscript authority for variant readings are Aristarchus contem-
poraries Callistratus and Crates. Didymus is the first author known to have
compiled anything in the nature of a critical apparatus. It is entirely unjusti-
fied to project his methods back onto Aristarchus or Zenodotus, or to assume
that all the various copies available to Didymus in the time of Augustus were
already part of the librarys holdings in the early third century.83 I feel I must
express some misgivings upon reading that it is entirely unjustified to proj-
ect his methods back onto Aristarchus immediately after the statement that
the first scholars known to have cited manuscript authority for variant read-
ings are Aristarchus contemporaries. It is simply begging the question to claim
that Didymus method is projected back onto Aristarchus or Zenodotus, for in
actual fact there is absolutely no evidence that Didymus was the first to apply
this method rather than having inherited it from his predecessors, unless it
be the preconception that Zenodotus, Aristophanes and Aristarchus did not
apply it at all.84
In M. Wests view, Zenodotus readings are to be regarded either as conjec-
tural emendations or as peculiarities of his base-text, but do not result from any
form of comparison among copies. Zenodotus is claimed to have lacked the
concept of variants, his only critical concern being the existence and identifi-
cation of spurious lines. It is Wests contention that The one kind of textual crit-
icism we know Zenodotus practised (my italics) was not concerned with choices
between variant readings but with the identification of spurious lines and pas-
sages. The one feature of his text that marked it out as a critical diorthosis was
the presence of obeloi in the margins (and perhaps brackets enclosing longer
passages) to signal the critics suspicion that certain verses were un-Homeric.85
This, however, obviously means that Zenodotus felt he had to tackle the prob-
lem of how to discriminate the authentic from the spurious: for whereas an
entire line held to be spurious could be eliminated from the text, a part of the
line (a word or an expression) cannot be removed without replacing it with
something else. Sometimes, by eliminating or accepting a line, the meaning
and syntax called for the alteration of some word or words before or after the
line itself; at times, the alteration of a word called for or allowed the elimina-
tion or addition of a line. But Zenodotus one kind of textual criticism and
the critics suspicion that certain verses were un-Homeric do suffice, for once
a critical approach towards the way in which the text presents itself has been
acquired, it is inconceivable for there to be a theoretical and essential separa-
tion which would discriminate between line athetesis and single word altera-
tion and would thereby justify the assumption that the philologists concern
focused only on athetesis of whole lines and not on shorter text alterations.86
In either case, the problem at hand for the philologist resides in the opposition
between authentic/correct vs. spurious/damaged and in seeking to identify the
proper text. By addressing the issue of the authentic text and how to devise the
critical-methodological tools to obtain it, Zenodotus achieved a major break-
through: it was a crucial intellectual step, which we identified above as residing
in the difference between correcting a single copy and editing a text.
M. West warns against a travesty of the situation: The misapprehension,
which goes back at least at the time of Wolf, is that Zenodotus, Aristophanes
and Aristarchus were all editors in the modern sense, who wanted to establish
a good text of Homer and who approached the task as a modern editor does,
by collecting manuscripts and comparing their readings.87 Now, if Zenodotus
had at least one kind of textual criticism, what is likely to have been his aim
in carrying out emendations on the Homeric text? Are we thus to believe
that Zenodotus had a conscious premeditated idea of modelling Homer
according to his own taste, i.e. Im going to set about reworking Homer and Im
going to make it the way I think it ought to be? This possibility is by no means
easy to accept, but actually this is the only alternative to the view that he
wanted to establish a good text, which is the natural goal of anyone who starts
working on and correcting a text. In effect the aim of the Alexandrian philolo-
gists cannot but have been to establish a good text, whatever the value of the
result according to modern scholars.
I fear that the misunderstandings arise from the fact that there is no clear
definition of the guidelines for our judgment on the work of the Alexandrian
philologists. By adopting our own point of view concerning the competence
on which they base their opinions and arguments, so that it can be ascertained
whether and when they are right or wrong in comparison with the truth
according to scientific philology, we risk producing unfounded and pointless
judgments. Naturally, evaluation of the quality of their choices is the proper
perspective for the interpreter and editor of Homer as a modern philologist;
on the other hand, maintaining conscious awareness of historical distance and
taking care not to overlay our criteria on their behavior is the proper perspec-
tive for the historian of philology as a cultural and intellectual phenomenon
and for the reading of Homer in ancient civilization. Perhaps it is hard to
conceive that Zenodotus aim (however incoherent and unsophisticated) was
precisely to establish a good text of Homer because the testimony that has
come down to us indicates that his text was far from good in fact it was dread-
ful, and incoherent seen through the filter of the requirements and knowledge
of modern classical philology. And even as regards the successors of Zenodotus,
or even the great Aristarchus, we can hardly claim always to agree with their
text choices. The viewpoint from which a modern Homeric scholar approaches
his task is the need to decide whether the text Zenodotus, Aristophanes or
Aristarchus judged to be the best is indeed the one to print in a present-day
critical edition, and whether their interpretations should be espoused as valid
in a scientific commentary. In contrast, the viewpoint from which a historian
of ancient philology starts out is that of seeking to understand their methods,
arguments, principles, knowledge in a word, their historical and intellectual
position. The tendency to scoff at the opinions of the Alexandrian philologists
in terms of modern Homeric studies should by no means translate into the
tendency to discredit their historical significance, which needs to be correctly
positioned and contextualized. It is mistaken to blur the distinction between
the two planes.88 It is impossible to escape the fact that by inventing the obelos
88 Janko [2002] seems to render this concept explicit rather more clearly. His position on the
methods of the Alexandrian philologists is not an extremist unilateral stance: he believes
that the majority of their readings are indeed arbitrary conjectures (by the Alexandrians
themselves or possibly of more ancient origin), but he does not go so far as to deny the
recourse to manuscripts and comparison among copies as part of their ekdosis work (for
Zenodotus, Janko [1992] 23: His caution was salutary, given the abundance of interpo-
lated texts; he certainly had MS authority for some omissions; for Aristarchus, Janko
[1992] 27, and Janko [2002]: This [i.e. sch. Il. 9.222, cf. above n. 75 and 84] certainly implies
that Aristarchus did check manuscripts for variant readings). On the one hand, Janko
argues, there stands the problem of the origin of their proposed text choices (subjec-
Ekdosis. A Product of the Ancient Scholarship 669
and setting himself the task of emending and restoring the text he had at hand,
Zenodotus had lit upon an idea which, however embryonic and crude it may
appear, would undergo further development among his successors, eventually
becoming the germ of the discipline we call classical philology. But even if one
were to suppose that he acted purely on the basis of conjectures, could it be
denied that conjecture is one of the emblematic and representative tools of
philology aimed at restoring the correct text?
A further comment by M. West is surprising. Consider what we know of
Aristarchus methods, for which we have plenty of material in the scholia. Of
course he had the text of his teacher Aristophanes before him. He also kept
an eye on [my italics] that of Zenodotus, and took up critical positions against
it. But the arguments he used, as reported by Aristonicus and Didymus, were
always based on the internal evidence of contextual coherence or general
Homeric usage. Not once does he appeal to the authority of manuscripts.89
SoAristarchus compared his working copy with that of Aristophanes and that
of Zenodotus; but the phrase kept an eye on is insidiously reductive, given
that the number of cases preserved by the tradition runs into the hundreds
and the tradition itself is patchy and incomplete. Be that as it may, the picture
delineated above implies he made a certain small comparison among copies,
but that he took great care not to let his eye stray onto any further copies: in
other words he did study and interpret Homer, but he painstakingly avoided
consulting any other exemplar than his own, together with that of Zenodotus
and that of Aristophanes, although these alone already presented him with not
tive emendation, comparison among copies) and therefore of their working procedures;
on the other, he points out, my own concern, as a Homerist, has always been whether
such readings are authentic. Perfectly clear: modern philologists can to some extent be
severe regarding the opinions of the Alexandrians, considering them to be fairly accept-
able or fairly unacceptable from their own point of view (Janko is very negative), but
they cannot downplay the fact that the ancient Alexandrians emended and compared
exemplars to correct the Homeric text, a method that combined interpretation of the
text with awareness of the history of the tradition. An extremely apt remark, perfectly
applicable to Alexandrian philology as well, is offered by Cassio [2002] 132, on the issue of
pre-Alexandrian criticism: The earliest scholarly approach to the Homeric text is totally
foreign to us...we do right to think along very different lines, but we should never forget
that it was the commonest approach to the Homeric text in the times of Socrates, Plato
and Aristotle. As a consequence, we ought to be wary of looking at it with a superior
smile, and ought to try to understand its motives in more depth instead.
89 M. L. West [2001a] 37: at least the mentioned case of sch. Il. 9.222 clearly contradicts this,
cf. above and nn. 75, 84, 88.
670 Montanari
90 M. L. West [2001a] 38: No doubt it would have been easy for him [scil. Zenodotus] to col-
lect several copies if he had taken the trouble. Are we to assume that Zenodotus was a
somewhat lazy philologist?
Ekdosis. A Product of the Ancient Scholarship 671
could be restored to its correct form either via conjecture or by choosing the
correct reading from among those offered by a non univocal tradition.91 The
recognition of transmission-induced damage that had affected the authentic
text, along with steps and procedures to restore it, is proof of how the mutual
dependency of textual criticism and textual interpretation became established
and operational.92
3 Conclusions
91 Naturally the works of Homer come to mind, but also of the tragic and comic poets.
92 See Pasquali [1920] (citation from the reprint of [1998] 26): a costituire un testo...occorre
la stessa preparazione che a interpretare...; costituire un testo e interpretarlo sono, in
fondo, tuttuno (constituting a text...requires the same learning and knowledge as
interpreting...; constituting a text and interpreting it are, ultimately, one and the same
thing).
93 See now Conte [2013] 4450.
672 Montanari
Homers grip on Greek literate culture gave him a dominant role in education,
scholarship and criticism of all kinds, and this predominance is reflected in the
centrality of the Homeric texts to the growth of critical practice and termino
logy, particularly as we can trace these from Aristotle onwards. This chapter
will be largely concerned with critical practices and ideas which flourished,
and in some cases arose, in the Hellenistic period, but it is important always
to bear in mind the classical, and in some cases, archaic roots of these pheno
mena. The modern study of ancient literary criticism has always suffered from
uncertainty as to what actually is being studied and where and how early the
relevant material is first found. Does one start, for example, with Odysseus
praise of the Phaeacian bard Demodocus (Od. 8.487491) and Alcinous praise
for the manner in which Odysseus himself tells his tale (Od. 11.363368),1 with
Pindars rich metatextual commentary on his own and others poetry,2 with
tragedy and satyr-play, where some of the richest reactions to Homer and
the Homeric ethos are to be found, even though the explicit dramatisation
of scenes from Homer is very uncommon in our surviving texts (Euripides
Cyclops, [Euripides] Rhesos),3 with Attic Old Comedyand, most notably, the
Frogs of Aristophanes, or with Plato?4 The concerns and critical practices of
the Frogs were certainly influential for centuries, texts such as the famous dis
cussion of an ode of Simonides in Platos Protagoras and of the expertise of the
1 The bibliography on Homeric poetics is of course daunting; Halliwell [2011] Chapter 2 offers
a thought-provoking guide through the maze.
2 For foreshadowings in Pindar of later critical ideas cf., e.g., Richardson [1985].
3 For the Cyclops as a reading of Homer cf. Hunter [2009a] Chapter 2.
4 Ford [2002] is an excellent guide to these issues and their bibliography. For origins and
beginnings of ancient scholarship see also Novokhatko in this volume.
5 On this aspect of the Ion cf. Hunter [2011] and [2012] 89108.
6 Cf. Usener [1990].
7 See Nnlist in this volume.
The Rhetorical Criticism of Homer 675
(late sixth century), who was the first to write about Homer.14 It seems likely
enough that, by the end of the fifth century at least, complex moral and physi
cal or cosmogonical interpretations of Homer circulated widely, though it is
not always easy to identify their authors or to form a clear sense of the outlines
of these interpretations.
The majority of our evidence for moralising and allegorical interpretations
comes, of course, from the post-classical period, and only very rarely are we
able to pick apart the various layers of interpretation which often survive in
summary form in later texts, notably in the scholia. Nevertheless, it is also
likely that some relatively simple kinds of interpretation persisted over cen
turies. The explanation of Athena as , for example, recurs persistently
throughout antiquity, very often in connection with the hero par
excellence, Odysseus.15 A particularly interesting manifestation of this con
cerns Odysseus indirectly, namely Homers representation of the development
of the heros son Telemachus in the early books of the Odyssey; this example
also neatly illustrates how ancient arguments and interpretations very often
foreshadow modern readings.
In one of the extant hypotheseis to Odyssey 1 (hypothesis c Pontani) we
read that Athena, in the guise of Mentes, advises Telemachus to visit Pylos
and Sparta, and then, This business of Athena going to Ithaca to encourage
Telemachus to make enquiries about his father hints at () nothing but
the fact that phronsis is called Athena, and Telemachus, who is a child ()
but then grows up and comes into wisdom (), is roused by Athena, that
is by his own phronsis, to make enquiries about his father.16 Heraclitus offers
a much extended version of this interpretation:
Mentes is said to be an aged friend of Odysseus. Grey hair and age are the
sacred harbours of our final years, a safe anchorage for men, and as the
strength of the body wanes, so the force of the intellect () increases.
Heraclitus, Homeric Problems 61
you should not continue in childish ways, for you are no longer a child
Homer, Odyssey 1.296297
17 De Jong [2001] 20; De Jong there gives helpful bibliography on the character of Telemachus.
678 Hunter
with the truth (vv. 412420).18 When we first saw Telemachus he was deep in
depressed thoughts as he wondered whether his father would ever return to
scatter the suitors (vv. 114116); the book closes with Telemachus again deep
in thought (427), but now it is about what he must do, and he spends the night
thinking over in his heart the journey which Athena had marked out (444).
The various different interpretative traditions which we glimpse through
the surviving scholia show how thin is the line between some types of allegori
cal reading19 and the non-allegorical: was Telemachus merely instructed by
Athena (cf., e.g., the scholia on 354) so that he becomes , or is Athena
actually herself ? is of course one of the standard scholiastic
glosses for , the formulaic epithet for Telemachus, which makes its
first appearance (v. 213) in the poem immediately after Athenas first address
to the young man, where it comes almost as confirmation of Athenas conclud
ing assertion of how like Odysseus Telemachus is. Another sign of how such
interpretative modes run together is the concern in the scholia, which seem
in this case to go back to Porphyry, with why Athena sends Telemachus away
from Ithaca at what looks like a moment fraught with danger and on a mission
which is essentially fruitless; this has of course also much exercised
modern scholars.20 Porphyrys long discussion21 adduces the fact that, brought
up by a woman on Ithaca and surrounded by hostile men, Telemachus could
never have learned the appropriate skills or had the appropriate experiences
to become like his father; he would therefore have either remained in this
impossible situation or launched an inevitably doomed attack upon the suit
ors by himself. Enquiries about his father are therefore the of the trip,
but the real is education, , which involves learning about his
father, and it is from this that Telemachus will acquire the which the
18 The exegetical scholia on v. 413 rightly note Telemachus strategy of keeping the suitors
relaxed.
19 I am aware that for the purposes of this discussion I have simply ignored some aspects of
some allegorical readings of Athena in Odyssey 1 which are less easily taken over. Thus,
for example, the allegorising scholia on v. 96 explain that Athenas lovely sandals denote
the powerful effects in action of phronsis and her spear its striking power ( ),
for through his own reason the phronimos strikes the unruly. There is, of course, a range
of allegorical, as of non-allegorical, views and gradations of detail within any one such
reading.
20 Cf., e.g., West [1988] 5355.
21 Scholia to 1.94 and 1.284. Some of the earlier history of Porphyrys arguments may be vis
ible at Philodemus, On the Good King according to Homer col. xxxiii Dorandi if, as (e.g.)
Asmis [1991] 38 suggests,
refers not to Odysseus, but to Telemachus.
The Rhetorical Criticism of Homer 679
Homeric Athena holds out for him (v. 95, cf. 13.422). Such an aim, as Porphyry
points out, is appropriate to Athena, in part (we should infer) because Athena
is associated with intelligence, education and . Part at least of this reading
of the Telemachy, which has of course considerable overlap with the standard
modern reading, will go a very long way back in antiquity;22 it seems first to
surface for us in Philodemus, On the Good King according to Homer, where, in
a recently restored, if still broken, column, the philosopher precisely discusses
Telemachus journey:23
...to be one who has constantly lived among guests not living according
to his will, since in addition it is necessary for him to be one who has
neither seen nor heard of many things and has had no experience of free
speech with equals ( ), and for the most part
even uneducated, for which reason...the poet...brings Telemachus to
Pylos and Sparta where he was to have dealings with such great men, for
he was certainly not going to achieve anything (more) concerning his
father, who was by then already on Ithaca...
Philodemus, On the Good King according to Homer col. 23, trans. J. Fish
22 Herter, RE V/A 351 suggested that it might have figured in Antisthenes Athena or
Concerning Telemachus.
23 Cf. Fish [2002] 193194, with discussion in Fish [1999], [2002] 213215, and [2004] 113114.
24 Cf. Hunter [2012] 6067.
25 Plutarch (Mor. 19e) notes that was the old term for what are now called
.
680 Hunter
26 For the bibliography and interpretation of the scholium cf. Luzzatto [1996], Pontanis
notes in Pontani [2007] 79, and Montiglio [2011] 2047.
27 Cf. Erbses note on schol. 9.682683, citing Porphyry.
28 Cf. Nnlist [2009a] 250 with n. 42.
29 Cf. Montiglio [2011].
30 Odysseus and his crew were, therefore, not Circes first Greek visitorsanother blow to
Odysseus list of achievements. A different view, adopted e.g. by Kidd in Waterfield-Kidd
[1992] 375 and Indelli [1995], is that Gryllus was in fact one of Odysseus companions;
this would necessitate a setting in the tenth, rather than the twelfth, book. Much might
seem to hang on the clearly disturbed text at 985e where the word or is
transmitted (Hubert adopts von Wilamowitzs deletion of the word), but this alternative
view seems to make nonsense of the opening exchanges with Circe and certainly destroys
any close link with the Odyssean narrative; so too Circes comment at 986a that if the ani
mals win the argument, Odysseus will be shown to have determined badly concerning
himself and his friends seems to support the view taken here. At 989e Gryllus claims
682 Hunter
that animals are actually more virtuous and live happier lives than men. The
dialogue is perhaps to be imagined as taking place immediately after Circe has
told Odysseus about his voyage home and warned him of the consequences
of interfering with the cattle of the Sun (12.137141); I think I understand and
will remember these things, Circe, but I would gladly hear from you..., begins
Plutarchs Odysseus, in one of what might have been a not uncommon ancient
game of writing new and often unusual scenes for the Odyssey.31 Odysseus
request is to know whether there are any Greeks among the metamorphosed
animals under Circes control so that, with Circes permission, he can restore
them to human shape from the pitiable and dishonourable existence they
now lead and take them back to Greece; this, says, Odysseus would bring him
with the Greeks. Circes response is very sharp:
,
.
This man thinks that his desire for glory () should, through his
stupidity, prove a disaster not just for himself and his companions, but for
complete strangers.
Plutarch, Gryllus 985d
Circe here produces a re-writing of the opening of the Odyssey, which is very
much not to Odysseus credit. This man, the famous , will be the undo
ing not just of himself and his (contrast Od. 1.5), but of many other
Greeks as well; his companions will not perish by their own reckless foolish
ness (Od. 1.6),32 as in Homer, but through Odysseus stupidity and desire for
glory, a force which drives him to pursue an empty form of goodness and a
phantom in place of the truth (986a). Such a reading takes its initial impetus
to have once seen Odysseus on Crete; the expression certainly does not suggest that he
was actually one of Odysseus companions (so, rightly, e.g. Russell [1993] 337), and in any
case Gryllus here turns Odysseus Cretan tale (!) to Penelope (Od. 19.221ff) back against
the hero himself. There is a useful account of the philosophical background of Plutarchs
essay by Ziegler in RE XXI (1951) 739743.
31 One thinks of the letter which Ovids Penelope writes to Odysseus (Her. 1).
32 Among the scholiastic glosses for are (D schol. on Iliad 4.409) and
(D schol. on Od. 1.7); Circes is a variant of this.
The Rhetorical Criticism of Homer 683
from Odysseuss assertion of his own at Odyssey 9.1920, but behind it lies
(again) a long tradition of interpretation and re-writing.33
Like , may have positive or negative connotations, and
(again) these differences may then bring differences of narrative with them.
, as the principal motivating force of Odysseus most of
men (986b), seems, for example, to have played a significant role in Euripides
Philoctetes, as this can be reconstructed from Dio 52 and 59.34 Odysseus
seems to have begun the prologue of that play by expressing his worries that
his reputation for cleverness may be undeserved, given the trouble he vol
untarily gives himself; that fear in fact comes true in the Gryllus when, in his
opening remarks, Gryllus observes that Odysseus reputation for cleverness
and surpassing wisdom will all have been for nothing (), if he will not
accept improvement, just because he has not given the matter proper thought
(986cd). The Euripidean Odysseus then proceeded to explain that good lead
ers such as himself are driven by and the desire for and to
undertake very difficult tasks; when Odysseus stated that nothing is as keen
for acclamation () as a man (fr. 788.1 Kannicht), it is hard not to remem
ber Gryllus criticisms of human folly. So too, the thesis that the life of a pig, a
life of all good things (986d), including deep, soft mud (989e), is much to be
preferred to the life of a man, particularly an Odysseus, overturns the very rich
mainstream tradition of interpretation of the Circe-scene and of the Odyssey
as a whole. This is particularly true of the insistent argument that pigs and ani
mals generally are not victim to the lustful desires of the flesh, unlike human
beings, whereas the normal interpretative view (cf., e.g., Hor. Epist. 1.2.2426)
is that men are precisely turned into pigs by their slothful lusts and the pursuit
of pleasure. The Gryllus is of course full of witty reworkings of the Odyssey and
its critical tradition:35 throughout antiquity, the critical interpretation and cre
ative mimesis of the Homeric text travelled hand-in-hand.
33 For Odysseus in the scholastic tradition cf. the scholia on Odyssey 5.401 (a
remarkable text) and 9.229.
34 Cf. frr. 7879 Kannicht, Stanford (1954) 115116.
35 No full account is possible here. At 986f Gryllus tells Odysseus that he once heard him
describing the land of the Cyclopes to Circe in the terms which Odysseus in fact had used
in his narration to the Phaeacians (9.108111); we are presumably to understand that the
conversation took place during the years stay with Circe (10.467468), but the apparent
misrepresentation of the Odyssey serves at least two purposes. Plutarch is shown to be
as wittily concerned as Homer with the How do you know? question which can always
be posed to a narrator (cf. Od. 12.389390), and there is a suggestion that the braggard
Odysseus used to bore Circe (presumably in bed) with the same tales which he told the
684 Hunter
The tradition of critical problems and their solutions which was illus
trated above from Platos Hippias Minor is one of the longest-lived ancient criti
cal practices:36 we can see it already in full (if satirical) swing in Aristophanes
Frogs,37 significant excerpts from Aristotles Homeric Problems are preserved
(and cf. also Poet. chap. 25), and the tradition is very much alive and well in
Alexandrian scholarship and, as we have seen, in the Homeric scholia.38 This
critical framework allowed scholars to appeal, inter alia, to change over time
in cultural practice, to both diachronic change and synchronic difference, for
example between dialects or in language usage more generally, and to the need
to pay close attention to shifting contexts, particularly rhetorical contexts, in
the course of a long poem. Aristotle is the crucial figure in establishing that
poetry had its own standards of correctness (), and that what matters
is not the existence of factual errors or inconsistencies per se, but rather the
quality and nature of those phenomena;39 such a realisation focused attention
again on the need for scholarly judgement as shown in decision-making, krisis
in both its senses. The challenge to Socrates which Plato puts in Protagoras
mouth was to foreshadow the principal thrust of Alexandrian criticism:
Protagoras test to see whether Socrates fits the bill then precisely concerns an
alleged inconsistency in a poem of Simonides. According to Aristotle in the
Phaeacians (and subsequently Penelope) and which she of course knew already anyway
(cf. 10.457459). For recent views of the literary form of the Gryllus cf. Fernndez Delgado
[2000] and Herchenroeder [2008].
36 See Novokhatko, Dubischar, and Nnlist in this volume.
37 Cf. Hunter [2009a] 2125.
38 Cf. Pfeiffer [1968] 6971, Nnlist [2009a] 1112, Slater [1982]. Some of Slaters conclu
sions need modification in view of Blank-Dyck [1984], but the continued importance of
problem-solving criticism is not in doubt.
39 Cf., e.g., Hunter [2012] 100103.
40 The pointed ambivalence of this phrase is marked by Socrates later observation of the
multivalency of (341a7b5). I have discussed some features of this passage of the
Protagoras in Hunter [2011] 36, and for a fuller account and bibliography of the discussion
of Simonides poem cf. Hunter [forthcoming].
The Rhetorical Criticism of Homer 685
Poetics, there are five standard grounds for identifying a problem requiring a
solution: that something is impossible, or irrational, or harmful, or contradic
tory, or contrary to artistic correctness (1461b2324). These categories recur
constantly in the critical traditions which came after Aristotle, but their roots
are deep and early.
The history and characteristics of textual and interpretative scholarship,
given perhaps their most authoritative expression in Rudolf Pfeiffers History
of Classical Scholarship of 1968, are relatively well understood (as are the many
areas of doubt and uncertainty) and offer a reasonably clearly defined area
of study.41 Literary criticism, under any definition, plays a major role in such
scholarship, and it is the scholia to Homer which offer probably our richest
sources for this.42 Any attempt, however, to separate readings and interpreta
tions of Homer from the history of reworkings of Homer within Greek litera
ture is bound to tie itself in unnecessary definitional knots, as well as almost
inevitably presenting a misleading view of the pervasive ancient engagement
with the epic texts.43 The matter is particularly acute when we reach the rich
prose literature of the Second Sophistic, in which revisions of Homer are a very
major presence, and into which a now long tradition of Homeric scholarship
is absorbed and re-used in new, often epideictic and/or paradoxical contexts.44
In the discussion which follows I focus (largely) on explicit indications, in
scholia and in rhetorical and critical treatises, of how Homer should be under
stood, rather than the implicit interpretations which Homeric reworkings in
creative literature, at both macro- and micro-level, bring with them. The forms
of expression in which reflection upon the epic heritage was couched were,
however, as varied as approaches to the texts themselves, and even the lim
ited case-study which follows can make no claim to do other than scratch the
surface.
Rhetorical criticism, that is the study of the strategies of both language and
substance which lead to the effective presentation of arguments and charac
ters, played a very prominent, perhaps indeed the dominant, role in the ancient
criticism of literature; literary criticism, as we might understand it, fellat
least in post-classical antiquitywithin the province of rhtorik.45 The basis
of much of the educational system, once the earliest preliminary studies were
completed, was the study of how speakers in the past, above all in epic, drama
and oratory, achieved particular aims, and how those achievements could be
replicated in the present;46 as with so much of ancient educated culture, the
seeds (and in some cases the full flowering) of virtually all rhetorical forms was
to be foundor so it was believedin Homer.47 The rhetorical turn was, as
has often been observed, also an important reason why, on the whole, ancient
critics seem less concerned with the meaning and interpretation of whole
works than with the study of parts, often detached from the context, a feature
of ancient criticism which has often puzzled their modern successors.
From the Hellenistic period on, an elaborate systematisation of rhetoric
and rhetorical education was developed, about which we are relatively well
informed from a large corpus of surviving rhetorical treatises and handbooks.
At the heart of this system lies the study of the texts of the past, what we would
call classical literature, and it is within the parameters of this system that
the foundations of a set of critical rules, amounting to no less than a body of
ancient literary theory, were established. If Aristotles Rhetoric represents a
sophisticated level of intellectual analysis never really reached again, the sub
sequent tradition is also at pains to explain the need for system and agreed
modes of analysis. Thus, for example, a treatise of perhaps the late second or
third century AD, wrongly ascribed to Dionysius of Halicarnassus and entitled
On the examination of logoi (i.e. literary works in general, though the focus
,
, 210
,
,
,
, ,
. 215
, ,
,
. 220
,
.
Homer, Iliad 3.209224
When they mingled with the assembled Trojans, Menelaos with his broad
shoulders rose above him as they stood, but when they were sitting,
Odysseus was the more distinguished. When they were weaving their
words and devices to all assembled, Menelaos indeed spoke fluently; his
words were few, but very clearly spoken, since he is not a man of many
words nor a rambler, and also younger by birth. When Odysseus of many
guiles leaped up, he stood looking down with his eyes fixed on the ground,
and he moved his staff neither back nor forwards, but he held it unmov
ing and seemed like an ignorant man. You would have said that he was
sullen and merely a fool. When, however, he sent forth his great voice
from his chest and words flowed like snowflakes in winter, then no other
mortal could compete with Odysseus, and then we were not so struck by
his appearance.
This passage was to become perhaps the most important foundational pas
sage for the later analysis of different styles of speaking and writing;50 Libanius
takes off explicitly from this passage to write the speeches which Menelaos and
Odysseus were supposed to have delivered on this occasion (5.199221, 228286
Foerster). For Libanius, this is a chance to show the different techniques of
compression and extension of the same material (5.200.37 Foerster), and the
result is that the speech of Menelaos, not a man of many words, takes twenty-
two pages in Foersters edition, and Odysseus fifty-eight. Such exercises were
a real test of the powers of (inuentio) for the orator, as there was no
Homeric text from which to work, and Libanius is not slow to point out to
50 The only other claimant to such an honour is Il. 1.247249 (Nestor); the scholia on Il.
3.212 match Menelaos-Odysseus-Nestor with Lysias-Demosthenes-Isocrates as the prime
representatives of the three styles. For further discussion and bibliography cf. Hillgruber
[1999] 370372.
The Rhetorical Criticism of Homer 689
his pupils just how successful he has been (5.228.5 Foerster).51 A related but
different challenge was the exercise of seeking to affirm () or dis
prove () the events of which poets, most notably Homer, had told.
Perhaps the most famous exercise of this kind, though it is in fact much more
than that, is Dio Chrysostoms Trojan Oration (11), in which Dio sets out rea
sons for wholesale rejection, not just of Homers account of the Trojan War, but
for much of the generally received story of Paris and Helen.52 A very powerful
weapon in such arguments was the appeal to probability (), and so it is
that the first in our collection of Libanian is That it is not probable
() that Chryses went to the Greek ships (8.123128 Foerster), and that one
of the is That the story of Achilles anger is probable (8.143150
Foerster); this latter speech contains much which functions as a rebuttal of the
about Chryses.53 One of the things which is most striking about these
exercises is the psychological depth and the level of calculation ascribed to
Homeric characters; this may be the fruit of rhetorical invention, but it is also
very instructive about how poetical texts were read and the sort of characters
that one expected to find there. In many ways, Libanius modes of argument
foreshadow some modern debates about character in literature, notably in
Greek drama, and what sort of intelligibility and motivation we are to ascribe
to poetic characters. Thus, for example, we learn that Agamemnon would not
have opposed the wishes of the majority, as Homer (Il. 1.2225) says he did,
because he knew that the security of his rule depended upon the goodwill of
those under him (8.126127), whereas on the other side it can be said both that
Agamemnon acted as a careful commander by throwing a potential Trojan spy
out of the Greek camp (8.146.69) and that the nature of Chryses subsequent
prayer (Il. 1.3742) makes perfect sense:
51 Despite this, Russell [1983] 110 claims that the speeches of the Libanian Menelaos and
Odysseus are not at all clearly differentiated; Libanius seems...to have been content
to give a very general impression.
52 Cf. Hunter [2009b], Kim [2010] Chap. 4, Minon [2012] (esp. pp. xlixlvi on the links to
rhetorical exercises).
53 Libanius two exercises have more than a little in common with the on show
in Dios account of Chryseis own motivation and calculations in Oration 61, cf. Drules
[1998] 7779, Kim [2008] 617620. That the opening scene of the Iliad should figure so
prominently in rhetorical texts is hardly surprising, given that this was probably the most
familiar piece of Homer, one known to every schoolboy. Kim [2010] 613617 rightly associ
ates the reading practice which fills in the gaps in Homers account of his characters
psychology and motivation with the grammarians interpretative principle of
, although that is usually used to explain apparent problems and omissions in
Homers presentation of facts, rather than of motivation.
690 Hunter
54 Cf. also Heracl., Quaest. Hom. 6.34, where it is claimed that the view that Apollo killed the
Greeks who had in fact urged respect for Chryses and spared Agamemnon is the result of
spiteful malice, Eust., Il. 37.610. The other reasons given by the exegetical scholia are also
predominantly psychological: Because the Greeks had given Chryseis to Agamemnon
after sacking Thebes [Il. 1.366369], because Agamemnon himself is included in the
Danaans, and because Chryses, being a barbarian, regards all Greeks as enemies.
The Rhetorical Criticism of Homer 691
The treatise of Aelius Theon of (probably) the early imperial period,55 one
of the principal witnesses for the preliminary exercises (progymnasmata)
which, as we have seen in Libanius, prepared students for the formal study of
rhetoric, offers a further helpful guide to the mindset which determined the
rhetorical approach. One kind of exercise which attracts particular notice in
the current context is paraphrase, the exercise of rewriting passages from clas
sical texts in your own words. Theon points out that, just as the same event or
material affects us in more than one way, so any which presents itself
to our minds can be expressed in a variety of linguistic modes, i.e. as questions,
prayers etc, according to the system of variations which pupils of the rhetori
cal schools followed. He evidences this claim by citing the fact that all classi
cal writers, poets and prose-writers alike, made excellent use of paraphrase,
by refashioning both their own work and that of each other (62.2325 Sp.),
and he then cites passages in which first Archilochus and then Demosthenes
and Aeschines might be thought to have paraphrased Homer, a passage where
Theopompus has paraphrased Thucydides, and several examples where one
Attic orator has used the words of a predecessor; finally he observes that
Demosthenes often paraphrases himself, not only by transferring what he
has said in another speech to elsewhere, but also by clearly saying the same
thing myriad times () in the one speech, although the audience do
not notice because of the variety of expression (63.2964.4 Sp.). The theory
of paraphrase, at the heart of which lies a distinction between what we say
and how we say it, the distinction expressed elsewhere as that between
and , is one first step along the road to a theory of what modern scholars
would call allusion, echo, even intertextuality, and that step is framed within
rhetorical education.
Theon is entirely typical in seeing the same rhetorical system governing the
writing of the ancients as is practised in his own day; the teacher of rhetoric
must first collect excellent examples of each exercise () from ancient
writings and instruct his pupils to learn them off by heart (65.3066.2 Sp.).
Thus the ancients supply the material for the rhetorical system, not merely
the , but are also themselves the principal examples of, and
hence authorities for, that system. In particular, as has already been noted, the
foundations of all rhetoric and rhetorical analysis were to be found in Homer.
Rhetorical criticism of, and illustration from, Homer shares with an approach
to literature through problems and their solutions an assumption that the
characters of literature have a familiar psychological depth which allows us to
55 For discussion of the author and date of the treatise cf. Patillon-Bolognesi [1997] viixvi;
Heath [20022003] proposes a radical re-dating to the fifth century AD.
692 Hunter
draw in our analysis of their strategies upon motivations and calculations not
made explicit in the text;56 when confronted by apparent anomalies, the best
interpretative strategy will usually be to give these characters the benefit of
the doubt. Working together with this fundamental assumption is the overrid
ing importance assigned to the notion of appropriateness ( ) and to
the shifting demands imposed by the particularities of any situation ( ).
Both are neatly seen together in an observation of Theon, in the context of the
rhetorical exercise of :
This analytical framework gave ancient critics a powerful tool for the analysis
of the speeches, and of the motives behind the speeches, in (particularly) epic
and drama. Rhetorical criticism is fundamentally the examination of why the
characters of literature act and speak as they do; it is not limited merely to the
formal analysis of speeches into their constituent parts. For a specific, though
not necessarily typical, example let us consider Agamemnons famous testing
of the troops in Iliad 2.
Zeus honours his promise to Thetis by sending a dream to Agamemnon
which (deceptively) leads him to think that the time for the capture of Troy is
at hand; Agamemnon calls a council of the Greek leaders,57 tells them of his
dream and then concludes:
, .
, ,
.
But come, let us see whether we can arm the sons of the Achaeans. I shall
first test them with words, as is appropriate (themis), and I shall urge
them to retreat in their many-benched ships; you however must use
words to restrain them, each in your various positions.
Homer, Iliad 2. 7275
56 On problems as a critical form cf. above, and Nnlist in this volume.
57 Dio 56.10 praises Agamemnon for wisely consulting the Greek elders before following the
advice of the dream.
The Rhetorical Criticism of Homer 693
The army was worn out from the plague and unmotivated because of the
length of time [of the war]; Achilles and his forces had withdrawn;
Agamemnon himself, when taking Briseis away in the assembly, had said
in order to frighten the others: anyone else should shrink from the idea
58 The various reasons for the athetesis given in the scholia on 2.76 do not necessarily (all)
go back to Aristarchus (cf., however, Lhrs [1992] 260261), though all are instructive
about ancient criticism: the poet should not have said that Agamemnon sat down after
his speech, because he did not stand to deliver it; Nestor has really nothing to say (
); it was silly to have Nestor say that they would only have believed such a dream
from the best of the Achaeans, because the powerful do not dream any differently from
the rest of us, and (finally), if the verses are deleted, then in v.85 will refer
to Agamemnon, as it should do, rather than to Nestor. Nestors speech has recently been
discussed by Nnlist [2012d]; the kinds of argument that Nnlist adduces to explain the
speech are in fact interestingly reminiscent of the kinds of rhetorical explanation that
we find in the scholia and in Libanius (cf. above). Schofield [1986] 29 calls Nestors obser
vation about the dream not being a deception because of who dreamt it ingenious as
well as tactful.
59 The apparent unclarity about the actions and knowledge of the commanders who had
heard Agamemnons speech has long been highlighted by those seeking to reconstruct
the creation of the text; cf., most recently, West [2011] 103105 (notes on vv. 7375, 192197).
60 The bT-scholia try to get around this interpretation of v. 75, but it seems inevitable.
61 Cf. McGlew [1989], Latacz [2003] 2930, 41, Cook [2003]; helpful guidance to older discus
sions in Katzung [1960].
694 Hunter
That Agamemnon did not really have any option but to test whether the war-
weary men were ready to go out against the Trojans is the conclusion of most
ancient discussion of the matter; the test, together with the precaution of tell
ing his colleagues how they are to act, is not, therefore, a sign of Agamemnons
weakness and mistaken leadership, but rather of his strategic good sense
(cf. also Eust., Il. 173.2433). In seeking to understand the text we must consider
the position which the character finds himself in and think out how he might
best handle that, even if these calculations are not made explicit in the text;
here modern critics of the peira have followed the ancient.63
Much ancient discussion of Agamemnons subsequent speech to the troops,
like much modern criticism, is focused upon the fact that a good part of what
he says seems designed to lead the troops to stay, rather than to go home,
which is the professed purpose of the speech. Later rhetoricians took very
great interest in this speech, for it seemed to be overtly arguing for an out
come which the speaker did not in fact want.64 The treatise On the method of
forcefulness, which is transmitted with the Hermogenean corpus, thus makes
Agamemnons speech the Homeric paradigm of accomplishing something by
arguing the opposite. In this figure the speaker will use arguments which are
easily refuted and contradictory and can be turned around:
62 Much of this has found its way into the scholia on Iliad 2.73; cf. also Dio 2.22, where,
however, Alexander (Aristotles pupil) misrepresents the events in Iliad 22, by omitting
Agamemnons peira speech entirely.
63 This was the basis of the discussion in Kullmann [1955], who saw the peira as a motif
dependent upon a situation of the Greek army known to the audience from the Kypria.
For a more recent attempt to explain the origin both of the peira and of the opening of
Book 9 (see below) cf. West [2011] 100105, 214215.
64 Cf. further Hillgruber [1999] 357359.
The Rhetorical Criticism of Homer 695
Homer did this. Agamemnon is testing the Greek army and wants them
to remain, while saying that they should not remain but should flee.
Through his whole speech he says things which are easy to refute and
turn around, thus giving openings to his opponents, and at the end he
says contradictory things. For to say
The timbers of the ships have rotted and the ropes are loose (2.135)
is very obviously opposed to Let us flee. How could they flee without
ships? This is the argument of someone who wants to prevent them from
sailing away, not an argument for doing what he is saying.
[Hermogenes] 437.14438.4 Rabe
65 Cf. Russell [2001a] 160163. Dentice di Accadia [2010b] is a recent and helpful attempt
to take the arguments of [Dionysius] seriously; the present chapter was drafted before
the appearance of that article, and I have not signalled the various places where our two
accounts agree or differ. Dentice di Accadia [2010a] is now the standard edition of the
treatises, and cf. also Schpsdau [1975], Hillgruber [1999] 357359, Heath [2003].
696 Hunter
Much ancient rhetorical training of course would have shown that the
implicit argument that emotion is a guarantee of sincerity was a very unsafe
assumption in oratory. Nevertheless, in his testing speech in reply to optimis
tic words from the steersman Tiphys in Book 2 of Apollonius Argonautica,
which clearly reflects not just the Homeric peira but also scholarly discussion
of that episode, Jason also begins in a distraught and highly emotional way
which would seem likely to assure anyone that it was grief, not design, which
prompted his words:
, ;
,
, 625
.
,
,
, . 630
,
,
. ,
, 635
,
.
,
.
But Jason answered him in return with soft words: Tiphys, why do you
offer me these consolations in my grief? I have erred; my wretched folly
offers no remedy. When Pelias gave me his instructions, I should have
immediately refused this expedition outright, even if it meant a cruel
death, torn apart limb from limb. As it is I am in constant terror and my
burdens are unendurable; I loathe sailing in our ship over the chill paths
of the sea, and I loathe our stops on dry land, for all around are our ene
mies. Ever since you first assembled for my sake, I have endured a cease
less round of painful nights and days, for I must give thought to every
detail. You can speak lightly, as your worries are only for yourself. I have
The Rhetorical Criticism of Homer 697
no anxiety at all for myself, but I must fear for this man and that, for you
no less than for all our other companions, that I shall be unable to bring
you back unharmed to Greece. So he spoke, testing the heroes, and they
all shouted words of encouragement.
Apollonius, Argonautica 2.620639
Jasons speech too has divided modern critics, particularly over its purpose (if
it is not simply an anguished retort to Tiphys optimism) and over whether
or not his attitudes are here in any way feigned;66 unlike Agamemnon, whose
opening appeal to at Jason echoes, Jason does not apparently have a specific
plan in mind (he does not propose that they now turn around), though the
two choices facing the crew are obviously, as in Homer, pressing forward or
abandoning the expedition. Like Agamemnon, Jason focuses on his own sit
uation and, like the Agamemnon of the later rhetoricians, some of his argu
ments would be very easy to refute; the rebuke to Tiphys of vv. 633634 had
in fact already been shown to be false by the narrative of Tiphys role in the
passage through the Clashing Rocks (2.581585), and it is patently absurd to
charge the steersman with concern only for himself. Over Jasons speech in
fact hovers the ubiquitous ancient parallel between the ruler and the steers
man, each responsible for the safety of the vessel under his command and
the people in it;67 Tiphys, no less than Jason, could claim that it is his duty
to bring the Argonauts safe back to Greece. From the perspective of ancient
rhetorical criticism (best attested for us, of course, in texts considerably later
than the Argonautica), Jasons speech would indeed be understood as a clear
example where the speaker says one thing and conceals another in his heart,
as Achilles rebuked Odysseus and Agamemnon for so doing (Il. 9. 313), in verses
which were indeed to become associated with .68 Whether
or not we receive Jasons speech as unprepared and unguided as his crew does
depends importantly upon the disputed meaning of the introductory
(2.621), but two observations are relevant here. Apollonius
has sought to make the effect of his peira more dramatic than Homers by
66 Frnkel [1968] 214221, arguing that in 2.638 means seeking to provoke, not
testing, has been an influential discussion; further observations and bibliography in
Hunter [1993] 2022.
67 Particularly striking when set against Jasons speech is Dio 3.6267 in which the good
ruler is first compared to the steersman battling a storm while all the other passengers are
idle and then to a general on campaign who must look after every soldier, whereas each
ordinary soldier only looks after himself.
68 Cf. Philostratus, fr. 542 Wright on Polemo.
698 Hunter
omitting any clear indication of what Jason is actually up to, so that, whatever
the meaning of 2.621, the external audience is placed more in the position of
the audience in the text than is the case in Homer, where we have been very
explicitly warned about Agamemnons real intentions; secondly, we may per
haps use this passage of the Argonautica to trace ancient discussion and rhe
torical analysis of Agamemnons speech in the Iliad back to a much earlier date
than that of the scholia and the rhetorical texts I have been considering.
The scholia on Agamemnons peira largely follow the interpretative patterns
already outlined: Agamemnon knows that the men are weary and depressed
at Achilles withdrawal and that his standing with them is fragile (Schol.AbT
2.73). A close engagement with the text again lies at the heart of interpreta
tion: Agamemnon calls the troops because such praise works
against any desire for flight (Schol. bT 2.110b); he says (v. 140), when
he might have said , so that the dishonourable word will have a neg
ative effect,69 and so forth. A further argument in the same scholium about
Agamemnons implicit calculations may be expanded along the following
lines: Agamemnons stratagem will help recover some of his standing, because
the ordinary troops do not like generals who are recklessly gung-ho about
fighting, and so he has nothing to loseif the men want to abandon the expe
dition, then the other leaders will dissuade them, and if not, then well and
good, and no one will be any the wiser about the stratagem, but the men will
know that he at least does not gamble recklessly with their lives. So too, the
pseudo-Plutarchan treatise On Homer, which absorbs and reflects a great deal
of the mainstream of ancient criticism, makes a rather similar point:
Does not Agamemnon...use rhetorical art, when he says to the mass the
opposite of what he wants, so that he can test their spirit and not become
hateful to them by forcing them to fight on his behalf? He himself speaks
in a way which will please them ( ), but one of those others with
the power to persuade them will turn them back and make them stay, as
this in truth is what the king wanted.
[Plutarch], On Homer 2.166
69 Some modern commentators (cf., e.g., the Basel-commentary on 1.173 and 2.74) observe
that does not necessarily have a negative connotation; it may suggest withdraw
from a position rather than flee. This may be true, but we must also recognise that, even
in formulaic epic style, the same words can resonate differently with different audiences.
The Rhetorical Criticism of Homer 699
Great men do not make mistakes; one merely has to try to understand
their stratagems. This portrait of a cunningly calculating Agamemnon, who
prepares the ground for the hostility of the troops to be displaced on to other
leaders rather than himself, is deeply rooted both in the analysis and debates
of rhetorical education and in the agonistic realities of ancient political and
oratorical struggle; modes of interpretation, then as now, unsurprisingly reflect
the culture that gave rise to them.
The final part of Agamemnons speech is particularly worthy of note:
,
, ,
, .
, ,
,
.
Nine of great Zeus years have gone past, and the timbers of the ships
have rotted and the rigging hangs loose. Our wives and young children sit
waiting for us at home, while the task which brought us here is utterly
unaccomplished. Come, let all of us do as I say: let us retreat in our
ships to our beloved native lands, for we shall never take Troy with its
broad streets,
Homer, Iliad 2.134141
our wives long for us (), or [we should not depart] because we have
accomplished nothing to match their expectations ().70 A modern
critic might observe that the juxtaposition of an explicit reference to the decay
of the ships over nine years to the waiting wives carries with it the powerful
implication that the wives too are not getting any younger. Some version of
this reasoning may in fact be reflected in the A-scholium on vv. 136137 which
warns against punctuating after , because that would make the meaning
; the point is presumably not merely the disrupted syntax which would
result, but also the low implication that the wives too, like the ships and the
ropes, are rotting and loose.71 This clearly is not what Agamemnon wants to
say, but the verses are indeed held together by the idea of the long passage of
timethe children will be no longerand behind the scholiastic
worry about punctuation lies a recognition of this resonance.
For many modern critics, the principal literary effect at work here is tragic
irony: Agamemnons (deceptive) claim to have been deceived by Zeus (vv.111
114) is more true than he knows, and when Agamemnon repeats some of the
verses, though without the ones which might seem most ambiguous in their
effect, in very different circumstances at the council (or assembly) which
opens Book 9, the full force of that irony hits home.72 It is an obvious ques
tion why (as far as we know) ancient critics too did not elaborate such an
approach to this passage. In fact, however, critical positions may indeed have
embraced something very like this, and ancient critics seem in fact to have
been divided as to whether or not Agamemnons speech in Book 9 (vv. 1728),
whichafter a different opening address (cf. further below)repeats verba
tim vv. 111118 and vv. 139141 from Book 2, was another test, parallel to that of
Book 2, just as some at least seem to have entertained the idea that his third
plea for withdrawal (14.7581) was also a test (bT-scholium ad loc). Pseudo-
Dionysius appears to take for granted that the speech in Book 9 is such a test (II
325.1416 U-R), but the absence of vv. 2325, which could be taken to suggest
that Troy might still be taken, from Zenodotus shortened version of the speech
and the athetsis of those verses by Aristarchus point to the other view; thus
the A-scholia on vv.2325 observe that [Agamemnon] is not making a test,
but he is speaking sincerely about withdrawal as Zeus has inflicted setbacks
upon them, and the bT-scholia consider it to hesitate on this sub
ject when in Agamemnons situation.73 This second view is not an expression
of tragic irony as such, but it draws upon the same contrasts as that modern
critical approach. In this difference of critical effortseeking to account for
the text as it confronts us or removing verses to produce the coherent text we
wantlies foreshadowed, of course, much of the history of Homeric criticism.
The bT-scholia on v. 17 produce interesting reasons for believing that the
speech of Book 9 is indeed another test:
He now makes this second test of the Argive leaders (i.e. not of the whole
army), because he fears lest the defeat and the rout inflicted by Zeus has
destroyed even their resolve. That Agamemnon is testing in this council
too is clear from the way in which he puts up with Diomedes rebuke,
when he did not put up with the speeches of Achilles who was a better
man, and from the fact that Nestor, a man who understands the kings
thought, praises Diomedes, though he had previously rebuked Achilles.
bT-scholium on Iliad 9.17b
The comparison with Iliad 1 reflects a proper critical sense, much echoed
(though not always with proper acknowledgement of ancient criticism) by
modern scholars,74 of how Book 9 acts as a kind of reprise of Book 1 and a reaf
firmation of Achilles withdrawal, here at another time of great crisis. In Book1
Agamemnon is known to have spoken without any of concealment;
whereas Achilles speeches there simply made Agamemnon angrier and more
determined, his (unrecorded) reaction, here taken for silent acquiescence, to
Diomedes speech in Book 9 professing enthusiasm for the fray shows that
Diomedes reaction, apparently the reverse of what Agamemnon was argu
ing for, was in fact just the reaction the king wanted. Agamemnons speech
was therefore a , a figured speech. As for Diomedes
73 Cf. also Eust., Il. 732.68733.2, where the contrast is between speaking and
speaking .
74 Cf., e.g., Lohmann [1970] 217218.
702 Hunter
himself, the standard critical position was that, whereas in Book 4 he had not
responded when rebuked by Agamemnon, he now feels free to attack because
of the authority given to him by his great martial deeds described in the inter
vening books.75 Ancient rhetorical critics start with the assumption that great
men know what they are doing and rise above circumstances. For both ancient
and modern critics, Agamemnons apparent silence after Diomedes speech
speaks volumes (cf. further below), but what it says may differ according to
critical idiom. Eustathius, for whom Agamemnons speech in Book 9 is sin
cere, perhaps has in mind arguments such as that of the bT-scholiast on v. 17
(above) when he observes that the king puts up with the rebuke both because
of the rule () of the assembly [cf. v. 33] and because the rebuke is not false,
but in this matter the hero speaks the truth... (Eust., Il. 733.22).
After Agamemnons first speech in Iliad 9 there is an awkward silence:76
,
.
So he spoke. They all remained silent and for a long time the sons of the
Achaeans were quiet and downcast. At length Diomedes, powerful in the
war-cry, addressed them...
Homer, Iliad 9.2931
75 Cf. bT-Schol. on 4.402, 9.31, [Plut.] Hom. 2.168, Plut. Mor. 29b. For a different view cf. [Dion.
Hal.] II 314.19316.14 U-R. Reeve [1972] 23 argues that vv. 3239 belong to a later stage of
the tradition than vv. 4049. It is interesting to compare the ancient accounts of Diomedes
speech with a modern account of its rhetoric, namely Martin [1989] 2425, and cf. also
125. Without reference to ancient views, Martin sees Diomedes speech as that of a novice
speaker who imitates, sometimes with inelegant repetition, phrases that he has heard in
the mouths of others; he made no reply in Book 4 because he needs time to compose his
reply (contrast, however, Martin [1989] 7172 on Diomedes cunning silence). Both the
ancient scholiasts and Martin account for the difference between Book 4 and Book 9 in
terms of Diomedes development, but do so in rather different ways; on the other hand,
there is more than a little in common between [Dionysius of Halicarnassus] account of
major Homeric speeches as and Martins account of Homeric flyt
ing in which the participants know the rules of the game. Cf. further Scodel [2008] 6061.
76 West [2011] 215 rightly notes that this pattern of silence after a speech which takes the plot
in a new direction is itself typical. This, however, is a particularly marked example: v. 30
occurs elsewhere in the corpus only as v. 695 of the same book, when vv. 2931 are (point
edly) repeated as the reaction to Odysseus report of the failure of the embassy to Achilles.
The Rhetorical Criticism of Homer 703
77 Cf. also Eust., Il. 733.2: The others are silent, fearing lest this speech of the king also is a
test of the Achaeans.
78 An exception is the discussion by Scodel [2008] 6869.
79 Cf. Edwards [1970], Hunter [1993] 141142, Nnlist [2009a] 316317, Hunter-Russell [2011]
106. The verbs which Plutarch uses are and ; the former is not
used in the scholia, whereas the latter is commonly used of the poet introducing/paving
the way for a character or later narration, but is also found in the narrower sense in which
Plutarch uses it (bT-scholium to Iliad 1.247248).
704 Hunter
80 Ancient critics, unlike for the most part their modern successors (but see von Wilamowitz-
Moellendorff [1920] 3334), were divided on whether the first gathering of Book 9 was of
the leaders only (e.g. bT-scholia on 9.11, 17, D-scholia on 9.12) or of the whole army (cf., e.g.,
Plut. Mor. 29c). Aristotle discussed the problem posed by 9.17 if one held that the whole
army was present and explained (fr. 382 Gigon) that the ordinary soldiers are entitled to
listen, but the leaders can also act.
81 The matter certainly deserves more discussion than it receives in Hainsworths note on
v.17; Lohmann [1970] 216 merely observes that the stolz und kriegerisch address of Book
2 is characteristically altered, without asking what the address in Book 9 actually means;
Griffins note on v. 17 acknowledges that that verse suggests that Agamemnon is talking
to the chiefs, but can only conclude that there is a lack of exact focus on the facts; West
[2011] 214 sees in this unclarity (11 and 17 suggest a meeting of leaders...the present gath
ering, however, soon appears as an assembly of the whole army (30, 50)) another argu
ment for his view that the opening of Book 9 was composed before, and was the model
for, the parallel scene in Book 2. De Jong [2004] 190 mistakenly makes 2.110 identical to
9.17, thus blurring the question of addressee.
The Rhetorical Criticism of Homer 705
Ren Nnlist
Introduction
1 Scholarship is the art of understanding, explaining, and restoring the literary tradition. It
originated as a separate intellectual discipline in the third century before Christ (Pfeiffer
[1968] 3). For the history of Hellenistic scholarship see also Montana in this volume.
2 Cf. e.g. Ford ([2002] with bibl., esp. 2 nn. 45).
3 See Novokhatko in this volume.
4 See Matthaios in this volume.
5 This second reason applies in particular to the large mass of scholia that cannot be attributed
to a particular critic. Such anonymous scholia are an important source of my book [2009a],
to which reference will repeatedly be made for practical examples in the notes on the
individual entries below. The entries themselves, however, are generally based on material
whose Hellenistic provenance is certain or very likely.
guess. Truly Hellenistic material normally reached posterity indirectly via other
sources,6 with the usual problems that accompany this type of transmission:
fragmentation, loss of context, misunderstanding, deliberate distortion. The
latter, in particular, looms large in the corpus of texts that, due to the recent
publication of much improved editions, has become a fundamental source of
literary criticism in Hellenistic times: the works of Philodemus.7 Leaving aside
the problems that an editor of these exceedingly difficult fragments faces, the
patent polemics of Philodemus arguments raises additional questions. How
reliable a witness is he? Can his selection of topics and critics be taken to be
representative of the Hellenistic period in general? And what was his aim?8
Antiquity did not recognise literary criticism as a field of its own, nor did
it attempt to define what a literary critic is or does. Consequently, the subject
of this chapter can be described as an essentially undefined grey zone that
overlaps with the domain of grammar, rhetoric, philosophy and education. In
spite of being in essence a modern construct, ancient literary criticism is not
a phantom. But the fact that it is a construct has a number of implications that
are worth repeating at the outset. An attempt to scrutinise the ancient mate-
rial for criteria that help identify what is literary criticism (and what is not)
is unlikely to produce meaningful results.9 For this kind of demarcation the
modern reader is fated to start from modern concepts of literary criticism.10
This necessity obviously entails the risk of anachronistic distortion, which can
be reduced, if not eliminated, by making every effort to see things through the
eyes of the ancient critic. Moreover, in the absence of a well-defined field it is
even less to be expected that ancient critics have something like a common
denominator (except, perhaps, for the very general observation that they all
deal in some form or the other with the understanding of texts). Consequently,
few, if any, of the approx. ninety concepts that will be discussed below can
safely be said to be characteristic or typical of the entire field of ancient literary
criticism. Most of the time, it could actually be shown that a particular concept
is either ignored or even openly contradicted by other critics. The selection
of topics was determined by the criterion whether the individual concept is
either frequent or remarkable enough to be worth treating here. The presen-
tation in the form of alphabetically listed catchwords with cross-references
attempts to combine several goals: to balance the selectivity imposed by the
limited space, to give an impression of the wide variety of the relevant con-
cepts and to enable easy reference.11
The fact that there is a considerable overlap specifically between literary
criticism and rhetoric has another consequence for the wording in this chap-
ter. The descriptions below will generally make no distinction between, on the
one hand, poet, prose writer and orator, and, on the other, reader, spectator
and listener. Instead the terms author and reader will often be used to rep-
resent each side. To a certain extent, this is an anachronistic distortion but it
has the advantage both of saving space and avoiding potentially cumbersome
phrases. More importantly, insights that were gained and thus described on
the basis of rhetorical texts recur in literary criticism and vice versa (see also
below on *rhetoric).
Several texts to which reference is made below had a strong pragmatic
component in that they provided an arsenal of interpretative tools. The pur-
pose was at least twofold. First, these texts aimed to instruct how to read and
understand literature, but they, second, also intended to teach how the various
features found in classical texts could be put to use. The idea perhaps was
not so much to instruct how to write poetry specifically, but how to use the
various features found in poetry and other literature for, mostly, rhetorical pur-
poses. At any rate, several texts that can be subsumed under the rubric poetics
(see on theory of *poetics below) or literary criticism were written as if
they were intended for an audience that planned to produce literature, write
speeches, etc.12
The preceding paragraph will also have made it clear that preference will be
given to texts that expressly and unambiguously address questions related to
literary criticism. Conversely, the immanent poetics of Hellenistic literature,
11 Even if there is enough space for the alternative, the writing of a coherent historical nar-
rative, it is not without problems, see Hunter [2009a] 8.
12 A striking exception is Aristotles Poetics (Halliwell [1986] 3739).
Poetics and Literary Criticism 709
Aesthetics
Greek poetry itself is rife with references to its own aesthetic qualities. It was
thus only natural that critics took their cue from the authors. They described
and judged literature by means also of aesthetic concepts such as beauty, ele-
gance or grace (and their respective opposites). This, however, is not to say
that they also developed a theory of aesthetics in the full sense. Even though
Aristotle held against Plato that art must be measured by its own standards
(poetic *licence) and recognised that the creation of aesthetic *pleasure was
fundamental to art in general and literature in particular, the theory expounded
in the Poetics had better not be called aestheticist (Halliwell [1986] esp. chs.
2 and 3). Moreover, neither Aristotle nor any of his successors went so far as
to postulate the full autonomy of art or the artist (see also *Lart pour lart,
*educative function). In this sense ancient critics did not develop a theory of
aesthetics, while they often applied aesthetic criteria in their interpretations.
The same holds true mutatis mutandis for the concept of taste, which was not
theorised as such by critics, but can be inherent in their aesthetic comments.14
Allegorical Interpretation
See *multiple meanings.
Allusion
The careful *analysis and *comparison of texts led critics, among other things,
to detect allusions. Their point of reference was in principle open. A particular
passage could allude to historical events or persons, *mythological stories, spe-
cific texts, etc. The latter are of particular interest because critics often identi-
fied or even quoted the relevant text, which otherwise might be lost to posterity.
An altogether different question is whether allusion was actually a separate
category (distinct from quotation, parody, etc.). Terminological considerations
13 Regular inclusion of the immanent poetics, though recommended by Classen [1995] 535
and attempted e.g. by Gutzwiller [2010], would have imposed an even sharper limitation
on the number of topics that could be dealt with.
14 On the history of aesthetics as a critical concept see Wiegmann [1992] 11341154, with
bibl., Halliwell [2002], for a professedly revisionist account Porter [2010]; on the history
of taste as a critical concept see Fick [1996] 870901, with bibl.; for practical examples see
e.g. Ps.-Demetr. Eloc. 67, 287.
710 Nnlist
induce caution. The most common term seems to be ainittomai (to allude,
hint at) with cognates, which, however, cannot be restricted to allusions in
the narrow sense. It can equally designate quotations, imitations, intertexts,
parodies, ridicule, etc. The term can also be used in connection with *multiple
meanings. Essentially the same holds true for the false friend emphasis (with
cognates), which normally does not mean emphasis in the modern sense but
allusion, insinuation, hint, adumbration, etc. Occasionally, its exact meaning
remains elusive. The terminological situation is, in a way, representative for the
entire subject; for it is difficult to identify general trends and the like.15
Alphabetical Order
Since it has become such a common feature of reference books, it is easy
to forget that the alphabetical order needed to be invented. Whether or
not Zenodotus can take credit for the introduction, his glossary was in any
case arranged alphabetically.16 And so were the individual subsections of
Callimachus *Pinakes (organised by *genre). The extant evidence strongly
suggests that it was the Hellenistic critics who introduced this ground-
breaking method.
Ambiguity (amphibolia)
In a very short section of Poetics chapter 25 (on *problems and solutions),
Aristotle deals with the problem of ambiguity (1461a2526). Context and
example (Il. 10.252) demonstrate that semantic ambiguity is meant, that is,
words and expressions that are open to more than one possible explanation.
Aristarchus regularly drew the readers attention to instances of ambiguity. He
would list the possibilities and offer arguments for his preferred solution (e.g.
sch. A Il. 2.629a Ariston., on reverse order, also known as ABBA pattern). On
the proviso that Aristonicus excerpts (*commentary) can be trusted in this
matter, some cases were actually left open, presumably because Aristarchus
thought there was not sufficient ground for a decision (e.g. sch. A Il. 16.561a1
Ariston.; *sense of proportion). Conversely, the concept of deliberate ambigu-
ity, popular among many modern critics, is unlikely to have appealed to their
ancient predecessors (cf. the warning in Arist. Rh. 1407a32). Likewise, discus-
sions of ambiguity normally circled around smaller linguistic units, words or
15 On the history of allusion as a critical concept see Hughes [1992] 652655, with bibl., also
Conte-Most [20124b], for practical examples Nnlist [2009a] 225237.
16 Pfeiffer [1968] 115, Alpers [1975] 116, cf. also Montana and Tosi in this volume.
Poetics and Literary Criticism 711
sentences, not whole texts. This is not to deny that critics also detected *mul-
tiple meanings, which, however, was not felt to be a problem of ambiguity.17
Anachronism
When critics detected a problem in a text, one of the attempted solutions
(*problems and solutions) was to see whether the passage perhaps reflected
a habit that was no longer current (cf. Arist. fr. 166 Rose, Poet. 1461a23, later
called lusis ek tou ethous, solution from the <earlier> habit). This interpreta-
tive principle was generalised by Aristarchus in such a way that readers of a
historical text must beware of possible anachronisms. The age of the Trojan
heroes (represented by Homers characters) was different from Homers
own, which again was different from subsequent periods (including that of
Aristarchus and his readers). For instance, the heroes did not ride on horse-
back, but a simile showed that Homer himself was aware of this technique
(*narrative voice).18 Conversely, Homer did not know yet the use of the crown
or the mixing of wine with water.19 Aristarchus probably had two reasons to
warn against possible anachronism. He took exception to the treatment of
the heroic age in post-Homeric literature such as tragedy, where characters
showed familiarity, for instance, with the concept of writing.20 More impor-
tantly, he saw that other scholars uncritically imposed the standards of their
own age onto the text and thus failed to do justice to it. He therefore advocated
the view that a text must be understood from the perspective of its own time
(see also textimmanente *interpretation).21 His basic argument is not so very
different from the one made by F. A. Wolf [1795], who laid the foundation for
modern Homeric studies.
17 On the history of ambiguity as a critical concept see Bernecker-Steinfeld [1992] 436444,
with bibl.
18 The most comprehensive collection of these examples is Schmidt [1976], on riding spe-
cifically [1976] 229; see also Nnlist [2009a] index s.v. anachronism.
19 Schmidt [1976] 215218 (crown), 261 (mixing of wine).
20 Schmidt [1976] 213.
21 This also means that Aristarchus did have a sense of (linguistic) history (Nnlist [2012a]).
22 Pfeiffer [1968] passim.
712 Nnlist
put the results on a much better footing. This improved method is particularly
easy to document with semantic studies. The entire text must be scrutinised
in order to determine, for example, that a word is attested only once, twice or
not at all.23 But the same thoroughness can be gathered from the comprehen-
sive discussion, for instance, of all the attestations of Mt. Olympus in the Iliad
or the systematic collection and marking of homonymous characters.24 More
generally, this approach allowed critics to determine how a particular author
normally dealt with a specific problem (*custom), for example, in *compari-
son with other authors. It is clear that the new institution of the *Museum,
together with the enormous library, provided unprecedented resources.25 It
thus made the task of systematic analysis easier or, in some cases, even pos-
sible. This (in the full sense) exhaustive analysis probably required critics to
work long hours. Though better attested in poetry, it is not improbable that the
lamp and long nights became symbols of the age in general.26
23 Once (hapax legomenon, for a collection see Lehrs [18823] 12, twice (e.g. sch. A Il. 14.463a
Ariston.), not at all (e.g. sch. A Il. 10.226 Ariston.).
24 Mt. Olympus: Nnlist [2011] 111, with bibl. in n. 17; homonymous characters: Nnlist
[2009a] 240242, with bibl. in n. 8; for more examples see index s.v. treatment, systematic.
25 See Montana in this volume.
26 Callim. Epigr. 27 Pfeiffer, cf. e.g. Horace Ars P. 267268. The critics long nights may have a
precursor in Aristoph. Ran. 931932 (Hunter [2009a] 24).
27 Zenodotus thought it was not appropriate, but Aristarchus replied that Aphrodite was in
the guise of an old servant, for whom it was (sch. A Il. 3.423a Ariston.).
28 On the history of appropriateness as a critical concept see Asmuth [1992] 579604, with
bibl., esp. Pohlenz [1965b], Rutherford (et al.) [1994] 423451, with bibl., for practical
examples see Nnlist [2009a] index s.v. appropriateness.
Poetics and Literary Criticism 713
Authenticity, Questions of
From a modern point of view, questions of authenticity are an aspect of textual,
not literary, criticism. It is, however, important to realise, especially when read-
ing *commentaries, that textual criticism was never far from an ancient critics
mind (literary *heritage). Questions of literary criticism regularly extended
into textual criticism because, for example, a particular literary device was
(not) common with the relevant author (*custom) and could therefore help
decide whether the passage was genuine or not. The same method applied of
course to topics other than literary criticism (linguistics, lexicography, etc.).
Even in the case of doubtful authenticity the relevant lines remained in the
text and were marked with a *critical sign (= to athetise in the ancient sense of
the word). Actual excision was rare (*sense of proportion).31
Biographical Criticism32
Starting probably as early as the sixth century BC, the authors biography
became a matter of keen interest.33 It can be found, for example, in Aristotles
dialogue On Poets, which throws an even sharper light on the fact that the
subject is virtually absent from his Poetics.34 The Peripatetic school, in par-
ticular, devoted considerable energy to biographical data, but the topic as such
was widespread. From a modern point of view, the prevalent method is not
entirely satisfactory, in that critics often seemed to be content with mining
an authors works for personal data. This method was rooted in the convic-
tion that there was an immediate correspondence between an authors work
and his life. A manly person wrote manly poetry, an effeminate person wrote
effeminate poetry, etc.35 According to this principle, an authors works allowed
readers to reconstruct his life.36 This kind of activity had a penchant for the
same type of anecdotal evidence with which ancient biographies are rife.37 It
did not necessarily help that critics regularly failed to differentiate between
*narrative voices in a text, in that a characters views were uncritically taken
for the authors. It is also worth noting that biographical criticism primarily
worked in one direction. The works were used as evidence for an authors life,
but there was little effort to find biographical data which was independent of
the works and could then be used to help illuminate them.
Book
As several entries (*Museum and library, *plagiarism) make clear, there is
some justification in regarding the book(roll) and thus the written word as
symbols for the dividing line that separated the Hellenistic era from the past.
This statement also applies to poetry itself. Poets of the early and classical peri-
ods with their emphasis on (feigned) orality are unlikely to picture themselves
as receiving Apollos inspiration with a writing tablet on their knees (Callim.
fr. 1.212 Pfeiffer).38 There is, however, a certain risk of exaggerating the book-
ishness of the Hellenistic period. Literature largely remained an aural form
of art (*enargeia). Not least because the format of a bookroll was not suitable
to quick reference, authors and critics alike kept relying on their memories,
which in many cases must have been formidable (cf. the anecdote reported
34 That On Poets discussed biographical topics is undisputed (e.g. fr. 76 Rose = 65 Janko, on
the life of Homer). The question whether it contained substantially more (Janko [2011]
317539) falls outside the scope of this chapter.
35 The clearest expression of the principle is perhaps Aristoph. Thesm. 148152, but it also
underlies, e.g., Arist. Poet. 1448b2427, 1449a26. Cf. Russell [1995] 162164.
36 Whether all the biographers believed this reconstruction to be authentic is another ques-
tion. Cf. e.g. the criticism expressed in Satyrus Life of Euripides (F 6 fr. 39 col. XVIII =
Schorn 2004: 108109, for the interpretation 48).
37 Lefkowitz [20122].
38 Note, however, that Callimachus too resorts to the motif of feigned orality (Harder [2004]
79, with reference to Hymns 5 and 6).
Poetics and Literary Criticism 715
in *plagiarism). They had been trained from an early age onwards. A standard
school exercise was to memorise and perform large chunks of poetry.39
Canons
The principle of singling out a particular group of authors because of their
exceptional quality is operative in the Frogs already. Whether or not this was
Aristophanes invention, Aeschylus, Sophocles and Euripides are given the
status as classical Greek tragedians that they occupy to this day. Likewise,
Aristotle and his entourage attributed the same status as classics to the epic
poets Homer and Hesiod. In the case of Homer, this also meant that the Iliad
and the Odyssey were severed from the poems of the epic cycle, which before
had mostly been considered to be his too (questions of *authenticity). The
Alexandrian critics continued this process of canonisation. First, they prob-
ably increased the number of classic epic poets to four or five. Second, they
selected an exceptionally large group of nine lyric poets: Pindar, Bacchylides,
Sappho, Anacreon, Stesichorus, Simonides, Ibycus, Alcaeus and Alcman. It is
interesting to note that they did not incorporate contemporary authors into
their canons (Quint. Inst. 10.1.54). Old Comedy presents a blurrier picture.
A triad Eupolis, Cratinus and Aristophanes can be found in Horace (Sat. 1.4.1
= Eupolis test. 23 Kassel-Austin) and later, but Pfeiffer ([1968] 204) has argued
that Eratosthenes and Aristophanes (sc. of Byzantium) regarded Pherecrates,
for instance, as its equal. The selection process was subsequently extended to
*prose, as seen in the canon of the Ten Attic Orators (led by their champion
Demosthenes). The relevance of this process can also be deduced from the
fact that none of Aristarchus numerous *commentaries dealt with an author
who was not canonical.40 The close correspondence with the writing of *lit-
erary history is particularly clear in the well-known chapter from Quintilian
(Inst. 10.1), which is expressly built on Alexandrian canons. A different but not
entirely unrelated selection process resulted in the compilation of the first lit-
erary anthologies.41
Censorship
Plato has a good chance of being the most vigorous and best-known advocate
of censorship because he notoriously banned all poetry from his ideal state
39 Cameron [1995] 65, Cribiore [2001] 166, 213, also index s.v. memory.
40 Pfeiffer [1968] 208.
41 On canonisation in Alexandria see Pfeiffer [1968] 203208, Easterling [20124] with bibl., on
its history as a critical concept Asper [1998] 869882, with bibl.; on anthologies Cameron
[1993].
716 Nnlist
(Resp. books 23, 10). He thus continued the ethically motivated objections
that Xenophanes of Colophon had raised against the anthropomorphism and
alleged immorality of Homers gods (21 B 1112, 15 D-K). Heraclitus went a step
further and suggested that Homer and Archilochus be clubbed and expelled
from poetic contests (22 B 42 D-K). A century or so before him, the tyrant of
Sicyon, Cleisthenes, had banned the performance of the Homeric epics for
political reasons (Hdt. 5.67.1). When Protagoras faced charges of atheism,
his books were collected from their owners and burnt in the agora (80 A 1
D-K). Ptolemy II (later called Philadelphus), though known as a book collec-
tor, apparently did not like being attacked by them. When Sotades criticised
Ptolemys marriage to his sister Arsinoe in graphic terms, he was, depending on
the source, either drowned (Hegesander ap. Ath. 14.620f) or incarcerated for a
long time (Plut. Quomodo adul. 11a). On the other hand, censorship was never
implemented in a systematic way nor does it seem to have played an impor-
tant role in criticism. When some anonymous scholars (refuted by others: sch.
Ar. Pax 778, with Holwerdas note) athetised the notorious story of Ares and
Aphrodites adultery (Od. 8.266369), it must be kept in mind that the disputed
lines remained in the text (questions of *authenticity). Nor is there evidence
that school texts were purged ad usum Delphini. One may compare Plutarch
(Quomodo adul. 14f15a), who preferred a good preparation of young readers
over censorship.42
Characterisation
Chapter 6 of Aristotles Poetics defines six qualitative parts of tragedy, the first
and most fundamental being *plot. Second comes characterisation (thos)
because the recommended *mimsis is that of an action, which requires
agents. The four other parts, thought, diction (*style), lyric poetry and spec-
tacle, receive conspicuously less attention in the Poetics than plot and charac-
terisation. Additional importance comes from the fact that, for Aristotle, the
characters (i.e. their speeches) ought to reveal a moral choice (prohairesis) and
display dispositions. In addition to bringing out the fundamental relevance of
character, Aristotle also determined which type of character should appear in
which *genre. Characters in serious genres such as tragedy or epic should be
serious themselves and better than the audience, characters in comedy should
be worse than the audience (better and worse primarily in terms of social
class). Next Aristotle (Poet. ch. 15) singled out four criteria of persuasive char-
42 On censorship in antiquity see Hornblower [20124] with bibl., on the history of the
concept Weller [2009] 14861500, with bibl.
Poetics and Literary Criticism 717
Chronology
In spite of pre-Hellenistic efforts to get to grips with chronological questions
(for literary topics see esp. Glaucus of Rhegium frr. 13 Lanata), the subject
was put on a scientific footing by Eratosthenes of Cyrene. Since he developed
a comprehensive chronology (built on the reckoning by Olympiads, which
proved to be very influential), it also included literary subjects (e.g. the dates
of authors and their works: *biographical criticism, *plot summary), although
this cannot have been his main goal. The chronology designed by Apollodorus
of Athens later superseded Eratosthenes, on which it was built.46
43 An exception is Philodemus (Pom. 4, cols. 106112 Janko), who attacked the Aristotelian
concept of character(isation) in general. Whether his target here is Aristotle himself or
his school is disputed (most recently Janko [2011] 220221, with bibl.).
44 Appropriateness e.g. sch. A Il. 1.133134 Ariston.; consistency: Nnlist [2009a] 249250,
with n. 42.
45 On Aristotles concept of character see Halliwell [1986] esp. ch. 5, for practical examples
Schironi [2009b] 290297, Nnlist [2009a] 246256, with bibl. in n. 29, also index s.vv.
character, characterisation.
46 Pfeiffer [1968] 163164, 255; for Eratosthenes of Cyrene and Apollodorus of Athens see
also Montana in this volume.
718 Nnlist
Clarity (saphneia)
Clarity (or lucidity) is arguably the most important virtue of *style for Aristotle
(Rh. 1404b12, cf. Poet. 1458a1819). Through the systematisation of his pupil
Theophrastus (virtues of *style), it had an enormous and lasting influence on
*rhetoric, including rhetorical criticism. Moreover, a poetological image such
as the pure and undefiled water from the holy spring (Callim. Hymn. 2.111112)
is likely to address the quality of lucidity too. In any case, Callimachus sharply
criticised Antimachus Lyde for its lack thereof (fr. 398 Pfeiffer). If clarity was
considered a virtue, it was only to be expected that obscurity (the trademark
of the philosopher Heraclitus) would be considered a vice.47 The clarity of the
text played a role in criticism at a more elementary level too. The insertion
of lectional signs (not just breathings and accents but also punctuation, divi-
sion of speakers and different *metres, etc.) made the text more user-friendly.
When discussing the pros and cons, for example, of different punctuations,
critics regularly preferred the one that made the text clearer (saphesteron).48
Classifications
The attempt to carve up the entire body of texts by means of classifications
is a recurrent phenomenon of literary criticism. Arguably the most common
and influential of them was the distinction between poetry and *prose. The
decisive criterion was *metre, in spite of the objections raised by Aristotle
(Poet. 1447b1323), who himself (1448a1924) reformulated Platos differentia-
tion between (a) pure narrative, (b) a mixture of narrative and speech (as in
Homer) and (c) pure speech (as in tragedy and comedy).49 The model as such
had a considerable echo, but a problem was to find actual examples of (a) pure
narrative. Consequently, the classification sometimes had two positions only.50
The bipartite model also had the advantage of squaring nicely with the other
47 Heraclitus 22 A 1a, 3a, 4, B 10 D-K; obscurity to be avoided: e.g. sch. A Il. 14.169a Ariston.
48 In this form the saphesteron argument (sch. A Il. 6.6869 Nic., etc.) is typical of Nicanor
(second cent. AD), but the idea is considerably older. Cf. Aristotles general discussion of
punctuation, the specific example being Heraclitus (Rh. 1407b1118 = Heraclitus 22 A4
D-K). On the history of clarity as a critical concept see Asmuth [2003] 814874, with bibl.,
on obscurity Walde (et al.) [2003] 358383, with bibl., for practical examples see Meijering
[1987] 224225, Nnlist [2009a] index s.v. clarity.
49 Halliwell [2009] has recently argued that Plato did not actually introduce such a tripartite
model in Republic 392c398b, but he agrees that this is how it was understood by subse-
quent critics from Aristotle onwards.
50 Nnlist [2009a] 94102, with bibl. in n. 2, add Janko [1984] 128130, Halliwell [2009], other
classifications [2009a] 109115.
Poetics and Literary Criticism 719
classification that Aristotle himself had used in the same context: *genre. A
rather different bipartition was the distinction between form and content,
usually said to originate with Plato (*formalism). It was taken up, for instance,
by the Stoic (Aristo of Chios?) who differentiated between thought (dianoia)
and composition (sunthesis). Perhaps this was a reaction against Neoptolemus
of Parion and his tripartite model poisis-poima-poits, the details of which
are not yet fully understood.51 See also *music.
Coherence
An axiom of the Poetics is the organic *unity of the *plot. Aristotle thus sharp-
ened the awareness of the fact that the individual parts of a literary text ought
to form a coherent body. The exact details of such an analysis were left for oth-
ers to develop (cf. *formalism). One result was a collection of terms and con-
cepts that allowed critics to examine and describe the narrative coherence of a
literary text. They show more than superficial similarity to the findings of mod-
ern narratology (e.g. proanaphnsis foreshadowing or prolepsis).52 Aristotle
also stipulated that the individual elements of the *plot should not follow
each other more or less randomly but according to the principle of cause and
effect (*plausibility). This is reflected in the countless notes that explain how
a particular scene motivates another. According to Aristotle, the coherence
of a tragedy ought to be more rigid, whereas epic poems, given their greater
size, are allowed to incorporate more episodic material (Poet. 1459b2831).
A similar distinction recurs in the comments which judge the coherence of
a literary text depending on its *genre. More surprising is perhaps the point
that the coherence could stretch over multiple texts. Aristarchus conviction
that the Iliad and Odyssey were the product of a single author induced him,
among other things, to detect a narrative coherence that linked the two epics
together.53
51 The models of Neoptolemus and Aristo are both known via Philodemus, who rejects them
(Pom. book 5, cols. 1320 Mangoni). For a (speculative) explanation of Neoptolemus
model see Gutzwiller [2010] 342346.
52 For proanaphnsis see Nnlist [2009a] 3442, with bibl. in nn. 36 and 38. On narratologi-
cal questions in general see Nnlist [2009b].
53 Nnlist [2009a] 3334, for more practical examples see index s.vv. (narrative) coherence,
motivation.
720 Nnlist
Comparison
To compare one author (or text) with another is likely to be an activity in
which readers of all times naturally engage. The first extant example of some
length and depth is the agn between Aeschylus and Euripides in the sec-
54 Pfeiffer [1968] 161; on commentaries see also 29, 175, 250, 276277, and on Aristarchus spe-
cifically 212225. On the history of the hupomnma in general see Eichele [1998] 122128,
with bibl. and Dubischar in this volume.
55 Brief summary of the reconstruction process in Dickey [2007] 19, and in this volume. For
glossaries of ancient grammatical terminology see Dickey [2007] 219265, for literary
criticism Nnlist [2009a] 368386.
Poetics and Literary Criticism 721
ond half of Aristophanes Frogs. It prepared the ground for all subsequent
comparisons between two or more authors (Alcidamas, Contest of Homer and
Hesiod;56 Dionysius of Halicarnassus, passim; Dio Chrysostom, Philoctetes
Bow; Plutarch, Comparison of Aristophanes and Menander, etc.). In addition to
these large-scale comparisons, there were countless smaller observations and
notes that compared authors in every conceivable respect. Aristarchus specifi-
cally differentiated between Homer and the neteroi (lit. younger), that is, all
the authors who postdated him (*anachronism).57
Conciseness (suntomia)
Conciseness was considered one of the virtues of *style. Often said to be Stoic,
the concept is actually found already in Isocrates (Artium Scriptores B XXIV 34 =
Quint. Inst. 4.2.31), where it formed a triad together with *clarity and *persua-
siveness. The ability to express something succinctly was a source of praise
from critics. Conversely, if they perceived unnecessary verbosity or bombast,
critics could reprimand the author, declare the incriminated line(s) superflu-
ous (perissos), etc.58 One might compare Callimachus slender Muse (fr. 1.23
24 Pfeiffer).
Consistency
One of Aristotles four critieria of convincing *characterisation was consis-
tency (Poet. 1454a2628), which was taken up by Hellenistic critics. A charac-
ter ought to act and speak in a consistent way throughout the entire play or
epic poem. The idea was not so much that characterisation must be done with
utmost rigour (*sense of proportion). Rather the various actions of a character
must be compatible with each other so as to provide a consistent and coher-
ent picture. Unwanted were utterances or actions that were patently out of
character.59 The principle of consistency clearly resembles that of narrative
*coherence, a prerequisite of the *unity of the *plot. It was thus only natural
56 The form of the Contest that has reached posterity is an amalgam. Its core, later expanded
by Hellenistic material, is generally agreed to go back to Alcidamas, who in turn may
depend on precursors (OSullivan [1992] esp. ch. 3). On literary contests in general see
Neumann [1992] 261285, Paulsen [2002], each with bibl.
57 On the history of comparison as a critical concept see Kneepkens [1994] 293299, with
bibl.
58 On the history of conciseness as a critical concept see Kallendorf [1994a] 5360, with bibl.,
for practical examples Meijering [1987] 147148, Nnlist [2009a] index s.v. conciseness.
59 Examples of praise and critique are given in Nnlist [2009a] 249252, with bibl. in nn. 39,
43, 47.
722 Nnlist
for critics to search for consistency when they analysed the *style or subject-
matter of individual texts and authors and to criticise its absence.
Contradiction
Thorough *analysis or downright *reading against the grain could also bring to
light that a passage contradicted another or was in conflict with an extratextual
fact (e.g. Aristotle, Poet. 1461b1518, cf. also 1455a2226). If the observed contra-
diction or inconsistency turned out to be real, critics entertained, among other
things, the possibility of a textual corruption (questions of *authenticity). It is,
however, remarkable how often they argued that a contradiction was not real
but apparent only (*problems and solutions).60
Critical Signs
The Hellenistic scholars put critical signs in the margins of their texts. Several
signs referred to questions of textual criticism but not all: Aristophanes
of Byzantium developed an elaborate method of marking metrical units
(*metre). Aristarchus used the dipl (>) to mark passages that were notewor-
thy and therefore discussed in the *commentary. These discussions regularly
addressed questions of literary criticism.61
Custom (ethos)
The systematic and comprehensive *analysis of an authors oeuvre allowed crit-
ics to identify features, literary techniques etc. that this particular author cus-
tomarily made use of.62 This type of research regularly included the question
whether or not there were exceptions (see also *contradiction). An answer in
the negative could lead to editorial interventions (questions of *authenticity).
Educative Function
With the exception of comparatively few critics (e.g. Eratosthenes, for whose
provocative denial see *emotions), there was widespread consensus among
readers and critics alike (formulated with different degrees of explicitness)
that one of poetrys primary functions was to educate the reader. This gen-
eral statement can be underpinned with observations of a more practical kind.
Emotions
(Early) Greek poetry regularly described the emotional impact it expected
to have on the audience (e.g. Od. 13.12). But this very emotive power was
among the most important objections raised by Plato against poetry (Resp.
602c607a). Aristotle, in turn, regarded the generation of emotions as a funda-
mental principle of tragedy and thus art in general. In this he was preceded by
Gorgias.67 Aristotles much-discussed definition of tragedy (Poet. 1449b2428)
circles around, specifically, pity (eleos) and fear (phobos). These emotions are
also the means by which tragedy produces the type of *pleasure (hdon) that
is conform to its nature (1453b1013), as does epic (1459a1821). Furthermore,
good literature generates wonder (to thaumaston, 1452a4) and amazement
(ekplxis, 1455a17). The latter might even help the author veil a passage that
Emulation
The imitation or emulation of the great masters of poetry such as Homer can
be seen at work throughout the entire history of Greek literature. Theoretical
statements, however, are not so easy to come by. They were written with a view
to the practical needs of an orator (e.g. Isoc. 2.41) and/or they postdate the
Hellenistic period (e.g. Horace and Ps.-Longinus).70 Neither qualification must
be decisive. Rhetoric and literary criticism often went hand in hand; criticism
of the Roman period regularly reflected Hellenistic views (leaving aside that
Horace urges the Romans to study the Greek models). The following paradox
is nevertheless worth pointing out. The same period which, by means of the
*canonisation process, was largely responsible for the concept of classical
literature appears to have been reluctant to express the view that the same
classical authors ought to be studied and emulated with particular attention.
Was the point too obvious to be made? Emulation did of course not mean
slavish adherence but a productive and original handling of the great models
(literary *tradition).71
Enargeia
Literature is primarily an aural type of art (e.g. Pl. Resp. 603b78; *euphonist
theory). This holds particularly true for the members of a society who were
accustomed to read aloud or have literature read to them.72 But it was equally
customary to compare literature with *visual forms of art. Such comparisons
implicitly underlay the numerous comments that highlight the visual quality of
(aural) literature. Arguably the most important term here was enargeia, usually
rendered in English with vividness, which, however, fails to express its decid-
edly visual quality (better captured, for instance, by German Anschaulichkeit).
A passage had enargeia when the wording was so gripping that the readers
could see (*visualisation) what was actually reaching them through their ears.
According to Plutarch (De glor. Ath. 347a), Thucydides tried hard to turn his
readers (i.e. hearers) into spectators. In Aristotles words (Rh. 1411b2426), the
author put (by means of *metaphor) before the readers eyes things which thus
had activity. By being brought before the readers eyes, these things, actions
etc. achieved actuality. The Greek word for actuality is energeia. Even though
Aristotle did not say it in so many words, in a sense he established an immedi-
ate connection between enargeia and energeia. This was bound to lead to trou-
ble. The regular confusion of the two terms in our manuscripts often makes
it difficult to decide which of them is actually at stake. Another point of con-
tact is less problematic. The same passage could be singled out for its *clarity
or enargeia. Likewise, praise for enargeia could be uttered when the relevant
scene seemed so real as if it were taken from real life (lifelike *realism).73
Ethics
See *appropriateness, *censorship, *educative function, *emotion, *fiction,
poetic *licence, *mimsis, *multiple meanings.
71 On originality within a traditional setting see Brink [1971] 208209. Ancient critics would
have been surprised by the modern craze about artistic originality. On the history of emu-
lation as a critical concept see Bauer [1992] 141187, Conte-Most (20124b), each with bibl.
72 The ongoing controversy around the beginnings of silent reading does not alter the fact
that to read aloud was in any case the rule (e.g. Johnson [2010]).
73 On the history of enargeia as a critical concept see Kemmann [1996] 3347, with bibl., for
practical examples Nnlist [2009a] 194198, with bibl. in n. 4.
726 Nnlist
Etymology
An interest in the etymology of words and, in particular, names was pervasive
among the Greeks. Homer and other poets subtly etymologised in their texts,
for instance, when Hectors name was derived from the verb echein (to hold,
Il. 5.472473), which made him the holder of the city of Troy. Similar expla-
nations could be found in philosophical and historical literature of the sixth
and fifth centuries, until the topic was examined with unprecedented zeal in
Platos Cratylus.74 Criticism of the Hellenistic and later periods, too, regularly
argued on the basis of etymological considerations. There was a widespread
consensus that etymology promised to lead to deeper insights into the word
or name under consideration. The principles that governed the practical appli-
cation of etymology were less rigid than in modern scholarship, where it has
become a branch of historical linguistics with its regular sound shifts, etc. A
recurrent characteristic is apt to illustrate the difference: some critics (Plato
included: e.g. Cra. 405a406a) saw no difficulty in offering multiple etymolo-
gies which for them were not mutually exclusive. This did not mean, however,
that an etymology could not be rejected by others as false (e.g. sch. A Il. 1.105a
Ariston.). In any case, etymology was a commonly practised reading strategy,
which justifies its inclusion here.75
Euphonist Theory
Poets in general and the composers of oral poetry in particular must always
have been aware of its sound effects. In *prose one might compare the wide-
spread avoidance of hiatus. Lasus of Hermione (fr. 704 Page) is known to have
written an entire poem without the letter sigma, no doubt because he did not
like its sound. Democritus (68 B 18a/b D-K) apparently wrote on euphony. The
*etymological explanations of Platos Cratylus explained several sound shifts
as being due to euphony.76 The ground was thus prepared for the striking nov-
elty of the so-called euphonists, whose theories must be wrestled primarily
from Philodemus works (see introduction above). Taking as their starting-
point the Platonic distinction between form and content (*formalism), the
euphonists felt that it unfairly favoured the content of literature. They there-
fore attempted to shift the balance. In so doing their main focus was on the
phonetic side of form specifically, which, among other things, made the ear an
important judge of poetic excellence. Some euphonists even went so far as to
argue that sound was actually more important than content. It is difficult to say
74 Pfeiffer [1968] 4, 12, 40, 612, see also Novokhatko in this volume.
75 On ancient etymology see the papers collected in Nifadopoulos [2003a].
76 Pfeiffer [1968] 64.
Poetics and Literary Criticism 727
whether euphonist theory, especially in its extreme form, was shared by many
outside their own circles (leaving aside the fact that they by no means agreed
among themselves). At any rate, Philodemus thought it important enough to
devote books 1 and 2 of On Poems to an elaborate description and rejection, in
which he insisted on the unity of form and content. Pieces of euphonist theory
popped up here and there. Aristarchus, for example, rejected Zenodotus text in
Iliad 6.34 with explicit reference to the phonetic effect (sch. A Il. 6.34 Ariston.).
Moreover, a substantial part of On Composition by Dionysius of Halicarnassus
provided a minute analysis of individual sounds. He also continued using the
euphonists testing method known as transposition (of words) (metathesis;
*word order). All in all, one may conclude that, in its extreme form, euphon-
ist theory remained on the margins of literary criticism. But the general topic
sound effects as such was far from being ignored by critics.77
Fiction
The word fiction is Latin (fingere). When Xenophanes criticised the poets, he
spoke of their fabrications (plasmata, 21 B 1.22 D-K = 1.22 West). His objections
were ethically motivated. Since all the Greeks without exception learned from
Homer (B 10, *educative function), it was unacceptable that he told stories
which were both morally objectionable and untrue. In his fundamental cri-
tique of *mimetic art Plato developed this idea further, in that he held poetry
against the criterion of (philosophical) truth or reality. The result was strictly
negative: poetry was a lie (pseudos) and thus to be banned (*censorship). In
his defence, Aristotle not only argued against Plato that poetry must be mea-
sured by its own standards (poetic *licence), he also made the much-quoted
point that Homer has taught the other poets how one must lie (pseud legein,
Poet. 1460a1819). Aristotle, however, did not mean to say that pseudos was
simply to be equated with poetry tout court. As the context of the quotation
reveals, the point was that a clever juxtaposition of true and untrue passages
allowed poets to get away with the latter. Eratosthenes, for his part, in reject-
ing the widespread view that literature had an educative function (*emotions)
allowed poets to resort to fiction (plattein) whenever it helped them achieve
their goal (fr. I A 19 Berger). He also exposed the excessive literal-mindedness
of those readers who tried to map Odysseus wanderings and cheekily told
them that they would succeed as soon as they found the cobbler who stitched
77 The point of reference for euphonist theory is Janko [2000], [2011]. His contention [2011]
229, however, that euphonic analysis had become at least as important (sc. as the
Aristotelian type of analysis) seems to me an exaggeration. On the history of euphony as
a critical concept see Umbach (et al.) [1996] 1022, with bibl.
728 Nnlist
together Aeolus bag of winds (fr. I A 16 Berger). The notion that Odysseus
wanderings take place in fictional space, though rejected by Polybius (34.2.4
4.8), Strabo (1.2.15) and others, found the approval of no lesser critic than
Aristarchus.78
Formalism
Plato is often credited with the differentiation between form and content in
poetry (e.g. Resp. 601b24, with Murrays note). As for Aristotle, the view that
in his Poetics he expounds a strictly formalist notion of unity as independent
of poetic meaning is better rejected.80 It is, however, true that the Poetics
paved the way for formalist approaches, which were further developed and
refined by subsequent critics (narrative *coherence, *narrative voice, *plot).
Aristotle himself built on a generic distinction that he had found in Plato
(*classification).
78 Lehrs [18823] 241245. On the history of fiction as a critical concept see Zinsmaier [1996]
342347, with bibl., for practical examples Nnlist [2009a] 174184.
79 On the history of figures of speech as a critical concept see Knape [1996] 289342, with
bibl., on the terminological confusion Schrader [1904].
80 Halliwell [1989] 156, with ref. to 1451a3032. This is not the place to discuss the ques-
tion whether there actually are formalist analyses that operate independent of poetic
meaning.
Poetics and Literary Criticism 729
Formulaic Language
No reader of Homer will miss the regular repetition of lines. The Alexandrian
critics consistently marked them in their texts (*critical signs). Since this rep-
etition clashed with the principle of *variety, which was particularly strong in
aesthetics of the Hellenistic period, critics cast doubts on the *authenticity of
repeated lines. Contrary to a widespread modern view, however, they made
no attempt to eliminate them systematically.81 In particular cases, for instance
when the messenger verbatim repeated the original instruction, the repetition
was even explicitly defended by critics.82 It is true, though, that formulaic rep-
etition was generally regarded as typical of the inferior poets of the epic cycle
(e.g. sch. A Il. 9.222a Ariston., cf. Callim. Epigr. 28 Pfeiffer; *canons). As Parry
[1971] 120124 himself acknowledged, Aristarchus essentially recognised the
nature of generic epithets. He also described two fundamental principles of
Homers type scenes: the individual components can be expanded or short-
ened, but their sequence must not be altered (for an example see *Realien). In
spite of these forays into what is now considered the territory of oral poetry,
critics seem to have taken it for granted that Homer had written his epics.83 In
Aristarchus case, this is at least worth mentioning because he insisted that
Homers characters did not write (*anachronism).
Genre
An interest in and the identification of literary genres clearly predate the
Hellenistic period. A good example is the list of genres in Plato (Leg. 700b),
not least because it gave prominence to lyric genres, a topic that Aristotle
rarely touched upon in his Poetics, which focused on defining and describ-
ing epic and drama.84 Callimachus *Pinakes were arranged by genre. And
so were the editions that the Alexandrians produced (literary *heritage), for
instance, the edition of Pindars works in seventeen books, which displayed
ten lyric sub-genres, as arranged by Aristophanes of Byzantium. Overall there
were multiple, mostly formal criteria for the assignment to a particular genre:
mode of performance (e.g. acting vs. narrating), *metre and *music, occasion,
Heritage, Literary
The express goal of the *Museum and library was to create a treasure-house
of the entire literary (and scientific) heritage that fulfilled certain quality stan-
dards.86 It was thus necessary (a) to collect and/or copy all the relevant works,
(b) to get or produce the best possible text for each (questions of *authentic-
ity), (c) to make them accessible, literally, by means of catalogues and other
means of support (*Pinakes), (d) figuratively, by writing exegetical works such
as *commentaries, *monographs or glossaries. The latter two had already
been in existence in pre-Hellenistic times, but the new emphasis on user-
friendliness can be deduced, among other things, from the forward-looking
invention to arrange glossaries in *alphabetical order. The same holds true for
the *critical signs.
Historical Perspective
See *anachronism, also *educative function.
Imagery
See *metaphor, *simile.
85 On the history of genre as a critical concept see Komfort-Hein Knoblauch [1996] 528
564, Conte-Most [20124a], each with bibl.; also Brink [1971] 160163.
86 Pfeiffer [1968] passim; see also Montana in this volume.
Poetics and Literary Criticism 731
Alexandrian exegesis of Homer in quite some detail.87 These are the immedi-
ate areas of influence. In addition, influential texts such as Horaces Ars Poetica
helped channel Hellenistic ideas through the (largely Greekless) Middle Ages,
until Greek studies were renewed in the Renaissance, which handed them
down to the modern era.
Inspiration
Hesiods meeting with the Muses on Mt. Helicon (Theog. 2234) is the earliest
known example in a long row of Dichterweihen. Theorising about divine inspi-
ration (also evident in the concept of the poet as seer, prophet, etc.) seems to
have begun with Democritus (68 B 1718, 21, 112 D-K) and Plato (Phdr. 245a,
Ion). They both regarded poetic inspiration as a form of divine possession
(enthousiasmos) or madness (mania). For Plato this was an ambivalent thing
because it was incompatible with true understanding. The poet might be pro-
ficient in the art (techn) of writing poetry but he could not succeed without
this madness. This last point was to become the basis for the age-long opposi-
tion between ars and ingenium (e.g. Hor. Ars P. 409411, Ps.-Long. Subl. 2, who
both argued for a fruitful combination of the two). Democritus saw things dif-
ferently. Not only did divine possession lead to results that were extremely
beautiful (kala karta, B 18), he also praised Homer because being endowed
with a nature open to divine influence he built a fair structure of all kinds of
words (B 21). The ars-ingenium controversy could also take the form of the
question whether a poet should drink wine or water.88
Intent, Authorial
There was a widespread, if largely implicit, consensus among critics that there
was such a thing as authorial intent and that it was the readers task to try to
grasp it. Ancient criticism was unaffected by the discussion around the inten-
tional fallacy which arose in the middle of the twentieth century. Even crit-
ics who advocated a form of *reading against the grain or between the lines
(*multiple meanings) essentially subscribed to the idea of authorial intent.
This additional meaning was there because the author put it there. The con-
cept of authorial intent was particularly strong in rhetorical criticism, which
saw the author in complete and conscious control of his material. This also
holds true for Aristotles Rhetoric, which makes it even more remarkable that
in his Poetics the notion of authorial intent is hardly ever called into play.89
Interpretation, textimmanente
The notion that each author was his own best interpreter with the consequence
that interpretation should be textimmanent is likely to be Aristarchus best-
known tenet (usually in form of the phrase to elucidate Homer from Homer).90
It is, however, important to keep in mind what Aristarchus point was. A fun-
damental concern of his was to avoid the various forms of *anachronism that
threatened an appropriate interpretation. In order to do so, it was best to anal-
yse first and above all how the relevant text itself dealt with the question under
consideration. This approach enabled the critic, for example, to establish the
meaning of a word in the Homeric epics (as opposed to later authors and/or
Hellenistic usage), to differentiate between variants of a myth (*mythology),
to reconstruct the specifically Homeric Weltbild, etc. Aristarchus, however, did
not advocate the view that the critic must completely ignore other relevant
data (*sense of proportion). First, his form of textimmanente interpretation
entailed a great deal of *comparison with other sources (primarily, but not
exclusively, texts, see *mythology): Homers *custom was regularly set off
against that of other authors. Second, Aristarchus was not opposed to using
other data, especially when Homer himself did not present an unambiguous
picture. For instance, he uses Hesiod as a witness (martus) for the physical
shape of the Molione (sch. A Il. 23.638642 Ariston.). The same poet is quoted
as an authority on Ascra in order to reject Zenodotus text (sch. A Il. 2.507a
Ariston.). Others notes support the point made with references to Stesichorus
(sch. A Il. 5.31d Ariston.) or Euripides (sch. A Il. 2.353a Ariston.). Likewise,
the explanation of certain *Realien (e.g. geography) was impossible without
recourse to other sources. Finally, the flexibility of Aristarchus approach also
transpired from the fact that he admitted exceptions, especially in the form of
things that occurred only once (hapax legomena, which was not an exclusively
lexicographic category).
89 On the history of authorial intent as a critical concept see Bernecker [1998] 451459, with
bibl.
90 The bibliography on the subject (see Montanari [1997o] 285286, with bibl. in n. 20) is
primarily concerned with the question whether or not the phrase Homron ex Homrou
saphnizein represents Aristarchus own wording. The point of the present entry is argued
at greater length in Nnlist [forthc.].
Poetics and Literary Criticism 733
Licence, Poetic
In his Poetics, Aristotle defended poetry against the fundamental objections
raised by Plato against all forms of art (*mimsis), among other things, by
holding that the content and meaning of mimetic works cannot justifiably be
tested against any fixed criterion of truth or reality (Halliwell [1989] 153) but
must be measured by its own standards (1460b815). He thus made a claim
on a large scale that on a smaller scale came to be known as poetic licence.
The main points, though not the term, were expounded already by Isocrates
(Euag. 910): unlike a historian or an orator, a poet was free to choose at will
both linguistic means and subject-matter, including fictional or even fantastic
story elements. The term poetic licence as such was regularly used in a defen-
sive manner, in order to reject objections raised by other critics. Aristarchus,
for instance, repeatedly defended Homer against the internal *contradictions
that quibblers had found. In addition, he and others granted poets a fair dose
of *fiction.91
Literary Dependence
The careful *analysis and *comparison of literary texts could also lead to the
insight that a passage directly depended on another (cf. already Glaucus of
Rhegium fr. 7 Lanata). Aristarchus *commentaries regularly identified such
cases. For instance, Hesiods story about Pandoras box (pithos) depended on
the Iliadic passage where Achilles speaks about Zeus jars (pithoi), which con-
tain good and bad lots.92 Likewise, Pindars unusual metaphor white breast
(mastos, of a hill, P. 4.8) was a productive adaptation of Homers udder of the
soil (outhar arours, Il. 9.114).93 The second example shows that scholars also
commented on the specific relationship between the two passages: quotation,
imitation, adaptation, *allusion, parody, etc.
Literary History
In chapters 4 to 5 of his Poetics, Aristotle gave a description of how trag-
edy and comedy had grown from other *genres and thus reached their
perfect form. The whole account is deeply rooted in his teleological philosophy
(cf. *unity). This is one reason why it proves so difficult to match it with actual
literary history (Halliwell [1986] chs. 3 and 9), leaving aside that, according to
91 On the history of (poetic) licence as a critical concept see Schmude [2001a] 253258, with
bibl., for practical examples see Nnlist [2009a] 174184, with bibl. in n. 1.
92 Sch. A Il. 24.527528a Ariston. with respect to Hesiod Works and Days 84104; Aristarchus
comment is triggered by the number of jars (Lehrs [18823] 189).
93 Sch. Pind. Pyth. 4.14.
734 Nnlist
Metaphor
It is no exaggeration to say that, to this day, metaphor has been a pet child of
literary criticism and the central *figure of speech.95 Aristotle set the point
when he declared metaphora (and foreign/strange words, glttai) typical
of poetic diction (Poet. 1458a2123) and the only thing you could not learn
from others (1459a67). Two points must, however, be borne in mind. First,
Aristotles term metaphora is roughly equivalent to figurative language and in
any case encompasses more than metaphor does. Second, Aristotle did not
advocate the view that metaphors were exclusive to poetry. Everybody uses
them (Rh. 1404b3435). He also put forward a fourfold definition of metaphor
(Poet. 1457b69) as a transfer (epiphora) (a) from genus (genos) to species
(eidos), (b) from species to genus, (c) from species to species, (d) by analogy,
each with examples. Moreover, he demonstrated that metaphors filled gaps in
the vocabulary, when there was no proper term yet for a particular phenom-
enon (1457b2532). He thus described the figure called katachrsis. The term
is not attested in his extant writings, but Cicero (Orat. 27.94) expressly refers it
to him. In any case, the term is common among Hellenistic scholars, who also
took up his fourfold definition, though the first two positions were thought to
be part of metonymy (or synecdoche), not metaphor. Another fourfold model,
built on the four possible combinations of animate (empsuchos) and inani-
94 On the Hellenistic origins see Nesselrath [1990], summary on pp. 186187.
95 On the history of metaphor as a critical concept see Eggs [2001] 10991183, with bibl.,
who, in particular, rejects (1103) the common view that Aristotle subscribed to the idea of
substitution (e.g. Silk [2003] 118).
Poetics and Literary Criticism 735
Metre
In spite of Aristotles objections (Poet. 1447b1323), metre remained the deci-
sive factor for the differentiation between poetry and *prose (*classifications).
The specific metre also helped answer the question to which *genre a poem
belonged. This applied in particular to lyric poetry (broadly understood) with
its wealth of subgenres (hymns, paeans, dithyrambs, maiden songs, encomia,
dirges, victory odes, etc.). This division also determined the editorial prac-
tice, in that the Alexandrian editions of lyric poets were organised by metre,
as were Callimachus *Pinakes. Furthermore, metre also helped differentiate
the various parts within a genre (e.g. choral odes and speech in drama). The
champion of metrical studies was Aristophanes of Byzantium, who marked
the ends of the metrical cola of lyric poetry with *critical signs. The separa-
tion of these cola, however, long believed to be his invention, apparently pre-
dates him, as the Lille papyrus of Stesichorus indicates (3rd c. BC, = fr. 222b
PMGF). Aristophanes work on the triadic structure of choral odes also helped
him recognise the principle of metrical responsion between the correspond-
ing parts, which had an immediate impact on textual criticism.97 Aristarchus
observed the rule that nowadays is called Hermanns Bridge.98 In addition to
developing an elaborate taxonomy, critics also discussed metrical anomalies
(e.g. mousetailed hexameters), collected examples of lines that were com-
paratively rare (e.g. purely spondaic hexameters) or otherwise remarkable (e.g.
three-word hexameters).99
Mimsis
Plato and Aristotle agreed that the essence of art was mimsis, though the term
did not denote exactly the same thing. For Plato it was a form of imitation and
he therefore rejected it (*censorship). The carpenter imitated the ideal form
of the table, while the artist imitated what was already an imitation, which
removed him even further from the ideal form. Moreover, Plato did not like the
96 This hypothesis (Lausberg [20084] 286) gains support from the fact that essentially the
same model is found in Aristarchus *commentaries (cf. esp. sch. A Il. 11.574a Ariston.) and
thus existed before Trypho.
97 Pfeiffer [1968] 185188.
98 Pfeiffer [1968] 229.
99 On the history of metre as a critical concept see Schmude [2001b] 12231232, with bibl., on
metrically remarkable lines Richardson [1980] 286287.
736 Nnlist
Monograph (sungramma)
The monograph was the other preferred format of Hellenistic criticism (cf.
*commentary).101 To treat a particular subject in the form of a monograph
had a long tradition which reached back into the early days of Greek prose. As
regards the primary subject of this chapter, the modern reader faces the source
problem described in the introductory section above. The Hellenistic critics
are known to have written countless monographs on various topics (*nick-
names), including literary criticism. Apollonius of Rhodes, for instance, wrote
a treatise On Archilochus, a comparatively early example of the peri-literature.102
None of these monographs, however, has been transmitted in its entirety.
The treatises that are still extant either predate or, more often, postdate the
Hellenistic period.103 In addition to reflections in these later sources, the con-
tent of the Hellenistic treatises themselves must be reconstructed on the basis
100 On the history of mimsis as a critical concept see Eusterschulte-Guthknecht [2001] 1232
1327, with bibl., Halliwell [2002].
101 Pfeiffer [1968] passim.
102 See Dubischar in this volume.
103 Pre-Hellenistic: e.g. Aristotle, Anaximenes of Lampsacus; post-Hellenistic: Dionysius of
Halicarnassus, Horace, Ps.-Longinus, Heraclitus (the allegorist), Cornutus, etc. The only
possible exception is On Style (peri hermneias, mistakenly attributed to Demetrius of
Phaleron in a tenth-century manuscript). But its date remains elusive (for a good sum-
mary of the principal questions see Innes [1995] 312321).
Poetics and Literary Criticism 737
of the mere title and/or fragments, usually small in number. This source prob-
lem seriously limits the quantity and reliability of the results.
Multiple Meanings
The term underlying meaning (huponoia) is indicative of the recognition
that, in addition to the literal meaning on the surface of the text, there was
a second level of meaning, for instance, an implied message.104 Arguably the
most important application of the term huponoia was in allegorical interpre-
tation.105 Said to have begun in the sixth century BC by Theagenes of Rhegium
(8 A 2 D-K), allegorical interpretation was one of the forms of defence that
the ethically motivated attacks on poetry triggered (*educative function). As
its proponents argued, it was the underlying second meaning of a text that
really mattered. The search for the huponoia was explicitly rejected as a suit-
able reading strategy by Plato (Resp. 378d, cf. Plut. Quomodo adul. 19e) and did
not engage Aristotles sympathy. The oldest extant example of some length is
the Derveni papyrus, which preserves substantial fragments of an allegorical
commentary on an Orphic theogony.106 Regarding the Hellenistic period spe-
cifically, the prevalent view among modern scholars is that there was a sharp
contrast between Alexandria on the one hand and Pergamon on the other:
Aristarchus strongly objected to a method that was practised by his Pergamene
rival Crates of Mallos. The picture may not be as clear-cut as that.107 In any
case, Alexandrian criticism paid little tribute to allegorical interpretation. The
following questions, however, are more difficult to answer and thus the subject
of ongoing scholarly debates: How important or widespread was allegorical
interpretation during the Hellenistic period (sc. outside of Alexandria)? How
narrow was its connection to Stoicism? Should it actually be regarded as a form
of literary criticism?108 Another type of multiple meanings was recognised in
what today would be called dramatic irony: the text had a second meaning,
104 The earliest attestation of underlying meaning probably is Xenophon Symp. 3.6; for
implied message see Aristarchus in sch. Pind. Isthm. 7.23a.
105 On the history of allegorical interpretation see Freytag [1992] 330393, Konstan [2005],
Trapp [20124], each with bibl.
106 For text and commentary see Tsantsanoglou-Parassoglou-Kouremenos [2006], Betegh
[2004]; on literary criticism specifically Henry [1986].
107 Aristarchus alleged anti-allegorism hinges on a single witness (sch. D Il. 5.385) which is
open to a much broader interpretation (Nnlist [2011]). Besides, Crates practised a very
peculiar and idiosyncratic form of allegorical interpretation (Porter [1992]).
108 These questions are too complex to be discussed in a short overview. For the divergent
views see e.g. Russell [1981] = [1995] 9596, Long [1992], Porter [1992], Dawson [1992],
Struck [2004], Konstan [2005] xiiixxvii, Gutzwiller [2010] 354359, Nnlist [2011].
738 Nnlist
which undermined the first and was understood by the audience but not the
relevant character(s).109
Music
A substantial portion of Greek poetry was not only accompanied by musical
instruments but either sung or chanted, with words and music forming a unity.
In the course of the fourth century, however, this unity was jeopardised when
some critics started focusing on the language alone, while Aristoxenus devel-
oped musicology as a discipline. Separation and specialisation may or may
not be responsible for the relative neglect of music among Hellenistic critics,
who would have been in a position to save this part of their *heritage too. And
although scholars like Aristophanes of Byzantium carefully studied the intri-
cate rules of *metre and his successor Apollonius (called the Classifier, eido-
graphos) assigned lyric poems to different musical classes (*classification),
music itself did not enter the mainstream of scholarly activity and was conse-
quently lost for ever.111
Mythology
Given that the subject-matter of much Greek poetry had its roots in the lit-
erary *tradition, critics took a natural interest in mythology, which for them
was at the same time a form of early history. The interpretation of individual
passages regularly required an explanation of the mythological facts (who is
who? etc.). Starting no later than with Ps.-Platos Minos (318de), it became
Narrative Voice
Due to their careful *analysis critics also identified instances of internal *con-
tradiction. Some of them could be shown to be apparent only. One possible
solution (*problems and solutions), attested from Aristotle onwards (fr. 146
Rose, Poet. 1461a78), was that the divergent passages were not spoken by the
same agent and therefore not real contradictions. The agents could either
be two different characters or a character and the narrator. The distinction
between narrator-text and speech could already be found in Plato (*classifica-
tion), who took exception to the impersonation that speech required (*mim-
sis). Aristarchus recognised that there was more to the distinction between
the voice of the narrator and that of his characters. On the one hand, this dis-
tinction helped him differentiate between their respective Weltbild (*anachro-
nism). On the other, he detected differences of *style and narrative technique
between narrator-text and speech. What is more, the pointed references to the
comments which the Homeric narrator expressed from outside in his own
voice (e.g. sch. A Il. 10.240 Ariston.) show Aristarchus awareness of the general
scarcity of such narratorial comments in the Homeric epics. He thus paved the
way for the Jamesian notion of showing vs. telling, the essence of which can
be found in Ps.-Demetrius On Style 288.114
Nickname
Ancient biographers were fond of anecdotes (*biographical criticism). One
result of this activity was that the biographical tradition reported nicknames
for several scholars too (on Zoilus see *polemics). The polymath Eratosthenes
was called bta (number two) because he was an expert in numerous fields
but the real champion in none. Another nickname, pentathlete, gave expres-
sion to the same idea. Given the large number of books that Didymus alleg-
edly wrote, he must have had brazen guts (chalkenteros). In fact he wrote so
many that he could not remember them all and was thus a bibliolathas (lit.
book-forgetter). Nicanors efforts to punctuate (gr. stizein) texts earned him
the nickname stigmatias, which was the word for a slave who had run away
and therefore been tattooed (also stizein).115 Ancient nicknames did not aim
for fairness either.
Performance
Like their modern successors ancient critics tried to reconstruct the conditions
under which the texts were performed.116 This type of criticism was particu-
larly prominent in commentaries on dramatic texts. As is well known, Aristotle
attributed comparatively little importance to questions of stagecraft (Poet.
1450b1620) and analysed the plays mostly from the point of view of a reader.
His Hellenistic successors paid more attention to this topic and recognised the
need to reconstruct the stage action in order to understand the text. They thus
deduced many stage directions from the bare bones of the text. Less expected
from a modern perspective is the frequency with which critics argued how a
particular utterance must be delivered (angrily, ironically, pleadingly, etc.). For
such instructions included non-dramatic genres and were therefore meant
for the reader (on the ancient habit of reading aloud see *enargeia; on the
importance of the phonetic quality see *euphonist theory; contrast the loss of
*music). The importance that the Hellenistic critics attributed to accentuation
and punctuation must be seen in the same light. Far from being a merely tech-
nical concern of the editor or textual critic, proper accentuation and punctua-
tion made it actually possible to read out the text in a meaningful way. This was
the first step towards a proper understanding of the relevant text.117
Persuasiveness (pithanots)
It is hardly a coincidence that the first two books of Aristotles Rhetoric, which
formed its backbone, dealt with the pisteis (means of persuasion) that an ora-
tor had at his disposal, since persuasion went to the very heart of *rhetoric.
Whether in court or in politics, it was essential to get the jury or the voters on
ones side; hence the continued emphasis on persuasion in rhetoric. For liter-
ary critics, too, it mattered whether or not the text under consideration suc-
ceeded in persuading the reader. The respective notes often went hand in hand
with questions of *plausibility or probability. Since critics did not abhor the
idea of *polemics, their notes could also comment on the question whether or
not the views of their colleagues were persuasive.118
Pinakes (Tables)
The literary *heritage stored up in the library of Alexandria (*Museum and
library) also needed to be organised in a meaningful way. By compiling his
Pinakes Callimachus set a trend. He divided the entire body of Greek literature
into several classes (*classification), some of which coincided with established
literary *genres such as epic, tragedy or comedy. Some categories (esp. lyric)
contained several sub-genres. Within each unit the sequence of authors was
*alphabetical. Each entry contained some biographical information, for which
Callimachus probably drew on Peripatetic sources (*biographical criticism).
Next there was a list of the authors works, perhaps in alphabetical order too. It
identified cases of doubtful *authenticity and marked works that were no lon-
ger extant. Quotation of the opening line (incipit) served as a welcome means
to identify individual texts, not least when they had no title (esp. lyric poems)
or one that was not necessarily authentic. The total length of an authors work
also gave an estimate of its cost (scribes were paid per line). This brief descrip-
tion is enough to show that, in spite of the similarities, the Pinakes should
not be reduced to a mere library catalogue. Callimachus interest in *literary
history can further be documented with another Pinax in which he listed the
dramatic poets in chronological order and from the beginning. The Pinakes
proved to be an important source and a model for future research. They were
later supplemented by Aristophanes of Byzantium (frr. 368369 Slater).119
Plagiarism
A concern for, anachronistically speaking, copyright issues can be traced back
to, for instance, Theognis of Megara and his seal (1923). It is nevertheless
appropriate to say that the foundation of the *Museum increased the aware-
ness of such issues and at the same time made it possible to investigate them
more thoroughly than before (*analysis, *comparison). It is therefore conceiv-
able that Aristophanes of Byzantium (fr. 376 Slater) introduced a novelty when
he compiled a whole list of passages that Menander had borrowed from oth-
ers. A thematically related anecdote (test. 17 Slater) shows that the ancients
stupendous memory still played an important role. It is, however, revealing
that Aristophanes not only caught the contestants in a public literary compe-
tition but also proved their plagiarism afterwards by unrolling innumerable
volumes in the library (Pfeiffer [1968] 191). Unless the anecdote projects cur-
rent (i.e. Augustan) practice back into the past, it vividly illustrates how the
Museum had changed the world.120
119 Pfeiffer [1968] 127134, Blum [1977] passim, summary: 231; on the English translation see
n.33. See also Montana in this volume.
120 On the history of plagiarism as a critical concept see Ackermann [2003] 12231230), Silk
[20124], each with bibl.
Poetics and Literary Criticism 743
solution). This should not conceal the underlying agreement among critics
that literature ought to conform to the relevant principle, which it usually did.121
Pleasure (hdon)
According to Aristotle, both to produce art and to watch works of art is a nat-
ural source of pleasure (*mimsis). The creation of aesthetic pleasure (and
other *emotions) was thus seen as an essential goal of literature, which was
achieved, among other things, by the *unity and *plausibility of the *plot, by
proper *characterisation, etc. Each *genre generated its specific type of plea-
sure (Poet. 1453a3536). While there was almost universal consensus among
critics that literature rightfully generated pleasure, the claim that this was its
primary or even sole purpose was made much less frequently (on Eratosthenes
see *emotions). The mixture, enshrined in Horaces pair prodesse and delectare
(Ars P. 333), is more representative of the ancient outlook (cf. *Lart pour lart).122
Plot
Put on the throne by Aristotle (Poet. esp. chs. 613), plot has been one of the
most durable and influential concepts of criticismancient and modern. As
the first and decisive qualitative part of tragedy (cf. *characterisation), plot
(muthos or sustasis tn pragmatn, construction of events) is the condicio sine
qua non of the Poetics, because it is the most important criterion whether or
not a work of art (exemplified by tragedy) reaches its goal. One might even
say that plot is its goal (1450a2223). The main characteristics of a good plot
are: (a)organic *unity: all parts fit together and form a self-contained whole,
no part needs to be added, none can be transposed or removed; (b) good size:
neither too small nor too big; (c) motivation or narrative *coherence: the indi-
vidual parts do not follow each other randomly but according to probability
or necessity (*plausibility), that is, they observe the principle of cause and
effect; (d) the solution to a tragic conflict should come from the plot itself and
not from outside (e.g. by means of a deus ex machina). The foundation was
thus laid. Rather than discussing Aristotles fairly theoretical model as such,
subsequent criticism concentrated on developing the interpretative tools that
121 On the history of plausibility etc. as critical concepts see Mainberger [1996] 9931000,
Steudel-Gnther [2003] 12821285, van Zantwijk [2009] 12851340, each with bibl., for
practical examples Schironi [2009b] 283288), Nnlist [2009a] index s.v. plausibility.
122 On the history of pleasure as a critical concept see Whrle [1994] 521523, with bibl.
744 Nnlist
enabled them to conduct an analysis of actual plots and their respective char-
acteristics (narrative *coherence, *formalism).123
Poetics, Theory of
Only Aristotles treatise and Horaces poem expressly declare themselves to
be poetics, neither of which is of a Hellenistic date, though Horace is known
to have incorporated much Hellenistic material. To focus on these texts alone
would plainly be absurd. There are plenty of other texts that are relevant to
the subject of this chapter. What is more, whether or not a critic wrote his own
poetics or made regular and explicit reference to one hardly matters. A case in
point is Pfeiffers well-known denial that Aristarchus followed the principles
of a theory of poetics [1968] 231. Pfeiffers own examples, which can easily be
added to, show that his scepticism is unfounded. Aristarchus no doubt had
to adapt Ax felicitous phrasea Poetik im Kopf, leaving aside the hermeneu-
tical question whether a text can actually be read without recourse to some
theory of poetics. And every critic who wrote on literary subjects automati-
cally (if often unexpressedly) took sides in the debate about ars and ingenium
(*inspiration).126
123 On the Poetics see Halliwell [1986], for practical examples Nnlist [2009a] 2368, with
bibl.
124 See Novokhatko and Montana in this volume.
125 Pfeiffer [1968] 190, 192195, citing older bibl., Bud [1977], Slater [1986] X, also 172 on fr.
434, Brown [1987], van Rossum-Steenbeek [1998] 3233.
126 On the history of theories of poetics see Till (et al.) [2003] 13041393, with bibl. According
to the original phrase, Aristarchus had a Grammatik im Kopf (Ax [2000] 107).
Poetics and Literary Criticism 745
Polemics
The tone of literary criticism could be sharp. A possible target of such polemics
was the author under consideration (cf. *censorship), as in the case of Zoilus,
who earned himself the *nickname Homromastix (lit. Homer-whip). More
often, however, scholars criticised each other with the familiar range of terms
that are apt to describe intellectual shortcomings. Aristarchus created a par-
ticular *critical sign in order to mark the passages where he disagreed with
Zenodotus.127
Prose
The differentiation between poetry and prose played an important role, with
*metre being the distinctive criterion (e.g. Gorgias 82 B 11.9 D-K), to which
Aristotle objected in vain (Poet. 1447b1323). Prose was usually considered
poetrys younger and less illustrious sibling. The relevant accounts assumed
that prose developed from poetry and came thus after it. Put in *stylistic terms,
poetry stepped down from its chariot (e.g. Strab. 1.2.6, cf. *sublimity). As one
Greek term for prose, pezos logos, indicates with perfect candour, prose walks
on foot. While it is true that Hellenistic critics devoted more energy to poetry,
prose authors were by no means ignored. Callimachus *Pinakes included prose
*genres, Aristarchus wrote a commentary on Herodotus.129 Studies with a
127 On the history of polemics as a critical concept see Stauffer [2003] 14031414, with bibl.
128 Gudeman [1927c], Pfeiffer [1968] 6970, Nnlist [2009a] 1112. See also Dubischar in this
volume.
129 An abridgment has been preserved on papyrus (P. Amh. II 12); see also Didymus com-
mentary on Demosthenes (Gibson [2002]).
746 Nnlist
Reader-Response Theory
Ancient literary criticism (just like pre-twentieth century criticism in gen-
eral) primarily focused on the production side of literature, but the reception
side was not completely ignored. In addition to the various *emotions that
literature created with the reader, critics also discussed the cooperation of
the reader and thus adumbrated aspects of modern reader-response theory.
Theophrastus (fr. 696 Fortenbaugh) recommended that one ought not to elab-
orate everything in detail, but leave some things for the listener, too, to per-
ceive and infer for himself; for when he perceives what you have left out, he not
only is a listener but also becomes your witness, and in addition more favour-
ably disposed. For he thinks himself perceptive, because you have provided
him with the occasion to exercise perception. Aristarchus applied the insight
of a rhetorical manual to the interpretation of literary texts and defended their
gaps as a regular feature, which was thus perfectly acceptable. Authors made
things clear not only explicitly (rhts) but also tacitly (kata to sipmenon).
The reader was required to understand (hupakouein) and thus supplement
the points that were not expressly stated. In other words, at least some critics
regarded the reader as an active participant in the process of making meaning.131
130 On the history of prose as a critical concept see Weissenberger [2005] 321348, with bibl.
131 On the history of reader response theory see Semsch [2005] 13631374, with bibl., for prac-
tical examples Nnlist [2009a] 157173, with bibl. in n. 1.
132 Konstan [2004].
Poetics and Literary Criticism 747
peculiar or even far-fetched that they are hardly compatible with the search for
the authors intention.
Reading, Close
Extensive reading (*analysis) went hand in hand with close reading. The *com-
mentaries, in particular, amply document critics attempts to analyse the text
with microscopic accuracy, wrestle sense from difficult or obscure passages,
discover particular subtleties, etc. Aristarchus expressly recommended: one
must look intently at the particular circumstances of the passage under con-
sideration (eis tn enestsan peristasin atenisteon, sch. A Il. 14.84a Ariston.).
No wonder he was hailed for his ability to divine (manteuesthai) the poems
meaning (Panaetius fr. 93 van Straaten = Ath. 634c). Extensive and close read-
ing also meant repeated reading, which only a truly great piece of art made a
rewarding experience (Hor. Ars P. 364365, Ps.-Long. Sub. 7.3).133
Realien
A general interest in antiquarian matters could already be found in the sophis-
tic movement (Hippias of Elis) and in the works of Aristotle.134 The writing of
running *commentaries required Hellenistic critics to deal with all kinds of
questions that the text under discussion posed. These included Realien, which
were thus regularly commented on. In connection with his concern for possible
*anachronism, Aristarchus made a great effort to reconstruct the living condi-
tions of the heroic age as accurately as possible because this was the condi-
cio sine qua non for a proper understanding of the text. The explanation, for
instance, that the Homeric shield was carried by means of a strap slung around
the shoulder (and not by means of a handle, as in later times) loses its apparent
insignificance as soon as one realises that it determines the correct sequence
of Homers type scene arming (*formulaic language). Aristarchus comment
(sch. A Il. 3.334335a Ariston., with test.) was meant to refute Zenodotus, who
had altered the sequence. The discussion of much scientific detail (geography,
astronomy, medicine, zoology, etc.) should also be read against the backdrop of
literatures *educative function.
Realism, Lifelike
Alcidamas perceived a direct relationship between literature and real life
when he referred to the Odyssey as a beautiful mirror of human life (quoted
by Aristotle Rh. 1406b1114). A similar correspondence, though not necessarily
133 On the history of close reading as a critical concept see Boone [1994] 257259, with bibl.
134 Pfeiffer [1968] 51, 7984.
748 Nnlist
one of exact identity, is the basis of the theories of art which circled around
the concept of *mimsis (cf. e.g. Aristotles recommendation that *characteri-
sation should fulfil the criterion of likeness, that is, general comparability of
characters and readers/spectators). A correspondence between art and life
also contributed to the latters *plausibility. Comparable to Alcidamas point,
a theory (probably of Peripatetic origin) saw comedy as imitation of life. It
was the springboard for Aristophanes of Byzantium famous question whether
Menanders comedies imitated real life or vice versa.135 The praise for lifelike
realism was, however, not bestowed on comedy alone (cf. *enargeia). A related
topic was the question what enabled an author to present scenes that were
so gripping. One possible answer was that he must have been present himself
(i.e. autopsy). Other critics, however, saw that it was enough to give the impres-
sion that he had been there himself (i.e. imagination, see *visualisation).136
Rhetoric
As mentioned in the introduction to this chapter, there was a considerable
overlap between rhetoric and literary criticism, in that the latter often had a
decidedly rhetorical outlook. The main difference was that rhetoric primarily
aimed at producing texts and literary criticism at interpreting them. One might
also say that the two subjects fertilised each other. Rhetorical handbooks often
illustrated the respective *figure of speech, literary device, etc. with examples
taken from poetry (most often Homer) and thus in turn influenced criticism.137
How this might work in practice can be documented, for instance, with the
famous line that foreshadows Patroclus death in the Iliad (11.603). In his hand-
book Trypho (III 203 Spengel) first defined proanaphnsis (prolepsis, fore-
shadowing) and then quoted the Iliadic passage as an example. Aristarchus
relevant note had used exactly the same term (sch. A Il. 11.604b Ariston.; for
another example see *style). More complex rhetorical notions such as stasis
theory equally recurred in criticism.138 The Patroclus example also demon-
strates that, although rhetorical analysis might a priori be assumed to have
a particular affinity to speech, it was unhesitatingly applied to the narrative
parts too.139
Scholar-Poet
The first part, in particular, of the period and subject under consideration is
characterised by the number of hommes de lettres who made their mark both
as scholars and poets.140 With Antimachus of Colophon as their precursor,
Philitas of Cos, Simias of Rhodes, Alexander Aetolus, Lycophron, Callimachus
and Apollonius of Rhodes are among the names that come to mind.141 Given
that they were all active and, in some cases (e.g. Callimachus), immensely pro-
lific scholars, it is no surprise that their poetry, too, displays intimate knowl-
edge of the questions that were discussed or even debated in scholarly circles
(*influence). Arguably, the so-called Alexandrian footnote owed its existence
to this fusion of scholarship and poetry (cf. Callim. fr. 612 Pfeiffer).
Sense of Proportion
Several entries above (*ambiguity, questions of *authenticity, *consistency,
*formulaic language, textimmanente *interpretation) make it clear that the
systematic and comprehensive *analysis whose goal was to uncover rules,
recurrent patterns (*custom) and the like did not automatically turn the critics
into pedants. All in all, their principal methods were applied with some flex-
ibility and a healthy sense of proportion.142
Simile (parabol)
Recognised as a distinctive feature of Homers poetic style, the simile received
much attention. Hellenistic critics identified the different parts (comparable
to Frnkels Wie-Stck and So-Stck) and debated whether the individual sim-
ile had a single or multiple points of comparison. Moreover, they produced
a wealth of interpretations in which they pointed out the similes aptness,
*appropriateness, *enargeia, *variety, etc. Not least they saw that the world
of the similes was different from that of the heroes (*anachronism) and thus
allowed a glimpse into Homers own world. The latter was by comparison more
domestic and humble, both conceptually and lexically (cf. *solemnity).143
Solemnity (semnots)
A text such as Aristophanes Frogs is indicative of a keen awareness that indi-
vidual poets or genres made use of different registers. Aeschylus, for example,
built the towering structures of solemn words (Ran. 1004). This topic had
deeper implications and should not be limited to a notion primarily of style,
though style was an important part of it (e.g. Arist. Poet. 1458a1822). Given the
general popularity of dichotomies, the relevant views were regularly expressed
in the form of a vertically oriented opposition of pairs: high vs. low, grand
vs. plain (cf. the theory of the three *styles), solemn (semnos) vs. base
(tapeinos) or mean (eutels). These opposing pairs could thus be exemplified
accordingly: poetry vs. *prose, tragedy vs. comedy, etc. In a similar vein critics
regularly discussed the question in what way a specific text displayed (or not)
the register that was typical of this particular *genre, author, etc. (cf. *simile).
Solemnity (semnots) represented as seen the high side and the correspond-
ing genres. It can therefore be aligned with similar concepts such as elevation/
sublimity (hupsos), which was later to receive its authoritative treatment in
the well-known monograph falsely attributed to Longinus.144
Style
Neither in the Poetics nor the Rhetoric did Aristotle put particular emphasis on
the importance of style as such. Both treatises, however, contained relevant
sections (Poet. chapters 2022 [on diction, the fourth qualitative part of trag-
edy: *characterisation], Rh. book 3 [originally a separate work?], chapters 112)
which proved to be very influential. The latter text, in particular, treated stylis-
tic questions in the framework of composition, a topic *rhetorical treatises had
a great deal to say about (e.g. Ps.-Demetrius or Dionysius of Halicarnassus).
Rhetorical handbooks also tried to systematise, define and illustrate with
examples the confusingly rich repertoire of *figures of speech, literary devices,
etc. Most examples came from classical poetry. The corresponding remarks
in literary criticism displayed noticeable similarities (both conceptually and
terminologically). For instance, the repetition of the half-line though his
(Achilles) hands are like flame in Iliad 20.371372 was a standard example of
epanalpsis or epanaphora in rhetorical handbooks and treated accordingly in
Aristarchus commentary.145 Apart from dealing with individual stylistic quali-
ties of a text (*clarity, *conciseness, *solemnity, etc.), an important branch of
144 On the history of solemnity and sublimity as critical concepts see Kallendorf (et al.)
[1994] 13571389, with bibl.
145 Rhetoric: see the test. collected by Hajd on Ps.-Herodian fig. 39; Aristarchus: sch. A Il.
20.372b Ariston. (cf. the test. collected by Erbse on 5.734736).
Poetics and Literary Criticism 751
Style, Virtues of
Taking up the ideas of his teacher Aristotle, Theophrastus (fr. 684 Fortenbaugh,
with comm.) codified a system which consisted of four virtues of style: (gram-
matical) correctness, *clarity, *appropriateness and ornamentation. This and
comparable lists (which included, e.g., *conciseness) proved to have a lasting
impact on *rhetoric, which in turn influenced literary critics. The latter were
less concerned about developing a canonical system as such, but the individ-
ual virtues frequently recur in their analyses of style. Not surprisingly, rheto-
ric also identified faults of style (e.g. the frigid, the affected, the arid and the
coarse in Ps.-Demetrius otherwise unorthodox theory of four styles), which
were equally taken up by literary critics.148
Sublimity
See *solemnity.
Synaesthesia
The entry on *enargeia demonstrates that it was very common to describe lit-
erature, which was aural, in terms of visual qualities. A notion comparable to
the modern metaphor taste can be found in an undatable note (sch. Ar. Plut.
515b) which argues that a line in Aristophanes last play Wealth already smells
(ozei) of Middle Comedy (*literary history). Even closer is the suggestive
comparison between reading and tasting food that opens Plutarchs treatise
How the young man should study poetry (14df).149
146 Sch. AbT Il. 3.212 ex. (omitted by Radermacher [1951] 69).
147 Pfeiffer [1968] 220. On the history of style as a critical concept see Sowinski [2007] 1393
1419, Mayer (et al.) [2009] 183, each with bibl., on the three styles Spang [1994] 921972,
wth bibl., for practical examples Nnlist [2009a] index s.v. style, esp. ch. 9, with bibl.
148 On the history of virtues of style as a critical concept see Hambsch [2009] 11431164, with
bibl.
149 On the history of synaesthesia as a critical concept see Ribicki Frhlich [2009] 344349,
with bibl.
752 Nnlist
Tradition, Literary
An important indicator whether an author had taken the subject-matter of his
text from tradition (h paradedomen historia, lit. the handed-down story)
or invented it himself were the names of the characters. Names known from
mythology (typical of epic and tragedy) pointed to tradition, unknown names
(typical of comedy) to invention (cf. Arist. Poet. 1451b1125). Like many other
scholars Aristarchus had a strong interest in the names and genealogies of the
characters. This allowed him to give a differentiated description of Homers
technique in these matters. While Homer took the names of his main char-
acters from tradition, he could also be shown to be a giver of names (ono-
matothetikos), with the clear implication that the respective characters were
a Homeric invention.150 The implicit premise of this distinction must be that
there was a tradition (i.e. texts) before Homer. Support for this interpreta-
tion comes from Aristarchus notes that either presuppose prior knowledge
among the audience (e.g. sch. A Il. 14.434a Ariston.) or argue that Homers
characterisation hinted at the traditional depiction (here of Odysseus: sch. A
Il. 11.430b Ariston.). Another point of interest among critics was the question
whether or not an author followed the traditional version of a particular myth
(*mythology).
Unity
The comparison of a literary text with a living organism had already been
made by Plato (Phdr. 264c69), but it was Aristotle who in his Poetics (esp.
chapters 78) developed it into the fundamental concept that a work of art
ought to possess unity. The fact that the unity was that of a living organism was
crucial because it tied in with Aristotles teleological philosophy. According to
this worldview, tragedy, Aristotles example for the best possible work of liter-
ary art, grew until it reached its perfection as shown by its organism-like unity,
size, proportions, balance, etc. This also meant that tragedy could produce the
*pleasure that was conform to its nature. The most important factor whether
a literary text fulfilled the necessary conditions was the unity of its *plot.
As is well known, the influential doctrine of the three unities (action, time,
room) was not developed until Aristotles Poetics was rediscovered in early
modern times. Only the first of them has truly ancient origins. It recurred, for
instance, in the Ars Poetica of Horace (141), who probably took his cue from
Neoptolemus of Parion. In practical criticism, the concept was put to use, for
150 For onomatothetikos see e.g. sch. A Il. 5.60a Ariston.; cf. Nnlist [2009a] 243244, with bibl.
in n. 20.
Poetics and Literary Criticism 753
example, when critics addressed the question whether or not a particular pas-
sage contributed to the texts narrative *coherence.151
Variety (poikilia)
Variety or, put negatively, avoidance of monotony was one of the most com-
mon and popular concepts of literary criticism. Critics almost universally rec-
ognised it as a virtue when an author broke away from uniform patterns which
might have exhausted the reader. A potential problem was that a stylistic prin-
ciple which was essentially Hellenistic could clash with texts that conformed
to other aesthetic standards, esp. the formulaic lines and other repetitions
in Homer. It should, however, be emphasised that the Alexandrian critics
accepted Homeric repetitions to a higher degree than is sometimes recognised
(*formulaic language). Needless to say, critics found plenty of opportunities
where they could praise his variety too.152
151 On unity in ancient criticism see Heath [1989], on Horace specifically Brink [1971] 7585.
152 On the history of variety as a critical concept see Celentano [1996] 15251527, Fekadu
[2009] 10061012, each with bibl., for practical examples Nnlist [2009a] index s.v. variety,
esp. 198202, with bibl. in n. 16.
754 Nnlist
of art.153 What is more, chances are that Aristarchus thought about the funda-
mental difference between linear and non-linear forms of art, that is, between
poetry and, say, painting.154 For his interest in visual arts see also *mythology.155
Visualisation
The concept of visualisation obtained both to the author and reader. First the
author: for Aristotle, thinking resulted in generating mental images (phanta-
siai), a process which he also called to put before ones eyes (pro ommatn
tithesthai, Mem. 1, cf. De an. 3.3). This type of visualisation was expressly rec-
ommended in the Poetics (1455a2226) because it allowed authors not only
to avoid *contradictions, but also enabled them to observe the principle of
*appropriateness. Given that poetry and rhetoric both intended to have their
audiences feel *emotions, each author aimed to transfer his mental images to
the audience because they enabled the generation of emotions.156 All three
aspects, the visualisation of the author, that of the reader and their interrela-
tion, were regularly discussed in Hellenistic criticism. As to the readers side
specifically, this included the observation that an aural sensation could be
turned into an image (cf. *enargeia).157
Word Order
The most extensive analysis of word order is the treatise On Composition by
Dionysius of Halicarnassus. A recurrent argumentative pattern of this treatise
is transposition (metathesis): Dionysius illustrates the different effects in that
he changes the original word order and juxtaposes the two versions. For the
modern reader the exact implications of these arguments can at times be dif-
ficult to follow. The same testing method (for which see also Ps.-Demetr. Eloc.,
e.g. 28, 48, 184185) could already be found in *euphonist theory, which used it
in order to explore the various sound-effects. More generally, as native speak-
ers ancient readers and critics alike could not help noticing when a literary
text departed from the natural word order. (Cf. Aristotles point that the iambic
153 On the carmina figurata see most recently Luz [2010]. The term technopaignia is not
ancient (Pfeiffer [1968] 90 n. 2).
154 Sch. pap. Il. 2.788 (p. 169 Erbse), with Lundon [2002a]. Needless to say, a similar idea is the
subject of Lessings Laokoon (1766).
155 On the history of comparing poetry with visual arts see Asmuth (et al.) [1994] 1030, with
bibl., Jacob [2009] 9971006, with bibl., for practical examples Nnlist [2009a] 195 with
n.3.
156 Meijering [1987] 1920, with reference to Rhetoric 1382a2122, 1385b1316.
157 On the history of visualisation/imagination as a critical concept see Beil [2003] 927943,
with bibl.
Poetics and Literary Criticism 755
158 Interestingly, Homeric tmesis was also considered a form of huperbaton (e.g. Apol. Dysc.
Synt. GG II/2, 447.17).
159 On word order see de Jonge [2008] ch. 5 and index s.v.
section 3.2
Grammar
chapter 1
It is especially the last reason which has important implications for the pres-
ent survey: in speaking of the parts of grammar in Greek Antiquity, we cannot
limit ourselves to the study of the organization of grammar books, but we have
to analyze the ideas of scholars about the parts of language, not only from
the period preceding the autonomization process of grammar studies, but also
from the times when grammar books were already available: discussion of the
structure of language is often found in philosophical and rhetorical investiga-
tions, i.e. in a philosophical or rhetorical context, but this does not detract from
their relevance for a study of ancient conceptions on the parts of grammar.
We will therefore discuss also views on units and parts of grammar formulated
in non-grammatical works, prior but also subsequent to the autonomization of
grammar as a discipline.
In Platos Sophist, the investigation concerns being and the moods of being,
and the issue of being vs. non-being reaches a crucial stage at the moment of
establishing the multiplicity of predication, and the communality of the
(genera), i.e. the sharing of properties of a higher order which rank above the
distinction between (sub)species. At this point, the Eleatic Stranger establishes
a comparison with language units: letters / sounds () show combi-
nations and restrictions on combinations. And the Eleatic Stranger points to
the fact that in order to know which combinations are possible and which are
excluded, it is necessary to dispose of an art or technical knowledge which
allows one to have adequate insight into this issue. The art in question is, as
Theaetetus states, the art of grammar.
Soph. 253a1
, ,
. ,
.
1 For the passages quoted from the Sophist we use (and at times, adapt) the English translation
of Fowler [1987 = 1921].
Description of the Constituent Elements of the Language 761
;
,
.
.
,
;
.
;
.
Str. Now since some things will commingle and others will not, they are
in much the same condition as the letters of the alphabet; for some of
these do not fit each other, and others do.
Theaet. Of course.
Str. And the vowels, to a greater extent than the others, run through them
all as a bond, so that without one of the vowels the other vowels cannot
be joined one to another.
Theaet. Certainly.
Str. Now does everybody know which letters can join with which others?
Or does he who is to join them properly, have need of art?
Theaet. He has need of art.
Str. Which art?
Theaet. The art of grammar.
Soph. 261e262c
.
;
, .
.
.
.
.
.
,
.
[...]
, ,
, .
;
,
,
.
,
.
, .
;
,
;
While these distinctions are insufficient for an adequate account of the gram-
matical structure of a language, they define a frame, from the lowest formal
764 Swiggers and Wouters
. , ,
,
.
,
(Poetics XX, 1, 1456b2021).4
The parts of diction in its entirety are as follows: (i) the element [i.e. let-
ter], (ii) the syllable, (iii) the conjunction, (iv) the name [i.e. noun or
adjective], (v) the verb, (vi) the joint, (vii) the case, (viii) the utterance.
(a) the list of (the properly said) parts of speech is still a very reduced one
(we find only , , and );
(b) Aristotle has no specific term for the word as a lexical-grammatical unit
(or better, type of unit), nor for the grammatical notion of word class
(part-of-speech).
after the elements which its demarcates, viz. the name and the verb), is the correct one. In
our opinion, the term could indeed include, in Aristotles view, propositional joints
of the type ; we would, however, not define such joints as a copula.
5 On the concept of in Greek Antiquity, see Ax [1986].
766 Swiggers and Wouters
(Poet. XX, 4, 1456b3133).
They differ according to the forms of the mouth, the places [in the mouth
where they are produced], aspiration, non-aspiration, length, shortness,
and also high, low or intermediate pitch.
6 Hence the twofold use of the term in the definition of the syllable:
(Poet. XX, 5, 1456b3436): A syllable is a
non-significant sound (= sound segment) composed of a consonant and [an element] which
has sound (= sonority).
Description of the Constituent Elements of the Language 767
(a)
7
,8
.
(Poet. XX, 6, 1456b381457a6).
(b) .
... ...9 .
(Poet. XX, 7,
1457a610).
7 We adopt the reading of Rosn [1990] 113. Kassel [1965] reads .
8 For this reading, see Dupont-Roc Lallot [1980] 323 and Laspia [1997] 85. Kassel [1965] reads
(referring to ).
9 Kassel [1965] prints . We keep here the readings of ms. A (cod. Parisinus
1741; saec. 10th/11th). Laspia [1997] 94 assumes that already the copyist of the hyperachetype
did no longer understand the examples for the term which in his days indicated the
article only.
768 Swiggers and Wouters
what way the correlate with (various) word classes (or sub-
groups of word classes). Because the text provides only for two of the four defi-
nitions a few (relatively certain)13 examples, , for (first
definition), for (first definition)it is not clear which word catego-
ries Aristotle had exactly in mind for each group. We can conjecture that the
first definition of would cover the connective (including disjunctive)
particles, the second definition of the (plain) conjunctions, both
grouping then the conjunctions and (trans)phrasal particles. The , in
its two definitions, would then cover articles, relative pronouns, prepositions
(and postpositions), and expletive adverbials or particles, as well as (trans-)
phrasal joints such as .
The two on the level of separate words are the
(noun) and the (verb).
, (Poet. XX, 8,
1457a1014).
, (Poet. XX, 9, 1457a1416).
The noun and the verb are defined as having an undivided meaning,14
and they are differentiated by a single secondary semantic feature, viz. refer-
ence to time, which is proper to the verb (Poet. XX, 9).
The concept of (or ) is used for the
reference (or indication) expressed by words or their components; it should
, , ,
,
(Poet. XX, 10, 1457a1823).
15 The latter aspect is at stake in Aristotles discussion of simple and compound nouns: com-
pound nouns can consist of parts which all have (lexical) meanings (Poet. XXI, 14). It
should be noted that this passage in fact implies an analysis into units corresponding to
what we would call morphemes, although Aristotle has no formal discovery procedure
for segmenting a string into morphemes.
16 For a detailed discussion, see DAvino [19751976].
Description of the Constituent Elements of the Language 771
(
, ,
, )
(Poet. XX 1112, 1457a2328).
17 Ax [1993] 29 proposes to translate the term, in its widest extension, as text. This is cer-
tainly in conformity with Aristotles view (according to which the entire text of the Iliad
could be called a ).
18 As Aristotle points out, the combination can be a verbless phrase (as is the case of nomi-
nal definitions), a noun-verb combination (i.e. a minimal sentence), or even a (very
long) text. Aristotle observes that the unit phrase can thus have a twofold semantic-
propositional status: it can either express one concept, or it can state a complex state-of-
affairs (and its changes through time). , ,
, , (Poet.
XX, 13, 1457a2830). An utterance can be single in two ways, either (a) by signifying one
thing, or (b) by a conjunction of several things. E.g. the Iliad is one by a conjunction [of
manythings], but the definition of a human being is one by signifying one thing (Janko
[1987] 28).
772 Swiggers and Wouters
Grammatical studies were part of the Stoics attempts to account for the
structure of utterances in Greek (cf. Frede [1978], [1987b]). The Stoics took up
Aristotles heritage and continued his reflections on types of sentences (declar-
ative utterances, problematic statements and fallacies, commands, prayers,
etc.), and on grammatical categories.19 However, grammar was always subordi-
nated to logic (or dialectics) in their view, and the general approach of gram-
mar was a semiotic-logical one: in dialectics, a study is made of how linguistic
expressions signify logical contents.
Our information on the Stoic contribution to grammar is to a large extent
secondary: apart from fragmentary testimonies (in the form of quotations),
we have at our disposalat least for the development of Stoic philosophy
the doxographical account of Diogenes Laertius (cf. Mansfeld [1986]) and the
polemically cast description of Sextus Empiricus. But in spite of the lack of
extensive and authentic source materials, it is clear that the Stoics made an
important contribution to grammatical studies.
This especially holds for Chrysippus (ca. 280207 BC), who wrote exten-
sively on various topics of grammar and semantics. Judging from the list of
writings attributed to him by Diogenes Laertius (VII, 189202), we can distin-
guish three main areas of grammatical research or, better, of research touching
upon issues relevant for the grammarian, such as:
Given the lack of direct source materials, we cannot say anything definite about
the Stoics view on the organization of grammar. Following Diogenes Laertius
account (book VII, more particularly in the sections devoted to Zeno), it seems
that the view of grammar held by the Stoics included (at least) three main
divisions:
774 Swiggers and Wouters
In addition, the Stoics seem to have dealt with stylistic aspects, with phraseol-
ogy, and with the distinction between prose and poetry.
In the field of the analysis of soundswhich the Stoics seem to have stud-
ied as part of a larger investigation of (the elements occurring in) nature20a
number of Stoic philosophers are said by Diogenes Laertius to have made a
contribution to this field: Archedemus (who seems to have written a treatise
), Antipater, and Chrysippus (who apparently dealt with sounds
in his ). Human voice was defined by the Stoics in terms of articula-
tion produced by a body; hence voice is also corporeal. At the level of the
sounds occurring in languagemore precisely, in the Greek language
the Stoics established 24 units or elements (). To the Stoics we owe
the clear distinction not only between spoken and written language
writing being the symbolization of meaningful sound(s), but also the dis-
tinction between (a) the minimal unit of language / speech, viz. ; (b)
its written representation, by a letter (); (c) its name (). The
Stoics also classified sounds (as represented by letters) into types. Diogenes
Laertius mentions the distinction between vowels (), viz. , , , ,
, , , and mutes (), consisting of the six plosive sounds , , , , , .
Although Diogenes Laertius does not mention any other subdivision, we may
assume that the Stoics also distinguished further subclasses for the remaining
11 sounds (including a liquid, a sibilant and a vibrant, nasals, aspirates and affri-
cate sounds). The Stoics introduced the view of multiple layering in language,
as distinct from sound production on itself: first, there is the layer of sounds
as belonging to a languagehere we enter into the distinction between
and ; second, there is the level of a verbal expression that has meaning
on its own, e.g. in the form of a word uttered in its quotation form (the nomi-
native singular case for nouns); third, there is the layer of saying something,
i.e.of making a statement about states of affairs. The latter are defined in terms
of contents that happen to occur. We can reconstruct the (early) Stoic view of
20 This can be inferred from the general definition of phon as quoted (and attributed to
Diogenes of Babylon) by Diogenes Laertius (VII, 55):
, . Now voice
is a percussion of the air or the proper object of the sense of hearing, as Diogenes the
Babylonian says in his handbook On Voice. We quoteoccasionally with some slight
modificationsthe English translation of Hicks [1991=1931].
Description of the Constituent Elements of the Language 775
,
[...]
, , , .
, . [...]
, ,
. , ,
, , .
. , ,
. (D. L. VII, 5557).
While the voice or cry of an animal is just (a percussion of) air brought
about by natural impulse, mans voice is articulate and comes forth from
the mind/intellect. [...].
Reduced to writing, what was voice, becomes a verbal expression, as day,
so says Diogenes. A statement or proposition is speech that issues from
the mind and signifies something, as it is day [...].
There is this difference between voice and speech that, while voice may
include just noise, speech is always articulate. Speech again differs from a
sentence or statement, because the latter always signifies something,
whereas a spoken word, as for example ,21 may be unintelligible
which a sentence never is. And speaking (in sentences) is more than just
uttering, since [in speaking] sounds are uttered, and things are spoken of,
which instantiate thought contents.22
As to word classes, this field of study was dealt with by the Stoics in the gen-
eral frame of semantic (and semantico-logical) contents expressed by lin-
guistic forms. The boundaries between what we today call morphology and
syntax are thus blurred, since the focus of interest for the Stoics was the
(verbal expression, or, better, sayable content). An overall distinction was
made between complete and incomplete ; the former can be sentences
or autonomously standing nouns or proper names (in the nominative singu-
lar form), the latter are predicates of which one or more places have to be
filled in (e.g. sleeps requiring a subject term; sees requiring a subject and an
21 This sequence became in later centuries one of the classical examples for a meaningless
word. Cf. Kotzia [1994].
22 Hicks translates this last sentence as: And to frame a sentence is more than just utter-
ance, for while vocal sounds are uttered, things are meant, that is, matters of discourse.
776 Swiggers and Wouters
object term). The Stoics elaborated the basis for a logical syntax, grounded in
the notion of (and its predicative content), and although we lack infor-
mation on the details of their views concerning the structure of propositions,
we can at least assert that
Also, the Stoicsin line with Aristotles analysis of types of speech acts (cf. Peri
Hermeneias / Poetics)went beyond a merely logical analysis of language: they
recognized, next to judgements, other types of sentences (or phrases):24 inter-
rogations, inquiries (which require a linguistically articulated answer), and
imperatives, adjuratives, optatives, and vocatives. These distinctions belong
primarily to the domain of rhetoric (and, secondarily, to that of grammar),
but the Stoics do not seem to have undertaken a properly linguistic analysis
of sentence types: judging from Diogenes Laertius account they approached
these sentences rather from a semiotic-discursive point of view, defining, e.g.,
an imperative as a signal conveying a command, and the vocative as an expres-
sion marking the fact that you address yourself to the interlocutor:
,
.
<> , ,
, .
23 Cases were analyzed by the Stoics in terms of their relationship with predicative contents;
see D. L. (VII, 6465).
24 The Stoics made a distinction between a signified content which is a judgement (),
and a signified content which is an object (). Judgements always have a truth value;
objects exist (or cease to exist). Interrogations and inquiries are always about judgement-
contents (but they are neither true nor false); the other types of sentences can be about
judgements or about objects.
Description of the Constituent Elements of the Language 777
A vocative utterance is something the use of which implies that you are
addressing someone, for instance: Most glorious son of Atreus,
Agamemnon, lord of men. (D. L. VII, 67)
,
.
.
Between the domain of the stoicheia and that of the aximata lies the domain
of the word. The Stoics probably used as the general term for word, in
line with Aristotles terminology. Although it is not clear whether the Stoics
took much interest in the classification of word classes, they are mentioned by
Diogenes Laertius for having made the distinction between proper name and
common noun, as separate parts of speech. Also, the Stoic author Antipater
framed the term for an additional word class, most likely the adverb.
Diogenes Laertius account can be found in his chapter on Zeno, although he is
reporting here on the doctrines of later Stoics:
,
, , , , , .
.
The splitting up of the former onoma-class into two word classes, viz. proper
name and (appellative) noun is fully understandable within the philosophi-
cal perspective of the Stoics: proper names are names of individual entities,
not named on the basis of a shared quality, whereas appellative nouns refer
to classes that are instantiated by objects sharing a common quality. Proper
names cannot be defined as to their content, whereas common nouns can be
defined in terms of a noetic content.
The division of the class of nouns into proper names and common nouns
is not only based on epistemological and logical grounds (with proper names
functioning as the normal instantiation of the subject term of an existential
proposition, and with common nouns corresponding to general [nominal]
predicates); there certainly was also a cultural (and literary) background to it:
proper names referred to individual heroes of the epic and dramatic literary
texts of the past, as well as to key figures in Greek history. Although the cul-
tural justification for the distinction does not appear from Diogenes Laertius
account, we may assume it played a role, given the Stoic interest in literary
style and in rhetorics (Diogenes Laertius mentions Rhetoric as one of the six
divisions of philosophical doctrine according to Cleanthes; cf. VII, 41).
The three remaining parts of speech distinguished by both Diogenes of
Babylon and Chrysippus (cf. D. L. VII, 57), viz. verb, conjunction ()
and article () were already present in Aristotles classification, but
contrary to the problematic text of Aristotles Poetics (cf. supra)the class of
the articles seems to have been narrowly defined by the Stoics;25 according
to Diogenes Laertius account, it comprised the (definite) article, although
one may suspect that the relative pronoun (as a postponed article) was also
included under the . The was defined by the Stoics as a linking
element between parts of a proposition, and given the Stoic interest in com-
plex propositions, one is justified to assume that elements linking one element
to another were also included under the class. As a matter of fact,
the term was used by the Stoics for referring to the conjunction in
conditional sentences, if Diogenes Laertius account can be read as a quotation
or summary from Chrysippus and / or Diogenes of Babylon:
25 As shown by Matthaios [1999] 614 ss., [2002f], the prepositions were first subsumed under
the (as reported also by Priscian, G.L. III, 501.10), but starting with Chrysippus the
class was narrowed down, and the prepositions were assigned to the
class, an evolution for which Apollonius Dyscolus (G.G. II, 1, 214) provides evidence.
Cf.Matthaios [1999] 616, [2002f] and Sluiter [1990] 117, [1997a] 234. This evolution
took place between Chrysippus and the period of activity of Posidonius of Apameia
(ca.13551BC).
Description of the Constituent Elements of the Language 779
,
,
.
Probably the most interesting aspect of Diogenes Laertius account is the fact
that he mentions the subsequent extension of the early Stoics repertory of five
parts of speech with a sixth class: the (mean / intermediate), a term
which Diogenes (VII, 57) attributes to the later Stoic philosopher Antipater of
Tarsos (ca. 150 BC).
This sixth part of speech was probably the adverb and the designation
may refer to the fact that, in terms of inflectional characteristics, the
780 Swiggers and Wouters
[...] cum adverbium Stoici [...] pandecten vocent, Nam omnia in se capit
quasi collata per saturam concessa sibi rerum varia potestate (Barwick
[1964] 252.2831).
[...] because the Stoics [...] call the adverb pandectes, all-receiver. For it
receives all things, as if they are collected without distinction, because it
has been granted power over all sorts of things (transl. Schenkeveld
[2004] 95).
26 Priscians account (in G.L. II, 54. 813) might reflect the later canonical Stoic position,
reverting to five word classes. Priscian explicitly mentions: nec etiam adverbium nomi-
nibus vel verbis connumerabant et quasi adjectiva verborum ea nominabant; the lat-
ter addition might constitute another explanation for Antipaters term : the
adverbs would be a class intermediate between noun and verb. For this interpretation,
see Schreiner (1954: 82).
27 Alexandria became a centre of learning in all fields of knowledge: mathematics, astron-
omy, philosophy, geography, and textual philology (see Montana in this volume). In fact,
the ancient notion of philologist is a very comprehensive one: it refers to a scholar, or
learned man in general. See in this volume, the chapter on Definitions of Grammar.
Description of the Constituent Elements of the Language 781
(a) Aristophanes of Byzantium may have been the first to posit the preposi-
tion as a separate word class, as can be concluded from Apollonius Dys-
colus treatise on syntax (G.G. II 2, 443.810):
, ,
.
,
, (Apol. Dysc. Pron., G.G. II 1, 4.1819).
They say that Comanus and his disciples called [the pronouns]
, because the general Greek term [for noun] is , and
not .
(c) One must reckon with the possibility that already in the period of Callis-
tratos and Comanos the participle was recognized as a separate part of
speech and was then the object of a closer study (cf. Matthaios [1999] 425
and Swiggers-Wouters [2007]).
(c) the manual was intended to help the 33 with putting grammati-
cal concepts and terms, of daily use in philological practice, into a
comprehensive frame, containing short, but useful definitions and offer-
ing a brief exemplification;
(d) the manual of Dionysius stands at the beginning of a tradition character-
ized by the presence of an art of grammar, i.e. of a grammar as a descrip-
tive object cast in an expository format. It is this format that we will focus
on in the following.
(a) at the macro-level, the grammatical analysis deals, successively, with the
following divisions within language structure:
prosodic signs / punctuation marks
segmental elements (stoicheia)
syllables
parts of speech
(b) at the micro-level, the analysis proceeds either by enumerating extant
types instantiating a higher-level division, or by enumerating categories,
which in their turn are then analyzed into their various realizations.
33 This is clear from the opening lines of the Techn which are in any case authentic. Grammar
is defined and divided into six parts, which are a perfect summary of the activity of the
(ancient) philologist studying and commenting Greek literary texts (see the contribution
Definitions of grammar). The Greek text reads as follows:
.
,
, , ,
, ,
(G.G. I 1, 56.10). Grammar is the empirical knowledge of the expressions com-
monly used among poets and prose-writers. Its parts are six [in number]: first, the skill-
ful reading in conformity with the prosody; second, the exegesis of the occurring poetic
phrases; third, the straightforward account of rare words and realia; fourth, the discovery
of the etymology; fifth, the establishing of analogical patterning; and sixth, the judgement
on poems, which is the finest part of all those [contained] in the art [of grammar]. See
the most recent comment on this definition by Callipo [2011] 9199.
34 The technique of division has its roots in Platos method of , which basically
operated on the basis of dichotomies; this method was also used by Aristotle (An. post. II,
13, 96b, 1597b, 23; Top. VI, 5, 6). The Stoics refined the terminological apparatus (adding
the terms and ) and they opposed conceptual analysis to division
of objects or substances.
784 Swiggers and Wouters
(a) noun
accidents: gender; species; figure; number; case
(b) verb
accidents: mood; diathesis; species; figure; number; person; tense; [the
conjugation class]36
(c) participle
accidents: gender; species; figure; number; case; diathesis; tense; [the
conjugation class]
(d) article
accidents: gender; number; case
(e) pronoun
accidents: person; gender; number; case; figure; species
(f) preposition
accidents: (there is only listing of monosyllabic and bisyllabic
prepositions)
35 The Techn ( 810 = G.G. I 1, 1720) specifies under which conditions each of these three
syllable types occurs. Common syllables are syllables that can be treated as long or short
by the poets.
36 This accident is, in fact, a differential characteristic within the verb class.
Description of the Constituent Elements of the Language 785
(g) adverb
accidents: figure; (semantic) species
(h) conjunction
accidents: (listing of different semantic-syntactic types).
How are the various word classes defined? In Dionysius Techn the definition
of each word class starts with assigning it to an overarching concepteither
word () or part of speech ( ), and then defines its nature
by summing up its essential formal and semantic properties. This results in a
string of minimal definitions for the eight parts of speech posited in the Techn.
37 In Swiggers Wouters [2011b] we have discussed in detail the mnemonic and didactic
virtues of this staircase-like model of grammatical description.
786 Swiggers and Wouters
, ,
, , ,
, . .
, , , , ( 12).
, ,
. , ,
, , , , , , ( 13).
.
( 15).
A participle is a word sharing the characteristics of both the verbs and the
nouns. It has the same accidences as the noun and the verbs, except for
persons and moods.
,
. , .
, , ( 16).
Description of the Constituent Elements of the Language 787
,
. , , , c,
, ( 17).
( 18).
, c
. ,
, ( 19).
. ,
, , , ,
, , ( 20).
788 Swiggers and Wouters
This format is exemplified by the Techn and by the other Greek and Latin
manuals of grammar in Antiquity; we deal here with a general format allow-
ing for a number of variations (depending upon individual grammarians /
grammar teachers, as well as upon the specific needs of the public aimed
at).38 Such variations occur, e.g., with respect to the number and order of the
accidents, the amount and specific nature of the examples given, the possible
39 For a detailed study of a case in point, viz. the treatment of the adverb in the Greek and
Latin tradition, see Swiggers Wouters [2002a] and Wouters Swiggers [2007].
40 dited by Wouters [2012].
790 Swiggers and Wouters
(2) The second point is that in the overall structure of the manuals we can
discern recurring patterns (cf. Wouters [1979] 4243); typologically speaking,
we find two main types or patterns, which in fact exploit, in a different way, the
layered organization frame referred to above:
Interestingly, these two types show that in the teaching of grammar two
opposed paths could be followed (and, in actual practice, both were used in
order to drill beginning students): either starting from the general and moving
to the particular, or proceeding from the particular and climbing up to the gen-
eral. The latter directionality was of course the one taken in dealing with con-
crete items found in texts (cf. the procedure of parsing, which reaches back to
the ancient exercise of partitiones).
(3) The third important point to be stressed is that these grammatical man-
uals are the (fragmentarily) preserved top of an iceberg, viz. of a much more
extensive body of (practical) grammatical writings, which are known to us in
the form of exercises, paradigm lists, conjugation exercises (not to mention
more elementary didactic texts) which have been conserved to a very limited
extent. This, again, shows that grammar was a subject for (systematic) instruc-
tion, and was, as such, exposed to variation and adaptation, in accordance with
Description of the Constituent Elements of the Language 791
The latter fact should be kept in mind before making too hastily a statement
about the dating and authenticity of Dionysius Techn, as if this text (trans-
mitted by textual testimonies that are, in any event, very late with regard to
the presumed original text) were an invariably transmitted codification of
Alexandrian grammatical knowledge. As a matter of fact, we still find in the
fourth-century P. Amh. 2.21 a discussion42 concerning the well-foundedness of
distinguishing nine or eight parts of speech (depending upon the treatment of
and as separate parts of speech or not).
Morphology was the central part of ancient Greek grammars, almost to the
point of exclusivity. While it is true thatgiven its importance for the rudi-
ments of Greek instructiongrapho-phonetic information was included in
(most of) the , this part was briefly dealt with and was in fact not linked
with the analysis of the structure of the Greek language, but with the (correct)
pronunciation and reciting of literary texts. As to syntax, it was not treated
43 Edited by Uhlig, in G.G. II, 2 [1910]. Translations by Householder [1981], Bcares Botas
[1987] and Lallot [1997].
44 Apollonius works on these parts of speech, dealing with their morphological and syntac-
tic characteristics, have been edited by Schneider, in G.G. II, 1 [18781902].
45 For this to be the case, it would have been necessary for Apollonius to have a clear view
of sentence-constituency (something which he lacks), and to have a grasp of (basic) syn-
tactic functions (for divergent opinions on this, see Bcares Botas [1987] 36 ff. vs. Lallot
[1997] i 6273).
46 See Lallot in this volume.
47 Edited by Kenyon [1909].
Description of the Constituent Elements of the Language 793
outlined the contents of the discipline in its more or less systematized form, as
it was elaborated by the Alexandrian grammarians. This outline has to be sup-
plemented with the discussion of a number of theoretically important issues.
The first issue to be raised is that of the evolutionary course of grammar in
Greek Antiquity. The above given outline, focusing on the contents of gram-
mar, has left aside the intricate question of the enduring relationship between
grammar and other disciplines, such as philosophy and rhetorics.48 The devel-
opment of the parts-of-speech system, the core part of ancient Greek grammar,
may give the impression of a rectilinear, irreversible evolution. In fact, this has
been for a long time the prevailing view (cherished by various authors, inde-
pendently of the opinion they held regarding the authenticity of the Techn
of Dionysius Thrax). This view can be summarized, following the important
overviews by Robins [1966], [1986] into a rectilinear sequence in which the sys-
tem of word classes developed from a binary one to one involving eight parts
of speech:
Tracing such a sequence would be, however, oversimplifying the actual history
of the parts-of-speech history. On the one hand, the above sketched sequence
ignores the essential role played by philosophical discussions (of different
backgrounds: Aristotelian, Stoic or even Platonic) in the development of gram-
mar during the last centuries BC and the first centuries AD; on the other hand,
as shown already above (cf. the discussion of the grammatical information
provided by the papyrological documentation), even after having received sys-
tematization and codification by the first generations of Alexandrian scholars,
the development of grammatical doctrine was characterized by variation in
contents and terminology.
52 As such, the case of Dionysius confirms the observations made above with respect to the
interplay between grammatical-philological studies and philosophical approaches. Cf. De
Jonge [2008] 165, with reference to Matthaios [2002f]): After Aristarchus, the Alexandrian
scholars were deeply influenced by Stoic theories. This Stoic influence resulted in a num-
ber of grammatical works that must have combined Alexandrian and Stoic ideas on lan-
guage. Most of these works are lost, but the few extant fragments of Dionysius Thrax and
Tyrannion show that they adopted Stoic views in their classification of the word classes.
796 Swiggers and Wouters
the object of a history (as retrospective account). This history of the second
degree took the form of doxographical accounts (cf. Swiggers-Wouters [2010]),
concerning the history, constitution and development of the word classes in
Greek (and later, Latin) grammaticography. For the Greek-speaking (or Greek-
writing world)53 we can mention two types of doxographical texts (cf. sets
A and B in the list below): the first set is that of full doxographies, i.e. texts
intended as an overview of the development of grammatical conceptions (on
the parts of speech) and as a brief discussion of the methodological problems
involved; the second set comprises doxographical statements relating either to
specific topics in the treatment of the parts of speech or to problems involved
in the analysis of a particular word class
(B) Doxographical statements (in Greek school papyri [4, 5], in Greek gram-
mars [6])
4) P. Heid. Siegmann 198 (Wouters [1979] n. 12, ll. 23) [3rd cent. AD]
5) P. Amh. 2.21 (Wouters [1979] n. 14, ll. 1317) [4th cent. AD]
6) Dionysius Thrax, Techn Grammatik (G.G. I 1, 23.13) [2nd1st cent. BC (?)]
This is not the place to go into a detailed study of the doxographical texts (most
of them are in Latin, but they cover the macro-evolution of the parts-of-speech
system in Antiquity). Here we want to stress their important (meta-)histori-
cal function, viz. as retrospective statements on the emergence and devel-
opment of grammatical doctrines (in the head of the doxographical writer),
and their methodological interest. All doxographical texts deal with division
into parts of speech, or with the rejection of a particular division proposed
by some ancient scholars; they all reflect a strategy of , division (of
the sentence / of language structure) into basic components. Differences
notwithstandinge.g., a more or less detailed discussion of the respective
divisions into word classes; the amount of argumentation found concerning
53 For a full overview and study of Greek and Latin doxographical texts, see Swiggers
Wouters [2010], [2011a].
Description of the Constituent Elements of the Language 797
54 For brief overviews of the history of the word-class system, of the concepts of word
and sentence, and of the two basic components of the sentence, viz. noun and verb, see
Swiggers Wouters [2009a], [2009b], [2014a], [2014b], [2014c].
Chapter 2
Lara Pagani
1 Introduction
2 The Concept of Language Correctness in Philosophy and Rhetoric
3 The Concept of Language Correctness in Hellenistic Scholarship
3.1 Aristophanes of Byzantium
3.2 Aristarchus of Samothrace
3.3 Crates of Mallus
4 Theoretical Reflections and Treatises on Hellenismos
5 Systematization of Language in the 2nd c. Ad and Beyond
5.1 Apollonius Dyscolus
5.2 Aelius Herodian
5.3 Linguistic Atticism
5.4 Later Authors
6 The Criteria of Hellenismos
6.1 Analogy
6.1.1 The Conditions of Analogy
6.1.2 The Analogy/Anomaly Controversy
6.2 Usage
6.3 Literary Usage and Tradition
6.4 Etymology (and Dialect)
7 Conclusions
1 Introduction
* English translation by Rachel Barritt Costa. I am very grateful to Albio C. Cassio for helpful
comments on a draft of this chapter.
1 Pagani [2011] 1721, with bibliography.
,
(Etym. Magn. 331.3739).
2 Ax [1982] 97; Ax [1991] 277278; Matthaios [1999], 1516; see also Barwick [1922] 227ff.;
Siebenborn [1976] 32ff. Cf. in this volume (Part 3.2) Valente and Swiggers-Wouters.
3 See Casevitz [1991]; Irmscher [1993] for formation and meanings of the term.
4 On these two aspects see the more detailed treatments by Valente and Probert, respectively,
in this volume.
5 Blank [1998] 204.
6 See e.g. Quint. Inst. 1.5.6; 1.5.16; Apol. Dysc. Synt. 3.8; Sext. Emp. Math. 1.210; D. L. 7.59.79 (who
quotes Stoic positions). Further examples in Blank [1998] 232234 and Hyman [2003]. See
also Lallot [1997] II 158159, 161.
7 See e.g. Arist. Soph. el. 3.165b 2021: Blank [1998] 232233.
8 Blank [1998] 206.
800 Pagani
,
(Sch. Lond. Dion. T. in GG I/III 446.1215).
In certain cases, the incorrect form or construction could be justified, for exam-
ple if this was due to the typical characteristics of a particular dialect or author.
Such cases were described as (schemata), which was another subject
frequently addressed in monographic treatises like the works by
Alexander Numenius, Lesbonax and Ps.-Herodian.
The criteria determining whether a construction should be accepted as cor-
rect Greek or rejected as a barbarism or solecism vary slightly among the differ-
ent sources. There was probably a lively debate about these criteria, enquiring
into what their merits were, their proper order of application and their mutual
interaction.9 In general, the main features were held to consist of analogy
a procedure of comparison, with varying degrees of refinement, between an
uncertain language form and one similar to the latter, used as a reference
modelas well as usage, literary tradition, and also etymology and dialect
considerations.10
In a diachronic perspective, delineating the concept of language correct-
ness in classical antiquity means considering the question of language study
in general, starting from the intersections with philosophyabove all Aristotle
and his school and the Stoicsand with philological-erudite research in
the Hellenistic period, and continuing up to the first theorizations from the
1st c. BC onwards, and thence to the great systematic codifications of Apollonius
Dyscolus and Aelius Herodian (2nd c. AD) and beyond. From at least the
1st c. BC onwards, the Greek model was transplanted into the Roman world,
where the concept of Latinitas, modelled on Hellenismos, was applied to the
Latin language in a wealth of specialist treatises.11
12 For this phase of ancient grammar, see Schmitter [1991a] 57272; Hennigfeld [1994] 4124;
Blank [2000] 400404; Schmitter [2000], with additional bibliography; Law [2003] 1351;
de Jonge-van Ophuijsen [2010]. A documented overview, albeit dated, is in Gudeman
[1912a] 17811791.
13 See Dahlmann [1928] 8; Fehling [1965] 218229; Siebenborn [1976] 2224; Coseriu [1996].
14 Pl. Cra. 389a401b.
802 Pagani
specific meanings of each of the different usages, in order to reach the cor-
rect use of words ( , test. 84 A 16 D.-K.).24 However, whether
the theoretical background underlying this practice was an aspiration towards
anartificialconstruction of a better language (on the presupposition of the
conventional nature of language) or an awareness that the fine nuances differ-
entiating presumed synonyms reflected differences in the situations described
by the individual terms (conception of language by nature) remains unclear.25
The concept of in the sense of the correctness of the rela-
tion between word form and the reality designated by a given word lies at the
basis of Platos (428/427348/347) dialogue Cratylus,26 where the two theories
by nature and by convention espoused by the two speakers Cratylus and
Hermogenes are both shown to be unsatisfactory as they are incapable of dis-
tinguishing true from false denominations, i.e. those that are adequate vs.
inadequate in portraying reality. Plato concludes that knowledge of the true
nature of things cannot be acquired through observation of their names, but
rather through observation of the things themselves. Examination of the possi-
ble correspondence between the form of a name and the reality denominated
is performed mainly in a lengthy etymological section (397a437d), where
words are traced back to presumed original forms, which are themselves then
decomposed into their minimal elements ().
Reflections on language correctness also influenced studies on rhetoric,
the field with the earliest known annotation on in the technical
sense, inquiring into what it is and how it may be achieved. The annotation is
owed to Aristotle (384322), who defined good Greek as a principle of style
( ) and held it to depend on compliance with five rules: 1) cor-
rect utilization of connecting particles; 2) use of specific rather than generic
terms; 3) rejection of ambiguous terms; d) proper distinction between the gen-
ders of words (with a reference to the precedent represented in this field by
Protagoras); 5) concord in number (Rh. 3.1407a 19ff.).27 Of these, only points 1,
4 and 5 involve the grammatical level,28 while the other two are more closely
concerned with (clarity) of style. On the specific issue of language
correctness, was discussed by Aristotle in the treatise On Sophistical
, , , , ,
.
[...]
,
. (Diogenes of Babylon ap. D.L. 7.59 = SVF 3.24).
29 See Stroux [1912]; Siebenborn [1976] 24; Schmitt [1982] 462; Innes [1985] 256; Blank
[1998] 208.
30 See Nnlist in this volume.
31 Siebenborn [1976] 24. See also Stroux [1912] 9ff.
32 The canonical reference work for the Stoics linguistic studies is Schmidt [1839]. See also
Steinthal [189018912] I 271ff.; Pohlenz [1939] and, more recently, Frede [1978]; Sluiter
[2000b], with bibliography. On Peripatetic influences, see Ax [1993].
Language Correctness ( Hellenismos ) and Its Criteria 805
33 Siebenborn [1976] 2527; Dalimier [1991] 21; Schenkeveld [1994] 281282; Blank [1998]
204, 208; Vassilaki [2007] 1124.
34 See Sluiter [1990] 1213, 1821; Allen [2005].
35 Cf. SVF 2.884, 2.886, 2.895 for Chrysippus celebrated discussion on the meaning of
(I).
36 Entitled IV libri (Four books on anomaly) according to Varro (Ling. 9.1);
(Four books on anomaly in expressions,
addressed to Dion) according to Diogenes Laertius (7.192).
806 Pagani
37 Lersch [18381841] I 51; Siebenborn [1976] 98100; Frede [1978] 73ff.; Ramelli-Lucchetta
[2004] 190ff.
38 See Montana in this volume.
39 See Pagani [2011] for a broader discussion of the problem and the related bibliography.
40 Gathered together in Nauck [1848a] and, more recently, in Slater [1986].
41 See Nauck [1848a] 264271; Steinthal [189018912] II 7377, 151 n. (cf. II 181 n.); Pfeiffer
[1968] 202203; Callanan [1987]; Ax [1990]; Schenkeveld [1990] 290298; Ax [1991]
277282.
Language Correctness ( Hellenismos ) and Its Criteria 807
reveals his conviction that nominatives of the same type are declined in the
same manner: he maintained that the Aeolians mistook the genitive
(from , old) for an - declension nominative and, as a result, that they
constructed the wrong dative plural .42 He also had knowledge of
accent shift in declension, as documented in the fragment conserved by sch.
Hdn. Il. 15.606b (p. 187 Sl.),43 which discusses the accent in the dative plural
in relation to the inflectional pattern to which the word belongs: if
the dative comes from the noun (woodland), the stress is on the anti-
penultimate syllabe (), the parallel form being (javelin), while
if it belongs to the declension of the adjective (dense), it is accented
on the penultimate syllable, as it happens in the analogical term of com-
parison, the adjective (sharp) (see below).44 More generally, the view
that Aristophanes formulated descriptive rules on the prosody of groups of
words which conformed to specific prerequisites, such as comparatives end-
ing in - (fr. 347 Sl.), remains on the level of hypothesis.45 As regards verb
inflection, we know that Aristophanes must have addressed the issue of use
of the dual in Homer, because in numerous cases he established the text with
a dual rather than the plural if the passage referred to two persons or objects
(sch. Did. Il. 6.121 [p. 178 Sl.]; 8.290c [p. 179 Sl.]; etc.).46 He also examined irregu-
lar verb forms, as emerges from two notes traceable to his collection of Lexeis
in which examples displaying the same phenomenon are arranged in groups,
suggesting a theoretical framework for verb inflection. Thus fr. 19 Sl. shows that
Aristophanes recorded the use of the - ending for some imperfects, explain-
ing it as a Chalcidian element, and fr. 28 Sl. testifies that the imperative forms
and were mentioned alongside the respective regular forms
and .47
Additionally, he is recognized as using the criterion of analogy,48 although
it is not possible to support the hypothesis, devoid of documentary source,
49 The hypothesis goes back to Nauck [1848a] 264271. For objections to this, see Steinthal
[189018912] II 151 n. (cf. II 181 n.); Pfeiffer [1968] 202203; Callanan [1987] 107; Ax [1990]
12; Ax [1991] 282.
50 See above, n. 43 for the possible textual problems raised by the passage.
51 On the problems raised by scholars concerning this evidence, see below 6.1.1.
52 Siebenborn [1976] 8183.
53 The passage, not mentioned at all by Callanan [1987], was pointed out by Ax [1990]: [...]
artufices egregii non reprehendundi, quod consuetudinem... superiorum non sunt secuti,
Aristophanes improbandus, qui potius in quibusdam veritatem [cod. Basiliensis et editio
princeps pro veteritatem] quam consuetudinem secutus?. See esp. pp. 711 for discussion
of the textual difficulties (presenting a parallel with Cic. Orat. 155162) and cf. also Ax
[1991] 277282.
54 Thus Callanan [1987] 103106 and passim.
55 Pfeiffer [1968] 197200; cf. Ax [1990] 1415; contra Callanan [1987] 7582.
Language Correctness ( Hellenismos ) and Its Criteria 809
56 Callanan [1987] 97102. Cf. Nauck [1848a] 268269; Pfeiffer [1968] 201 and n. 4, 260; Slater
[1986] 19.
57 See Callanan [1987] 97102 for the examples.
58 In contrast to Reitzensteins proposal [1897] 184. The passage from Varro (fr. 372 Sl.) more
likely refers to inflection rather than derivation of one word from another. Slater [1986]
138; Callanan [1987] 9798; cf. Schenkeveld [1990] 297298 and Blank [2008] 54ff. for the
problem of the meaning of declinatio in Varro.
59 On partial collections, an important breakthrough has been achieved by Matthaios
[1999], with collection and study of all the fragments concerning the doctrine of parts of
speech.
60 Erbse [1980]; Ax [1982]; Ax [1991]; Schenkeveld [1994] 283287; Matthaios [1999] 2832.
Cf. already Ribbach [1883]. Decidedly more reductive positions are found in Steinthal
[189018912] II 112113; Siebenborn [1976] 3031, 6878; Frede [1977] 76. On prosody and
orthography, see in this volume (Part 3.2) respectively Probert and Valente.
61 Siebenborn [1976] 81; Blank [1998] xxxvii.
62 Siebenborn [1976] 70ff.
810 Pagani
forms and widespread usage, a kind of poetic licence:69 for instance, the use
of a passive instead of an active verb form, a nominative in place of a genitive
for a noun, a simple case instead of a prepositional construction, an adjective
instead of an adverb and so on.70
Homers language was also invoked as a standard of correctness to be applied
to Greek in general,71 because in Homer everything concerning Hellenismos is
perfect (fr. 125 Matth.).72 Accordingly, in addition to considerations of syn-
tactic tolerability and inflectional analogy, Aristarchus rejected the compound
forms of the third person plural reflexive pronoun, which he regarded as incor-
rect, in favor of the periphrastic forms.73 Some centuries later, Sextus Empiricus
(Math. 1.202ff.) would reflect ironically on such a criterion, in open opposition
to the position adopted by Aristarchus pupil, Ptolemy Pindarion (see below
4), and would ridicule the curiosities that might be said if Homers language
were taken as a standard of Hellenismos.74
In Aristarchus doctrine, the criterion of analogy apparently interacted not
only with Homeric auctoritas but also with contemporary language usage,
as emerges from the evidence of Varro (Ling. 9.1), who held that Aristarchus
exercised the rules of similarity in derivation within the limits allowed by
usage (quoad patiatur consuetudo).75 This assessment may include some
Aristarchean interventions on the Homeric text that were later judgedby
Aelius Herodian or other grammariansas contrary to analogy (,
).76
69 According to the definition of Sch. Lond. D.T. in GG I/III 456.2326, were errors
committed deliberately by an author as an artistic flourish or for the delight of using an
uncommon form.
70 Cf. the collection of Fragmenta schematologiae Aristarcheae by Friedlnder [1853] 135
(who wrote ... explicare conabor, quibus finibus circumscriptam esse poeticam licen-
tiam quibusque proprietatibus Homeri sermonem a stabilito recentiorum usu differre
statuerit grammaticus nobilissimus); Erbse [1980] 242243, with examples; Ax [1991] 284;
Blank [1998] 230231; Matthaios [1999] 382ff.
71 On literary use as a criterion of Hellenismos, see below 6.3. In particular, on use of
Homer as a linguistic model, see Pontani [2011c].
72 Siebenborn [1976] 31; Ax [1982] 105; Blank [1998] 228.
73 Fr. 125 Matthaios. For a discussion on this point, cf. Wackernagel [1876 = 1979] 55;
Siebenborn [1976] 3031; Ax [1982] 105106; Lallot [1997] II 146; Matthaios [1999] 479
480; Pagani [2010a] 117122. The importance of this passage is minimized by Schenkeveld
[1994] 286.
74 Siebenborn [1976] 97; Montanari [1995b] 4549; Blank [1998] 225232; Boatti [2000]
8795.
75 Cf. Siebenborn [1976] 3031; Ax [1982] 109.
76 Examples in Erbse [1980] 241. Cf. also Ax [1991] 287.
812 Pagani
77 Siebenborn [1976] 3031; Schenkeveld [1994] 289. Further examples in Ribbach [1883] 89
and Siebenborn [1976] 29.
78 Ax [1991] 287288.
79 On linguistic aspects, cf. e.g. Steinthal [189018912] II 121126.
80 Cf. Broggiato [2001 = 2006] xxxvxxxvi and Montana in this volume.
81 Fehling [1956] 223ff., 266; Broggiato [2001 = 2006] 267.
82 Cf. Fehling [1956] 268269; Pinborg [1975] 109110, 112; Schenkeveld [1994] 286.
83 Mette [1952], esp. 911 and 3145 for Hellenismos. Observations on the nature of this work
can be found in Broggiato [2001 = 2006] xiiixiv.
Language Correctness ( Hellenismos ) and Its Criteria 813
greater parts of Varro and Sextus Empiricus than those definitely ascribable
to Crates, Mette credited Crates with formulating the expression
(observation of linguistic usage) and traced the methodological
foundation of Crates linguistic theories to the empirical medical school.84 But
the terminological considerations underlying both hypotheses have proven to
be far from decisive.85
In the two Varronian fragmentsbesides that which alludes to the
disputewhere Crates and Aristarchus are presented as holding opposite
views, it is said, firstly (fr. 102 Brogg.), that the Pergamene erudite wondered
why the names of the letters of the alphabet, unlike all other nouns, were not
declined, in contrast with the predictions of the rules of analogy, to which
the Aristarcheans answered that the names of the letters were not subject to
declension because of their Phoenician origin (thus they were not Greek).86
Furthermore (fr. 103 Brogg.) Varro asserts that Aristarchus opposed Crates
on the possibility of considering three proper names which, although having
identical endings in the nominative, differed in the vocative, as being analo-
gous.87 These passages thus appear to reflect a debate on the criteria and limits
of application of analogy.88 Interpreted in the perspective of the controversy
mentioned in the third Varronian passage, the argument would appear to sup-
port the above-described traditional opinion. But it is worth considering the
proposal by D. Blank, who suggests that Varro or his sources may have cited
both Crates and Aristarchus as representatives of analogy in a typical argument
from disagreement, using the divergence between the two great grammarians
on the application of analogy to discredit the entire analogical method.89
Recourse to etymology is fairly extensive in the Cratetean fragments,90 but it
aims mainly at solving problems involving interpretation of the Homeric text.91
84 Mette [1952] 3148. The parallel between medicine and grammar is studied extensively by
Siebenborn [1976] 116135.
85 Broggiato [2001 = 2006] xxxviixxxviii with references to the previous bibliography. Cf.
also Siebenborn [1976] 118ff.; Blank [1982] 15; Blank [1994] 153155.
86 Ax [1991] 292; Blank [1998] xxxvi; Broggiato [2001 = 2006] 265266; Blank [2005] 225228.
87 Cf. Fehling [1957] 94; Ax [1991] 292 (who speaks of Vokativthese); Blank [1998] xxxvii;
Broggiato [2001 = 2006] 266; Blank [2005] 228233.
88 Broggiato [2001 = 2006] xxxvxxxvi.
89 Blank [1998] xxxivxl; Blank [2005], who goes so far as to hypothesize that the opposi-
tion between the two saw Crates defending analogy more strenuously than Aristarchus
(pp. 227228, 232, 238).
90 See Broggiato [2001 = 2006] 329, s.v. etimologia.
91 Broggiato [2001 = 2006] lxiii, with observations also on the concomitance of etymological
explanations with allegorical interpretations (frr. 3, 21, 130 Brogg.); Blank [2005] 222.
814 Pagani
Starting from the period between the 2nd and 1st c. BC, the issue of language
correctness became the object of theoretical speculations, partly within the
systematization of the branch of knowledge termed grammatike, but also
supplying material for definitions, abstract and general considerations, or
monographic studies. On the first point, two elements employed as criteria
of Hellenismos, analogy and etymology, are cited as parts of grammar in
the celebrated definition of Dionysius Thrax92 (17090 approx.), a pupil of
Aristarchus93 who at times seems to have applied the analogical method to
exegesis of the Homeric text even more stringently than his master.94 Moreover
Dionysius interest in orthographic problems emerges from the grammatical
treatise preserved in the scriptio inferior of the palimpsest Lipsiensis gr. 2 (olim
Tischendorfianus II) and identified by K. Alpers as the commentary by Oros
on the Orthography of Herodian,95 where the name of Dionysius appears in
a list of grammarians who dealt with orthography (22v ll. 1826). As a general
expression, Hellenismos appears in the systematic description of the gramma-
tike of Asclepiades of Myrlea (2nd1st c. BC), an erudite of Pergamene orienta-
tion whose work was probably utilized by Sextus Empiricus
far more extensively than merely in the three points that explicitly cite
Asclepiades (Math. 1.4748; 1.7274; 1.252253).96 Asclepiades devoted the part
of the defined as technical to study of letters, parts of speech,
orthography, Hellenismos and related aspects97 (Sext. Emp. Math. 1.92, 1.252).
His mention, alongside Hellenismos, of one of the issues actually addressed
in Hellenismos, namely orthography, shows he awarded it particular impor-
tance, and this fits well with the above-mentioned testimony of the palimps.
Lipsiensis gr. 2, where Asclepiades can be recognized among the grammarians
listed as having orthographic interests (the name is in lacuna, but the integra-
tion, based on the ethnic designation, is traditionally considered certain).
Another pupil of Aristarchus, Ptolemy Pindarion (2nd c. BC), reformulated
the theoretical bases of analogy in response to the criticism that it was point-
less frippery necessarily appealing to (linguistic usage) for exam-
ples establishing regularity and rules. Pindarion neutralized this objection by
arguing that the usage underlying analogy was not the fickle parlance of the
speech community but rather the codified language of the Homeric poems,
which was guaranteed by the primacy and antiquity of such works (Sext. Emp.
Math. 1.202ff.: fr. 12 Boatti). Thus Pindarions reformulation, far from represent-
ing a compromise,98 reasserted the validity of analogy by capitalizing on the
idea, present in Aristarchus, that Homeric language constituted a reference
point for Hellenismos. Furthermore, that Pindarion had rigorously analogistic
leanings is also indicated not only by other fragments but also by the epithet
(analogist) assigned to him by Apollonius Dyscolus (Conj. 241.13
Schneider: fr. *14 B.), unless this refers to the homonymous grammarian of
Ascalon (see below).99
Among Aristarchus followers, Dionysius of Sidon (second half of 2nd c. BC)
was noted by Varro (Ling. 10.10) as having recognized the incredible number of
seventy-one conditions of analogy (or forty-seven, if limited to conditions on
nouns, but see below 6.1.1 for the exorbitance of these numbers). The same
passage of Varro also cites Parmeniscus (2nd1st c.), for whom the Homeric
scholia conserve fragments on prosodic topics, and Aristocles (second half of
1st c. BC), who possibly composed a work (On dialects) and is
the putative author of a couple of fragments of prosodic nature and a defini-
tion of analogy, described by Varro as obscure, albeit without citation (Ling.
10.7475). According to Varro, Parmeniscus established eight conditions of
analogy, Aristocles fourteen. Varro criticized Aristocles definition of analogy,
but also that of Aristeas (1st c. BC?), an erudite for whom a fragment on pros-
ody can be mentioned, and that of Aristodemus (1st c. BC), identifiable with
the grammarian and rhetor of Nisa.100
From the 1st c. BC onwards, evidence has survived of treatises on, or on
parts of, Hellenismos, such as orthography,101 or on criteria for determining
98 Thus, among others, Lersch [18381841] I 75; Steinthal [189018912] II 154; discussion of
the problem in Montanari [1995b] and Boatti [2000] 9395, with further bibliography.
99 Cf. Boatti [2000] 96103 and Pontani [2011c] 93 n. 21, who argues in favor of attributing the
epithet to Pindarion.
100 On the above-mentioned grammarians, see the LGGA, s.vv.
101 Schneider [1999].
816 Pagani
Hellenismos, like analogy. For the period between the 1st c. BC and the 1st AD,
Strabo (14.2.28) spoke of (handbooks on Hellenismos)
as if this were a known and widespread subject.102 However, little information
on the content of such works can be obtained. The available documentation
suggests they addressed such issues as the correct meaning of words, prosody,
choice among phonetic-orthographic variants, use of etymological and dialec-
tal considerations, as well as the pursuit of linguistic regularities on the basis of
analogical reasoning, whereas no reference attests that these works also dealt
with syntax.103
The most ancient monograph known to have focused on Hellenismos is that
of Philoxenus of Alexandria (1st c. BC), of which only two small fragments
survive (frr. 288 and 289 Theodoridis, cf. test. 1), both pertaining to correct
word meaning obtainable through recourse to etymology.104 An etymological
approach is also traceable in fragments of what was probably Philoxenus main
work, (On monosyllabic verbs; test. 1, frr. 1*215
Th.), founded on the theory that the Greek lexicon ultimately derived from a
core of monosyllabic verb roots.105 The same holds for
(On reduplication, test. 1, frr. 219*284 Th.), where terms were said to have
derived from their original roots through syllabic reduplication, and for
(On derivatives, fr. 330 Th.) and (On the verb, frr. 354
400 Th.). Additionally, Philoxenus investigated dialects, as testified by works on
Ionic dialect (test. 1, frr. 290*310 Th.), Roman (frr. 311*329 Th.)Latin being
seen as a derivation from Aeolic106, Syracusan (test. 1 Th.), Laconian (test. 1
Th.). Finally, he did not disregard prosody, as shown by (On
prosodies), apparently with reference to the Homeric text (frr. 407411 Th.).107
A treatise on Hellenismos is also attested for Tryphon of Alexandria, an eru-
dite who made a decisive contribution to the development of grammatical
thought in the second half of the 1st c. BC,108 and was considered an authority
worthy of frequent mention in the monumental systematization of grammar
115 Thus Mller [1891] 49; Siebenborn [1976] 49 and n. 5, who, however, misunderstands the
meaning of the passage from Athenaeus. On Seleucus see also Razzetti [2002b]. On lin-
guistic Atticism see below 5.3.
116 On this phenomenon according to the ancient grammarians, see Lehrs [18823] 300325;
Ribbach [1883] 910 and, more recently, Schironi [2004] 507510.
117 Further examples in von Velsen [1853] 9.
118 Baege [1882] 26 and others after him: overview of the studies in Boatti [2000] xviiixxi,
8083 and in Razzetti [2003e].
119 Baege [1882] 1112.
120 Baege [1882] 911.
121 Baege [1882] 1521, with the indispensable updates by Palmieri [1984].
122 Numerous examples in Baege [1882] 2730.
123 Baege [1882] 3031; Boatti [2000] 101103.
124 Schmidt [1854] 345349.
Language Correctness ( Hellenismos ) and Its Criteria 819
aboveand reaches the conclusion that since common usage is replete with
anomaly, the rules of analogy should be abandoned in favor of attention to the
forms of common usage ( 240). He likewise rejects the view that the crite-
rion for speaking good Greek rests on etymology, employed by grammarians to
determine not only correct spelling but also correct word meaning.138 Sextus
report is clearly polemical and presents an extreme picture of the grammar-
ians, who, as we have seen, by no means made exclusive recourse to analogy at
the expense of usage as a criterion of Hellenismos.
138 Sextus does not explicitly say how etymology was used in determination of Hellenismos:
Schenkeveld [1994] 289.
139 See Matthaios in this volume.
140 The main recent reference points on Apollonius are: Blank [1982]; Sluiter [1990]; Kemp
[1991] 316331; Blank [1993]; Blank [1994]; Lallot [1997]; Blank [2000]; Dalimier [2001];
Brandenburg [2005], all with further references. A constantly updated bibliography is on
the Internet website edited by A. Schmidhauser (http://schmidhauser.us/apollonius/).
141 Kemp [1991]; Blank [1993] 713; Schenkeveld [1994] 293298; Matthaios [2003]; Lambert
[2011].
142 See the list of Sud. 3422. Cf. Schneider [1910] viix; Blank [1993] 712713; Kemp [1991] 318;
Blank [2000] 414.
822 Pagani
150 Based on this comparison, D. Blank felt it was hardly surprising that the Syntaxis does not
offer a systematic treatment of all possible constructions for each part of speech, treating
only specific problems relating to doubtful constructions (Blank [1982] 710, with bibliog-
raphy; Blank [1993] 720; Schenkeveld [1994] 293; Blank [2000] 415).
151 See Frede [1977] 353ff.; Blank [1982] chapters 3 and 4; Blank [1993] 724ff.; Blank [2000] 415,
with additional bibliography.
152 On these concepts, see Blank [1982] 2728; Blank [1993] 724725.
824 Pagani
recognizes that it has entered into common usage and proposes an explana-
tion for this development, yet without branding it as incorrect. Usage becomes
the reference point in Apollonius approach, and is studied with the rational
tool of analogy. Finally, as can be inferred from the reference to the
, the literary tradition was a further element considered in Apollonius
system.153
In this context, he too awarded major importance to Homers poetry, from
which he took the overwhelming proportion of his examples.154 But unlike pre-
vious grammarians and his son Aelius Herodian (see below 5.2), Apollonius
did not seek solutions to textual or linguistic aspects of the Homeric poems:
rather, he used them to support grammatical rules.155 He made use of poetic
syntax to explain prose constructions (Synt. 3.166),156 or to invoke syntactic
rules found in prose to demonstrate that poetic syntax was a corruption of
these rules (Synt. 1.57):157 accordingly, Apollonius often explained unusual or
aberrant Homeric forms in relation to poetic license (cf. Synt. 1.62), thereby
absolving Homer of suspected solecism. Moreover, he sometimes defined typi-
cally Homeric forms as more ancient (), revealing his awareness
of diachronic development of language (Pron. 44.1113; Synt. 2.90).158
Apollonius great theoretical construction thus reconciles everyday lan-
guage usage with that of Homer and, more generally, of literature, and with
analogy seen as rational explanation of language phenomena.
153 Blank [1982] 61; Schenkeveld [1994] 298; Pontani [2011c] 99ff.
154 Lallot [1997] I 18.
155 Erbse [1960] 311370; Blank [1993] 718; Pontani [2011c] 99, 101.
156 Erbse [1960] 355; Blank [1993] 717718; Lallot [1995a] 119120; Lallot [1997] II 258259.
157 Erbse [1960] 360361; Blank [1993] 718; Viljamaa [1995] 176; Lallot [1997] II 3839.
158 Pontani [2011c] 101; Lallot [2011] 247248.
Language Correctness ( Hellenismos ) and Its Criteria 825
,
,
, , ...
, ... (GG III/I 24.1525.3).
159 The title is attested in two passages only: Apoll. Dysc. Synt. 4.10 (fr. 1 Cohn) and Ammon.
Diff. 336 Nickau (fr. 4 C.). It appears from the latter that the work probably comprised
more than one book.
160 It should be borne in mind that the content of A. Lentzs edition in the Grammatici Graeci
series is, similarly to that of Herodians other works, an erudite attempt to reconstruct
the grammarians thought rather than a genuine critical edition. It includes integration
of parts of other works of Herodian himself and even of other authors such as Stephanus
of Byzantium, Theodosius, Choeroboscus etc. (Lentz [18671870]). For an overview of the
problems posed by this method, cf. Dyck [1993a]. In addition to the above-mentioned
epitomes, there are another two that preserve portions of books 57 (ms. Vindob. Hist.
gr. 10: Hunger [1967a]) and of book 5, respectively (P.Ant. II 67: Wouters [1979] 216224).
161 Additionally, Hunger [1967a] 1415; cf. Dyck [1993a] 786.
826 Pagani
162 Herodians approach to the Homeric text, as compared to that of Apollonius, is studied by
Erbse [1960] 311370, esp. 363; cf. Siebenborn [1976] 77; Dyck [1993a] 784.
163 Cf. Broggiato [2003] 6769; Schironi [2003] 7577; Blank [2005] 222.
164 Lentzs reconstruction is based mainly on Choeroboscus scholia to the Canones of
Theodosius and on the material presented by Cramer [18351837] IV 333338. To this one
can add at least the epitomes contained in ms. Vindob. gr. 293 and in P.Flor. inv. 3005
(Wouters [1979] 231236): cf. Dyck [1993a] 789.
165 Cf. Siebenborn [1976] 73.
166 The most recent discussion on the content of this work is in Sluiter [2011], with bibliogra-
phy. Cf. also Siebenborn [1976] 84, 90, 108 and Dyck [1993a] 790791.
Language Correctness ( Hellenismos ) and Its Criteria 827
. (GG III/II 910.68).
Let this criterion of lexical singularity be frequent use among the ancients
and everyday usage, which sometimes represents a usage comparable to
that of the ancient Greeks.
167 This concept was already used by Varro (Ling. 10.82; cf. 9.53): see Fehling [1957] 67 and
Sluiter [2011] 294.
168 Lentzs distinction between this work (GG III/II 166388) and a
(Commentary of Didymus On linguistic changes: GG III/II 389) has long been
regarded as artificial. Furthermore, the insertion of numerous fragments of varied and
sometimes uncertain provenance, and even of authors other than Herodian, together with
an ordering of the material not matching that of the original, make it particularly difficult
to deal with the content of this treatise (cf. Dyck [1993a] 786788, with bibliography).
828 Pagani
dependent on the analogical method, and the behavior of certain terms was
quoted for comparison, in order to justify the behavior of others regarded as
analogous. Despite the ecdotic problems affecting the fragments of this work
(see n. 168), an examination of the vocabulary Herodian used for describing
word modifications suggests his system was more complex than the traditional
four-part system of addition, deletion, permutation and substitution.169
A work on solecism and barbarism, the two main defects of diction, is attrib-
uted to Herodian, but it is spurious (see below 5.3), as occurs for many other
works.170
169 On the system known as quadripertita ratio see Ax [1987]; for a study on Herodians
system, see Nifadopoulos [2003b] and [2005].
170 Reitzenstein [1897] 379ff.; Dyck [1993a].
171 Dihle [1977]; Swain [1996] 1764.
172 On the different meanings of this term in the Greek world over the centuries, see Versteegh
[1987].
173 On the socio-political aspects of Atticism and the relation between the rise of this ten-
dency and the spread of Latin, see Swain [1996] 3342, with previous bibliography.
174 Cf. Broggiato [2001 = 2006] xliixlvi.
Language Correctness ( Hellenismos ) and Its Criteria 829
175 For Aelius Dionysius and Pausanias see Erbse [1950], for Phrynichus see Fischer [1974], for
Julius Pollux see Bethe [19001937], for Moeris see Hansen [1998], for Philemon see Osann
[1821] 285301 and Reitzenstein [1897] 392396, for the Philetaeros see Dain [1954], for the
so-called Antiatticist see Bekker [18141821] I 75116 and III 10741077, for Oros see Alpers
[1981]. Additional bibliography is available in the respective LGGA index cards.
176 Sicking [1883] 2.
177 Swain [1996] 5152.
178 On the use of Homer as a linguistic model, see Pontani [2011c].
830 Pagani
179 Pontani [2011c] 96, with the associated Excursus on the textual tradition, pp. 102103.
180 Swain [1996] 5556.
181 Pfeiffer [1968] 228; Pontani [2011c] 9192, with further bibliography.
182 Probert [2011], with examples.
183 See Matthaios and Pontani in this volume.
184 Hilgard [18891894]; Robins [1993] 111115; Robins [2000] 418419; Matthaios [2002a];
Pagani [2006b]; Dickey [2007] 8384 (with bibliography). See also Dickey and Matthaios
in this volume.
Language Correctness ( Hellenismos ) and Its Criteria 831
The same topic was addressed, again with a teachers approach and style,
by Gregory of Corinth (12th13th c.) in his
(On discourse syntax, or on not making syntactic mistakes).193
His production further included a study on dialects ( ),194
which drew on the works of Tryphon (see above 4) and John Philoponus on
this topic.
It is far from easy to present a synchronic account of the criteria used in the
ancient world to determine language correctness, because they vary (as noted
above 1) depending on the source and were probably the object of extensive
debate; furthermore, their use undoubtedly changed over time. In this chapter,
observations on the rules of Hellenismos are outlined in the historical descrip-
tion (see above 25): they will now be explored individually, in order to
acquire an overall and systematic view.
6.1 Analogy
One crucial criterion was analogy,195 taken not only as the mechanism capable,
with a greater or lesser degree of refinement, of deriving information on the
correct form of a word (from the prosodic, orthographic or inflectional per-
spective) through comparison with another regarded as similar, but also as
the more general principle of rational regularity thought to preside over the
general language system.
In the Greek world, the concept of analogy originated in mathematical
and geometric studies, where it indicated a proportion between magnitudes
or terms.196 Regarding its application to grammar, the Greek sources attest to
over a dozen definitionsmainly of late dateand a further twenty or so are
documented by the Latin tradition.197 Indeed, the form of Varros reference to
the definitions of analogy provided by the grammarians Aristeas, Aristodemus
and Aristocles (Ling. 10.7475, see above 4) may suggest he was quoting the
beginning of a list, alphabetically ordered, conceivably much longer in the
eius [sc. analogiae] haec vis est, ut id quod dubium est ad aliquid simile de
quo non quaeritur referat, et incerta certis probet.
Concretely, this heuristic tool based on associating similar terms had various
realizations. It could involve a simple comparison among two members if the
doubt concerned prosody, morphology or grammar, taking as model another
word classified as analogous according to specific criteria (see below) and free
from any doubtful aspect; the accidents of the original term are then defined
greater restrictions on the ending, with identity required both in the nomina-
tive and vocative (Varro Ling. 8.68; 9.43; 9.91).211
A general overview of the situation between the 2nd and 1st c. BC can be
gleaned from Varros passage on genera similitudinis (Ling. 10.910), although
he may mistakenly have grouped together unrelated aspects. Varro contends
that for Dionysius of Sidon the conditions of analogy numbered seventy-one
if all possible distinctions were taken together, but only forty-seven if atten-
tion was limited to words having case, namely nouns. Aristocles is said to have
established fourteen conditions, Parmeniscus eight (see above 4), other
scholars various different counts. Dionysius number of seventy-one (or forty-
seven) conditions appears excessive, suggesting that Varro was referring not
to conditions of analogy but to identified through analogy, i.e. groups
of words sharing certain characteristics and inflectional rules.212 Moreover,
Varros statement regarding Aristocles could suggest he was identifying four-
teen classes of nouns, distinguished by their final syllable, rather than four-
teen conditions of analogy.213 On the other hand the number proposed by
Parmeniscus does correspond to the subsequent canonical count of analogy
conditions.
This number recurs in Caesar (fr. 11 Funaioli), where the conditions regard
nouns and are subdivided into those concerning accidents and those related
to syllable characteristics. Six are classified among accidents: qualitas (taken
as difference between noun and appellative), comparatio (alluding to degrees
of comparison characteristic of the adjective), genus (gender), numerus
(number), figura (distinction between simple and compound words), casus
(case); additionally, there is exitus syllabarum (ending) and paenultimarum
ratio (letters and quantity of the penultimate syllable).214
Varros position on the conditions of analogy may well have reflected the
number eight, but the passage in question (Ling. 10.2126) is affected by lacuna
at the end of 23, making interpretation of an already rather ambiguous text
somewhat uncertain.215 Varro mentions, firstly, four conditions: genus (gen-
der), species (distinction between noun and adjective, normally indicated by
qualitas), casus (case), exitus (ending) (Ling. 10.21); there follows a disquisi-
tion on the doctrine of inflection, totally unrelated to the problems of anal-
ogy conditions (Ling. 10.2223), and interrupted by the lacuna. The text then
resumes midway through a new list of conditions, which refers to number ( 24)
and cites figura (here indicating possible modifications at the beginning, end
or in the middle of a word, affecting the verb rather than the noun, 25) with
the related criterion sunt animadvertendae ... etiam quae proxumae sunt
neque moventur (that is to say, the syllables surrounding the modified ele-
ment should also be considered, 26). The lacuna eliminates the introduction
to the second sequence, possibly obscuring the reasons why Varro presented
two separate lists; it may also have obliterated the eighth condition. Equally
unclear is the reason behind Varros modification in the criterion of the figura,
resulting in a duplicate of exitus, at least for nouns, which show modifications
only in the final part.216
Some have argued217 that Varro drew on a source common to Caesar and
Aelius Herodian. Herodian provided a list and discussion of criteria for group-
ing nouns into classes when determining the rules for inflection, in a work
entitled (On noun inflection).218
, , ,
,
.
216 Fehling [1957] 7475, who explains: Aber auch hier hat Varro es nicht lassen knnen,
durch eigene Variationen und Erweiterung das in seiner Quelle Vorgefundene zu vern-
dern, und wiederum fhrt jede Vernderung zu Unstimmigkeiten und Fehlern (p. 74). Cf.
Siebenborn [1976] 74.
217 Fehling [1957] 75; cf. Fehling [1956] 246247.
218 Cf. Fehling [1956] 246247 n. 2; Siebenborn [1976] 73.
Language Correctness ( Hellenismos ) and Its Criteria 837
224 Siebenborn [1976] 23; cf. De Marco [1957]; Broggiato [2001 = 2006] xxxviii.
225 Ax [1995].
226 Several additional passages, likewise relevant to the theme, were cited above 3.2 and
3.3 and still others are found in Ax [1991] 291.
227 According to Siebenborn [1976] 97102, even anomaly understood in this sense (and not
as inflectional irregularity) is not unconnected with an opposition to the Alexandrian
analogy, since the latter tended to be applied, at least at the beginning, to terms similar
not only by virtue of formal but also semantic considerations. Besides, it is not clear from
Varros passage exactly what Crates position was (cf. Blank [1998] xxxv and n. 84; Blank
[2005] 236237).
Language Correctness ( Hellenismos ) and Its Criteria 839
6.2 Usage
Common usage (),233 taken in a schematic and manichean vision of
ancient tradition as a criterion antithetical to the analogical method, was nev-
ertheless employed together and in interaction with the latter.234 One of the
criticisms of the analogical method lay precisely in its recourse to usage for
examples in devising its rules. The polemical conclusion of Sextus Empiricus
was that analogy thus became a pointless artifice, to be abandoned in favor of
observing speakers common usage (see above 4). Grammarians, however, or
some grammarians, reconciled the two methods by combining them together.
Such was apparently the approach of Aristarchus of Samothrace, who
according to Varro (Ling. 9.1)applied the rules of analogy within the lim-
its allowed by common usage. Predominant attention to observing usage
( ) for determination of correct Greek is traditionally
attributed to Crates of Mallus, but this may have formed part of reflection on
applicability of the analogical method without necessarily involving its rejec-
tion (see above 3.3 and 6.1.2). The value of usage in assessing Hellenismos
228 The bibliography on the topic has become very extensive: here mention is made only
of the main contributions, to which the reader is referred for in-depth information and
further references.
229 Lersch [18381841].
230 Steinthal [189018912]; Colson [1919]; Barwick [1922].
231 Fehling [19561957]; cf. the interpretation of Ax [1991] 293294.
232 Collart [1963] 129ff.; Pinborg [1975] 106112; Blank [1982] 14; Taylor [1987] 68; Schenkeveld
[1990] 293; Ax [1991] 294295; Blank [1994] 149158; Schenkeveld [1994], with caution; Ax
[1996] 117118.
233 According to Versteegh [1987] 260264, who recognizes a situation of diglossia in the
Greek world, as also in the Latin and Arabic world, this term was used by grammarians to
indicate the standard use, on an elevated level, of the language of their time and not the
colloquial usage of the common people.
234 Siebenborn [1976] 5355, 9092.
840 Pagani
239 Examples from Scipio, Antonius Gnipho, Caesar and Varro in Siebenborn [1976] 111115.
240 Siebenborn [1976] 116.
241 On the complex relations between the two parameters, cf. Siebenborn [1976] 2731,
8589; Sluiter [1990] 6061; Viljamaa [1995]; Pontani [2011c] 99.
242 Siebenborn [1976] 8591.
243 Siebenborn [1976] 9091.
244 Naturally, mentions of other authors are not lacking: thus Tryphon (see above 4), in
seeking to determine the meaning of certain words, included not only the authority of
Homer but also Pindar (fr. 105 V.) and Xenophon (fr. 106 V.): cf. Siebenborn [1976] 89. On
Homer as the reference point of Hellenismos, cf. Pontani [2011c].
842 Pagani
correctness are perfectly accomplished (fr. 125 Matthaios, see above 3.2).
Apparent oddities in Homers language compared to common usage were jus-
tified by the grammarian as poetic licenses (). Similarly, Aristarchus
pupil Ptolemy Pindarion challenged empirically-oriented criticisms against
analogy, objecting that the use accepted by everyone on which analogy was
founded was that of Homers poetry. But this argument, according to Sextus
Empiricus, who handed down and criticized Pindarions position, did not solve
the problem of whether the reference point for correct Greek should be usage
or analogy: rather, it further complicated the problem by introducing the addi-
tional parameter of Homeric usage. Moreover, it involved the risk of ridiculous
modes of expression, by reproducing peculiarities of Homeric language fallen
into disuse (Math. 1.205ff.).245 The idea that Pindarion genuinely intended to
promote Homeric usage in everyday life is undoubtedly an excess due to the
provocative tone of Sextus argument,246 which, on the other hand, is emblem-
atic of the overlapping and possible conflicts among the criteria of Hellenismos
(see above 4).
A position akin to Pindarions also emerges from a problematic passage of
Ps.-Herodians De soloecismo et barbarismo.
, ,
, ...
Some say that Hellenismos is the Poet, others that it is the common lan-
guage which arose when the Greeks gathered at Troy, still others etymol-
ogy (311.510 Nauck).
The statement that Hellenismos is the Poet seems to allude to the predomi-
nance of Homeric authority as the criterion of good Greek, presumably with-
out implying adoption of Homeric language in everyday usage.247 However, the
text of the manuscripts has raised doubts among scholars. J. F. Boissonade con-
jectured (sc. ) (the art of poetry), instead of ,
while A. Nauck proposed, albeit doubtfully, the alternative (the
[language?] of the poets).248 In both cases the criterion of literary authority
appears to be invoked generically, without explicit reference to Homer. Likewise
interesting is the idea that Hellenismos resided in the common language of the
Achaean heroes at Troy. That the Greek warriors all spoke the same language,
unlike the case of their enemies on Trojan territory, is repeatedly underlined
in the scholia. This concept lies on a different plane in comparison to the view,
itself widespread in antiquity, that Homer as a poet combined different vari-
eties of Greek into a common language.249 However, there is some suspicion
that the redaction of the paragraph in question, which is not present in manu-
script evidence from before the beginning of the 15th c., is of a late date250 (see
above 5.3).
General reference to literary usage and to Homer in particular also clearly
emerges in the work of Apollonius Dyscolus,251 who, having no interest in
solving textual or linguistic problems in works of the ancients, used the latter,
especially the Homeric poems, as a source of examples to support grammati-
cal rules. The predominance of Homer is evident in Apollonius repartee to a
grammarian of the Augustan age called Habron, who supported a linguistic
position by quoting examples from Plato diverging from Homeric usage: but,
according to Apollonius, the existence of a use in Plato is no more trustwor-
thy than its absence in Homer (Pron. 72.1519). Furthermore, following the
path indicated by Aristarchus, Apollonius reacted to the objection against the
adoption of Homer as a model of Hellenismos by justifying the unusual forms
or constructions in Homers language either as poetic license () or due to
their antiquity (see above 5.1).
A phenomenon in which, between the 2nd and 3rd c. AD, the literary tradi-
tion played a crucial role was linguistic Atticism, whereby the idea of language
correctness was equated, tout-court, with adoption of a literary model. Here
the reference point consisted of a canon covering a certain number of classi-
cal Attic authors, of which numerous versions existed; it sometimes included
Homer as an example of a proto-Attic author (see above 5.3).
The corresponding Latin criterion was identified as auctoritas,252 with the
most ancient description dating back to Varro (fr. 268 Funaioli). Analogously
to the Greek sphere, auctoritas was understood as the final rule of language
correctness, to be invoked when all others failed. Varro reasoned that this was
because auctoritas derived from nothing other than an opinion founded on a
reading of the ancient writers, who, challenged to justify their language choices,
259 On the utilization of etymology in Cratylus see Sedley [1998b] and [2003a], Aronadio
[2011] 83183.
260 Cf. Ax [1987].
261 See Sluiter [1990] 1213, 1821; Allen [2005].
262 Thus Schironi [2003].
263 For observations on philologists use of etymology see Lallot [1991a] and [1991b] and, spe-
cifically on Aristarchus and Crates, Broggiato [2003] and Schironi [2003].
846 Pagani
267 On the status of etymology in Varros thought, see above all Blank [2008].
268 Siebenborn [1976] 140, 146.
848 Pagani
of methods, yet without achieving anything more than the most hideous
absurdities.269 He thus held a decidedly disenchanted attitude, which, how-
ever, was destined to be superseded in later centuries when the Late-Antique
and Byzantine world experienced a flourishing of studies and anthologies of
an etymological nature.
7 Conclusions
In what cases should one or the other parameter be made to prevail? This
dilemma was an easy butt of empirically inspired polemics against the quest
for Hellenismos (Sextus Empiricus, see 4): why appeal to criteria such as
analogy and etymology when, demonstrably, the reference point of both of
these was usage, and why invoke the literary authority if this results in forms
or constructions so distant from usage as to appear ridiculous? The art of
combining the different aspects without grotesque distortions was a tribute
to the grammarians intelligence. It was successfully achieved, perhaps, by the
brilliant intellect of Apollonius Dyscolus, who devised a system which, while
solidly grounded in rational knowledge of language and its rules (explicitly
rejecting a purely practical approach: Synt. 1.60), admittedalbeit, not with-
out painstaking reconstructionsirregular forms that had been accepted
into general usage; at the same time he assumed a constructive and far from
uncritical attitude towards the auctoritas of Homer and the literary tradition
(see above 5.1). Furthermore, the coexistence of rules and exceptionsof
what can be explained rationally according to the rules, alongside what is right
because thats the way its usedis a quandary that inevitably confounds the
approach to any language system. Here we have endeavored to show the kind
of answers offered in this regard by the classical world.
Chapter 3
Syntax
Jean Lallot
1 Introduction: -
1.1 Syntax and Grammar
1.2 Apollonius, Obligatory Reference
1.3 The Programme of the Syntax (Synt. I 1)
1.4 /
2 Congruence ()
2.1 Conjoined Signifieds ( )
2.2 Solecism
2.2.1 Error of Syntax and Error of Deixis
2.2.2 A Limit of Incongruence: The Formal Coincidence ()
3 The Scheme of Functions
3.1 Noun and Verb: A Natural Harmony
3.2 Accompaniment and Replacement
3.3 Transposition ()
3.4 Syntactic Relations
4 The Syntactic Domains
4.1 The Domain of the Noun: Article and Pronoun
4.2 The Domain of the Verb
4.2.1 The Programme
4.2.2 The Infinitive, The Most General Mood; Compositionality
4.2.3 Diatheses
4.2.4 Syntax of the Cases
4.3 Uninflected Words
4.3.1 The Adverb
4.3.2 The Preposition
4.3.3 The Conjunction and the Complex Sentence
5 Syntax Between Logic and Philology
5.1
5.2
5.3 Norm and Figure
1 Introduction: -
(...)
2. the discipline which treats the phonology, morphology and derivation of a language, also
the corresponding manual; 3. a work which presents the rules describing the phonological,
morphological and syntactic structure of a language. It is pleasant to state that this last
definition is illustrated by an example: The Neohelllenic Grammar of M. Triantafillidisa
work which does not include syntax!).
3 We will of course occasionally refer to what the authors other works tell us about syntax:
the treatise On Conjunctions ( ), valuable for the syntax of complex phrases,
the treatises On the Pronoun ( ) and On Adverbs ( ), certain
theoretical parts of which complete the content of the Syntax.
4 Priscian repeatedly calls Apollonius the greatest (summus, maximus) Greek grammarian;
cf. IG III 24.6: Graecorum doctissimi..., et maxime Apollonius, cuius auctoritatem in omnibus
sequendam putaui. the most learned of the Greeks..., and above all Apollonius, whose
authority I have thought should be followed in all matters.
Syntax 853
, ,
,
.
5 This symmetrical relation between grammar and syntax clearly emerges from the passage of
the Techne (chapter 11) where we find definitions of word (, a term freely varied with
in the Techne) and sentence ():
. A word is the smallest
part of a properly constructed sentence ( , lit. according to the [correct]
arrangement). A sentence is a combination (, lit. put together) of words in prose
conveying a meaning which is complete in itself (Kemps translation). We will return later
to the coexistence, in this double definition, of and .
6 This epithet designates the systematic part of the grammatical discipline, which is recorded
in the works, called , composed by the . These works, which aim at the
methodical description of the constituent elements of (what we call) the language, are
distinguished from another category of grammatical works, those devoted to the study of
texts (typically, the of Aristarchus). For definitions of Techne grammatike and
typology of grammatical treatises see, respectively, Swiggers-Wouters (section II) and Valente
(section II) in this volume.
7 See Montana, Swiggers-Wouters (section II), and Pagani in this volume.
854 Lallot
1.4 /
The word , which gives its name to Apollonius treatise, and which sur-
vives as a borrowing in modern languages, is found once in competition with
. In chapter 11 of the Techne (see n. 5), and seem to
coexist in free variation. The scholia on this passage are at a loss. Sometimes
they condemn an improper usage of , which they say should be
applied only to the composition of nouns, the correct term for the juxtaposi-
tion of words being (Sch. Dion. T. 355.24). Elsewhere they construct
an (ad hoc?) opposition between two terms, where refers only to the
agreement of parts of speech while refers to the structure of thought:
(ibid., 214.9). It is more probable that the two
definitions which match (part of ) and (formed from ),
perhaps inherited from distinct sources, both retain the traces of a former syn-
onymy, which later usage eliminated in principle by specifying in the
sense of nominal composition. Despite this, a variation in the use of the two
terms has persistedsee, in Apollonius himself, Synt. III 14, 280.3, Conj. 214.7,
221.19. We recall also that Dionysius of Halicarnassus treatise on stylistic com-
position is entitled .
etymologically includes the idea of order or rank (), specified by
the associative prefix -, a little redundant in this case, which adds the sense
that the order governs the arrangement of a whole. Apollonius did not invent
the word. As an action-noun derived from the verb put in order
together (LSJ), is well attested in the classical era in various senses
(military, political, literary, etc.), all related by the common seme of ordered
arrangementthe formation of an army, institutional structure of a rgime,
organisation of the constituent parts of a text, and so on. The Stoic philoso-
pher Chrysippus had written a treatise in four books entitled
, without any doubt devoted to the logical analysis of proposi-
tions. A similar use of in a linguistic sense prepared the word for its
8 This rooting of grammar in philology doubtless explains a good part of why Alexandrian
grammar was exclusively a grammar of Greek. The questions to be resolved were posed by
Greek texts, and the scholars who attempted to answer them, even if they were polyglot in
Hellenistic and Roman Alexandria, never dreamed of leaving the strictly Hellenic field which
defined their corpus.
9 See Pagani in this volume.
Syntax 855
2 Congruence ()
10 De constructione sive ordinatione is the title given by Priscian to Books XVII and XVIII
of his Grammatical Institutions, in which he treats a great part of the material of the
of Apollonius. The more faithful Latin calque of would have
been coordinatio, but one does not find this word among the Latin grammarians; the
introduction of coordination into French grammatical vocabulary, notably in the syntagm
conjonction de coordination, is very late (1888).
856 Lallot
no choice but to find out what it means for ourselves. Many modern interpret-
ers have understood it as an altered form of the Stoic or expressibles.11
Although plausible in itself, this connection is only of limited help to us, since
it does not explain why Apollonius might have substituted the one word for the
other, abandoning the Stoic usage of a technical term of dialectics which has
no place in grammar.12 On the other hand, Apollonius himself somewhat clari-
fies the term by inserting it, a little later, into the syntagm
the which is the signified conjoined to
each word (Synt. I 2, 2.10). What is he talking about?
Without any possible doubt, he is referring to the lexical signification of
words constructed to form an intelligible utterance. But also, at a more abstract
level, to the categorial signifieds attached to the parts of the sentence such as
they are. The partition of the effectively produces lexical entities which
are differentiated functionally and predisposed to play a certain role in syntac-
tic construction. Thus it is because two words are a noun and a verb that they
can combine to form a complete proposition: (a) man fell.
Again, it is because is a postpositive article (i.e., a relative pronoun) that it
can connect a noun to a secondary verb which follows it:
, a grammarian has arrived who has argued (Synt. I143) it is because
is a preposition that it can precede a noun, either in juxtaposition:
, against (the) law, or in composition, , illegal (Synt. IV13),
and so on. No congruence is possible if we are unaware of the functional pre-
dispositions of the parts of the sentencefor instance, if we put a prepositive
article in place of the postpositive: * , * a gram-
marian has arrived the has argued.
But this is not all: the signifieds conjoined () to words include
also the grammatical significations linked to the accidents () of
the parts of the sentence, and notably those which are inflected. The inflecting
form indicates number, case and gender for the constituents of the nominal
group (noun, participle, article, pronoun), and person, number, tense, mood
and diathesis for the verb (and partially for the participle). Bearing these
11 Cf. Frede [1987a] 354f.: In translation, Apollonius would say: corresponding to each word
there is an element in the lekton; in putting the words together we put the elements of
the lekton together, i.e., construct a lekton. Whether we get a syntactically proper sentence
depends on whether the lekton we construct satisfies the syntax of the lekta.
12 is absent from the Syntax. Apollonius uses it sometimes in his other surviving
treatises, always in the same kind of context: when he underlines that a phonetic accident
(typically, aphaeresis of an initial vowel, for instance in for ) affects only the
form () of a word, never its sense ().
Syntax 857
significations, the inflected forms when constructed together must obey the
rules of what we would call co-occurrence. Although the distinction is not
made by Apollonius, it may be a question of the rules concerning either, to use
the terms of Port-Royal (Arnauld-Lancelot [1660] chap. 24), the syntax of con-
venance (agreement in case and number between an adjective and the noun
which it determines, or in person and number between a verb and its nomina-
tive term (subject); restriction in the usage of adverbs of time depending on
the verbal tense;13 etc.) or the syntax of rgime (distribution of the oblique
cases assigned to verbal or nominal determination; governance of preposi-
tions, etc.). The fundamental cause of incongruence (
), Apollonius tells us (Synt. III13, 279.5), lies in the incorrect com-
bination of inflected forms.
2.2 Solecism
This incorrect combination is called solecism ().14 The verbs
and have been known since Herodotus (III 57, IV 117),
where they mean to speak (Greek) badly. In Aristotle and
also denote vices of language, the former an abuse of rare words
(Poet. 1458a26), the latter an incorrect, barbarous utterance (Soph. el. 165b20).
It is apparently with the Stoics that the two words formed a contrasted couple
(cf. D. L. VII 59), later adopted by the grammarians. Barbarism is when a lexical
form deviates from good usage, solecism when a proposition is constructed in
an incongruent manner: (Diog. Bab. ap. D. L.,
l.c.), the Stoic formula faithfully reflected in Apollonius (Synt. I8, 7.12):
.
Reflection on the solecism leads Apollonius to observations of the great-
est importance for syntactic theory. The one concerns the limits of syntax, the
other the limits of incongruence.
13 In this last case, the congruence is established between verbal inflection and adverbial
lexemes.
14 Cf. Pagani in this volume.
858 Lallot
Incongruence and congruence lie not in the referents, but in the con-
struction of words which must adapt their form to the necessities of cor-
rection (275.6)
(...) ,
, .
15 Hardly convincing in itself, this explanation loses all relevance when the verb is intransi-
tive and so forbids any interpretation of the neutral form as a complement of the object:
= (s)he runs the children. Thierfelder [1935] n. 23 reproaches Apollonius
for wanting at all costs to explain a construction which is completely irrational in the eyes
of a scholar ignorant of comparative grammar. One imagines that Apollonius, ignorant of
comparative grammar, could ask the modern grammarians, who are familiar with it, if it is
fully satisfying to explain a disconcerting common turn solely by comparative diachrony.
16 Apollonius studies a long series of them, Synt. III 2234.
Syntax 859
,
.
.
The object of syntax is, as we have seen, the study of the (understood as
an assemblage of words to form a complete sense) (literally that
which is its own end, or in other words, that which has no place for any sus-
pense, that which is fastenedthe metaphor of closure often designating
syntactic completeness, for instance at Synt. I14, 16.13, on the minimal form
of noun (subject) + verb: without these
(two constituents) no is fastened). Thus described, the corre-
sponds rather well to what we call a sentence, A series of words in connected
speech or writing, forming the grammatically complete expression of a single
thought (OED). Hence the translation chosen here of as part of
the sentence, rather than part of speech, the traditional calque of the Latin
pars orationis.
We have seen above that the parts of the sentence are not a raw material,
but rather that each is associated with what we labelled earlier functional pre-
disposition. We must now attempt to define how the Alexandrian grammarian
860 Lallot
there is between the noun and the verb a natural harmony which does
not involve conjunction (...); just as, in effect, form and matter are linked
to each other without conjunction, so the noun has a natural harmony
with the verb (Sch. Dion. T., 515, 36)
( (...).
,
).
Given this ontologically primary and natural link, the entire syntax of a sen-
tence is described as the result of a process of enrichment by accretion, in
which all other parts of speech find their usage and justification.
17 The term function is used here as a handy and familiar generic word for the modern
reader. But we must point out that no single term corresponds to it in the Alexandrian
metalanguage.
18 This metaphor is not at all anachronistic: ancient grammarians happily compared the
sentence to a material artefactnotably to a ship, in which they distinguished the parts
properly called (the hull, the rudder, the tackle) from the joining materials (pitch, oakum,
nails): likewise the sentence includes parts properly called and conjunctive elements
(Sch. Dion. T., 515, 21).
Syntax 861
Since all the other parts of the sentence depend on the construction of
the verb and the noun (...), we must examine the usage of each, whether
it is to accompany or replace them, or both, as for instance, pronouns can
replace nouns and accompany them, or again, participles can replace
verbs and accompany them, and so likewise for the others parts of the
sentence. (Synt. I36, 33.9)
(...),
,
,
, .
adverb predicates (, Adv. 120.19) the forms of the verb, that is,
brings a complement to the main verbal predicate.
The function of replacement is fulfilled by the participle (), a form
with nominal inflection constructed on a verbal root, which puts its noun-
like morphosyntax in the service of verbal expression. If Agamemnon
fought (), and then vanquished (), a syntactic rule for-
bids juxtaposing the two verbs in asyndeton *
*Agamemnon fought vanquished, but the participle, by replac-
ing the first verb, solves the difficulty:
Agamemnon, having fought, vanquished,20 for nothing prevents the
juxtaposition of two declined words in asyndeton:
the blond Menelaus fought.21 We can also see that, although
it came about to replace the verb, the participle, thanks to its nominal
inflection, acquired the ability to accompany the noun (
) and the verb ( )simultaneously; this latter
capacity of the participle is indicated succinctly in Synt. I 36, cited earlier:
the participles can replace the verbs and accompany them.
At the last count, all these parts of the sentence appear subject, in their syn-
tactic employment, to a small number of distributive principles: the noun and
the verb are naturally predisposed to associate with each other to form the
in the description of the adverbial function highlights the great relevance of the
syntactic criterion when trying to identify the adverb.
20 The example is taken from Sch. Dion. T., 415.27. On the service rendered by the participle,
see also Synt. I 136137.
21 This last example highlights in passing that there is, among the nouns, a sub-class which
has the purpose of accompanying other nouns and that, therefore, while Alexandrian
grammar did not isolate the adjective to make it an autonomous part of the sentence, it
nonetheless recognised a difference in function between two kinds of noun, later called
substantive and adjective noun. It is only because is of the second kind that it
can be juxtaposed with , which is of the first. It will be noted that substantive,
as a designation of a kind of noun, is foreign to Alexandrian grammar, and it is only for
convenience that we here use this anachronistic term.
Syntax 863
predicative kernel of the sentence; the participle and the pronoun serve the
function of replacement; and the article, adverb and preposition function as
accompaniments. All, that is, except one: the conjunction stands alone, its
specific function being not to assemble words into sentences, but to join sen-
tences together (Synt. I14, 17.5).
3.3 Transposition ()
The functional scheme just presented may give the impression of a somewhat
mechanistic analysis of syntactic functioning: one place for each word (part of
the sentence), one word for each placethe taxonomy () conditions
the syntax. But a corrective to this mechanism is manifested when the syntax
dislodges the , that is, when one word changes its class as a result of its
construction. A typical case is that of the article deprived of the noun which
it normally accompanies, for instance Il. I 12
he [lit. the] came to the swift ships of the Achaeans, where, according to
Apollonius (Synt. II31), the ellipsis of the name from the base phrase
...(lit. the Chryses came...) has caused a transposition
() of the article into a pronoun. At 33 of the same book, Apollonius
enumerates other cases of parts of the sentence which, abandoning the con-
struction which is proper to them ( ) to assume
the [functions] proper to another, receive the appellation of the latter: nouns
(adjectives or appellatives) becoming adverbs very well, in
private, around in a circleparticiples becoming nouns
destinyetc. The transposition thus introduces an element of fluidity into
the syntactic mechanism.
22 Rather unspecific notions expressed by various verbs and little or not at all differentiated
from a semantic perspective cf. Lallot [1997] I, 67.
23 The plan of Book IV of the Syntax is another example. The list of possible sequences
prposition + such or such parts of the sentence shows clearly that, by construction,
one must understand here the material contiguity of the preposition with the word that
follows, even if it is an article (IV 54), without asking if any kinds of rapport other than
contiguity might be established between two contiguous words, and, if so, of what nature.
Syntax 865
despite the fact that the pronoun is substituted for the noun so that, in
being connected to the verb in turn, it holds the sentence together. An
illuminating explanation would be that the pronouns were invented to
agree with the verbs. (I19, 20.3)
.
,
.
This is an allusion to the account given in Synt. II 4044 of the origin of the
pronouns, invented to replace nouns incapable of entering into with
The letters in both the direct case and the accusative can be employed
either without or with the article.
1) Without the article, when we say , this is
(an) alpha, that is (a) beta, with the direct case understood, as when
we say this is (a) man (nomin.),
that is (a) horse (nomin.). This we learn also from the agree-
ment of the verb, which goes with the direct case. Now the accusa-
tive: , the teacher calls
this alpha, that beta. Here the transitivity of the verb applies to
that beta and the letter is conceived in the accusative, as
if we should say: the
teacher is indicating this letter (acc.).
2) With the article. When we say the alpha admits of
two quantities, , the alpha is a
final letter of feminines and neuters, and, in the accusative:
the child rubbed out the alpha, this corresponds to
he rubbed out the letter (acc.). (Synt. I46)
. , , , ,
, , .
, .
, , ,
, ,
Syntax 867
. , ,
, , ,
, , .
more than the proper name, which is vulnerable to homonymy (I 121). The
power of the pronoun is diminished, however, when the person it designates
is absent; this is the case, for instance, in a letter, where the referents of I and
you are identifiable only by anaphoric recourse to the proper names appearing
in the initial address Dionysius to Tryphon, greetings! (II 11). On the other
hand, thanks to its characteristic double inflection, both tonic and atonic, the
pronoun allows us to distinguish absolute reference he struck
mefrom contrastive () reference its me
that he struck. Pronominal syntax also permits the expression of reflexivity,
when the agent and patient of a process are conflated. Classical Greek uses
composite pronouns (-/-/-) for this, although Homeric Greek lacks
such forms. An examination of Homeric passages involving reflexive construc-
tions prompts Apollonius to deny the existence of any specifically reflexive
form in the poet; the technical study of this question (which has been difficult
and controversial since antiquity) provides a nice example of the close overlap
between philological procedure (the interpretation of written forms, problems
of accentuation) and syntactic analysis (see Synt. II 91102). Apollonius again
highlights an interesting function of the pronoun, in its anaphoric form: that
of introducing some form of connectedness into a story. He gives the example
of Il. 13. 1 and 3:
, (...)
... ... (Synt. II 8)
When Zeus let the Trojans and Hector approach the ships, (...) he turned
away (from them) his shining eyes...
,
, .
.
Syntax 871
Here the grammarian goes beyond the topic of simple intraphrasal congru-
ence to concern himself with what we would call text grammar.
27 Morphology also plays a part: the invocation of the augmented forms and is a
strong argument in favour of the verbalstatus of and (Synt. III 73).
28 The subjunctive is not a mood like the others: it consists in the product constructed
() by a conjunction (especially or ) with indicative forms, which next
Syntax 873
in all the tenses attested for the indicative, while the imperative (103115)
does not exist for all persons. Why? The responses are here of a pragmatic
nature: wishes and commands are not easily compatible with the past, and
commands can only be addressed to a second personhence the temporal
defectiveness of the optative, and the temporal and personal defectiveness
of the imperative. But that is only a first, rough approximation of the answer.
In the domain of tense, the morphology contains optative and imperative
forms constructed on the temporal stems of the perfect and aorist, stems
which, in the indicative, refer to the past. Apollonius admits, then, that there
are orders and wishes in the past (), but in paraphrasing these
forms, he highlights their perfective () aspectual component; thus
his description of the paradoxical combinations mood + tense leads him
to the category of aspect, although still imperfectly and without naming it.
Another paradox is offered by injunctions in the first and third persons:
let us flee, ... Well! let me say... (Il. 9.60),
let him say. This last form, which in Greek is morphologically an
imperative, is as it were reduced to the norm by Apollonius, for whom it
combines two persons, a third and a second, the latter inherent in the
imperative:
,
,
.
4.2.3 Diatheses
The verbal category of diathesis is not an invention of the grammarians. It is
tempting to follow Benvenistes suggestion ([1966a] 67 f.) that this category
underpins the three Aristotelian categories of to make (active),
to suffer (passive), to be in a position (middle). Among the
Stoics (D. L. VII 6465), it forms the basis of the classification of the predicates
into direct (, ex. , , ), reversed (, ex. ,
), neuter (, ex. , ), reflexive (lit. anti-passive
, ex. he shaves (himself)). It will be noted that, even if
Greek morphology has a certain place in this classification (/ ),
it is not the basis of it (!), as the descriptions recorded by Diogenes
confirm: the direct predicates are those which are constructed with an oblique
case, the reversed with the passive particle,30 the neuters with neither of the
two, and the reflexives are those which being reversed, are active. Except the
reflexive, all these descriptions are syntactic. We will not be surprised, then,
that Apollonius, whose debt to the Stoa nobody doubts, devotes a section of
his Syntax to diatheses.
The term itself, diathesis, merits some consideration. For Aristotle (Cat., e.g.
8b35), is the substantive corresponding to the middle intransitive verb
to be in a state. The word thus designates a state, a disposition (cor-
poral posture, intellectual or moral dispositions, etc.). But disposition, Aristotle
also tells us (Cat. 6b2), is relative; thus knowledge is knowledge of something
or someone, and position is the position of something or someone (
). To put it another way, disposition is a
qualified relation31 involving one or more terms: knowledge is the knowledge
of someone about something, position is that of a body in position. Although
this may seem a truism, it is in fact of capital importance for understanding
the grammatical category of diathesis. It appears in effect that, for Apollonius,
dsignates32 the semantic content of the verb, insofar as it establishes
the relation between the participants of the signified process, called persons
().33 Whereas in the Categories, we find beside only the middle
stative verb , and never the active transitive to dispose, for
Apollonius this verb, attested in both the active and the passive, is a key term
of verbal syntax. For the grammarian, every verb form possesses, in correlation
to its semantic content, a diathetic schema which organises the dispositions
of the persons involved in the signified processin other words, the syntax
of the nouns (or pronouns) designating the actants. The base couple is here
that of the persons respectively disposed and disposing (
, Synt. II 141, 237.4, cf. II70, 177.15, Pron. 45.19, etc.), corresponding
referentially to the patient and the agent of a transitive process. This binary
schema allows Apollonius to describe the active transitive construction (
I thrashed you) and its symmetrical passive ( You
were thrashed by me), as well as the particular cases represented by 1) the
reflexive construction, where the two persons become one (fastened transi-
tivity: / I struck myself / I was struck by
myself) (Synt. III 141), and 2) the intransitive construction which produces
a complete sentence without any second actant ( Trypho
walks) (Synt. III 155). The diathesis is at the same time that of the verb
whether active or passive in form ( / ), that of
the disposing / disposed persons (/ ), and that
I 137, 112.8: the oblique cases are connected with the direct cases by the
intermediary of a verb whose diathesis, originating in the direct case
associated with it, passes to the oblique.
,
.
II 29, 148. 3: as for the oblique cases, they form the end of a construction
which begins with the direct cases, the intermediary verbs indicating the
diathesis of each.
,
.
The syntactic schema here described is the following: two case-marked actants
are related to one another by an intermediary verb (cf.
) which is a sort of conveyor of diathesis between the originating
actant, in a direct case, and the receiving actant (cf. ), in an oblique
case. Between the verb and the direct case, we discover, expressed here by the
verb (cf. ), the privileged relation of
signalled earlier ( 3.4). An active transitive sentence, for instance
, Dion strikes Dionysius, thus reflects in a quasi-iconic way a sce-
nario of action, oriented dynamically between an active protagonist (Dion who
strikes) and a passive protagonist (Dionysius who receives the blows); between
the two is the process with its source in the first, crossing over to be applied to
the second (cf. ). The verb in the active voice thus shows
that it shares the diathesis of the actant in the direct case (). The pas-
sive sentence Dionysius is-struck by Dion, in which the Stoics saw a reversed
() predicate, presents the same scenario inverted (as if seen in a mirror),
but with the same syntactic cohesion: the verb in the passive diathesis still
shares the diathesis of the actant in the direct case, in this instance the man
who receives the blows.
There is little to say about the middle () diathesis. Its name signifies
that it occupies an intermediary position between the active and the passive.
But what does this mean? It should be noted that Alexandrian grammar for-
Syntax 877
34 A conventional designation, by the endings of the 1st sing. of the present indicative,
above all a paradigm which evidently presents a great variety of other forms. What is
important here is that the morphological analysis of the ancients (and largely also that of
the moderns) assigns on principle one diathesis and one alone to a given personal ending:
thus, for the 1st sing., an ending in - (, ) is as a ruleactive, a form in -,
-() (, , ) as a rulepassive.
35 There is a great temptation to compare this Stoic formula to that of Benveniste
([1966b]172): in the middle diathesis, (the subject) lies within the process of which it is
the agent. That said, the Alexandrian grammarians, unlike Benveniste, did not see in the
middle an internal diathesis opposed to the active external diathesis; they say nothing
about pairs like to establish the laws to give oneself
laws and admit without qualms that the middle says nothing more than the active.
878 Lallot
The senses receive influences from without, so that a sound which pen-
etrates the ear affects also our entire bodyeven against our will: the
noises of saws or thunder are hardly sounds that the ear accepts willingly!
As we have said, the construction in the genitive gets close to passivity,
but we are not dealing with a construction with [by], since an activity
also exists which results from the diathesis. The sense of touch involves
an action, and a passive counter-diathesis upon contact with hot, cold or
other [objects]. It is the same for the senses of smell and taste... (Synt.
III170, 417.3)
,
. ,
,
,
.
, ...
As in the passive turn with , but less decisively, the genitive characterises
an actant which participates in the active mood and thus introduces into the
36 In the Greek tradition we catch a glimpse of speculations where the presence of the geni-
tive with in the passive construction is invoked to establish a proximity between the
genitive and the direct case, and so to explain its place immediately after the direct case
in the canonical list of cases (Sch. Dion. T. 548.37).
880 Lallot
The verbs which refer to a symmetrical diathesis, that is, to the equal
activity of two persons directed towards each other, take the dative. This
is the case with I fight with you (Synt. III185, 427.9)
,
, , .
because the passive diathesis has already been assigned to the genitive.
Thus, the construction of these verbs will have to give up the two cases
[sc. accusative and genitive] and admit only the dativewhich is ready
[to express] the reciprocality of the diatheses (Synt. III186, 428.13)
.
,
, .
Here we note the two kinds of reasons invoked by Apollonius. The first, which
is somewhat surprising to us, is that to express symmetrical diathesis we have
no choice but to settle for the last available case in the system; syntax is con-
ceived as the customer of a morphological shop whose limited resources must
content its clientele! But its wares are well made, and this is the second reason:
the dative is precisely the case of rendering service to another (, cf.
Synt. III177178), and so it may easily () be used to express the reci-
procity () characteristic of symmetrical diathesis.
Armed with this interpretive schema (which not only explains to him the
entirety of case syntax, but is intended to account for the case of the second
actant in the transitive scenario), Apollonius examines a great variety of verbs
whose constructions fit it more or less well. This gives him the opportunity to
reveal fine nuances in the semantic content of the verb, for example in distin-
guishing the diathesis of two verbs to love, manifested in their construction:
+ accusative designates an active love with no implication of reciprocity,
while + genitive implies that the lover is the object of a counter-diathesis
from the loved one (Synt. III172). Among the verbs of perception, the original-
Syntax 881
By contrast (to the other senses), the diathesis of the verb to see is
highly active and has a stronger transitivity, as witnessed by the verse
(Il. 23.477):
...and your eyes do not throw forth from your head a very piercing
glance.
Our sight is not easily affected by counter-sensations from without, for
we can shut out this supplementary diathesis by closing the eyes. (Synt.
III171)
,
,
... ( 477).
,
.
As this last example clearly illustrates, Apollonian syntax was part of a global
explanatory procedure, where Homeric philology and physical theories of
vision came together. Indeed, the grammarian does not see language as an
abstract system: its schemas (especially syntactic) always exist more or less in
a mimetic relation to the objects and scenarios that it describes.
37 They are found, among the Latin grammarians, in a separate class, that of the interiectio.
Syntax 883
(...) ,
, ,
, ,
, ,
.
Incidentally, again, at the start of the Conjunctions (214.4), he recalls that the
Stoic Posidonius addressed the semantic difference between to give
freely, to offer and to give back, return, or between to
demand back and to bega difference that Apollonius does not
dream of being in doubt, as shown by his own expression (Synt. IV 15, 448.5):
lit. the condesignatum of the prepositions
which is associated with the relational value of the verb (
).
It should be noted that neither the semantic value of prepositions, nor the
phenomenon of consignification, hold Apollonius attention when he devotes
a book of his Syntax to the construction of prepositions. Instead, this book
offers a methodical and detailed examination of questions about cutting the
884 Lallot
The list of the seventeen classes appears neither logically orderedwe can
find no commentary which justifies its ordernor subdivisible into specified
sectionsand in particular, no distinction is made between coordination and
subordination. As far as the total number of conjunctions, there is no consen-
sus among the grammarians, although the mention of dissent (Sch. Dion. T.
440.15) indicates at least that a calculation would be legitimate. The inventory
of the conjunctions, then, occupies a middle place between the strictly-closed
inventory of the 18 prepositions and the open inventory of adverbs.
We have seen that the function of was for Aristotle to unify the
many ( ). This idea is revived by Apollonius, Synt. I 10, where
the conjunctive function is put in parallel with the composition of nouns,
which unifies two words into one:
, , ,
(Od. 10.251252)
41 In this inventory only the most typical conjunction is indicated as an example of each
class.
Syntax 887
( 70) (
527).
,
.
, , , ,
( 2512).
.
(Il. 1.68)
42 The scholia on the Iliad witness the vigilant attention of the Alexandrian philologists to
asyndeton. Meanwhile, rather than censuring the Poet like Apollonius, the scholiast often
finds justifications for asyndetic expressions: the insertion of a maxim (e.g. A 2168.8
Erbse), the language of a ruler ( , e.g. B 8c.3), the effect of anger (e.g.
53.1), etc.
888 Lallot
,
( 68). ,
( 68),
, .
43 One consequence of this statement is that the subjunctive appears as a mood with a
premorpheme, and that the conjunctive character of the premorpheme was eventually
forgotten: Synt. III 57.
Syntax 889
constructions: the relevance or not of the order in which the simple conjoined
sentences appear (the terms and seem to denote the same real-
ity from two different angles, the first connoting the order of the words, the
second a more semantic aspect of consequence, cf. Sch. Dion. T. 102.27). The
order is irrelevant for p and q and p or q, but not for p if q. These considerations,
although derived from treatises on formal logic (and repeated in certain gram-
matical scholia, e.g. Sch. Dion. T. 103.15), sometimes conflict with philologi-
cal data: Apollonius, doubtlessly following many others, finds fault with the
hysteron proteron of Homer, as in Od. 12.134
having thus raised them and brought them into the world, condemning any
inversion of two verbs. This is one example among others of the inevitable ten-
sion between logical norms and philological realities.
One singularity, in the eyes of modern scholars who are otherwise largely
in debt to the Alexandrians on the subject of conjunctions, lies in the fact that
what we would call temporal (e.g. when) and comparative (e.g.
as) conjunctions are not classed as conjunctions. It seems strange to us that p
when q or p as q should not be described as conjoined propositions under the
same heading as p because q. One cause of this situation is certainly that the
grammarians did not find the schemes p when q, p as q in the Stoic list of non-
simple utterances. Moreover, the fact that q or q can be exchanged
for adverbs of time or manner like or most likely inclined them to
give an adverbial status to invariant words like and themselves. But
whatever the reason, that is what they did, notwithstanding the (in our eyes
decisive) fact that the function of adverbs is that of the proposition and not
that of the connector. By this example it may be seen that the Alexandrian
analysis of complex sentences occurred within a framework of thought materi-
ally different from that imposed upon our own grammar; the abstract concept
of propositional subordination, in particular, did not exist for the Alexandrian
grammarians.44 Instead, semantic data carried for them a weight which
appears to us unjustified. A single example may be given here, inverse and
complementary to the foregoingwords like , , constructed with a
nominal genitive in the sense of because of, are classed with the conjunctions
for the simple reason that their semantic field relates to causation, that is, to
an idea which uses conjunctive expressions (the of the Stoics).
Equally, to our surprise but nonetheless for the same reasons, the prepositional
44 Another consequence of this conception of syntax is the absence of individual status, for
Apollonius, of the constructions that we call completive, whether they are conjunctives
() or infinitives; on this point, see Lallot [1996], [1999] and Ildefonse [2000], [2004].
890 Lallot
5.1
Whatever one might think of these kinds of odditieswhich happen to be
rather infrequentour impression of the Apollonian procedure is that of a
constant rationalising effort, of a desire to explain the constructions he comes
across. It is even a sort of leitmotif of the Syntax that the mere accumulation
and exhibition of examples () offers no benefit to syntactic theory
(). This is not to say, however, that Apollonius constructs, or even has
any intention of constructing, a formal and abstract system. As has already
been suggested, his goal of is part of his broader project, which he
never dreams of denying, to account for as much as possible of the empirical
data,45 that is, the corpus of the Greek literary inheritance, which he calls the
. This , diversified over time (diachronic changes, imperfec-
tions in copying) and over space (variations in dialect), is multiple and varie-
gatedand the grammarian must both respect and master it. Again drawing a
parallel between syntax and morphology, Apollonius writes:
There are competing forms among which tradition does not distinguish
[the correct from the incorrect]. [...] It seems that here the coherence of
the theory will allow us to rule out bad forms. The same will be true for
the following examination [sc. of the study of syntax]: in the event of
doubt, the application of the theory, taken together with a consideration
of natural data, will allow us to eliminate incorrect constructions. (Synt. I
61, 51.12)
,
(...) .
.
In this passage, with its neatly normative focus (the correction of texts,
, is one of the enduring tasks of the Alexandrian grammarian),46 the
keywords are and . The latter, tradition, should be under-
stood in the double sense of a manuscript transmission of texts and the tra-
dition of grammatical teaching itself, with its scholarly benefits, but also its
uncertainties, its polemics, its contradictions. At any given moment in history,
the concrete product of this is typically the group formed by the
available manuscripts of a literary oeuvre (for instance, the Homeric poems),
together with the grammatical writings which refer to themexegetical com-
mentaries and technical treatises produced by earlier and .
Since antiquity this group has offered the professional grammarian, for any
given text, a mosaic of concurrent lessons and explanations to choose from
(cf. ), since, by definition, the does not take a
stance (cf. ).
5.2
In the face of such diversity, the decision comes down to the . This, too,
has many facets. On one level, it designates an analogical reasoning which, in
confronting the unknown with the known, the doubtful with the certain, helps
to normalise the language and to correct texts. Of constant use in the study of
morphology, analogical reasoning is often designated as such by the adjective
, the adverb or the derived noun .47 One example
among many: is , the Homeric form of the first-person pronoun in the
genitive case correct? Perfectly so (), for it stands in exactly the
same relation to as does to (Synt. II 119, 218.10).48 The same rea-
soning is valid in syntax: in and (Il. 2.135), where
after a plural subject the verb is singular in the one case, plural in the other, the
(more) regular form () is the second, because we say (and
not *) the men (masc.) speak (and not *speaks), and we do
The norm, generally speaking, viz. that which should be the case ( ,
e.g. Synt. 107.18, opposed here to (Homeric) usage, ), as opposed to
its literal contrary ( , e.g. Synt. 52.5); or, by a well-worn medical
Syntax 893
metaphor, that which is healthy (, e.g. Synt. 31.19, opposed here to the
deviant, )
that which conforms to a regularity (), as opposed to that which is
irregular (), e.g. Synt. 63.17
that which is congruent (), as opposed to the incongruent
();
that which is complete and integral (, ), as opposed to that
which is elliptical (), defective () or, on the contrary, pleonas-
tic (), e.g. Pron. 38.23, Synt. 53.8, etc.;
that which is in (good) order ( ), as opposed to that which has suffered
a displacement ( ), e.g. Pron. 41.3.
, (Il. 3.277)
and you, sun (nomin., instead of voc.) who see all things,
, (Od. 17.415)
and similar [turns of phrase] are innumerable. We might say that they
illustrate the Attic figure, since, as we have said, the vocative demands
another ending. We find the opposite situation when the vocative is used
instead of the direct case, following the Macedonian or Thessalian usage,
as our predecessors have established, for instance:
(Il. 2.107)
lit.: in his turn the (nomin.) Thyestes (voc.) (handed over the sceptre) to
Agamemnon,
Here the article also betrays the deviation in case. (Synt. III 34, 300.8)
, ,
,
. ,
<..........................> ,
Syntax 895
This passage clarifies, using the example of hypallage of case, the procedure of
the Alexandrian syntactician as it has been presented here. Let us distinguish
the principal contributions: 1) The vocative is normally the case of address.
2) The corpus, both Homeric and classical, offers examples of the nominative
functioning as a vocative, and vice versa. 3) Is this a fault or a figure of speech?
The venerable age and considerable quantity of occurrences of the phenom-
enon leads us to believe that it is a figurenamely, hypallage. 4) This figure
is considered Attic in the nominativus pro vocativo case; in the inverse case,
we may speak of other dialectal usages. The dialectal details do not concern
us here; what is important, rather, is that the deviation of case (
) should be identified and recovered for use. The syntactic norm
remains safe, while deviations are given as figures, attributable to dialectal
diversity.50
This chapter has offered a depictiona simplified one admittedly, but hope-
fully faithfulof the characteristic features of Alexandrian syntax, as rep-
resented by Apollonius Dyscolus: a discourse on the construction of words
together (as the masters of Port-Royal would say), whose rationalist orienta-
tion is continually compromised by empirical datathe whole together aim-
ing to clarify the logic of the language, , and to explain poetry,
(to put it like Apollonius).
50 Idiolectal in other cases: among other singularities, Homer is accustomed to elide his
articles ( , Synt. I 42, 38.5).
chapter 4
1 Introduction
2 Ancient Etymology: Denkform and Discursive Practice
2.1 Anchoring Practices: Etymology, Mythology, Genealogy
2.2 Discourse Characteristics
2.3 Emphasis on Causality and Motivation
2.4 The Successful Etymology
3 A Case Study: Platos Cratylus on the Name of Apollo
3.1 Illustration of Discursive Principles
3.2 Etymological Technique
4 Functions of Etymology
5 Final Adhortation
1 Introduction
This chapter will deal with the ancient scholarly and poetic practices of
etymology.2 Rather than providing a historical overview, its main focus will
be on the cultural and historical embedding of these practices, an analysis of
the type of discourse they represent, and their cognitive and rhetorical func-
tions. These aspects of etymology remain important throughout antiquity and
1 Parts of this chapter are based on Sluiter [1997b], and on the etymology dossier in Copeland-
Sluiter [2009] 339366 (esp. 339344). I am grateful to Christopher Pelling, Philomen
Probert, and Stephen Halliwell, and to the other colleagues and students in Oxford (where
I was allowed to give the Nellie Wallace Lectures in the Spring of 2010 on Thinking with
Language); parts of this material were also presented in Leiden, Utrecht (OIKOS), and the
Department of Christiane Reitz at Rostock; it also informed some of my contributions to
the team studying textual practices, including etymology, at the Max-Planck-Institut fr
Wissenschaftsgeschichte in Berlin, brought together by Lorraine Daston, Anthony Grafton
and Glenn Most in the Summer of 2012. I would like to thank them, and also Gregory Nagy
and his wonderful staff and librarians at the Center for Hellenic Studies for providing me
with the peace of mind to write this chapter.
2 For historical overviews, see e.g. Amsler [1989], Lallot [1991a], and the introductions to
Buridant [1998] and Nifadopoulos [2003b].
the Middle Ages and they connect technical, poetic, and general (rhetorical)
uses, even though at the same time there is a development in the technical
disciplines to use etymology for the more specialized purpose of thinking
about morphology and lexicon by organizing words into clusters with a family
resemblance.3
Even there, though, it has virtually nothing to do with our modern academic
practice of etymology. The first thing to clear out of the way, then, is the pos-
sible confusion of ancient etymology and the modern form that has given us
our etymological dictionaries, the most recent one for Greek by Robert Beekes
[2010]. Such dictionaries ultimately go back to the linguistic discoveries by the
Junggrammatiker of the 19th century, based on the comparison of different
languages, and in particular the realization that a number of them, including
ancient Greek and Latin, are related as Indo-European languages. These lan-
guages all derive from a reconstructed common ancestor, Proto-Indoeuropean,
and they diverge from that common stock in accordance with strictly defined
and strictly conditioned phonological changes (sound laws). These laws
describe the situation before and after the sound changes, including the pho-
nological contexts in which at a given moment all phonemes under the scope
of the law underwent its influence. Exceptions need to be explained either as
the result of later sound changes or on the basis of processes of analogy.
The ancient discursive practice of etymology, on the other hand, is simply
a different kind of language game. In antiquity, to the extent that rules are for-
mulated, they are mostly ad hoc4 and as it were after the fact, the fact being
a preliminary semantic observation, leading to an interpretive relationship
between the explanandum and the explanans. This is to say that etymologies
are mostly put forward to corroborate a specific view of what a word really
means, probably even where they are presented as a tool to find the meaning
of a word.5 There are some attempts to systematize, but as we will see, they are
designed to allow maximum amplitude in relating words to other words. This
observation is not in any way meant as a disparagement of the ancient prac-
tice. Quite to the contrary, its aim is to allow us to value and appreciate that
ancient practice for what it really is and purports to do, rather than trying to
3 Philoxenus (1st c. BC) may be our first source to move in this direction, see Lallot [1991b]. We
see a similar development in Herodian, the Greek lexicographical tradition (the Etymologica),
and in the Middle Ages, where it is the branch of grammar called ethimologia that subsumes
the study of morphology, see Law [1985].
4 There were also some general principles guiding these practices; see below 3.2.
5 For this heuristic function, see Maltby [2003] and below 4.
898 Sluiter
make it conform to what we consider the correct, even the only scientific, way
of talking about language.
The differences that we can observe are connected with the different pur-
poses of ancient and modern etymology. Modern etymology is interested in
the systematic nature of language change and is a historical discipline relating
words to their past forms (the Proto-Indoeuropean roots). Although this may
also be useful as a general background to the study of semantic developments,
this form of etymology cannot be used reliably to explain the actual usage of a
word at any given point in time. It is usually made very clear to students that
we should not fall into the trap of confusing diachrony with synchrony: syn-
chronic semantics (and syntax, and phonology) can be described as a system
without reference to the developments that led to any given state of that system.
Diachronic linguistics, on the other hand, needs knowledge of the successive
synchronic states to construe the development that led from one to the other.
De Saussure used his famous comparison with a game of chess for this pur-
pose: we can completely and adequately describe the positions of the pieces
on the chessboard without knowing or caring what moves created those par-
ticular positions on the board.6
Ancient etymology, on the other hand, is all about synchrony, even though
it invokes a discourse that references the past. It is about the relationship
between words and their semantic explanation or definitionit wants to
know why anything is called what it is called, the reason for the name, and
what motivates the namegiverand the explanations it comes up with are not
intended to give us insight into the past, into the historical processes and devel-
opments leading to the present situation; rather, and importantly, (ancient)
etymology is about understanding the present.7 So whereas modern etymology
does not provide an immediate insight into the contemporary semantics of a
word, that is actually precisely what ancient etymology is meant to do. Ancient
etymology is primarily about the present, modern etymology is about the
6 De Saussure [1916 (1974)] 124127. Scholars have pointed out various infelicities in this com-
parison (e.g. Willems [1971]), some of which were already anticipated by De Saussure himself
(notably the fact that playing chess is an intentional activity, whereas language change, apart
from analogical change, is an evolutionary process (cf. [1916 (1974)] 127). However, the main
point referenced above is still an important one.
7 This is true both in technical and non-technical forms of etymology. For the near-absence
of considerations relating to the history of the Greek language in the Alexandrians, cf. Lallot
[2011] passim, here at 248; for the same point specifically about etymology, and for etymology
as Benennungsgrund, see Herbermann [1991].
Ancient Etymology: A Tool for Thinking 899
8 For the importance of synchrony and interpretation, see Peraki-Kyriakidou [2002] 480
482. Socrates position in Platos Cratylus is exceptional, but the positions he is arguing
against are the typical ones. What is new in Socrates position is that he considers ety-
mology a way to reconstruct the namegivers thoughts and considerations in producing
specific names for specific things. This would make etymology a historical type of inves-
tigation, leading to knowledge about a situation in the past; Socrates attempt to disqual-
ify etymology from contributing relevant arguments to investigations of contemporary
issues is virtually unique in antiquity, see below 3.
9 For the linguistic notion of common ground, see e.g. Clark-Brennan [1991]; tools for
thinking: Dennett [2000]; [2013], where the equally appealing term intuition pump is
also used. This label is applied primarily to thought experiments by Dennett, but it seems
readily applicable to the exploratory character of numerous ancient etymologies.
10 The term language talk to describe the various, often informal, discursive practices that
take language itself as its starting point and object is inspired by the unpublished Leiden
dissertation on ideas on language in Euripides by Christaan Caspers [2011], whose first
chapter is about talk.
900 Sluiter
is in part similar to the general use, and in part more specifically tailored to
talking about issues of morphology and lexicography.11 In the list of the tasks
of grammar by Dionysius Thrax, the fourth item is specified as
, the invention of etymology.12 This means that etymology is now (2nd
c. BC) a canonical part of grammar, but at the same time, the formulation sug-
gests a link with rhetorical inventio and the argumentative role of etymology
which is part of its general and poetic use. The reason why it came to be sub-
sumed under the field of grammar is probably precisely because it plays such
an important role in poetrypoetry after all is the primary study and teaching
material of the grammarian.
itself became the object of reflection, and was re-etymologized and connected
to , woof, and robe, mantle; this etymology would have been an
impulse or mnemonic support to generate the story of a heroine who spun a
robe by day and undid her work by night.15 Of course, the alternative is that
the myth was there first, and that a suitable name for its heroine was subse-
quently devised: this is a chicken-and-egg question, but however that is, there
is an undeniable link between the etymology of the name and the mythologi-
cal story.
In the Odyssey, the name of Odysseus, too, is etymologically connected to
the role and character of the hero; Odysseus is etymologized many times
(Rank [1951] 5163).16 The most explicit instance links the choice for baby
Odysseus name to the verb (Od. 19.406ff.). , to hate, to be
mad at, characterizes, it is said, the relationship between Odysseus grandfa-
ther Autolycus and the world, and it is projected onto the new baby, who gets a
name that fits his grandfather.17 Two implicit references come in the words of
Athena to Zeus in Odyssey 1.62, where Odysseus has grown into his name and
carries it in his own right, for she asks Zeus: why, Zeus, are you so mad at him?
( , ;) The same passage also hints at a link with
to lament, when Athena says that Circe is holding back poor Odysseus, who
is lamenting his fate: (Od. 1.55). Odysseus is
not only, like here, frequently in the position of having cause for lamentation
himself, in the Odyssey he is obviously also the object of the lamentations of
those who miss him, notably Penelope, Telemachus, and, in a striking passage,
Eumaeus.18 In this case, we have both a link with the story and a link between
etymology and genealogy: Odysseus gets the name that fits his grandfather and
only subsequently does that name become appropriate to the man Odysseus
as well. The three discursive practices of genealogy, mythology and etymology
are all useful in helping to create a mental roadmap of reality, to give people a
sense of where they are in the world.
18 Penelope, e.g. Od. 14.129f.where notice the context of the absent husband:
/ ,
and the tears fall from her eyelids, while she weeps, as is the way of a woman, when
her husband dies afar (trad. Murray); Eumaeus, a little further on in the same passage,
claims that not even his absent parents arouse such weeping and longing in him as does
his absent master Odysseus (Od. 14.142ff., where evokes the name of , a
name Eumaeus states he feels socially inhibited using): notice how the explicit reference
to naming may be considered a clue to the presence of etymologizing (
,... . / , ,
/ ... yet it is not for them that I
henceforth mourn so much; instead, it is longing for Odysseus, who is gone, that seizes
me. His name, stranger, absent though he is, I am ashamed to pronounce; ... instead I call
him honored friend (trad. Murray-Dimock).
Ancient Etymology: A Tool for Thinking 903
19 See OHara [1996] 60 and 75ff. on naming constructions as etymological signposts (in
Vergil); we just saw an example in the Eumaeus passage Od. 14.142ff. (see n. 18). Another
example: Ov. Fast. 3.725ff., is about explaining the causae, the reasons why the vine-father
summons (vocet) the people to his cakes. The combination of causa and vocare is enough
to prime an ancient audience for the presence of etymologies: there follows a connection
between liba and Liber, but the real connection comes at 733736: Liber explains the
name libamina, and then states liba are so called, because [again causal language calling
attention to the etymology] part of them (i.e. of the libamina) is dedicated. This must be a
playful etymology: liba forms part of the word libamina, and that fits the actual sacrificial
procedure. For the phrases or as signposts, see Peraki-Kyriakidou [2002] 482.
20 These are never clearly distinguished in antiquity; the term or can cover
both or either. is usually reserved for (inarticulate) sound.
904 Sluiter
21 This is what Herbermann [1991] calls the Benennungsgrund. For the earliest reflections
on words origin and meaning in the Greek world, see Novokhatko and Pagani in this
volume.
22 For this concept (without the name) applied to genealogy, see West [1985] 11; Fowler
[1999], 2 n. 7.
23 See Culler [1988] 11 and 13, pointing out this importance of the urge to motivate. Precisely
because the linguistic sign is arbitrary, discourse works incessantly, deviously to motivate.
For undoing the arbitrariness of the sign through metaphor and poetic language, see
Conte [1986] 45, who uses Platos Cratylus as a parallel for this process. Etymology makes
poetry out of language, i.e. it makes language substantially motivated (ibid.). OHara
[1996] 3 also adopts this view of Conte in thinking about poetic etymologizing (cf. Conte
[1986] 50).
Ancient Etymology: A Tool for Thinking 905
Cicero is one of our sources stating this causal principle quite clearly, both
when speaking about the Academics and the Stoics (the latter in a very critical
passage):
They commanded the explanation of words, i.e., why each thing was
called by its particular name (they called this etymology). Later they used
some of them as arguments and deployed as it were the signs of things
as guides to prove and show conclusively that which they wished to have
explained.
24 Some examples: Hom. Il. 9.562; Hymn. Hom. Ap. 3.372ff.; Hes. Theog. 144;
cf. Pl. Cra. 407b; : e.g. Heracl. Gram. Quaest. Hom. 55 Hermes stands for ,
Leto is opposed to him: ,
Leto fights all reason, being as it were a letho [forgetfulness] if one changes one
element; quasi etc.: see Isid. Etym. passim, e.g. I v 3 oratio dicta quasi oris ratio; (combined
with quod): I iii 2 litterae autem dictae quasi legiterae, quod iter legentibus praestent
litterae [letters] are called as it were legiterae, because they show readers (leg-entibus)
the way (iter); cf. further phrases such as Isid. Etym. I xvli 2 (on metrical feet) ipsi autem
pedes habent speciales causas nominum quare ita vocentur. Pyrrhichus dictus est quia ...
the (metrical) feet themselves have special reasons for their names, why there are called
what they are called. The Pyrrhichus is called that because ...; an example of an allusive
etymology, betraying knowledge of the Greek tradition is Vergils trunca pedum devoid
of feet, as flagged by the grammarian Sacerdos (GL 6.477.16): apes quasi quod sine
pedibus nascatur, sicut Virgilius de his [Georg. 4.310] trunca pedum apes bee is as it were
a-pous feet-less, because it is born without feet, as Vergil says about them devoid of feet.
906 Sluiter
The indications or signs of reality (rerum nota) are used as guiding principle,
to argue and to underpin whatever explanation is offered.25 This explanation
of etymology as the Benennungsgrund and motivation for names is clearly
expressed by qua de causa quaeque essent ita nominata, and cur quidque ita
appellatum sit causas explicare.
Since the normal order of cause (here: the semantic explanation) and effect
(here: the word under discussion) is precisely that, we also see the frequent use
of () ; (ducere) a(b) etc.
-
-
;
, , -
.
who can have given a name so altogether truewas it some power invis-
ible guiding his tongue aright by forecasting of destiny?who named
that bride of the spear and source of strife with the name of Helen? For,
it was conspicuously as a Hell to ships, Hell to men, Hell to city that she
sailed the sea, stepping forth from her delicate and costly-curtained
bower (trad. Weir Smyth/Lloyd-Jones, adapted).
.
, , ; (in a work not dealing
with etymology) Artem. 1.4: a dream of a hostel called the camel was explained as
announcing that the dreamer would break a leg:
(sc. ),
; and
: Aesch. Supp. 315 on the name of Epaphus, derived from (see vs. 313):
Epaphus, and truly named from laying on of hands
(trad. Weir Smyth).
28 The word etymology is absent from Platos Cratylus and was apparently coined by the
Stoic philosopher Chrysippus in the 3rd c. BC. for etymology is first used by
Plutarch, e.g. Mor. 278c ... the etymology of the word is ....
908 Sluiter
i.e. she behaved in a way appropriate to her name, as if she was somehow
socially expected (, ) to do the right thing by her name. She was
certainly seen to be doing what her name might suggest. Notice how the
clause motivates the appropriateness of the name.
and mean that things are as they ought to be, that regularity
and order are preserved.29 comes from the semantic field of aesthetics,30
of reasonableness,31 and of correctness, rightness according to
a (straight) rulethis of course is the word that will become the 5th-century
catchword for correctness of speech.32
In the 5th and 4th centuries it became increasingly fashionable to explore and
exploit the notion that language itself can somehow be of direct and instru-
mental use in illuminating the relationship between reality, thought, and lan-
guage itself, that there is a satisfying fit between language and reality, and that
this relationship can be expressed as right, just, true, correct or beautiful, as
fits the context. In Platos Cratylus, written in the 4th c. BC, but with a dramatic
date in the 5th, Socrates is made to address this fashion in an attempt, in his
case, to disqualify language as a direct route to philosophical truth. As in several
other dialogues, Socrates dismantles the etymological method only after hav-
ing proven his unrivalled excellence at this form of discourse.36Plato always
makes sure that the sour grapes argument will never affect Socrates: whenever
In Aeschylus, we also encounter the terms and clearly: these are terms that
refer to the auditory domain, the shrillness and clarity of sounds; they are less relevant here.
33 For , used for a name in so far as it relates to something else, see Sulzberger
[1926], Sluiter [1997b] 157. ends up in the technical tradition as a subclass of
nouns (Dion. T. GG I 1 38.3 , ,
, an eponym, also called
di-onym (double name), is the name that is used for a single referent together with
another word that is the proper name, e.g. Poseidon is (also) Earthshaker, and Apollo is
also Phoebus.
34 E.g. Hom. Od. 19.409 therefore let Odysseus be his
(significant) name, the name by which he is called; Soph. Aj. 430ff. (Ajax speaking) .
/ ; /
Alas! Who ever would have thought that my name would come to harmonise
with my sorrows? For now I can say Alas a second time (trad. Lloyd Jones); this is a
case of Ajax having grown into his name, where the assonance with the interjection of
lament aiai has suddenly become meaningful; cf. the relationship between and
in Eur. Bacch. 367 (without a term like flagging the etymology).
35 This example was also discussed with a slightly different focus in Sluiter [1998].
36 Barney [1998] for the competitive nature of Socrates performance; on Cra., see further
in particular Baxter [1992]; Silverman [1992]; Barney [2001]; Sedley [2003b], and the
commentary by Ademollo [2011].
910 Sluiter
37 E.g. forensic rhetoric in Ap.; different types of epideictic rhetoric in Menex., Symp., Phdr.,
Prt.; sophistic discourse in Tht., and Euthd.; Other types of discourse, not always rejected
for philosophical purposes: symbouleutic rhetoric in the preambles of Leg.; cosmological
discourse in Ti. (with Timaeus as speaker), historiographical discourse (again not with
Socrates as speaker) in Ti., Cri., Leg. III; legal discourse in Leg. See also Nightingale [1995].
38 The modern discussion about taking the Cratylus seriously or not is probably not quite
on target: the use of etymology as a vehicle for philosophical discussion is explored
quite seriously; the outcome that it should not be so used is equally serious. None of this
precludes a certain playfulness on the way. In antiquity the Cratylus was sometimes seen
as originating etymological theory, e.g. Dion. Hal. Comp. 16:
,
Plato the Socratic was the first to introduce the theory of etymology, in many other
places, but in particular in his Cratylus.
39 Cf. Pagani, Probert, and Valente (section III.2) in this volume.
Ancient Etymology: A Tool for Thinking 911
.
. [
] ,
, , []
(d) , , ,
,
, - -,
, (e) ,
.
[], (406a) ,
, , , , .
For no single name could more aptly indicate the four functions of
the god, touching upon them all and in a manner declaring his power
in music, prophecy, medicine, and archery...(b) In accordance, then,
with his acts of delivering and his washings, as being the physician of
such diseases, (c) he might properly be called Apoluon [, the
washer], and in accordance with his soothsaying and truth and simplicity
(haploun)for the two are identicalhe might most properly be called
by the name the Thessalians use; for all Thessalians call the god Aplun.
And because he is always by his archery controller of darts [] he
is ever darting [ ]. And in accordance with his music we have
to understand that alpha often signifies together, and here it denotes
moving together both in the heavens about the poles, as we call them,
and with respect to (d) harmony in song, which is called concord. For, as
the ingenious musicians and astronomers tell us, all these things move
together by a kind of harmony. And this god directs the harmony, making
them all move together, among both gods and men. And so, just as we call
homokeleuthon (him who accompanies), and homokoitin (bedfellow), by
changing the homo- to alpha, akolouthon and akoitin, so also we called
him Apollo who was Homopolo, (e) and the second lambda was inserted
because without it the name sounded of disaster. Even as it is, some have
a suspicion of this, because they do not properly regard the force of the
name, and therefore they fear it, thinking that it denotes some kind of
ruin. But in fact, (406a) as was said, the name touches upon all the quali-
ties of the god, as simple, ever-darting, purifying, and accompanying.
[trad. Fowler, slightly adapted]
912 Sluiter
(1) First of all, here are four etymologies that are clearly meant to give us, col-
lectively, a picture of the roles of Apollo in 5th4th-century Athens: roles
in music, divination, medicine, and archery. The different etymologies do
not exclude, but rather supplement each other. None of them is supposed
to offer the single true historical derivation of the name, but each of them
reveals an aspect of the god. They are simultaneously true.
(2) Second, each of the four gives a Benennungsgrund, they motivate the
name of the god; each time the Benennungsgrund is different in accor-
dance with the different roles of the god. The etymologies are marked by
the use of causal language or the suggestion of causal connections.40 The
etymologies are evaluated, in this case primarily by means of terms such
as and (405c).41
(3) Apart from the causal language, the vocabulary used draws explicit atten-
tion to the presence of names (e.g. in the very first line of this excerpt: )
and the practice of naming (various forms of are used throughout
this text). In the Cratylus, with its explicit focus on etymology, this may not
cause wonder, but as noted above, such discourse elements may signpost
etymologies also in texts that are not explicitly about etymology.
All the points mentioned so far can readily be paralleled in the poetic tradi-
tion, for instance in the multiple explanations for the name of Ion in Euripides
Ion; in the prologue by Hermes we learn that Apollo will make sure that he will
be called by the name Ion throughout Greece: as future founder of the Ionians
(Ion 7475), he will be their eponymous hero, i.e. they will be named after
him; and in fact, Hermes proceeds immediately to call him by that name he is
yet to get (8081). The actual naming is based on Apollos oracle to Ions new
father Xuthus: whoever encounters Xuthus on his leaving the temple (,
535) will be his son, says Apollo, and Xuthus converts the fact that Ion is the
40 In 405b may mean no more than in accordance with, with reference to, but the
implication is clearly that etymology and domain are in accordance with each other; 405b
+ ptc. because; in 405c, again the use of , especially , and again ; 405d ;
supplemented with a principle of analogy (405d ... ...).
41 The qualification is obviously important in Cra. given its theme of
. But the other commendations also play a role, e.g. (vs. ~ )
Pl. Cra. 404d.
Ancient Etymology: A Tool for Thinking 913
first person he saw into the motivation for his name.42 Leaving or going
(out) is thus something done by the father. When the chorus reports the nam-
ing incident, they seem to transfer the going to the sonof course, meet-
ing is something done mutually.43 Each time, the etymology is signposted by
the vocabulary of names and naming, and a causal relationship is suggested
between name and motivation. The name also related to what happens in the
story. This takes us back to the Cratylus example again.
(4) In the Apollo example, the etymologies together make up a narrative: they
are a story about Apollo, in fact, the passage has a neat ring-composition
that strengthens that effect (406a picks up 405a). It has frequently been
observed that etymology is an important tool in allegory, without com-
pletely coinciding with it.44 Etymology can provide the building blocks,
often based on establishing individual interpretations (or motivations)
of names. Put together, these can constitute allegorical narratives. This is
what we see happening, for instance, in the 1st c. AD work by Cornutus.
However, as we see, etymology by itself also has narrative potential.45
(5) Fifth, the last etymology, with its reference to people who believe that the
name of Apollo is somehow related to the verb to destroy, is
polemical in tone. It is claimed that the insertion of a second lambda was
in fact done on purpose to prohibit such an association. In fact, however,
our poetic tradition does indeed offer such an etymology, for instance
in Aeschylus Agamemnon 10801082,46 where Cassandra calls on Apollo,
alpha may not just be an privans but may also indicate a relationship of
togetherness. And he uses a principle of analogy: the relation between
and is the same as that between and .
To say that etymology is a very different language game from our disci-
pline of historical grammar is definitely not to say that there is not a great
amount of linguistic observation and knowledge feeding into it.50
(c) This is also our first extant text in which the avoidance of homonymy is
explicitly invoked as a reason for linguistic changein later grammatical
theory we will encounter the phrase in order to avoid
coincidence; refers to the coincidence of forms (co-in-cide
is actually a calque of --).51
(d) Another issue of the technique of etymology is phonetic bridging, the
phonetic transition technique that takes us from one word-form to the
next. This is connected with the set of rules, also going back to the Craty-
lus that is associated with etymology in antiquity. These rules are asserted
quite confidently by Socrates in the passage in the Cratylus in which he
claims to be under the influence of a strange inspiration. If we wish to
understand why a word is called whatever it is calleda clear enuncia-
tion of the ancient mission statement of etymology, he says, we should
fully focus on the semantic aspect. Ultimately, that is the only thing that
counts. The word-form can undergo all kinds of changes, which will not
ultimately affect the meaning. Socrates distinguishes four kinds of change
or operations:
50 In that sense the criticism of Nifadopoulos [2003b] of the observation that ancient
etymology will use anything that will create the desired result is misguided. The
recognition that ancient etymology is a particular tool for thinking in its own right
rather than a precursor of historical grammar is perfectly compatible both with taking
it seriously, and with acknowledging that observations of linguistic regularities may feed
into it (cf. also Pagani, Probert, and Valente [section III.2] in this volume).
51 See Sluiter [1990] 125139.
52 See above n. 17.
916 Sluiter
So perhaps the man who knows about names considers their value and
is not confused if some letter is added, transposed or subtracted, or even
if the force of the name is expressed in entirely different letters. (trad.
Fowler)
53 Note again that like all ancient thinkers, Socrates does not distinguish between sounds
and letters.
54 See Pagani in this volume.
55 See OHara [1996] 59 and 60ff. on paronomasia (the poetic linking of words of similar
sound).
Ancient Etymology: A Tool for Thinking 917
that part of the ancient tradition that is highly critical of etymology (such crit-
ics of etymology notably include Aristotle, Cicero and Galen). In Ciceros De
Natura Deorum, for instance, it is put like this:56
The criticism is put quite clearly here: if Neptune can be derived from nare,
any word can be linked to any other by having just one letter in common: one
letter will suffice to explain its provenance (una littera explicare unde ductum
sit). This criticism, too, would be long-lived. It is the basis for Mark Twains
famous dictum on the derivation of the name of the village of Middletown
from Moses, by dropping oses and adding iddletown.57
The Cratylus passage has provided examples of semantic bridging, of the
use of any kind of linguistic observation, of the argument from linguistic econ-
omy, and of phonetic bridging with its application of the four categories of
change. These will remain important instruments of etymologists throughout
antiquity and the Middle Ages.58
4 Functions of Etymology
take their cue.63 And there is a second important feedback loop connecting
the language disciplines (grammar, rhetoric, dialectic, philology).
It is in technical grammar (but also in rhetorical contexts) that etymology
is also usedor at least presentedas a heuristic tool, an intuition pump for
assessing the meaning or orthography of a word.64 This presupposes that the
etymology is easy to follow. Varro complains about an etymology in Ennius
that presupposes knowledge of Greek to an extent that makes the etymology
itself highly obscure.65 In the technical grammarians Apollonius Dyscolus and
Herodian (2nd c. AD), etymology plays a rather minor role.66 However, two
passages from Herodian may illustrate the range of its usage.
This is the usage traditionally labeled heuristic: the etymology of the name
Hephaestus is used as an argument to settle the question of whether the
opening vowel should have a rough or a smooth breathing.67 Since has a
rough breathing, so should . However, the etymology itself is offered
quite apodictically.68 There is no argument or motivation for it, i.e. the name
Hephaestus is not motivated through an explicit semantic link with the verb
to touch. The second example takes a different approach still:
The fragment relates the word for sister (soror) etymologically to seorsum
separate or apart. A reasonable guess would be that Labeo used this etymol-
ogy to argue for the legal status of sisters: by nature, that is in natural law, they
would be expected to leave the house and the jurisdiction of their fathers when
they got married and to go over into the manus of their husband. This natural
state of affairs appears from their name soror, and it means that natural law
and positive law are in agreement. This argument would have appealed to the
Stoa and may in fact have been inspired by them.73
Finally, there may be a more basic mnemonic function than the one
we started out with: etymologies are a helpful support for memory, simply
because they can be delightful, clever, and easy to remember. We will end this
overview with two examples from the Middle Ages, where yet another type
of etymology becomes popular: the syllabic one.74 This leads to etymologies
such as cadaver = ca-ro da-ta ver-mibus (flesh given to worms) or fenestra
= fe-rens n-os extra (taking us outside).75 These etymologies are funny and
72 E.g. Arist. Rh. 1400b1725; Top. 112a3238; Cic. Top. 3537 cum ex vi nominis argumentum
elicitur when an argument is drawn from the meaning of a name. Cicero experiments
with different translations, but rejects the literal veriloquium (a calque of )
for notatio (quia sunt verba rerum notae because words are symbols of reality, relating
this choice to Aristotles . Cic. De or. 2.256257 provides more examples of
paronomasia and rhetoric based on etymology (in spite of the philosophical objections
raised in Nat. D., see above at n. 56).
73 See Allen [2005] on Stoic etymology.
74 The Di-ka ~ Dios Kora example discussed above at n. 30 is an early version of this.
75 To be found in Petrus Helias, Summa super Priscianum I 2 (see Copeland-Sluiter [2009],
351); the gloss on Priscian Promisimus (Copeland-Sluiter [2009], 356).
922 Sluiter
5 Final Adhortation
1 Introduction
2 Branches of Prosody
3 Breathings
4 Vowel Quantities
5 Types of Accent
6 Positions for the Word Accent
7 Barytone Words
8 Classification of Words into Groups with Similar Accentuation
9 Base Accent and Case Accent
10 Basic Word Accent and Accents due to Enclitics
11 Concluding Remarks: Prosody as a Part of Ancient Grammar
1 Introduction
The term theory is used in a number of different senses. In a fairly weak sense,
theory involves not just the statement of known facts but some generalisation
over those facts. In a somewhat stronger sense theory is used of the system-
atic statement of generalisations over a large and in some way complete body
of facts. The second edition of the Oxford English Dictionarys definition of
theory, sense 5, begins, In the abstract (without article): Systematic concep-
tion or statement of the principles of something; abstract knowledge, or the
formulation of it:....
A systematic treatment of the facts of ancient Greek prosody was achieved
by the second century AD, when Herodian2 produced a large work entitled
On prosody in general. Before Herodian, Heraclides
of Miletus had composed a work entitled , and there-
fore presumably also a systematic treatment of the field, in the late first or
early second century AD. But we have too scant remains of this work to have
1 I am very grateful to Eleanor Dickey and Stephanie Roussou for helpful comments and
corrections.
2 See Matthaios in this volume.
a clear idea of its scope or structure (see the fifteen fragments assigned to the
work by Cohn [1884a] 3744).
Neither Herodians nor his other treatises survive
complete, with the exception of the On exceptional
words,3 but we have two epitomes of the . One of
these is attributed to Arcadius in some manuscripts and to Theodosius of
Alexandria in others; we shall refer to it under the name [Arcadius]. The other
is ascribed to Johannes Philoponus of Alexandria and may be an abbreviated
version of an epitome Philoponus produced. We also have a body of Homeric
scholia deriving from Herodians On the prosody of the
Iliad and On the prosody of the Odyssey4; and
numerous citations of Herodian in later authors, and other passages whose
content can be attributed to Herodian with more or less certainty.5
Herodian was dependent on Alexandrian discussions of prosody beginning
in the early second century AD (see below), and his Alexandrian predeces-
sors are explicitly mentioned in some of our surviving sources. Aristarchus, in
particular, is very frequently mentioned in the Homeric scholia deriving from
Herodian.6 But we have much fuller information on Herodians works relating
to prosody than we have for those of any of his predecessors. For this reason, the
following discussion will focus on the concepts and categories Herodian used
in formulating the regularities of ancient Greek prosody, and in stating excep-
tions to these regularities, but will attempt to show which concepts and cat-
egories are known to have been used already in early Alexandrian discussions.
2 Branches of Prosody
breathings thus made up the field of prosody. All these subjects have in com-
mon that they pertain to aspects of the pronunciation of Greek which were
not represented in writing, or in the case of vowel quantities not fully repre-
sented in writing, in ordinary (non-scholarly) written Greek of the Hellenistic
or Roman periods. An interpolated passage of uncertain date appearing in
two 16th-century manuscripts (Par. gr. 2603 and 2102) of [Arcadius] ([Arc.] 211.
8216. 2) explains that the marks for accents, vowel quantities, and breathings
were invented in order to disambiguate words, and to indicate the singing
qualities of the voice:
,
, . ([Arc.] 211. 812)
The (marks for) quantities and pitches and breathings, which Aristophanes
[of Byzantium] created, were devised for the purpose of disambiguat-
ing an ambiguous reading and for the singing of the whole voice and the
melody, as if we were to sing along to our speaking.
Although the antiquity of the material in this passage is unclear,8 the etymol-
ogy of (approximately singing along) suggests that in origin the
were considered to be features of speech accompanying the basic
Dyck [1993a] 778779, with bibliography). Our knowledge of Herodianic doctrine on breath-
ings is consequently the poorer, and work remains to be done in distinguishing Herodians
views from those of his successors (see Egenolff [1903] 3961; Dyck [1993a] 779), but Lentz
[1860] provides a most useful collection and discussion of ancient doctrine on breathings
in general, while Egenolff [1903] 3961) provides a very useful critique of Lentz [18671870]
reconstruction of Herodianic doctrine on breathings. Apollonius Dyscolus refers to a work
on breathings of his own (called at Synt. 458. 10, at Pron. 57.6, and
at Adv. 198. 2627), and we have a few fragments of a
by Trypho (see von Velsen [1853] 510). Theognostus refers to a by Herodian
(An. Ox. ii. 19. 33), but the relevant part of the twentieth book of the
may be meant. For Herodianic doctrine on vocalic quantities we are more fortunate in that
Herodianic material on vocalic quantities is transmitted as a separate treatise
On vowels capable of being either long or short, possibly excerpted from the
(see Dyck [1993a] 783 with n. 58 and bibliography, and cf. Dyck [1993a] 778779).
8 For some of the controversy see Pfeiffer [1968] 179 with n. 1 and Nagy [2000a] 1516 n. 21,
both with bibliography. The part of the passage I print here occurs in both Par. gr. 2603 and
Par. gr. 2102 (see Lameere [1960] 91), so whatever its origin it was not composed by Jacob
Diassorinus (in whose hand Par. gr. 2102 is written) since, as Nagy [2000a] 16 n. 21 emphasises,
the consensus is that Par. gr. 2603 is independent of Par. gr. 2102.
926 Probert
sequence of vowels and consonantsa notion not very different from the
modern notion of suprasegmental features.
It is possible that the term was already used to include breathings
in one passage of Aristotle,9 but otherwise the attested uses of the term before
the Roman period refer to accents only (when the term has anything to do with
what we would call phonology).10 In the first century BC Trypho applied the
term to breathings as well as accents, if we can trust a passage of [Ammonius]
crediting him with an explanation of the term smooth prosody
(i.e. smooth breathing):
,
, . ([Ammonius]
521 Nickau = Trypho fr. 108 von Velsen)
, ,
,
, ,
When we learn the letters, first we learn their names, then their shapes
and values, then similarly the syllables and the changes that happen in
them, and after this words and their accompaniments, I mean lengths
and shortnesses of vowels and prosodies and things similar to these.
(Dion. Hal. Comp. 25. 41)
The first really clear indication, then, that applied to the whole triad
consisting of accents, breathings, and vowel quantities is the inclusion of all
three of these in Herodians . After Herodian, the defi-
nition of was extended still further, and eventually the term came to
be usable of almost any feature of a text for which a diacritic mark was avail-
able. The term appears in this sense in the already-mentioned interpolation in
the text of [Arcadius]:
, , , , ,
, , ( Schmidt) , , , ,
, , , , , ,
, , , .
, , .
([Arc.] 216. 411)
12 However, Robins [1986] 18, [19974] 48 observes that this use of the term groups
together a range of phenomena remarkably similar to those of central concern to Firthian
prosodic phonology.
928 Probert
3 Breathings
The Ionic alphabet that was normally used for writing Greek in the Hellenistic
period did not include a letter for the sound h. Breathing marks are an inven-
tion of Alexandrian scholarship (whether or not they are actually due to
Aristophanes of Byzantium,13 as the above-quoted passage implies), and
appear sporadically in more scholarly Hellenistic papyri. An interest in estab-
lishing regularities for the aspiration of vowels arose in the context of the
textual criticism of Homer and other already ancient authors, since correct
pronunciation depended on knowing which vowels were aspirated, but not all
words were still heard in the Hellenistic period and the texts which had come
down to the Alexandrian period did not indicate aspiration on vowels.
Although the transmitted texts did not indicate aspiration on vowels, they
did show whether a voiceless stop was aspirated or not when it occurred
immediately before a word-initial vowel (often as a result of elision), and in
such cases the aspirated or unaspirated status of the stop provided evidence
for the aspiration or non-aspiration of the following vowel (see Lentz [1860]
647649; Wackernagel [1916] 41 n. 1). Thus, Sch. Il. 6. 239c (A), deriving from
Herodian, shows that the phrase in a line of Aeschylus (fr. 281a. 28
Radt), and in a line of Euripides (fr. 1014 Kannicht), were transmitted as
and respectively; Herodian argues from the transmitted
(rather than ) of and the K (rather than ) of that the word had
a smooth breathing, not a rough breathing.
But not all words whose aspiration or non-aspiration needed to be estab-
lished were attested in contexts where preceding voiceless stops would pro-
vide evidence. For other words general rules were established on the basis of
words whose aspiration or non-aspiration was known. Some of the criteria that
featured most prominently in these rules were:
(a) The sequence of vowels and consonants with which the word began (see
Lentz [1860] 649650, 697700, 718776). Rules about which sequences
received which breathing were produced at least as early as Trypho (see
e.g. Etym. Magn. 148. 412 = Trypho fr. 1 von Velsen:
,
Trypho says in the that before is short and has a
rough breathing). For an example deriving from Herodian see e.g. Sch. Il.
6. 348b (b(BCE3)T), stating that should have a smooth breathing
Inevitably, many of the rules governing words beginning in certain ways were
valid only by chance for the majority of relevant words with known aspiration
or non-aspiration, and the use of such rules to predict the aspiration or non-
aspiration of obsolete words was a risky enterprise. Furthermore, when differ-
ent criteria for determining aspiration or non-aspiration came into conflict,
there was no consistent system for deciding which criterion should prevail
(see Lentz [1860] 659661, 697). However, many of the ancient observations
on breathings are genuinely insightful. Furthermore, two of them foreshadow,
respectively, an important result of modern comparative philology and a point
which has received much discussion in modern Homeric scholarship: the
observation that (with certain exceptions) no vowel preceding an aspirated
stop is aspirated (and nor is a stop followed by vowel plus aspirated stop),14 and
14 For the generalisation applying to vowels before aspirated consonants see Lex. spir. 211.
1720 and Etym. Gud. 573. 13 Stefani (both reading every
vowel before an aspirated stop has a smooth breathing). Related but more restricted
generalisations appear e.g. at Apol. Dysc., Adv. 209. 22 (
vowels preceding never have a rough breathing) and the Herodianic
scholia Sch. Il. 12. 260 (A) ( every vowel before has a
smooth breathing); Sch. Il. 12. 391a1 (A) (... ...
...every vowel before an aspirated stop plus has a smooth
breathing... apart from ). See further Lentz [1860] 652653, with bibliography.
For observations on the avoidance of aspirated consonants in successive syllables, see
[Herodian], 21. 1621 La Roche, referring to a rule
930 Probert
4 Vowel Quantities
In the Hellenistic period the Greek letters , , and represented both long
and short vowels; these vowels were termed anceps, i.e. capable of
being long or short. Alexandrian grammarians were interested in establish-
ing the correct lengths of these vowels in Homeric and other literary words;
the Homeric scholia frequently record the views of Alexandrian scholars on
vowel quantities, although usually without an indication as to the reasoning
the Alexandrians used to arrive at the correct quantity.16
By Herodians day the distinction between long and short vowels had at
least begun to break down, though it is controversial when this distinction was
,
a disyllabic word is not usually in use with the first syllable
beginning with an aspirate and the second with an aspirate); approximately the same
rule at Choer. Th. 2. 146. 3335; see also the discussion at Choer. Th. 2. 146. 16148. 4 and cf.
Choer. Th. 2. 327. 1719; 2. 327. 2526. Gramann [1863] demonstrated that aspiration on
a vowel or stop was lost in the prehistory of Greek when an aspirated stop began the next
syllable. For more recent discussion, see Collinge [1985] 4761.
15 The Herodianic scholion Sch. Il. 9. 6b1 (A) suggests that this observation was already made
by Aristarchus or his circle (cf. Lentz [1860] 691, 696); cf. Sch. Il. 1. 576 (A), also deriving
from Herodian. For modern discussion see Wackernagel [1916] 4052; Wathelet [1970]
218221; Chantraine [1988] 184188; West [1988] 163.
16 E.g. Sch. Il. 2. 53c1 (b(BCE3E4)T); Sch. Il. 6. 268b1 (A). Cf. Herodian, . 13. 1617, with an
observation from Aristophanes of Byzantiums Attic words on the long
iota of comparatives in - in Attic.
Ancient Theory of Prosody 931
17 The loss of distinctions in vowel quantity appears to have progressed at different rates
in different areas of the Greek-speaking world. Orthographic mistakes involving vowel
quantity do not become common in Attic inscriptions until about 100 AD (see Threatte
[1980] 385387), but for Egypt the evidence of papyri suggests a significant loss of dis-
tinction in vowel quantity already in the Hellenistic period (see Mayser-Schmoll [1970]
117119; Gignac [1976] 325 with n. 2). The dating of the completion of the process is, how-
ever, controversial for all parts of the Greek-speaking world because there is no consensus
as to the weight that should be given to apparently relevant spelling mistakes at periods
when they are still rare (see Dickey [2009] 151). For an early dating for the loss of distinc-
tive vowel length even in Attic, see Teodorsson [1974] 218219, but cf. Ruijgh [1978] 84.
18 The accent was also restricted to the last two syllables if the last syllable was closed by a
consonant cluster, as in rich in springs (see Steriade [1988] 273275), although
not many Greek words end in a consonant cluster. The statement ,
, at Io. Philop. Ton. praecept. 5. 56 may
implicitly recognise that the restriction to the last two syllables applies to words with a
final consonant cluster as well as words with a long vowel in the final syllable, since the
expression needs to be taken with understood, and a final syllable
was considered long if it contained a long vowel or was closed by a consonant cluster
(see e.g. [Arc.] 160. 1415).
932 Probert
, , , , .
[ ] .
(Herodian, . 10. 58)
In the following example, the vowel in question does not belong to any par-
ticular suffix, but its shortness is predictable from the form of the comparative
suffix that follows:
, ,
, , , , .
. (Herodian, . 13. 1113)
, ,
, , ,
, ,
, , .
. (Herodian, . 11. 1015)
,
,
, , , , , , ,
.
, , , , , ,
. (Herodian, . 12. 1217)
Not all the rules apply to the ends of words. Some apply to word-initial vowels
in contexts that we would label purely phonological, not also morphological
(e.g. . 17. 911, on word-initial followed by plus stop). The following rule
applies to a prefix, alpha privative, and is noteworthy for its metrical explana-
tion of the long initial - of and , which provided the starting
point for extensive modern discussion of the long - of these words:20
, , , .
,
. (Herodian, . 15. 2830)
But the central branch of ancient prosody, to which we now turn, was the
accent, the branch of prosody which yielded the largest number of statable
20 See Schulze [1892] 140142, with Hoekstra [1978] 3 n. 10 and 2526, with bibliography.
934 Probert
5 Types of Accent
long vowel, so that over the whole vowel a falling pitch was perceived.24 One
might expect the other kind of accent on a long vowel or diphthong, marked
with an acute, to constitute the word accent on the second mora of the long
vowel or diphthong, and it is clear that grammarians understood a long vowel
or diphthong marked with an acute to occur under circumstances where one
would expect the second mora to be accented. Thus, if two vowels contract
then (other things being equal) the resulting vowel has a circumflex if the first
vowel in the sequence was accented before contraction, and an acute if the
second vowel was accented before contraction: uncontracted light gives
contracted , but uncontracted alive gives contracted .25 What is
less clear is how the accent represented by an acute on a long vowel or diph-
thong was pronounced. The assumption that a rising pitch was heard is per-
haps the easier one, given the occurrence of this accent where a rising pitch
might be expected.26 However, the choice of the same accent mark for this
accent and for the accent on a short vowel suggests that what was perceived
was rather similar in the two cases. Possibly the word accent on a short vowel
was also perceived as a rising pitch, so that the accent that really stood out as
sounding different was the accent on the first mora of a long vowel.
In the Hellenistic system the grave accent mark or desig-
nated lack of accent; every syllable that did not have an acute or circumflex
accent was considered to carry a grave, and could be marked with the grave
accent. In the system of writing accents that had become regular by the
Byzantine period, however, the grave is never used when a word is written by
itself. Rather, a word that has a final acute when written in isolation is written
with a final grave before another non-enclitic word, as long as punctuation
does not intervene. Thus, the genitive singular of Zeus has an acute on
the final syllable when written in isolation or before punctuation or an enclitic;
in the expression son of Zeus, however, the final syllable of is
written with a grave. This use of the grave accent mark is not regular in the
Hellenistic and early Roman periods, but examples are found, and there is
24 See Io. Philop. Ton. praecept. 6. 1920, and cf. (in the interpolation already mentioned)
[Arc.] 212. 14213. 4.
25 See Io. Philop. Ton. praecept. 6. 1929; Sch. Il. 5. 887a1 (A), with the parallel passages cited
by Erbse ad loc.
26 Support for a rising pitch has been seen in fragments of ancient Greek non-strophic
music with words, where a long vowel or diphthong with an acute accent is sometimes set
to two notes of which the second is the higher, but only very rarely to two notes of which
the second is the lower (Devine-Stephens [1994] 193). However, perusal of the Hellenistic
and probably non-strophic compositions in Phlmann-West [2001] suggests that a long
vowel with acute or (in modern notation) grave accent was not set to a rising sequence of
notes any more often than a long unaccented vowel.
936 Probert
,
<>
,
, . (Apol. Dysc. Pron. 36. 15)
For an enclinomenon [a word with a lulled accent] just lulls its own
accent:
But an enclitic, in addition to extinguishing its accent, also makes the
grave before it acute:
, .
... , 28
, ,
. ,
, .
. ,
([Arc.] 160. 20161. 4)29
27 A large literature on this question was motivated by the controversial thesis of Laum
(1928) that the modern use of the grave was, in most contexts, a late and purely graphic
convention. The papyrological evidence is well collected and presented by Mazzucchi
(1979b), with further bibliography.
28 I print the variant rather than (with Schmidt) in composition. The
sense requires , which is also found in a parallel passage at [Arc.] 199. 6.
29 The two discussions of enclitics at [Arc.] 159. 4162. 9 and (immediately afterwards) [Arc.]
162. 10169. 23 are thought to be composed (to a greater or lesser degree) of Herodianic
Ancient Theory of Prosody 937
... since the rule says that every oxytone word often lulls its acute to a
grave in connected speech, apart from , as in but ;
but ; but . The words apart from are
added because this word keeps the acute, as in . And
one needs to add to the rule, unless punctuation or an enclitic follows.
For then the acute is not lulled to a grave, as in .30
< >: ,
. (Sch. Il. 11. 51b (A))
material, but enclitics (and the accentuation of connected speech more generally) were
originally discussed in a supplement appended to the (see
[Arc.] 5. 114 and Cohn [1895b] 1156)).
30 Cf. the fuller discussion of the same material, ascribed to Choeroboscus, at Sch. Dion. T. 127.
31128. 10, and see also [Arc.] 5. 610; 199. 69; Io. Philop. Ton. praecept. 3. 234. 1; 6. 510.
938 Probert
than oxytone words with lulled accent. Evidence from fragments of ancient
Greek music also suggests that a lulled acute was higher in pitch than an
unaccented syllable but lower in pitch than an ordinary acute.31 The conclu-
sion that a lulled accent was a reduced rather than neutralised high pitch
may be supported further by the avoidance in accented papyri of either acute
or grave accents on syllables with lulled accents.32 If oxytone words within
the sentence had a reduced high pitch rather than a complete neutralisation
of the accent, this point was never quite explicitly recognised in grammatical
discussions, but the grammarians treatment of oxytone words in the sentence
certainly displays a concept of contextual variation. Oxytone words are treated
as having a basic or underlying acute on the final syllable, subject to a rule lull-
ing the accent in connected discourse. The lulling rule does not apply before
punctuation or enclitics. Alternatively, enclitics are treated as waking up an
already lulled acute (so Apol. Dysc., Pron. 90. 19; [Arc.] 167. 1214).
From the Hellenistic period onwards, word accents were classified accord-
ing to the syllable on which the accent fell, counting syllables from the end of
the word, and whether the accent was an acute or a circumflex. The following
terms, which have given rise to modern equivalents, go back to the Hellenistic
period:33
31 See Wackernagel [1896]; Ehrlich [1912] 252; West [1992] 199; Devine-Stephens [1994]
181183.
32 On the other hand, the practice of avoiding any accent marks on these syllables could, in
principle, be due to awareness of an underlying word accent, on the one hand, and of this
accents lack of phonetic realisation, on the other.
33 Most Hellenistic scholarship on the accent is not preserved in words which we can be
confident are the original ones, but see e.g. Sch. Il. 15. 656b (b(BCE3)Til) = Dion. T. fr. 22
Linke, where the text of the b manuscripts suggests that Dionysius Thrax is being quoted
directly. From the first century BC, we have fragments of Trypho in which this terminology
appears to be due to Trypho himself (see e.g. [Ammonius] 405 Nickau = Trypho fr. 15
von Velsen).
Ancient Theory of Prosody 939
The first two terms in this list can also be used of individual syllables within the
word: any syllable with an acute accent may be called , and any syllable
with a circumflex may be called .35 When applied to whole words,
all the terms in the above list take the final syllable of the word as a reference
point: a word described as is one whose final syllable is ; a
word described as is one whose final syllable is ;
a word described as is one whose pre-final syllable is ;
and so on. This classification of positions for the accent recognises an impor-
tant fact about Greek accentuation: positions for the accent need to be reck-
oned from the end of the word in order to show how words with a similar
morphological structure often turn out to be accented in the same way. Thus,
nominative singular masculine forms of adjectives with the suffix -- are
almost all oxytone: expressing character, Aeolic;
scientific, mathematical. If the position of the accent is calculated instead
from the beginning of the word, is accented on the third syllable,
on the fourth, and on the fifth; but the reckoning of the
position of the accent from the end of the word shows how all these words
share a position for the accent. We shall return further on to the importance
given to the terminations of words in giving rules determining how individual
words are accented.
7 Barytone Words
34 In addition to these adjectives, associated active and passive verbs are used: = puts
an acute on the final syllable; = is accented with an acute on the final syllable;
= puts a circumflex on the final syllable, = is accented with a circu-
mflex on the final syllable; = puts an acute on the penultimate syllable; etc.
35 See e.g. Sch. Il. 9. 529d (A).
940 Probert
normally allow: two syllables from the end of the word if the final syllable has
a long vowel or is closed by more than one consonant, and three syllables from
the end otherwise. Words with an accent as far from the end as these limits
allow are called recessive in modern terminology. The concept of the reces-
sive accent allows one to express clearly what most finite verb forms, for exam-
ple, share in their accentuation: they are recessive. Thus in the paradigm of
deliberate the forms , , , ,
, and are all recessive, although the first three are par-
oxytone and the last three proparoxytone.
Herodian was aware of the limits on the position of the Greek accent (see
especially Io. Philop. Ton. praecept. 4. 286. 5), but the concept of the recessive
accent as a useful category is only partly developed in ancient grammar. A word
which might be described as , , or
is often described instead as . In Philoponus epitome of Herodians
, the term is explicitly defined as referring
to a word with neither an acute nor a circumflex on the final syllable:
,
, , 36 ,
.
, (Io.
Philop. Ton. praecept. 6. 1017)
And one must know that for every word we put either an acute or a cir-
cumflex on one syllable, and on the other syllables a grave. So in
[] the second syllable has an acute, and the rest have a grave,
and in [] the middle syllable has a circumflex, and the
first and third have a grave. This is also why paroxytone and proparoxy-
tone and properispomenon words are called : because the last
syllable of these has a grave.
Not all words which are by this definition are recessive. Thus,
Philoponus example of another kind is but not recessive,
as is (for example) father (acc. sg.). Occasionally, the term
is indeed used for such non-recessive properispomenon and paroxytone
.
,... ([Arc.] 203. 1922)
37 Occasionally, or the associated verb is also used of oxytone words with a lulled
acute: so [Arc.] 167. 12.
942 Probert
, ,
, 38 ( ). ([Arc.]
7. 1517)
Words in - with one of the voiced stops before the are oxytone, when-
ever they have an anceps vowel in the penultimate syllable, or a vowel
long by nature, as in (tripod).
The following rule covers words with a certain termination (-) that are not
compounds, and with an aspirated or unaspirated voiceless stop (, , , , , )
before the termination:
,
, ,
. ([Arc.] 6. 1013)
However, not all these criteria were regarded as equally important for grouping
words likely to be accented similarly. The overall organisation of Herodians
, preserved in [Arcadius] epitome, divided words
first of all into major groups depending on their part of speech: (i)
or nominal forms (both adjectives and nouns), a group which occupied the
first fifteen of the twenty books; (ii) verbs; (iii) participles; (iv) pronouns; (v)
articles; (vi) prepositions; (vii) adverbs; and (viii) conjunctions.39 Within most
of these groups, there are major subdivisions into words with similar termi-
nations. Thus, the citation forms of masculine and feminine nominal forms
with more than one syllable are treated in the first twelve books, of which the
first deals with words ending in -, the second with words ending in -, -, and
-, the third with words ending in -, and so on.40 The main divisions within
these books are also into words with the same termination. Thus, [Arcadius]
first book deals in order with words ending in -, -, then more specifically
and , -, , , , , , , , , , , then
more specifically , -, -, , -, -, -, -, -, -, various
, ,
,
, , , , ,
,... (Io. Philop. Ton. praecept. 5. 510)
41 See e.g. Sch. Il. 8. 355 (A) (Aristarchus made perispomenon like ); Sch. Il.
2. 755b (A) (Aristarchus makes oxytone like ). However, the meaning
Aristarchus intended in using the formula X Y in such instances is controversial: see
Matthaios [1999] 2830, with bibliography.
42 See e.g. Chandler [1881] 56, although the term persistent is not used; Probert [2003]
5051, 54. For a somewhat different conception, see Bally [1945] 2930.
944 Probert
When the final syllable is long,43 the acute never falls three syllables from
the end. Thence, although the nominative, accusative, and vocative [sin-
gular] of are proparoxytone (, , ), the geni-
tive and dative [singular], and , are paroxytone because of
the long final syllable.
The principle of persistent accentuation itself is left implicit here, but in order
to see why the accent of and requires any explanation at all, we
are expected to understand that if it were not for the long vowels in the final
syllables of these forms, they would be accented on the first syllable, like the
nominative and accusative. In the part of Philoponus epitome corresponding
to Herodians fifteenth book (Io. Philop. Ton. praecept. 9. 1920. 34),44 persis-
tent accentuation in most nominal paradigms is recognised in that the accen-
tuation of most forms is related to that of the nominative singular in the same
paradigm (or sometimes the nominative of the same number as the form being
discussed). Thus, most accusative singulars and vocative singulars are said to
be accented on the same syllable as the nominative singular in the same para-
digm (Io. Philop. Ton. praecept. 12. 713. 1; 13. 812). The accentuation of most
genitive singular forms is also stated in terms of that of the nominative singu-
lar, with allowances for the need for the accent to shift in paradigms such as
that of , gen. sg. (Io. Philop. Ton. praecept. 9. 3410. 2; 10. 1434).
But relationships between pairs of non-nominative case forms are expressed
too. For example, any dative singular is said to be accented in the same way as
the genitive singular in the same paradigm, provided the two forms have the
same number of syllables (Io. Philop. Ton. praecept. 11. 612). This statement
captures a significant generalisation even about paradigms that do not display
persistent accentuation: even where the genitive singular is not accented like
the nominative singular, as in third-declension words with monosyllabic stems
such as nom. , gen. foot, the dative singular is accented like the
corresponding genitive singular if the two forms have the same number of syl-
lables: so gen. , dat. foot.
In addition to recognising the phenomenon of persistent accentuation,
Philoponus (and probably Herodians) method of formulating regularities for
the accentuation of oblique cases is an example of the ancient word and para-
digm approach to morphology, which has found advocates among linguists of
our time for its ability to capture significant generalisations about the struc-
ture of paradigms.45
The concept that oxytone words have a basic acute accent on the final syllable,
which is lulled in some contexts within the sentence and may be woken up
in others, has already been mentioned above. Not only oxytone words, how-
ever, but all words are considered to have a basic accent, a or .
Their accentuation may be altered not only by the rules applying to oxytone
words within the sentence but also by the rules applying to words followed
by enclitics.
Enclitics, like other words, were considered to have a basic accent of their
own; in fact all enclitics were considered to be either oxytone or perispome-
non ([Arc.] 162. 1112). The normal rules for enclitics apply only to monosyl-
labic enclitics and to disyllabic enclitics containing not more than one heavy
syllable;46 the basic accent of these enclitics was considered to be thrown back
onto the preceding word in certain contexts:
The enclitic throws its accent back onto the preceding word (to give an
acute on the last syllable of that word) if the preceding word is
(a) proparoxytone: as one might now express the idea, underlying
becomes
(b) paroxytone with a trochaic ending: underlying becomes
47
(c) properispomenon as long as the last syllable is not closed by more
than one consonant:underlying becomes 48
(See [Arc.] 160. 517)
45 See Robins [1959] and Matthews [19912] 185205, who proposes a modernisation of the
ancient model. On the ancient model see further Roussou (2012).
46 A category of disyllabic enclitics with two heavy syllables was also recognised: see [Arc.]
160. 15, and cf. Probert [2003] 150151.
47 This rule is not observed in most modern editions of ancient texts, and is generally
suspected of having been stated too generally in the grammatical tradition. For discussion
and bibliography see Probert [2003] 148150.
48 The restriction to words whose last syllable is not closed by more than one consonant
is necessary because words such as and are properispomenon but do not
receive an accent on their final syllables when an enclitic follows (so , not
* : see [Arc.] 160. 1417).
946 Probert
In other contexts, the enclitic was considered to lose its basic accent while the
preceding syllable kept its own basic accent:
The enclitic loses its basic accent while the preceding word keeps its
basic accent if the preceding word is
(a) oxytone: underlying becomes
(b) perispomenon: underlying becomes
(See [Arc.] 160. 17161. 3)
,
. ([Arc.] 161. 1921)
For these [enclitics] neither can keep their own accent, nor do they throw
it back onto the preceding word.
49 An exception is made for paroxytone words followed by enclitic pronoun forms beginning
with -: see [Arc.] 161. 1316; 166. 1922; and cf. Probert [2003] 150.
Ancient Theory of Prosody 947
mental phonology (on which see Goldsmith [1990]). Furthermore, the idea
that there is implicit or mental enclisis in a sequence such as ,
with the enclitic unable to keep its own accent or to throw it onto the preced-
ing word, is rather comparable to the notion of floating tones in autosegmental
phonology: tones that exist underlyingly but fail to surface in environments
where there is no syllable on which the language in question allows them to
surface. It would, however, be a mistake to regard Herodian or [Arcadius] as
having developed autosegmental phonology in any detail.
,
. ,
, , . (Apol. Dysc.,
Conj. 256. 29257. 1)
And the following is most clear: that words that have lost a consonant or
acquired a consonant have the same accentuation, unless a general rule
opposes this. Therefore did not become , nor did become by the
addition of the . For the one has a circumflex, the other an acute.
Breathings, too, could not simply come and go at random when a word was
affected by a . Some affected breathings, but they did so according
to rules that could be stated (see Lentz [1860] 664697). For example, word-
initial - normally had a smooth breathing, and when a turned word-
initial - into -, the resulting - received a smooth breathing even if the - of
the basic form had a rough breathing. Apollonius Dyscolus appeals to this reg-
ularity in the course of an argument that the form is derived from
by the deletion of word-initial -, a that causes a rough breathing:
,
, , ;
. ,
. (Apol. Dysc., Synt. 153. 21154. 4)
For when does the diphthong have a rough breathing, considering that
even aspirated vowels become unaspirated when it appears? (So)
(becomes) , (becomes) . But has a rough breathing
because of . Therefore would not otherwise have a rough
breathing, if it didnt have a reason in the deletion of -.
Orthography*
Stefano Valente
1 Introduction
2 Ancient Origins: From the Beginnings to Aristarchus
3 The Rise of Greek Orthography: Asclepiades of Myrlea, Trypho and
Apollonius Dyscolus
4 Herodian and the Sistematization of the Greek Orthography
1 Introduction
Greek orthography was a well-defined part of ancient grammar from the 2nd
half of the 2nd century BC onwards. According to the standard definition, ini-
tially formulated by Trypho and then repeated with slight modifications by all
the Greek grammarians, orthography () is both the correct spell-
ing of a word and the account of its correctness. It does not address all the
spelling problems, as modern orthography does, but only particular
(inquiries):1 as Quintilan stresses (Inst. 1.7.1), the orthographical art has all its
subtlety in dubious points (totam ... subtilitatem in dubiis habet)2 and exam-
ines those cases in which the pronunciation does not help to determine the
correct spelling of a word.3
There are three fields of research within Greek orthography: syntax, quality
and quantity.4 , syntax (or , division) deals with problems
of syllabification, especially with the boundaries of syllables; , qual-
ity, concerns doubts in the spelling of consonants, and , quantity,
* I would like to express my gratitude to Marco Ercoles, Leonardo Fiorentini, Camillo Neri,
and Renzo Tosi for their valuable suggestions and critical appraisal of my paper, as well as to
Rachel Costa and Christopher Sheppard for revising my English.
1 See Wendel [1942a] 1438; Blank [1998] 195.
2 See Ax [2011] 309f.
3 See Apoll. Dysc. Synt. 7.614 (see below, section 4); see also Wendel [1942a] 1437f.; Blank
[1982] 9.
4 See Wendel [1942a] 1554f.; Blank [1998] 199; Hunger [1978] 18; Schneider [1999] 4ff; see also
Pagani in this volume.
5 See Wendel [1942a] 1456; Siebenborn [1976] 5355, 159161; Hunger [1978] 18; Pagani in
this volume.
6 See Di Benedetto [2007] 417.
7 See below, section 2, and Pagani in this volume. On the history of this word, see e.g.
Canfora [1987].
8 See Wendel [1942a] 14421454; Hunger [1978] 1822; Alpers [2004] with further bibl.
9 See Dickey in this volume.
10 See Egenolff [1888] 3f.; Wendel [1942a] 1440; Hunger [1978] 18; Alpers [2004] 2f.
11 See M. Victor. 4.58 (p. 80 Mariotti = GL 6.17.13) orthographia Graecorum ex maxima parte
in littera consistit (Greek orthography consists for its most part in the letter iota;
see Wendel [1942a] 1444; Blank [1998] 198). The development of the diphthong epsilon-
iota to the sound ||, already (but rarely) attested in the 4th century BC, is quite typical of
the Hellenistic period (see Schwyzer [19593] 193; Meillet [19657] 273, 307f.; Threatte [1980]
Orthography 951
Prior to the modern age the Greek language had never undergone any spelling
reform that would adapt changes in pronunciation to the written language.
However, for centuries grammarians had aimed to preserve the traditional
spelling and its rules,12 thus opposing the historical evolution of the Greek lan-
guage and formulating new and ever stricter rules and canons of correction.
190207; Allen [19873] 70; Adrados [2005] 192f. with bibl.). However, grammarians seem to
have begun to handle this phenomenon only in the 2nd century BC (see Wendel [1942a]
1438). It must be stressed that Greek orthographical treatises could represent important
(but sometimes neglected) witnesses of the phonetic evolution of the Greek language.
12 Such views have contributed to the increasing dichotomy between the spoken and writ-
ten language, which brought about the modern distinction between demotik and kathar-
evousa (see Adrados [2005] 291ff. with bibl.).
13 See Wendel [1942a] 1438; Jeffery [1961] 142.
14 See Threatte [1980] 2651; Erbse [1994] 82f. with bibl.; Ruijgh [2001] 269f.; Smith [2003]
318. For instance, some passages of Platos Cratylus are based on speculations about the
introduction of the Ionic alphabet in Athens (e.g. 398de, 410c: see Lebeck [1969] 61).
15 See Marrou [19656] 7486; Morgan [1998] 921. For the history of orthography in the clas-
sical age, passages like Xen. Mem. 4.2.20, where the knowledge of reading and writing cor-
rectly () seems to be limited to primary school education, do not help to recover the
rules designed to correct the written (and spoken) language, if indeed such rules existed.
16 See Maas [1922] 1396; Page [1953] 59f., 66f.; Meillet [19657] 96f.
17 See Barwick [1922] 202f.; Fehling [1965] 216; Siebenborn [1976] 1424; Pfeiffer [1968] 37, 42,
53, 280f.; see also Novokhatko and, especially, Pagani in this volume. For the distinction
between correct spelling and correct diction, see for instance Vel. Long. 9f. Di Napoli (= GL
7.71f.): see Barwick [1922] 203, 208f.; Siebenborn [1976] 36f.).
952 Valente
18 Praxiphanes (fr. 9 Wehrli) and Aristoteles (see Rh. 1407a20) are cited in Prol. Vat. Dion. T.
164.26ff. = sch. Lond. Dion. T. 448.13ff., where the origin of this doctrine is referred back to
Theagenes of Rhegion (6th century BC, fr. 1a D.-K.): see Pfeiffer [1968] 11, 158 with bibl.
19 See Diogenes of Babylon, fr. 24 Arnim (SVF 3.214.13 = D. L. 7.59); see Pohlenz [194819552]
1.53, 2.31; Siebenborn [1976] 24.
20 See Montanari [1995b] 42ff.
21 See also Morgan [1998] 152ff.
22 On the notion of dialect in the Classical age and in Hellenistic scholarship, see e.g. Latte
[1925]; Pfeiffer [1968] 12ff., 41ff.; Morpurgo Davies [1987]; Tosi [1994b] 209; Colvin [1999]
3989; Ascheri [2005] 440 n. 2 with further bibl.
23 See Siebenborn [1976] 27.
24 See Matthaios [1999] with bibl. See also Montana in this volume.
25 See Fehling [1956] 260; Erbse [1980]; Callanan [1987] 2225, 97102; Schenkeveld [1990]
291f.; Matthaios [1999] 588; Ax [2000] 100102; Di Benedetto [2007] 421f.
26 On the idea of , see Tosi [1994a] 234f.; Montanari [1995b] 42ff.; Sluiter [2011] 303f.
27 See Siebenborn [1976] 2731; Ax [2000] 107, 138; Pagani in this volume.
28 See Siebenborn [1976] 27f.; Taylor [1986] 186.
Orthography 953
in the scholia and epimerisms to Homer and in Byzantine lexica with regard
to orthographic (and more generally grammatical) problems are more likely
due to later grammarians,29 who used them as auctoritates both in bonam and
in malam partem, shifting their criteria for textual criticism to a more theo-
retical and normative perspective. In this respect, some good examples
are provided by the Alexandrian discussions on the right spelling of
(Skiron)which involved at first Callimachus and Aristophanes30and on
the spelling of (shore) and (nose).31 On these two latter problematic
cases, the primary source is an entry of the Etymologicum Gudianum (77.1621
de Stefani s.v. ), which preserves the doctrine of Herodian (GG 3.2 431.111)
through the intermediation of Choeroboscus Orthography:32
...
,
,
. ,
.
.
.
Aristarchus thinks that (shore) and (nose) are rightly spelt with
the diphthong epsilon-iota and, following the etymology, he said that
comes from (to be struck), because the waves strike on and hit
the beach, or from (to run), because the waves run on the beach;
comes from (to flow), because the head secretions flow and run
down from the nose. Thus, following such an etymology, he says that
these words are spelt with the diphthong epsilon-iota, while the para-
dosis knows them with iota.
29 See Wendel [1942a] 1439; Desbordes [1990] 165f.; Schneider [1999] 850f.
30 See below, section 4.
31 See Lentz [1867] XCIII; Wendel [1942a] 1439; Siebenborn [1976] 31.
32 See Choerob. Orth. 167.15 (unde plenior Etym. Gud. 76.13 de Stefani). The direct tradition
of Choeroboscus Orthographya work in alphabetic order and , that is to say,
notes taken at his lectures or classespreserves only an epitome of the work, while a
fuller version was the source of the Byzantine Etymologica (see Alpers [2004] 31ff. with
bibl.; Valente [2010a] 639642). On Choeroboscus sources, see Alpers [2004] 33ff. with
bibl.; Dickey [2007] 80f. with bibl. and in this volume.
954 Valente
33 On this (par)etymological spelling, see Apollon. Soph. 86.24 (~ Hsch. 591 Latte; see also
Epim. Hom. 34D1,2 Dyck with the testimonia ad ll.; Eust. Il. 109.4f.). See also Schironi [2004]
276279.
34 See van der Valk [1964] 145, who deals with the Aristarchean creation of the reading
instead of the correct (terrible) in Il. 20.259, stressing that we need no doubt
that originates with Arist[archus]. The latter coined a new word in order to make
the text more interesting, and in n. 282: we must not forget that a number of similar
instances can be mentioned which illustrate Aristarchus mentality in this respect. Thus
he explained and by and .
35 See Pagani in this volume.
36 Nevertheless, in all the grammatical works up to Herodian the interpretation of texts
remained basic and still performed, as can be seen, for instance, in the fragments of
Asclepiades of Myrlea, Tyrannion, and Philoxenus (see Montanari [1993b] 256).
37 See Tischendorf [1847] 5456; Tchernetska [2000] 100f.; Ascheri [2005] 412ff. with further
bibl.
38 Formerly, Reitzenstein ([1897] 299ff., [1901] 84ff.) ascribed this text to Herodian (followed
by many, like Wendel [1939c] 1179; [1942a] 1441). See also Ascheri [2005] 417420 with fur-
ther bibl.
39 I print Reitzensteins text [1897] 302: |[ ......] |20
[ ] , | [***] |[ , ]
|[ .............] | [, ]
Orthography 955
,
. , ,
,
, ,
,
(...) ,
,
,
.
They say that orthography is threefold divided into quantity, quality, and
division. Quantity is when we inquire whether iota must be adscribed to
datives, and whether (well-brided) and (fruitful)
50 Asclepiades is probably the source of the examples in Sextus (Blank [1998] 198; see also
Calboli [1962] 147ff.; Di Benedetto [2007] 477).
51 See also Barwick [1922] 227f.; Siebenborn [1976] 5355; Baratin [1989a] 204, [1989b] 215;
Matthaios [1999] 1416; Ax [2000] 128f.; Matthaios [2007] 13f.; Pagani in this volume.
52 See above, p. 954f. Pagani [2007a] 136 rightly reasserts that the ascription of an autono-
mous orthographical treatise to Asclepiades is groundless (see also Mller [1903] 29; con-
tra Barr Reid Forbes-Saks [19963] 187).
53 I print the text of Mau [1954], except for Dorvilles (see Egenolff [1888] 4) instead of
mss . See also Blank [1998] 195201 and Pellegrin [2002].
Orthography 957
must be spelt only with iota or with the diphthong epsilon-iota. Quality
is when we inquire whether (scalpel) and (Smyrna)
must be spelt with zeta or sigma. Division is when we are puzzled by the
word (strong), wondering whether beta begins the second syl-
lable or ends the first, and, in the name (Aristion), we are
unsure to which syllable sigma must be assigned [...]. We are not injured
according to whether we spell the dative case with iota or not, and
(scalpel) and (Smyrna) with sigma or zeta, and whether we
divide the word (Aristion) assigning sigma to the first syllable
or to the following one.
54 See Hiller [1871] 613f.; Blank [1998] 197201. This tripartition can also be detected in the
orthographic chapter of Quintilians Institutio oratoria (1.7.19; see Colson [1924] 92).
55 See Allen [19873] 8488. See also Quint. Inst. 1.7.17 with Ax [2011] 323f.
56 See above, section 2.
57 See Threatte [1980] 510, 547549; Allen [19873] 45f. In Charax Orthography (see below,
section 4), the spelling of is not only a problem of quality, but also of syntax (see
Egenolff [1888] 10). See also Blank [1998] 199, who rightly quotes Luc. Jud. Voc. 9 as useful
parallel.
58 Syllabification was a typical exercise in Classical and Hellenistic primary school: see
Guraud-Jouguet [1938]; Marrou [19656] 229242; Threatte [1980] 6473 for a survey of
syllabic divisions in inscriptions; Cribiore [1996] 269, no. 379 with bibl.; Morgan [1998]
14, who rightly cites Pl. Plt. 277e-278c, and 164. See also Callias Alphabet Tragedy (test. *7
K.-A. ap. Ath. 7.276a, 10.453c: Smith [2003] with bibl.). On the word division (),
see also sch. Lond. Dion. T. 447.25f. (where the focus is on Hellenismos); see also comm.
Melamp. seu Diom. in Dion. T. 31.69 (see Blank [1998] 199 n. 185) ~ sch. Marc. Dion. T.
316.31317.2. See also Quint. Inst. 1.7.9 (for haruspex and abstemius) with Axs commentary
([2011] 315).
59 See Heinicke [1904] 71. At least analogy, etymology and usage were already part of
Hellenismos (see for instance Sext. Emp. Math. 1.189, 1.241247). See Barwick [1922] 214,
259; Siebenborn [1976] 54; Baratin [1989b] 213ff.
958 Valente
]|15 [ ]| ,
]| [ ]| -
[, ]|20 , [ ]|
[ .
60 We know little of the development of orthographical doctrine after Asclepiades: besides
the names given in the Leipzig palimpsest (see above, p. 954 n. 39), we are aware that
other grammarians (like Philoxenus and Diocles, that is to say Tyrannion of Stratonikeia)
dealt with orthography (see Ascheri [2005] 434 and nn. 37 with further bibl.).
61 On Trypho and the different aspects of his grammatical studies, see Dickey [2007] 84f.
with bibl.; Ippolito [2008] with bibl; also Montana in this volume.
62 It is attested in the biographical entry of Hesychius of Miletus in the Suda lexicon ( 1115
Adler; on the relation between Hesychius of Miletus and the Suda lexicon, see now Alpers
[2009b = 20132] 151158 with bibl.).
63 I print Reitzensteins text [1897] 303.
64 See below, section 4.
65 Mller [1903] 36 and Wendel [1942a] 1442 ascribe this change to Herodian. A syntax prob-
lem handled by Trypho is probably attested by sch. Marc. Dion. T. 446.1619 (spelling of
, trickle, with or without iota mutum). On the term , see also Lambert [2011].
Orthography 959
is predictably missing.75 Once again, Scaurus seems to have drawn the doc-
trine of orthographical canons correspond to regulaenot from
Varro, but from an unidentified Greek source.76 The translations of
(etymology) with originatio and of (analogy) with proportio seem
to have been well documented, and likewise their use,77 but the criterion of
historia within the grammatical field seems to be otherwise unattested in Latin
authors.78 Thus if Scaurus source was Trypho, we would have a primary wit-
ness as to his lost work. Moreover, Scaurus order of the orthographical can-
ons, i.e. history, etymology and analogy, may be the same as that of Apollonius
Dyscolus79 and thus may echo that of Trypho. Overall, the first canon of cor-
rection remains history, or paradosis, which therefore retains a central role as
in the case of the Alexandrian scholars.80
In short, the general conclusions of Schenkeveld ([1994] 281) on the devel-
opment of Greek grammar in the Hellenistic age also apply to the first steps of
Greek orthographical doctrine:
75 See Fehling [1956] 252 with n. 1; Siebenborn [1976] 92; Biddau [2008] 73; contra Usener
[1913] 296f.
76 See Biddau [2008] 72.
77 See Biddau [2008] LIIf.
78 See Strzelecki [1950] 98; Siebenborn [1976] 54, 92; Schmid [19361942] 2837f. (II.B); Biddau
[2008] 73.
79 See below, section 4.
80 See Wendel [1942a] 1456.
81 On the orthograpic works of Didymus, Alexion, Soteridas, and Dracon of Stratonikeia, see
Schneider [1999] 852ff. with bibl.
82 See Wendel [1942a] 1437f., 1440; Schneider [1999] 863867; Dickey [2007] 7375: 74; and
Matthaios in this volume.
Orthography 961
,
,
,
(...) ,
.
83 The title is indicated by Apollonius himself (Synt. 388.8, see below n. 90). See Schneider-
Uhlig [1910] 7 (Suda 3422 Adler is quoted, where such work is missing); Wendel [1942a]
1437f., 1440.
84 Especially if he is the Apollonius who begins the list of orthographers in the canon of the
Par. Coisl. 387 (see Kroehnert [1897] 7; Wendel [1942a] 1440; Alpers [1981] 144; Schneider
[1999] 872).
85 See Reitzenstein [1897] 302; Schneider-Uhlig [1910] 7; Blank [1982] 8, 69 n. 83; Schneider
[1999] 867.
86 See Blank [1982] 810; Lallot [1997] 2.14.
87 See also Schneider [1999] 864.
88 For the idea of correction in the orthographic field, see for instance Apol. Dysc. Conj.
213.1114 and Wackernagel [1876] 48; Dalimier [2001] 223f.
89 See Lallot [1997] 2.14; Schneider [1999] 865. On (reason) used in an analogi-
cal context, see also Sophronius excerpta from Charax commentary to Theodosius (GG
4.2 431.1921 Hilgard).
962 Valente
for the use of analogy to solve syntactic problems by introducing once again
the parallelism with orthography within the doctrine of Hellenism:
,
, ,
.
,
, , ,
.
.
90 Synt. 388.48
,
,
. See Schneider-Uhlig [1910] 6f. (with the quotation
of Choerob. in Theod. 219.1824); Lallot [1997] 2.238. See also Schneider [1999] 866, who
quotes the so-called Anonymous Crameri, AO 2.311.512 Cramer.
91 See Siebenborn [1976] 54f.; Blank [1982] 24.
92 ,
,
, (see Lallot [1997] 2.149).
Orthography 963
93 ,
, ,
. ,
.
, ,
(for already even the so-called first, undivided matter of the ele-
ments of the alphabet [i.e. ] demonstrated this long before, insofar as it did not make
combinations of elements in just any manner, but only in the construction which was
according to the necessity, whence they virtually took their name [i.e. ]; and the
syllable, going beyond the element, has accepted this same thing, since the constructions
resulting from syllables and filled-out according to the necessitys law complete the word.
And it clearly follows from this that the words too, since they are part of the sentence
which is syntactically complete, accept the regularity of syntax, tranls. Blank [1982] 30).
94 See Lallot [1997] 292.
95 See also Dalimier [2001] 34f.
96 See Apol. Dysc. Synt. 9.2 (see Dalimier [2001] 34f.); see also e.g. [Timotheus of Gaza]
AP 4.241.10 Cramer (Hdn. GG 3.2 393.28 and 395.25).
97 On dialect, see Apol. Dysc. Conj. 213.1114 (see above, n. 88), Pron. 94.1417 (together with
paradosis; see Siebenborn [1976] 150), on etymology see Adv. 153.8 (quotation of Trypho
fr. 67 von Velsen).
964 Valente
98 See Dyck [1993a] with bibl.; Alpers [2004] 1ff. with bibl.; Dickey [2007] 7577 with bibl.;
Matthaios in this volume.
99 On the title of this work, see Egenolff [1888] 7f.; Schneider [1999] 770, 776f.; the standard
but not unchallenged edition is that of Lentz [19671970], see Dyck [1993a] 788f., Alpers
[2004] 1f. with bibl.
100 On Charax, see Alpers [2004] 19ff. with bibl.
101 See Egenolff [1888] 4ff.; Alpers [2004] 7f.
102 I print the text of Alpers [2004] 6f. with few changes in punctuation: moreover, I adopt
Graux instead of of the manuscripts, and
proposed by Egenolff ([1888] 20 n. 14) instead of . See Valente [2010a] 647ff.
Another complete preface to orthography (very close to that of Charax) is preserved in
the Grammar falsely ascribed to the grammarian Theodosius (see Uhlig [1883] XXXVII;
Egenolff [1888] 1013; Alpers [2004] 23ff. with bibl.): this work, a hotchpotch of gram-
matical materials from many different and heterogeneous sources (some of them still
unidentified), was badly edited for the first time by Goettling [1822] (see Alpers [2004]
2326) on the basis of Parr. gr. 2553 and 2555; the orthographic chapters are contained on
pp. 6179. The preface on orthography of Ps.-Theodosius (pp. 61.2262.26) was also edited
by Bekker [1821] 1127f. n. * on the basis of Vat. gr. 1370 (see Egenolff [1888] 11; Alpers [2004]
8, 24). The relationship between the latter text and Charax is complex and still unclear:
it is generally assumed that they do not derive one from another, but independently use
Herodian as the direct source. In particular, Ps.-Theodosius pp. 61.2262.26 shows a strik-
ing coincidence with a passage of Choeroboscus Epimerisms on the Psalms (89.530, unde
Etym. Gud. 499.2640 Sturz: see Valente [2010a] 642f.), and thus could possibly come from
Choeroboscus lost preface to his Orthography (see Hilgard [1894] LXXXI; Alpers [2004] 32
n. 130; Valente [2010a]).
Orthography 965
<>
,
. ,
.
,
, .
, , .
,
, ,
, ,
, .
, ,
.
, ,
. ,
, ,
, .
, ,
, ,
, ,
,
, ,
, .
103 ,
.
103 The correct spelling would be (see Hamm [1957] 20, section 27), but is shared
also by Etym. Magn. 816.52 and therefore must not be corrected: see below, pp. 969f.
966 Valente
In the first section Charax gives a definition of orthography which recalls that
of Trypho, although he shows slight but significant differences:105 in Tryphos
definition the formula word correctly spelt ( )106
appears twice, namely in both parts. On the other hand Charax (or better still
his source Herodian) defines orthography first as the exact spelling according
to the word ( ), introducing Tryphos formula
104 In the translation of these passages I take into account that of Blank [1998] 197.
105 See Reitzenstein [1897] 303; Wendel [1942a] 1437.
106 The same definition occurs in Ps.-Theodosius (p. 61.23f. Goettling [1822], and Bekker [1821]
1127 n. *): Valente [2010a] 644f.
Orthography 967
107 Here, Ps.-Theodosius definition (p. 61.24f. Goettling [1822], and Bekker [1821] 1127 n.
*) is close to that of Charax: { } (Bekker : om. Goettling, seclusi)
, the demonstrative canon of the word cor-
rectly spelt (Valente [2010a] 644f.). A quite different definition of orthography is testified
to by the ms. Vindob. phil. gr. 240 (see Egenolff [1888] 13; Wendel [1942a] 1449):
.
, , ,
, (this orthography is the written manifestation of let-
ters revealed through the voice. It is divided into three parts: syntax, quality and quantity;
it has four canons: analogy, history, dialect and etymology). On the influence of Greek
definitions of orthography on Latin grammarians, see Wendel [1942a] 1437.
108 See above, section 3.
109 ...
. On the
relationship between this work and Herodian, see Dyck [1993a] 777 and nn. 23f. with bibl.;
Dickey [2007] 81f. with bibl.
110 See Lentz [1867] XCIX, 7 app. l. 19f.; Schneider [1999] 781.
968 Valente
118 On spelling problems concerning consonants in Attic inscriptions, see Threatte [1980]
434643.
119 See Siebenborn [1976] 39 with n. 6, who rightly quotes M. Victor. 4.53 (p. 79 Mariotti = GL
6.16.6) and 4.70 (p. 82 Mariotti = GL 6.19.12); Threatte [1980] 595597.
120 See sch. Il. 8.441b1 Erbse (~ Etym. Gen. 614 Lasserre-Livadaras, unde Etym. Magn. 1048
Lasserre-Livadaras = 81.15 Gaisford) (fr. l.) ,
(Od. 7.100). ,
, , (Il. 10.298).
(Hdn. l.l.). See also Threatte [1980] 616f., 620623.
121 See Lentz [1867] CI; Hiller [1871] 614.
122 On spelling problems concerning vowels in Attic inscriptions, see Threatte [1980] 120434.
123 See also M. Victor. 4.58 (p. 80 Mariotti = GL 6.17.13: see Siebenborn [1976] 40 n. 2).
124 See above, section 3.
125 See sch. Il. 1.129a1 Erbse (Hdn. Orth. fr. 48, GG 3.2 419.19); see Erbse [1960] 91.
126 See also Choerob. Epim. Ps. 89.530 Gaisford: see above, n. 102.
127 Hilgard [1894] LXXXI points out that this entry may come from the lost preface to
Choeroboscus Orthography (see Choerob. Orth. 275.19 ~ Etym. Gud. 566.2636 Sturz
[Hdn. GG 3.2 604.30]); see Alpers [2004] 8 n. 26, 32 n. 130; Valente [2010a] 641 n. 9; how-
ever, the similarities with Charaxs introduction [see above, n. 102] could suggest a dif-
ferent interpretation). Alpers ([2004] 8 n. 26) suggests that the similar definitions of
the four canons in sch. Lond. Dion. T. 454.14 (dealing with the 12 canons of ,
970 Valente
,
, ,
, ( <>
, 128
, ,
)
(i.e. )129 .
reading, see Fehling [1956] 251) may likewise come from Choeroboscus (see also
Siebenborn [1976] 159).
128 See above p. 956f.
129 The meaningless of Etym. Magn. should probably be understood as
on the basis of sch. Lond. Dion. T. 454.14.
130 See Siebenborn [1976] 67, 159ff. See also the anonymous excerpt in Reitzenstein [1897]
384386 (first edited by Aldus in the Thesaurus cornu copiae et horti Adonidis, Venetiis
1496), falsely ascribed to Choeroboscus (see Hilgard [1894] LXXXII; Reitzenstein [1897]
384 n. 1).
131 See Fehling [1956] 219ff.; Siebenborn [1976] 56ff.; Sluiter [2011]; Pagani in this volume.
Orthography 971
zusammen und bildet daraus die Regeln.132 Analogy allows one to build para-
digmatic canons: the enunciation of the canon ( )133 suffices
to find the correct spelling of a word. Thus analogy can be the demonstrative
canon itself ( ),134 corresponding to the second definition
of orthography;
ii) dialect is a special form of a language ( ).135 Dionysius Thrax
may well have used this criterion to solve the spelling problem of the adverb
,136 but it was probably only with Trypho that it became one of the four
orthographic canons.137 For example, the spelling of is accounted for
by the Aeolic form : this concise account can be explained on the basis
of Herodians wider canon (GG 3.2 302.612), in which some Aeolic and Attic
forms of corresponding words are compared to justify the correct spelling;138
iii) etymology139 is the concise and true demonstration of the matter of
inquiry according to its genuine sense, that is, its true origin, or the mean-
ing that has the maximum degree of plausibility (
, ,
).140 In orthography, etymology is used to analyse a
132 See Comm. Melamp. seu Diom. in Dion. T. 15.1117, sch. Marc. Dion. T. 309.9f. (see
Siebenborn [1976] 67 n. 1).
133 See sch. Lond. Dion. T. 454.17f.
(explanation of the canons which comes from the juxtaposition of what is similar). For
(juxtaposition of what is similar) indicating the criterion of anal-
ogy, see Sext. Emp. Math. 1.199, 1.236 (Siebenborn [1976] 63 with n. 3; Schenkeveld [1994]
282f.; Sluiter [2011] 294f.).
134 See sch. Vat. Dion. T. 169.26f. ~ sch. Marc. Dion. T. 303.22f.
135 The same definition is given by Gregory of Corinth (p. 9 Schaefer): see Morpurgo Davies
[1987] 7f. As Siebenborn [1976] 91f. demonstrated, Usener [1913] 622 and Barwick [1922] 258
n. 3 are wrong in trying to identify dialect with usage (), a canon of Hellenismos.
136 See above, section 3.
137 See above, section 3. See also sch. Lond. Dion. T. 470.2225
,
,
(see Siebenborn [1976] 147).
138 Etym. Magn. 582.3450 (brevius [Zonar.] 1342.37) ~ Choerob. Orth. 242.1527 (unde Etym.
Gud. 413.4452 Sturz [Hdn. Orth. GG 3.2 557.412]; see also Heracl. Mil. fr. 19 Cohn; Eust. Il.
1340.18). See also Siebenborn [1976] 147.
139 See Usener [1913] 297; Barwick [1922] 208f.; Fehling [1956] 252; Schrter [1960] 2537;
Siebenborn [1976] 141ff.
140 On this twofold definition, see Siebenborn [1976] 160, who rightly cites Quint. Inst. 1.6.28ff.
(see Ax [2011] 272ff. with further bibl.). On the other hand Charax has only the first (with
the probably inferior variant readings and ).
972 Valente
word in order to discover its true origin, not for a theoretical but for a practi-
cal goal: correct spelling. As Schrter stresses ([1960] 35), die Erkentniss des
Ursprungs verhilft zur Erkenntnis der richtigen sprachlichen Form, der Lesung
und Schreibung. im zweiten Teil bedeutet etwa Kennzeichnung,
gleichsam den Umri oder Abdruck geben wie bei einem Siegel oder einer Mnze.
Es kann wohl Bedeutung und Formales enthalten. Wichtig ist die Einschrnkung
ihrer Gltigkeit am Schlu: sie mu berzeugend sein. This is the case, for
instance, of and ,141 as well as of the challenged spelling of
and ,142 which involved Aristarchus and Herodian in different perspectives
(textual criticism for Aristarchus, grammatical-orthographical for Herodian);143
iv) history144 is the usage of the ancients, which is also the textual tradi-
tion ( ,145 ). The distinction between
usage () and the textual tradition or paradosis () is not an idle
question,146 as can be confirmed by an entry of the Etymologicum Magnum
(791.49792.10),147 which offers a good example of an ancient orthographical
inquiry and comes from Herodian (who discusses a spelling of Aristarchus)148
through Choeroboscus:149
141 For example, as regards , see Steph. Byz. 69 B.-Z., Choerob. Orth. 208.2527 (Hdn.
Orth. GG 3.2 500.2426) ~ Choerob. in Theod. 250.37 and Etym. Gud. 419.19f. de Stefani
(cf. [Hdn.] Epim. 48.9); on , see e.g. Choerob. Orth. 204.510 (Hdn. Orth. GG 3.2
499.34500.4).
142 See above, section 2.
143 On the use of etymology in the field of syntax, see the case of , thief (Quint. Inst.
1.7.17 with Ax [2011] 323f., Hdn. Mon. Lex. GG 3.2 946.4f., Etym. Gen. 95 Alpers [Hdn.
Orth. GG 3.2 545.4f.], Comm. Melamp. seu Diom. in Dion. T. 40.4). For the use of etymology
in a Roman context (in the field of syntax), see e.g. Scaur. 2.5 (p. 7.10ff. Biddau) and 9.1
(p. 51.12ff. Biddau), Quint. Inst. 1.7.9 (see Siebenborn [1976] 144; Ax [2011] 315).
144 See Siebenborn [1976] 85ff. Before becoming one of the criteria of Hellenismos, it was
originally used for textual criticism by Alexandrian scholars (see van Groningen [1963] 1;
Siebenborn [1976] 86).
145 See Hdn. Mon. Lex. GG 3.2 910.7f., 911.11 and 24, 919.6, 920.1, 935.6 and 8. See also sch. Lond.
Dion. T. 470.4f. and 22f.
146 On this opposition originally belonging to textual criticism, see Etym. Magn. 815.1621;
Rutherford [1905] 374; Siebenborn [1976] 86ff.
147 See Schironi [2004] 205f.
148 See Schironi [2004] 205211 with bibl.
149 See also Epim. Hom. 47 Dyck with testt., Choerob. in Theod. 27.25
,
, ~ 328.69
, ,
, , , , and
Orthography 973
History is not only the evidence of the paradosis, which is the starting point
of any orthographical inquiry, as the Leipzig palimpsest confirms (ll. 1517
[] [ ]| [ , trying to find
the canons of the transmitted writing). History is also the (literary) usage,151
namely the literary tradition ( , evi-
dence given by ancient and famous men, or , the usage
of the ancients152). Thus history is the constitutive matter of orthography
( ). At the same time, it is also an orthographic
tool ( ... ) to be used as extrema ratio in the correction
(), if a spelling could not be corrected with the other three canons.153
This specification provides greater insight not only into the explanations given
for the spellings of , and ,154 but for that of as well:155 its spell-
ing with iota could not be explained on the basis of dialect, because, since
the Aeolic form is ,156 the expected spelling should have been .157
Therefore, it is now evident why history and dialect are often contrasted with
each other ( or
).
Moreover, the entry of the Etymologicum Magnum is useful to clarify the
equivalence of history and paradosis, briefly hinted at with the spelling prob-
lem of . It was a Hellenistic vexata quaestio involving Callimachus and
Aristophanes of Byzantium with regard to the challenged spelling /
(Etym. Gen. AB s.v. ):158 while Callimachus in the Hecale (fr.
296 Pfeiffer = 59 Hollis) used the (correct) spelling with iota for the name
159 On the correct spelling with iota, see Pfeiffer [1949] 273; Threatte [1980] 193; Nelis [1994]
931f., especially no. 106 (Berlin, Staatl. Mus. F 2288); Hollis [20092] 210.
160 Deest in Nauck [1848a] and Slater [1986], see Callanan [1987] 24 n. 20; Theodoridis [2009] 397.
161 See Callanan [1987] 24.
162 Sch. Vat. Dion. T. 165.1624 ]
(see Slater [1972] 317 n. 2 with bibl.)
<> (...)
(Siebenborn [1976] 160 identifies with )
, ,
, ,
. (~ sch. Lond. Dion. T. 448.1926 ~ Sext. Emp. Math. 1.60f.). See also
Theodoridis [2009] 397 n. 14.
163 See Lentzs apparatus to Hdn. GG 3.2. 881.22 and the addendum of E. Plew in GG 3.2 1259.
164 See Wendel [1939c] 2151; Pfeiffer [1949] 273; Schneider [1999] 253 with n. 140; Theodoridis
[2009] 397 with n. 13.
165 See also Egenolff [1888] 6. Herodians Orthography was not alphabetically arranged, like
the works of Timotheus of Gaza, Orus and, later, Theognostus (see Alpers [2004] 819,
2950 with bibl.).
976 Valente
to Orus revision (except for the alphabetical arrangement) with any certainty,
it provides a good example of an ancient orthographic investigation:166
[] [ ],
[. ] (fr. 22 Berndt) (fr. 619 Theodoridis)
[ ] [ , (Il. 18.446)]
[ ] [ (Od.
23.331). ] [] [ ]
[ ] [ ]
(Il. 2.833) (Il. 6.407?) [ . ] ,
[ , ] <> <>[ ,
]. [ ] [ .
166 I print the text of Theodoridis [1976] 362 (see his apparatus for a complete survey of
orthographical loci paralleli); see Reitzenstein [1897] 307, [1901] 84f.; Alpers [2004] 47.
167 In the 3rd century, the same scholarly activity applied to a contemporary text can be found
in Porphyrys correction of autographs of Plotinus (Plot. 7
and 24 ), who wrote [...] paying no attention to the orthogra-
phy (Plot. 8 [scil. Plotinus] ... .).
Orthography 977
168 See Wendel [1942a] 14411454; Siebenborn [1976] 40; Hunger [1978] 1822; Alpers [2004].
169 See Alpers [2004] 3.
170 See e.g. Valente [2010b] on Stesichorus presence in a Byzantine orthographical canon,
possibly due to his mention in the canon of the lyric poets and in the Epistles of Phalaris.
Section 3.3
Philological and Linguistic Observations
and Theories in Interdisciplinary Context
chapter 1
1 Introduction
2 Three Rhetoricians: Demetrius, Dionysius, and Longinus
3 Linguistic Units in Rhetorical Treatises
4 Letters, Elements of Speech, and Euphony
5 The Parts of Speech and Stylistic Composition
5.1 Dionysius: The Parts of Speech and the Types of Composition
5.2 Demetrius: The Parts of Speech and the Types of Style
6 Word Order
6.1 Natural Word Order versus Hyperbaton
6.2 Demetrius on Word Order and Style
7 The Grammatical Figures of Speech
7.1 Longinus on the Grammatical Figures
7.2 The Grammatical Figures and the Sublime
7.3 Change of Number
7.4 Change of Tense
7.5 Change of Person
7.6 Sublime Linguistics between Grammar and Rhetoric
1 Introduction
There are many ways in which the ancient disciplines of grammar and rhet-
oric interact, intertwine and overlap.1 It is neither possible nor desirable to
draw clear boundaries between the two disciplines, as Quintilian already
implies when he describes the activities of the grammarians and rhetoricians
1 This contribution builds on work that I have published earlier, especially in de Jonge [2007],
[2008], and [2011]. De Jonge [2014] deals with the relationship between grammar and rhetoric
in a more condensed form. While summarizing some of the most important results of those
publications, the present article presents also new material in a comparative discussion of
three Greek rhetoricians and their views on grammar and style, viz. Demetrius, Dionysius,
and Longinus.
in Rome.2 Grammar and rhetoric were the pillars of the traditional system of
Hellenistic and Roman education.3 The rhetorician would formally start his
teaching where the grammarian had finished, but Quintilian records that
grammatical teaching in many cases anticipated the lessons of the rhetorician
(Inst. 2.1.413). On the other hand, the teacher of rhetoric devoted much atten-
tion to grammatical doctrine, thereby introducing his pupils to deeper levels of
linguistic understanding. Since both grammar and rhetoric deal with language
and linguistic communication, there are inevitably many topics where the two
disciplines meet. The most important domain where grammar and rhetoric
come together is the study of style (, , , elocutio), understood
as the expression of thoughts in words. The correct use of language (,
Latinitas)4 was regarded as the first of the so-called virtues of style (
). It is therefore only natural that rhetoricians make extensive use of lin-
guistic categories, employing, adapting and elaborating the theories that were
developed not only by grammarians, but also by philosophers and theorists
of music: all these disciplines contributed their share to the body of linguistic
knowledge that we find in ancient rhetorical treatises.
A central activity in the rhetorical teaching of the Hellenistic and Imperial
periods was the close reading of poets, orators and historians of archaic and
classical Greece: the works of Homer, Lysias, Demosthenes, Thucydides, Plato
and the tragedians were the models that Greek students continuously had to
read, analyze and imitate. In the education of Roman students, Greek texts
were combined with Vergil, Cicero and other Latin classics. How could a power-
ful sentence in a speech by Demosthenes, a narrative passage from Herodotus,
or a few Homeric hexameters inspire new writing? And in what way could a
student imitate such classical examples without presenting himself as a mere
epigone of the ancient writers? It was the task of the teachers of rhetoric to
demonstrate the quality of the classical texts, and to guide their students in the
eclectic and creative imitation () of these models.5 Although the rheto-
ricians admired the stylistic models of the past, they needed to be pragmatic
as well. In many cases, teachers had to warn their students that a text was actu-
ally not the best model to be imitated, for instance if the choice of words was
archaic or the syntax too complex. Dionysius of Halicarnassus identifies such
characteristics in passages from Plato and Thucydides, which he considers less
2 See esp. Quint. Inst. 2.1.413. On Quintilians grammar (Inst. 1.48), see Ax [2011].
3 See Marrou [19656], Bonner [1977], Morgan [1998].
4 See Pagani in this volume.
5 On the concept of in ancient rhetoric, see Russell [19952] 99113. Dionysius of
Halicarnassus, On Imitation survives in fragments and an epitome: see Aujac vol. 5 [1992].
Grammatical Theory and Rhetorical Teaching 983
6 For Dionysius criticism, see esp. Dion. Hal. Ep. Amm. 2, with de Jonge [2011].
7 Rhetores Graeci: ed. Walz [18321836] and Spengel [18531856]. For the system of classical
rhetoric, Lausberg [20084] is indispensable. Kennedy [1994] and Pernot [2000] are useful sur-
veys of the history of rhetoric in antiquity.
8 In this chapter Demetrius, Dionysius, and Longinus are cited according to the editions of
Innes [1995], Aujac [19781992], and Russell [1964]. Translations are adapted from Innes
[1995], Usher [19741985], and Fyfe-Russell [1995].
984 de Jonge
rhetorical theory, but they are also properly characterized as treatises of liter-
ary criticism, for they combine prescriptive instruction with the analysis and
evaluation of passages from classical Greek literature. Their thorough inter-
est in stylistic expression stimulates these critics to reflect on grammatical
categories and to formulate their views on the sounds of human speech, the
use of connectives, articles, and other parts of speech, word order and gram-
matical figures of speech. Before we examine their rhetorical applications of
these linguistic categories, the three main characters of this contribution will
be briefly introduced.
Demetrius is the conventional name of the author who wrote the treatise On
Style ( ).9 Although various dates between the third century BC
and the second century AD have been suggested, it seems most plausible that
the work was written in the second or early first century BC.10 The doctrine
of different styles or types of style was a very productive branch of ancient
stylistic theory. A basic dichotomy was recognized as early as Aristophanes
Frogs (405 BC), which portrays Aeschylus and Euripides as representing the
grand style and the plain style respectively. From the first century BC, rhetori-
cians generally employed a system of three styles (plain, middle, and grand or
elevated).11 Demetrius however presents a system of four styles, which most
probably belongs to an earlier period in the tradition of stylistic theory. His
four types of style ( ) are the grand (), the ele-
gant (), the plain () and the forceful (), each of which is
treated under three headings: content, diction, and composition (,
, ). It is in the discussion of composition () that Demetrius
includes grammatical observations on syntax, word order, and the use of the
parts of speech. As we will see below, this rhetorician has a special interest in
the category of connectives (), which can be used with various effects
in different styles. Since the date of Demetrius is unknown, it is difficult to
9 Edition and translation: Innes [1995]. Marini [2007] provides a useful commentary.
It is plausible that the author of the treatise On Style was called Demetrius: the tenth-
century manuscript P ascribes the work first mistakenly to Demetrius of Phaleron
(superscription), and later just to Demetrius (subscription): see Schenkeveld [1964]
135148, Rhys Roberts [1902] 4964, Innes [1999] 312321, and Marini [2007] 416.
10 See the overview in de Jonge [2009].
11 Rhet. Her. 4.1116 and Cic. Orat. 2021.
Grammatical Theory and Rhetorical Teaching 985
establish his place in the history of grammar and rhetoric. But he clearly stands
in the Peripatetic tradition: he was influenced not only by Aristotles Rhetoric
and Theophrastus On Style, but also by the linguistic ideas of Praxiphanes of
Mytilene, who was a student of Theophrastus.12
Dionysius of Halicarnassus was active in Rome under Augustus (end of
the first cent. BC).13 Apart from a history of early Rome, he wrote a number of
critical essays and treatises on style. His works include On the Ancient Orators
(with separate essays On Lysias, On Isocrates, On Isaeus and On Demosthenes),
On Thucydides and several letters that he addressed to intellectual friends and
colleagues. Many of these works contain grammatical observations as well
as linguistic analyses of passages from Plato, Thucydides and Demosthenes.
An important work for our purposes is On Composition (
), the only extant treatise on the arrangement of words to survive
from antiquity.14 The ancient theory of style generally distinguished between
two separate procedures, viz. the selection of words ( ) and
the combination of words ( ). On Composition deals with the
latter topic, following the organization of a systematic handbook. Starting
from a definition of composition (), the treatise discusses the activi-
ties () of composition, its two aims (attractiveness and beauty), the four
means of attaining these aims (melody, rhythm, variety and appropriateness)
and the three composition types or harmonies (). These are the aus-
tere composition ( ), smooth composition ( )
and well-blended composition ( ). The concluding discus-
sion of the work (Comp. 2526) deals with the relationship between prose
and poetry. Since composition is defined as a certain process of arranging the
parts of speech (Comp. 2.1), grammar forms the starting point for Dionysius
views on word arrangement, although he has also much to say on the musi-
cal aspects of . Dionysius makes use of linguistic doctrines from sev-
eral traditions.15 The influence of Stoic philosophy on his work is significant,
and he himself refers to Chrysippus work On the Syntax of the Parts of Speech
( , Comp. 4.20).16 He also mentions the
work of Alexandrian scholars like Aristophanes of Byzantium (Comp. 26.14),
28 For an overview on the topic see Swiggers-Wouters (section III.2) in this volume.
29 See Matthaios [2002f]; de Jonge [2008] 91104.
30 On Dionysius as historian of linguistics, see de Jonge [2008] 168183.
31 Dion. Hal. Comp. 2.13; Quintilian (Inst. 1.4.1721) presents a similar history of the theory of
the parts of speech, but there are some interesting differences between the two accounts:
de Jonge [2008] 168183.
Grammatical Theory and Rhetorical Teaching 991
Two figures in the grand style are concerned with the use of case ().
The first one is anthypallage (Eloc. 60: ), which is the substitution
of one case for another. In (Hom.
Od. 12.73: the two rocks, one of them reaches up to the wide heaven), the geni-
tive ( ) would have been usual (), but anything
usual is trivial, Demetrius asserts. The parallel discussion of substitutions of
cases ( ) in Apollonius Dyscolus Syntax casts light
on the connections between rhetorical and grammatical theory.47 The gram-
marian explains that these constructions, although they are strictly incongru-
ent, can be accepted as figures if they have the authority of ancient usage.48
A second figure related to is described as not staying in the same
case (Eloc. 65: ). Demetrius cites an exam-
ple from Thucydides (4.12.1), where the grammatical subject of the first verb,
accompanied by a participle in the nominative, becomes the subject of a
genitive absolute: ,
...(The first to step on the gangway, he
fainted, and in his falling on the oars...). Similar observations on Thucydides
syntax of cases can be found in Dionysius Second Letter to Ammaeus as well
as the Thucydides scholia.49 We have already seen above that Dionysius char-
acterizes the austere harmony as flexible (quick-changing) in its use of cases
(Comp. 22.6: ); Thucydides is an important repre-
sentative of that type of composition type.
The elegant style ( : Eloc. 128189), which covers the
charm () typically associated with Sappho, the urban wit () of
Lysias, and various other forms of elegant expression, is again treated in terms
of diction, composition and subject matter. Here Demetrius has much to say
on the use of rhythm, euphony, composition and word order (on which see
below), but technical grammar recedes into the background, only to return in
the discussion of the plain style.
The plain style ( : Eloc. 190239) takes its topics from every-
day life; it makes use of common words and a clear syntax. Clarity ( ,
Eloc. 192203) involves a number of linguistic characteristics. The use of
connectives is essential, as sentences that are unconnected and disjointed
throughout are always unclear (Eloc. 192). Asyndeton, on the other hand,
47 Apol. Dysc. Synt. 3.34, G.G. 2.2, 300.8302.2. Apollonius Dyscolus cites some examples
from Homer and Sophocles, in which a nominative is used instead of a vocative.
48 See Lallot [1997] II 173174.
49 Dion. Hal. Ep. Amm. 2.11. Cf. de Jonge [2011] 468471. Dion. Hal. Ep. Amm. 2.4 comments
on the same passage from Thucydides (4.12.1).
996 de Jonge
for which the term (dissolution) is used here, is said to fit the disjointed
style ( ), also known as the performative or acting style
(). In Menanders , , , (fr. 456 Kassel-
Austin: I conceived, I gave birth, I nurse, my dear) the emotion () is due
to the lack of connectives, as Demetrius points out.50
One might wonder how these observations on in the plain style
precisely relate to the discussion of connectives and polysyndeton in the grand
style (see above, esp. Eloc. 59: connectives give grandeur to the composition).
Apparently, connectives are important both in the grand and the plain style.
In the first instance this might seem to be a little confusing, but we should
realize that the focus in the discussion of the plain style (unlike that of the
grand style) is consistently on the contribution that can make to
clarity. For example, epanalepsis (Eloc. 196) is the resumptive repetition
of the same connective in the course of a long sentence. Demetrius cites an
example (author unknown) in which the particle (on the one hand) is
repeated for the sake of clarity. The passage cited in Eloc. 53 (Antiphon fr. 50
Blass, mentioned above under the grand style) had a similar repetition of ,
but that passage illustrated a different point: there the point was not so much
the repetition of the particle , but the lack of correspondence between the
particles and , which fits the imprecise character of the grand style.
The plain style avoids (Eloc. 198), which scholars interpret either
as the use of dependent constructions or as the use of oblique cases.51 The
example cited from Xenophon (An. 1.2.21) does not decide the matter. It starts
with ...(and that he had heard that tri-
remes were sailing etc.), whereas Demetrius own rewriting of the same pas-
sage not only begins with the nominative, but also omits , the conjunction
that introduces a dependent statement: ...(triremes
were expected etc.). The discussion of word order that immediately follows
this passage, however, suggests that the correct interpretation of is
the use of oblique cases: Demetrius points out that in narrative passages one
should start with the nominative (Eloc. 201: ) or accusa-
tive ( ), whereas other cases cause obscurity ().
The forceful style ( : Eloc. 240304) is especially associated
with the strong emotion aroused by speeches of Demosthenes. Brevity in
composition and harsh collocations of sounds are some of the characteristics
of this style. Whereas the elegant style strives for euphony, the forceful style
50 Cf. Dion. Hal. Comp. 22.6 (see above): the austere harmony is sparing of connectives
().
51 See Innes [1995] 467 and LSJ s.v. See also Marini [2007] 214215 on Eloc. 104 ( ).
Grammatical Theory and Rhetorical Teaching 997
6 Word Order
proves that nature is wrong, so that Dionysius decides to reject the rules that
he formulated at the beginning of his experiment. In the remaining part of
his treatise he adopts a more musical approach to word arrangement, based
on considerations of melodic sound, rhythm, appropriateness and variety.
Although Dionysius did presumably not copy his discussion of natural word
order (Comp. 5) from a Stoic source, as some scholars have thought in the past,
it is plausible that the experiment on natural word order was inspired by Stoic
philosophy, in particular by the doctrine of categories.
A more pragmatic account of natural word order ( ) is pre-
sented in Demetrius On Style (199201).56 This rhetorician argues that in
the plain style the topic ( ) is mentioned first, and then what it
is ( ), as for example in ...(Thuc. 1.24.1:
Epidamnos is a city...). But Demetrius, who immediately acknowledges that
the reverse order is also possible, adopts a rather flexible attitude in his discus-
sion of word order. Clear communication is a central concern in his discussion
of the plain style. This is especially obvious in his advice on the use of cases.
In narrative passages, he points out, one should begin either with the nomina-
tive or with the accusative: use of the other cases will cause some obscurity
() and torture () for the actual speaker and the listener. While
Demetrius terminology echoes that of Aristotles Rhetoric, his views on the
order of the topic and what it is can be seen to anticipate the theories of
modern scholars on topic and focus constituents in Greek word order.57
Whereas Demetrius is interested in the lucid presentation of the plain
style, the author of the treatise On the Sublime concentrates on the surprising
and unexpected use of language that elevates or overwhelms the audience.
One of the figures of speech that can contribute to this effect is hyperbaton
(Subl. 22.1), which is defined as the dislocated ordering of words or thoughts
out of the logical sequence (
).58 The term (sequence) is also used in the works of Dionysius
of Halicarnassus.59 As we have seen, the austere composition type is charac-
terized as in many cases neglecting the logical order (Comp. 22.6:
The grammatical figures of speech, which are also known as linguistic changes
or alterations (, , etc.), clearly form a bridge between the
disciplines of grammar and rhetoric.61 This group of figures is discussed in a
number of rhetorical treatises of the Imperial period, including Greek hand-
books On Figures by Caecilius, Alexander, Tiberius, Zonaeus and Phoebammon,
as well as Quintilians Institutio oratoria.62 The present discussion will focus on
the discussion of the grammatical figures in the treatise On the Sublime, which
will be compared with Caecilius of Caleacte and Quintilian. These Greek and
Roman rhetoricians were able to build on the work of Hellenistic scholars, who
had paid due attention to the concept of linguistic change in their philologi-
cal commentaries on Homer and classical literature. Aristarchus frequently
commented on the variations or substitutions that he found in Homers lan-
guage, like the change of voice: the use of active for passive forms and vice
versa. By claiming that such variations were characteristic of Homers linguis-
tic usage, Aristarchus was able to explain textual difficulties in the Iliad or to
defend his reading of the text.63
Caecilius of Caleacte (first cent. BC) may have been the first rhetorician to
offer a systematic discussion of the grammatical figures in stylistic theory.64
Caecilius On Figures ( ) has not survived, but fragments of
this influential work have been preserved in the works of later rhetoricians,
especially in Tiberius treatise On Figures in Demosthenes (probably third or
fourth cent. AD).65 According to Tiberius, Caecilius also introduces the fig-
ure of change, and says that it occurs in nouns, cases, numbers, persons and
tenses.66 Apart from a work On Figures ( ) and several pamphlets
on Atticism, Caecilius of Caleacte wrote a treatise On the Sublime ( ).
The extant work with the same title presents itself as a polemical reaction to
that earlier treatise (Subl. 1.1). It is thus plausible that Longinus knew Caecilius
theory of figures, either from his work On Figures or from the treatment of fig-
ures in his On the Sublime.67
Tiberius summary suggests that Caecilius discussed only five subtypes
of , but in later theory the list of grammatical figures was quickly
extended. Quintilian (Inst. 9.3.227) presents a total of sixteen different altera-
tions. Apart from the change of nouns (or gender: see below), case, number,
person and tense, which we find in Caecilius, Quintilian mentions several other
categories, such as the change of voice, the alteration of mood, and the confu-
sion of the word classes themselves, like the use of verbs for nouns and parti-
ciples for verbs. The terminology of these grammatical changes was rather
flexible: in the rhetorical handbooks many terms are used, including ,
, , , , , and the Latin mutatio.68
Longinus (see below) prefers the term . In early theory, the grammat-
ical figures are not explicitly distinguished from the other : Demetrius
(Eloc. 60), as we have seen above, mentions or change of case as
one of the figures used in the grand style. But as the tendency towards sys-
tematization increases, the grammatical figures acquire a fixed position and
separate status within the rhetorical system.69
Since grammatical figures were considered to be changes or deviations
from regular or natural usage, they could easily be mistaken as linguistic errors.
Quintilian (9.3.11) points out that there is a figure corresponding to every kind
of solecism.70 Normally, the use of present instead of past tense would be con-
sidered a fault, but if there was a literary precedent, for example a passage
in Vergils Aeneid, the same confusion of tenses could be regarded as a figure.
Apart from literary authority, the criteria for the distinction between figures
and mistakes are antiquity, usage and logical principle.71
,
;
But what of changes of case, tense, person, number and gender? How do
they vary and excite the expression?
Longinus here mentions five accidentia of nouns and verbs, and he then
illustrates the variations () that occur in the use of three of those
accidentia. The change of number (Subl. 2324) concerns the substitution of
plural for singular and of singular for plural. The discussion of change of tense
(Subl. 25) focuses on the use of present for past tense. The change of person
(Subl. 2627), finally, can be the use of second person for third person, the
use of first person for third person, or the turning away from one addressee to
another.
A comparison of the lists of grammatical figures in Caecilius (fr. 75 Ofenloch =
fr. 15 Augello) and Longinus shows that their treatments of this category are
closely related, despite some terminological differences. Longinus refers to
these variations as , whereas Caecilius calls the same figure
72 It is possible, however, that this is the (later) terminology of Tiberius, who is our source of
this fragment.
73 Caecilius fr. 75 Ofenloch (= fr. 15 Augello):
, . They change nouns
by adopting the feminine or neuter instead of the masculine, or by using the masculine
instead of both. The first example (Thuc. 1.6.1) concerns the expression
(all Greece) for (the Greeks), where a change of gender is combined with a
change of number (as Caecilius observes).
74 Quint. Inst. 9.3.27: Haec schemata aut his similia...et convertunt in se auditorem nec
languere patiuntur subinde aliqua notabili figura excitatum, et habent quandam ex illa vitii
similitudine gratiam, ut in cibis interim acor ipse iucundus est.
Grammatical Theory and Rhetorical Teaching 1005
change of person actively involve the audience in the narrative. Such effects
of surprise, emotion and active engagement are closely related to Longinus
concept of sublimity ().
Longinus presentation of the grammatical figures is different from that of
other rhetoricians, due to his focus on the sublime. While Caecilius (as far as
we know), Quintilian and later rhetoricians present a list of grammatical fig-
ures adding one or more literary examples for each of them, Longinus selects
only three grammatical figures for discussion (the changes of number, tense
and person), because they are especially relevant to the topic of his treatise.
It is instructive to compare Longinus views on these three categories with the
discussions of the same figures in Caecilius and Quintilian.
75 Subl. 23.2: ,
.
76 Tib. Fig. 26 reports that Caecilius of Caleacte cited the same example (Soph. OT 1403) in his
discussion of (repetition), which Tiberius himself prefers to call .
77 Here we might compare Demetrius advise (Eloc. 63) that the repetition of the same
connective in the grand style suggests infinite numbers, as in
(see above).
78 Subl. 24.2: , ,
,
.
Grammatical Theory and Rhetorical Teaching 1007
,
. ,
,
, .
.81
79 Modern scholars disagree on the precise interpretation of the historical present. While
it is a common assumption that the present tense verbs in a narrative mark the events
described as vivid, lively or dramatic, recent scholars describe this phenomenon in
terms of actuality, immediacy and the involvement of the reader. Different approaches
to the historical present in Thucydides are presented in Lallot-Rijksbaron-Jacquinod-
Buijs [2011].
80 Caecilius has a second example of the change of tense, which modern scholars would
rather consider a change of verbal aspect (Dem. 59.34): the rhetorician interestingly
claims that Demosthenes uses (present participle) instead of (perfect
participle).
81 Boter [2009] proposes to read instead of .
82 The transmitted text of Xenophon (Cyr. 7.1.37) slightly differs from Longinus citation:
1008 de Jonge
By using a present tense for a past tense, Xenophon presents past events as
happening at the present moment ( ). In other words, the
distance in time between the narrative and the moment of narration is anni-
hilated, and it is as if the reader becomes an eyewitness of the events in the
narrative. The story is no longer a (narrative) but an ,
Longinus states. Fyfe and Russell translate these words as a vivid actuality, but
the term seems to have a more specific meaning in ancient literary
criticism.83 It refers to a text (style, figure) that actively involves the audience
in the narrative.84 Where a (narrative) keeps the audience at a cer-
tain distance from the events that took place in the past, the historical present
draws the reader into the text. This idea of active involvement, which Longinus
associates with sublime moments in literature, anticipates the modern con-
cept of immersion, introduced by scholars of literary theory.85 The important
word (actively involving) in fact connects the change of tense with
the third grammatical figure that Longinus examines.
...
.
When the narrator suddenly uses a second person instead of the third, the
audience is directly drawn into the scene. In other words, both the change of
tense (the use of the historical present) and the change of person (the unex-
pected address of the second person) can dislocate the audience and involve
them in the narrative, as if they are in the middle of danger (Subl. 26.1:
).
Longinus adds some more examples of the same type.86 Having cited a pas-
sage from Herodotus (2.29), he asks: Do you see, friend (, ) how
Herodotus takes you along with him through the country and turns hearing
into sight? This passage is interesting for two reasons. Firstly, we notice that
Longinus himself applies the figure that he is discussing, by suddenly address-
ing his own reader. In this way the rhetorician involves his audience in his own
text just as Herodotus invites his reader to follow him on his travels: the reader
is drawn into the treatise, as Too rightly observes.87 Secondly, Longinus points
out that the use of the second person has the effect of turning hearing into
sight: the reader or listener becomes an eyewitness of the events in the text.
This grammatical figure is thus closely related to the concept of visualization
(), an important source of the sublime that Longinus discusses else-
where in his treatise (Subl. 15).88
There are two further types of the change of persons that Longinus presents
(Subl. 27). On the one hand, there is the use of the first person for the third
person. The narrator, who is talking about a character in the third person (
) suddenly changes into the person himself (
). Longinus first example is from the Iliad (15.346349):
,
,
.
90 See Too [1998] 194202, esp. 199: [I]t becomes impossible to distinguish between the
author and the audience. In this case, the author merges not with his reader but with a
character in his narrative.
91 Dem. 25.2728; Hom. Od. 4.681689: Penelope, speaking to the herald Medon, suddenly
addresses the suitors.
92 Caecilius adds an example from Euripides, Or. 720722.
93 Quintilian (9.3.2122) adds a couple of examples from Cicero, where the orator talks about
himself in the third person: de nobis loquimur tamquam de aliis (we speak of ourselves as
of other people).
Grammatical Theory and Rhetorical Teaching 1011
Walter Lapini
1 Language
1.1 Sophistic Methods prior to the Sophistic Movement
1.2 An Unreformed Language
1.3 Nomen et nominatum
1.4 Language as Added Value
1.5 Focus on Aristotle
1.6 Language and Hellenistic Philosophies
2 Philology
2.1 Alexandrian Scholarship as a Cultural Paradigm
2.2 Apologetic Philology: Philology and Philosophical Schools
2.3 The Formal Philology of the Epicureans
2.4 Two Philologists Greater than Their Time: Panaetius and Galen
2.5 The Aristotelian Commentaries
1 Language
* I would like to thank F. Ademollo and R. Chiaradonna, whose valuable suggestions con-
tributed greatly to this paper; naturally, final responsibility for any errors in the content lies
exclusively with me.
1 See F. Decleva Caizzi [1974], who composed a classic study on fr. 34 of Xenophanes. I am
inclined towards the pessimistic interpretation.
2 The text of the fragment is controversial (cf. Dorandi [2010a]; Mansfeld [2010]), but the
overall meaning is clear.
certainty. So fully convinced are they of their assertions that they hardly ever
find it necessary to support their opinions with arguments and to assess the
grounds in favor and against their views.3
In the iconography of the ancient scholar, intellectualbut also physical
isolation is a recurrent element. Take the case of Heraclitus: in principle, he
does not seem to be averse to engaging with the overall community, but his
aspiration is that the community should conform to his ideals. Since such an
outcome is not accomplished, he withdraws disdainfully and turns to playing
dice with some small boys (D. L. 9.3). However, isolation cannot give rise to
debate,4 and if debate does not come into being, then neither does science;
moreover, if science does not come into being, neither does a scientific lan-
guage, nor the interest in creating one. But what is a scientific language? First
and foremost, it consists of speakers willingness to agree on the meaning of
certain terms.
In a passage from Politics (1261a1521), Aristotle writes:
I refer to the ideal of the fullest possible unity of the entire state (
... ), which Socrates takes as his funda-
mental principle. Yet it is clear that if the process of unification advances
beyond a certain point (... ), the city will not be a city
at all; for a state essentially consists of a multitude of persons, and if its
unification is carried beyond a certain point, city will be reduced to fam-
ily and family to individual, for we should pronounce the family to be a
more complete unity than the city, and the single person than the family;
so that even if any lawgiver were able to unify the state, he must not do
so, for he will destroy it in the process.5
But Socrates (i.e. Plato) never uttered the statement attributed to him here.
What he asserts is that the good city must be one, but in the sense of unitary.6
Aristotle, on the other hand, insists on the meaning of one as homogeneous,
3 A quest for consensus and support was present among the archaic thinkers as well (see
Obbink [1992] 196), but it was not systematic.
4 This is Cornfords well known argument [1952]. According to a widespread tradition,
Heraclitus is said to have written his biblion in a deliberately obscure style, in order that
none but adepts should approach it, and lest familiarity should breed contempt (D. L. 9.6;
transl. Hicks [1925] II, 413).
5 Transl. Rackham [1932] 7.
6 Resp. 5.422e423a, and above all 462ab: Do we know of any greater evil for a state than the
thing that distracts it and makes it many instead of one, or a greater good than that which
binds it together and makes it one? (transl. Shorey [1930] 469).
1014 Lapini
If these are the Hesiodic lines referred to by Heraclitus,9 the polemic cannot
but be based on the twofold meaning of day, which indicates both the 24-hour
span and also the (roughly) 12 hours of daylight. Heraclitus feigns that Hesiod
is talking about astronomy and that he makes a mistake involving astronomy.
It is quite clear, however, that Hesiod is speaking as a poet, and that Heraclitus
is striking a blow at a target he himself has built.
In both examples, one from the period following the age of , and
one from an earlier period, the method is the same: singling out a word from
its context and interpreting it in a manner different from that intended by the
author. The Sophists legalized this method and turned it into an element of
the paideia of a successful man. But they by no means invented the method. It
is attested in every era of ancient thought, and bears direct responsibility for
the fact that in Greece a specifically philosophical language was late in arising,
and perhaps never came to exist. In any case, the Greek philosophers, includ-
ing the physikoi, drew considerable advantage from being able to shift tacitly
from the commonly accepted meaning of a term to its technical significance.10
We will return to this question later.
Lapini [2013] 161ff.). Be that as it may, it can also be noted that the philosopher who
criticized other authors failure to eliminate ambivalence did not himself feel equally
under the obligation to express himself without ambiguity.
11 On this passage, see now Ioli [2013] 239ff.
12 Thus e.g. Guthrie [1971] 271, wrongly in my opinion.
13 For linguistic reflections of the Sophists cf. Novokhatko in this volume.
14 See Pagani in this volume.
1016 Lapini
And upon this basis you will enquire whether knowledge and perception
are the same thing or different things. But you will not proceed as you
did just now. You will not base your argument upon the use and wont of
language; you will not follow the practice of most men, who drag words
this way and that at their pleasure ( ), so making
every imaginable difficulty for one another.18
Since Protagoras wrote works on the origins of the mankind, it was probably in
precisely this context that he discussed the nature of language, and it is quite
possible that he took up a position with regard to the nature/convention alter-
native (/).19
Several doctrines formulated by Protagoras are known. We know from
Diogenes Laertius (9.53) that he divided the parts of discourse into four,20 i.e.
entreaty (), question (), answer (), and command
().21 This subdivision evidently underlies the passage from Aristotle,
Poetics 19.1456b1519:
It would thus appear that Protagoras reproof of Homer was due to the poets
use of (Il. 1.1 ) instead of or . Additionally, Aristotle,
Rhetoric 3.5.1407b6, attests that Protagoras distinguished nouns according to
the genders: masculine, feminine, and neuter.23 A further item of informa-
tion on Protagoras gender-based subdivision is again preserved by Aristotle
in Sophistical Refutations 14.173bff., from which it emerges that Protagoras
(i.e. )
,
24 See the observation by Dorion [1995] 312313, and by Fait [2007] 168169.
25 Not by chance, Diels included ll. 658ff. of the Clouds among the Protagorean imitations
(C 3 D.-K.). It is worth recalling the examination on the Homeric that the bad boy
has to face in Aristophanes Banqueters: Pfeiffer [1968] 1415. In the competition depicted
in the Frogs it is, once again, names that constitute the topic of discussion (Dover [1994]
2437). On the general question, see Classen [1959b].
26 See Novokhatko in this volume.
27 On Cratylus, see now the monumental commentary by F. Ademollo [2011].
Philological Observations and Approaches 1019
, ,
.
I for my part think that the names also of the arts have been given them
because of their real essences; for it is absurdnay impossibleto hold
that real essences spring from names. For names are conventions, but
real essences are not conventions but the offspring of nature.28
28 Transl. Jones [1923] 193, who, however, following Gomperz, prints after .
29 On the close relationship of On the Art with Sophistic debates, see now Mann [2012] 17,
and passim, in the commentary.
30 As underlined by Momigliano [1969] 156, is a much stronger expression than ;
but it cannot be categorically ruled out that it may have been used by Democritus. See
Bertagna [2007] 395, with the important footnote 9.
1020 Lapini
to communicate, they will create names and languages in various ways that
differ from group to group and from population to population.31
It is also worth mentioning (though an in-depth analysis would be beyond
the scope of this paper) a passage from the commentary by Olympiodorus
on Platos Philebus (12c); this is a passage where Democritus is credited with
the surprising definition of the names of the gods as
(B 142 D.-K.), which one could translate as statues endowed with words or
talking images or vocal effigies of the gods.32 Strictly interpreted, the frag-
ment implies a natural correspondence between nomen and nominatum; this,
on the other hand, would be in contradiction with the idea that language is
, or (B 26), having no necessary relationship with . On the
basis of A 127 D.-K.
, Haag believed that in Democritus vision even the nomen was
constituted by atoms.33 This would imply a physical and material correspon-
dencean extremely firm bondbetween name and object. Furthermore,
Haag also maintained that the Democritean doxa could even be seen as having
influenced the section of the Cratylus that investigates the relation holding
between names and reality. But ingenious though Haags position may be, it is
evidently an over-interpretation: is more likely to have the
simpler meaning that names reflect in an immediate manner what men say, i.e.
what men think, about the gods: accordingly, names should be seen as cultural
products that have undergone a particular process of elaboration, and not at
all as a product of physis. In short, it does not seem appropriate to distance
Democritus from Hermogenes positions in order to bring him as close as pos-
sible to those of Cratylus.
31 According to Pohlenz [19552] I, 36, the first thinkers to recognize this phenomenon
were precisely the Stoics. In actual fact, however, Democritus had already suggested the
existence of a relationship between the variety of languages and the variety of .
The same phenomenon would subsequently be recognized by the author of [Aristotle]
Pr. 10.38.895a6 , (cf. Hist. an. 4.9.356b).
32 Graham [2010] I, 613.
33 Haag [1933] 22.
Philological Observations and Approaches 1021
, ,
,
,
, ,
.
It is in reality only his36 body that lives and sleeps in the city. His mind, hav-
ing come to the conclusion that all these things are of little or no account,
spurns them and pursues its winged way, as Pindar says, throughout the
universe, in the deeps below the earth ( ) and in the
heights above the heaven; geometrising upon earth, measuring its sur-
face, astronomising in the heavens ( ); tracking down by
every path the entire nature of each whole among the things that are, and
never condescending to what lies near at hand.37
The two short quotations from Pindar (which constitute the fr. 292 Maehler)
are not intended to stand in for words that are beyond the scope of normal
prose (quite the opposite: the concept is extremely simple), but rather to warn
the reader that the continuation of the sentence is not to be taken in the literal
sense. The things in the deeps below the earth and in the heights above the
34 Ademollo [2000] 72 has very aptly defined the etymological section of Cratylus as a
sort of encyclopaedia of Greek culture. And alsoI would addas a vast anthology of
argumentative models; cf. Amsler [1989] 39: Etymologia as a strategy for reading becomes
a strategy for writing other texts which encode the technical, historical, and exegetical
understandings produced by etymological explanations and interpretations. Amsler is
speaking of later authors, but the same line of reasoning is already valid for Plato. Cf.
Sluiter in this volume.
35 See e.g. Trabattoni [2002] 91.
36 I.e. of the sophos.
37 Transl. Levett [1990] 301.
1022 Lapini
38 The first time the term phylakes appears in the plural and with a technical meaning is in
Resp. 374d .
39 The division into three is developed in depth only in books VI and VII.
Philological Observations and Approaches 1023
that is bravest and most intelligent, that would be least disturbed and
altered by any external affection?40
The dianoia of the rulers cannot change because if they are true rulers, they
will ensure that their government conforms to eternal and immutable prin-
ciples. But neither can the dianoia of the soldiers change. Were it to change,
this would be as if dogs protecting a flock were to turn into wolves. Thus there
exists a virtue held in common by the highest classes, namely the virtue of
preservation (, ).
From the point of view of expression, Platos line is the same as that of the
physikoi: innovating as little as possible, accepting conventional language.41
But anthropology is often more delicate, more complex than the physike, and it
requires a richer Wortschatz. Accordingly, after , , etc., even ,
, , , , etc. take on, when necessary, a technical
value. As always, the great store of names is found in the world of the arts and
sciences: horse-riding, divination, medicine and so forth.
The osmosis between philosophical language and everyday language pro-
duces two effects: on the one hand it guarantees that philosophy will not
become isolated from common people, while on the other it provides thinkers
with the formidable resource of having at their disposal a set of double-faced
words, which can be used in the technical or non-technical sense according to
the requirements of the given case. This averts the risk that research may be
blocked by the impossibility of reaching absolute precision on certain points.
It thus becomes clear why, in Charmides 163d and in Politicus 261e Plato states
that distinctions like those of Prodicus are of no help in doing philosophy.42
Rigid pursuit of the correctness of names would soon lead to paralysis.
In 4th century BC linguistic speculation, an important role must be awarded
to Antisthenes, who was first a rhetor, and then a philosopher. Titles such as
On Expression, or Styles of Speaking and On Talk show that he devoted great
attention to theoretical questions of language. One of his works on this subject
has come down to us in sch. Od. 1 l1 (pp. 79 Pontani),43 through the media-
tion of Porphyry. The scholium bears witness to a debate, probably between
Socrates and Hippias, on the term used by Homer in the proem
of the Odyssey. Can be considered a term of appreciation? At first
glance, the answer would appear to be negative, given that it is the contrary
of , simple. Having an extensive array of ways (tropoi) of entering
into relations with others points to duplicity, a tendency towards deceitful-
ness. Besides, no other great heroAchilles, Ajax, Agamemnon etc.is ever
described in this manner.
But it is conceivablethis is the argument put forward by Antisthenes
that Odysseus was called not because , but rather because
. And here there follows a detailed episkepsis of the word tropos and its
numerous meanings. The concept of tropos, Antisthenes points out, can be
applied both to character and to ways of speaking. Thereafter, however, the
line of reasoning becomes uncertain. The assertion that the responsibility for
the tropos of the logoi lies with the would be a good conclusion if we
knew exactly what is meant by and if the correctness of the text were
not somewhat doubtful.44 But in any case the conclusion would seem to be the
following: that the polytropos is not necessarily a cheat or a traitor: he may also
be one who knows how to adapt the logoi to his addressees. Men are of many
different kinds, and therefore a , a unilateral statement, does
not always achieve its aim. And so one has to be like doctors, who adapt the
treatment to the patient. There is a significant sentence:
(sch. Od. 1 l1, p. 9
Pontani), the multiformity of the logos and its constantly changing use proves
to be the only form suited to constantly different ears. From the ethical point
of view, it is true that the authentic agathos is monotropos; but, paradoxically,
its unity of character is manifested precisely in its ability to adapt to each indi-
vidual. In short: to achieve ethical monotropy one must necessarily act through
polytropy of behavior.
43 On this very extensively debated passage see Brancacci [1990] 5860; Luzzatto [1996],
Brancacci [2002a], as well as the sensible overview by Pontani [2005b] 30 n. 29.
44 Pontani [2007] 8 wisely prints the cruces: .
45 Cf. Novokhatko in this volume.
Philological Observations and Approaches 1025
Poetics and in book III of Rhetoric, but the most extensive and in-depth inves-
tigations are found in the works of the Organon, in particular in the treatise On
Interpretation ( ). It is in the Organon that a tight-meshed bond
was established between language and logic, and between grammar and phi-
losophy, a bond that was destined to be long-lasting46with its positive and
negative aspects. The positive aspect is that logic is of aid in understanding
the imperfections of language, especially since language is not fully overlap-
ping with things; logic thus helps to identify more clearly the pitfalls of the
. Furthermore, if logoi and pragmata are not fully overlapping, it fol-
lows that casting doubt on language does not necessarily imply casting doubt
on things, e.g. science, the institutions, ethics, etc.
The negative aspect, at least from the modern point of view, is that the
language/logic association is not always temporary: it can become a cohesive
and enduring unity. In this perspective, Swiggers and Wouters went as far as
to introduce the concept of endogenesis of grammar from within a philo-
sophical theoriaalways fed by the close familiarity with the great literary
texts.47 And a third aspect should also be considered: once language becomes
an object of study, it becomes secularized, losing forever its archaic sacral com-
ponent. Aristotle states very clearly, in a celebrated passage, that utterances
( ) are the corresponding elements of thoughts (
, or ) (Arist. Int. 16a34);48 this explains why he had little sympa-
thy for the manner in which the physikoi formulated their theories. At times,
his criticism is highly trenchant: e.g. in Meteorology 2.3.356a24, ridiculous
is the term he uses in reference to the Empedoclean definition of the sea as
the sweat of earth (B 55 D.-K. ): On the poetic levelAristotle
saysEmpedocles wording may have been adequate (for metaphor is indeed
appropriate for poetry), but, as far as knowledge of nature is concerned, he did
not express himself adequately. Or consider Generation of Animals 4.8.777a7,
where Aristotle disparages Empedocles definition of milk as white pus (B 68
D.-K. ), and suggests, speculatively: As for Empedocles, either he
46 See e.g. Blank [1994] 149150; de Jonge [2008] 147ff. One of the reasons for the failure
to separate philosophy and grammar was the use of a substantially shared range of
vocabulary: , , , etc.
47 Swiggers-Wouters [2002c] 14. See also, on the grammar-logic-philosophy interaction,
Swiggers-Wouters [1996b] 124, [1997] 38ff., and [2005] 3ff., chapter Les rapports entre
grammaire et philosophie.
48 See E. Montanari [1988] 31ff.
1026 Lapini
, ,
(...).
.
nobody believes there existed pure grammarians. Not even the Alexandrians can be
classified as pure grammarians. The Alexandrians/Stoics opposition is another school-
based framework which, while of practical aid, is also weak: on these issues see the short
but highly effective overview given by Blank [1982] 45, and Montana in this volume.
55 Cf. Sedley [1973] 17ff., with the additions by Tepedino Guerra [1990].
56 For a study of the passage (not only of the part anthologized by Diels in 68 B 26), see
Ademollo [2003].
57 Plato says something similar in Cra. 422d423a: If we had no voice or tongue, and wished
to make things clear to one another, should we not try, as dumb people actually do, to
make signs with our hands and head and person generally? (...) If we wished to designate
that which is above and is light, we should, I fancy, raise our hand towards heaven in
imitation of the nature of the thing in question; but if the things to be designated were
below or heavy, we should extend our hands towards the ground; and if we wished to
mention a galloping horse or any other animal, we should, of course, make our bodily
attitudes as much like theirs as possible (transl. Fowler [1939] 133).
1028 Lapini
First of all, Herodotus, we must grasp the ideas attached to words (...).
For this purpose it is essential that the first mental image associated
with each word should be regarded, and that there should be no need of
explanation.58
truth and error, aversion for the art of the possible. They by no means elimi-
natedon the contrary, they accentuatedthe overlapping between logic
and grammar.62 Their affinity with Aristotle is evident, but some differences
can also be noted. One of the main divergences lies in the Stoics belief that
language is a product of physis: Origenes, Against Celsus 1.24, raised the ques-
tion of whether, as Aristotle thinks, names were bestowed by arrangement, or
as the Stoics hold, by nature; the first words being imitations of things, agree-
ably to which the names were formed, etc.
The general doctrine can be summarized in the following manner: words
are, in origin, imitations of things, i.e. onomatopeic. Over time, onomatopeic
terms undergo alterations and distortions, sometimes to the point of a radical
change in appearance. But if one succeeds in removing the encrustations, the
original truth reappears. Whoever looks at language in this way cannot fail to
regard it as insidious and ambiguous. But also fertile: for ambiguity can cre-
ate unexpected correspondences, just as the appearance of material objects
changes with a change in vantage point. A famous case in point (see D. L. 7.62)
was that of , which, according to how it is divided up, can
mean a dancing-girl has fallen ( ) or a house has three times
fallen ( ).63 Obviously, here we are dealing simply with word-
play. But in the treatise How to Study Poetry 31e, Plutarch states that Cleanthes
divided the Homeric invocation (Zeus, lord of Dodona) in
such a way as to obtain , i.e. which provokes the exhalation
of the vapour from the earth. Plutarch adds that this case belonged to the cat-
egory of ironic interpretations; but it is an irony that reflects reality,64 and
therefore it is no longer merely a joke.
Whereas Aristotle mainly uses everyday language, the Stoics introduced
numerous innovations (cf. Cic. Acad. post. 1.41: Zenon used plurima nova
verba), not so much in order to fill the blank spaces of the anonymia, as, rather,
to Stoicize the concepts. Consider the example of . Greek is rich
in words meaning to understand, to comprehend (e.g. , ,
etc.). So why adopt and the derivatives ,
and so forth? Evidently because and belong to everyone,
We are to ignore the Cynics or such Stoics as were virtually Cynics who
scold and mock us because we consider it scandalous to use the terms for
acts which are not shameful, and yet we call by their proper names those
which are. Robbery with violence, swindling, adultery are all disgrace-
ful when performed (re turpe est), but we speak of them without inde-
cency (sed dicitur non obscene), whereas the effort of begetting children
is honourable but a filthy expression (liberis dare operam re honestum est,
nomine obscenum).67
But further insight into the Stoics attitude can be gleaned from the follow-
ing anecdote reported by Cleanthes, taken from D. L. 7.172: in response to a
young man who was determined to put him on the spot with a play on words,
Cleanthes answered: Similar words do not necessarily indicate similar situa-
tions ( ), thereby
showing he was fully aware that language is not univocal and that this non-
univocality can give rise to misunderstandings and deceit. Another passage
worthy of attention is in D. L. 7.20:
,
,
, .
The second speaker must not be heard whether the former speaker
proved his case (for then the inquiry is at an end) or did not prove it (for
that is tantamount to his not having appeared when summoned or to
having responded to the summons with mere gibberish); but either he
proved his case or he did not prove it; therefore, the second speaker must
not be heard.70
The Stoics strongly proclaimed the primacy of philosophy over grammar, and
this aspect of their doctrine gained great popularity: Philo, On Mating with the
Preliminary Studies 146 (III, 102.15 Wendland) declares that philosophy laid
the bases for the traditional sciences. Equilateral and scalene triangles were
discovered thanks to geometry. But this was not a genuine act of discover-
ing (): it was merely perfecting the discovery (),
because the fundamental concepts on which geometric figures are based,
for instance the point where there are no parts ( ), the infinite
line, the surface without the third dimension, are all things forming part of
the objects of study of philosophy. And the same is true of grammar. There
exists a basic grammar, which teaches reading and writing, and a specialist
grammar, which is concerned with explaining the works of poets and prose
writers ( and ). But when it is a question of identifying func-
tions and relations, one has to turn to philosophy. It is philosophy that clarifies
what is meant by conjunction, noun, verb, the question form or exclamations.
This primacy that the Stoics assigned to philosophy emerges from an
interesting passage of the commentary by Ammonius on Aristotles On
Interpretation 42.30ff. Busse (= Praxiphanes *34 Matelli, pp. 234237). The
question concerns whether the nominative is to be called simply onoma, as
Aristotle prescribes, or ptosis, as asserted by the Stoics and grammarians (note
the association Stoics-grammarians). The Peripatetics argue that one cannot
speak of because the nominative is not a grammatical case, i.e. it does
not descend from something else. The Stoics answer was that the nominative
does indeed fall from the we have in our soul. The nominative
falls from the that resides within us. The supporters of this theory
expressed an underlying aspiration: to find something superordered, some-
thing truly basic and primitive, for the elements of conversation. In a sense,
even modern grammar seeks the aid of . For us the nominative is a ptosis,
the most primitive among the ptoseis,71 but not the most primitive in abso-
lute terms. The line of argument embodied by , as reported
by Ammonius, could be simplified as follows: the nominative is ptosis of the
so-called stem; is the ptosis of -, which does not exist other
than as . This confirms once more that, for the Stoics, linguistics was
not an autonomous discipline, with methods of its own; instead, it formed part
of the conceptual structure of the school. The system of declensions is a hin-
drance to abstraction, to the of the word. We speakers of a modern
language, devoid of declensions or with weakened declensions, have the word
in itself, for example Zeus. The ancients had , and , all forms of
one and the same word, yet at the same time independent words. And this cir-
cumstance opens up great scope for etymology. As Chrysippus would later say,72
is because he gives life (), but he is also because
.73 Etymologies of this kind are not errors but theories, the extreme and
71 Not only for us: in 1.139, Herodotus says that all Persian names ended with the letter the
Dorians call san and the Ionians sigma. But this is not true. Herodotus did not check his
sources; however, he was so sure of his statement that he transformed the god Mit(h)ras
from masculine to feminine: 1.131 ,
, . Quite apart from the mistake, it is evident that for Herodotus
name means name in the nominative.
72 And before him Aesch. Ag. 1485ff.
73 See Stob. Ecl. 1, p. 31.11 W.
Philological Observations and Approaches 1033
2 Philology
a sensible man who, before stepping outside on a stormy night, provides him-
self with a lantern and equips it with lantern-sides as shields against the vari-
ous winds.79 Line 4 continues in the following manner: (shields) that protect
against the gusts of gale-force winds. The syntax of line 3 is difficult, and the
line has numerous obscure words. Line 4 clarifies its meaning, repeating the
same concept in a more easily understandable manner.
In addition to using their own poetry to interpret themselves, the poets also
use it to interpret other poets (almost inevitably Homer and Hesiod). Consider
another example, Empedocles B 35.6 D.-K.
, Not immediately, but coming together from different directions
at will.80 In the context, (which Wright translates as at will) is the
opposite of and therefore it has been taken in the sense of at a leisurely
pace, slowly, little by little, , probably bearing in mind Hesiod,
Works 118 | , And they themselves, willing,
mild-mannered, shared out the fruits of their labors,81 where seems like
a selfgloss.82
Looking beyond the sphere of the poets, the first name encountered is
Theagenes of Rhegium (VI century BC). He would subsequently be followed
by Hippias of Thasus, Metrodorus of Lampsacus and Stesimbrotus of Thasus;
and later Democritus, Alcidamas, Antisthenes and many others.83 But exegesis
conducted in a religious background should not be overlooked: consider for
example the interpretation of oracles, extensively represented in Herodotus,
its extension and influence being testified by Aristophanes in his parodies
(Eq. 125ff.).
Furthermore, in the context of Aristophanes, it is worth noting that precisely
his comedies offer the first example of extended exegesis.84 The reference here
is to Frogs and the famous contest between Aeschylus and Euripides (ll. 830ff.).
The accusations Aeschylus and Euripides launch against each otherarchaic
features, fixity or, alternatively, freedom, luxuriancehave aspects that verge
on caricature, almost as if the one were trying to describe the other through
the eyes of the man on the street. But the two characters also address pre-
cise criticisms to each other, of a conceptual nature, which the average spec-
tator probably would not have grasped. Stinging criticisms are voiced against
individual words, discussions focus on their semantics. And improper uses are
called errors (psogoi, hamartiai: ll. 1129ff.).
Another example to be considered is the passage from Platos Protagoras
in which Socrates analyzes a poem by Simonides (fr. 19 Page = F 260 Poltera).85
Socrates states that its one thing to be good, but quite another to become
good. He maintains that means bad. Even an insignificant is
brought into the discussion. Socrates applies hyponoia, i.e. the quest for hid-
den meanings, which is an archaizing element. Overall, it is a rather contrived,
specious analysis, designed not with the idea of facilitating comprehension of
the actual text, but for a didactic and moral purpose. However, a complex array
of instruments is present, together with the remarkable method of compari-
son, to the point that H. Baltussen went as far as to wonder whether this might
not be a quite early version of the Homerum ex Homero principle, which holds
that an author can be best explained internally, or from (ex) his or her own
words, and thus whether the passage, despite its confused structural arrange-
ment and the continuous intermingling of the words of the poet with philo-
sophical considerations,86 might not perhaps constitute a sort of prehistory
of the commentary.87
Once we reach the Derveni Papyrus, dated to earlier than 300 BC, we are
no longer in pre-history. Its text is a line by line commentary which, very
clearly, is a forerunner of the Alexandrian hypomnema.88 The unknown author
awards great importance to the linguistic element, making use of tools that
were already in use such as metaphors, similes, allegories, etymologies etc.89
But some novelties can also be observed: for instance, the interpretation of
as good and not as his own,90 an interpretation the author put forward
on the basis of Homeric usage: This isas Madeleine Henry rightly stated
philological criticism of the oldest and best sort.91
At this point, a strictly chronological overview of the subject-matter dis-
cussed here would now require focus on the Alexandrian school.92 But this
would take us beyond the bounds of our specific perspective, as the interests of
the various scholars such as Zenodotus, Aristophanes and Aristarchus did not
basically center on philosophy. Perhaps, however, the roots of this school may
have been less foreign to sphere of philosophy.
It is well known that Pfeiffer regarded the Alexandrian school as a sort of
absolute incipit of philology.93 Subsequent studies, intensified by discoveries
and connections, have underscored the momentous force of this experience.
In a series of seminal studies, F. Montanari argued that the importance of the
Alexandrian philologists did not consist merely in the results they obtained,
but alsoand above allin the fact of creating a new cultural paradigm,
describable as a strong unity between exegesis, collation and text criticism
aimed at restoring the exact wording.94 This was a radically new paradigm,
a mental horizon previously unexplored. But a question naturally arises: did
this great innovation spring from nowhere, or was it a fire that blazed up from
sparks that had long been glowing? Certainly, if we consider scholarship in
the narrow sense, that is to say, exclusively as the production of editions, then
Alexandrian philology has no precedents at all; but if scholarship is construed
as reflection on the text, then one has to go much further back in time, at least
to Aristotle and his school.95
Aristotle, Parts of Animals 1.5.644b22ff., asserts that while knowledge of
eternal things, such as the stars, is more pleasant (), perceptible situa-
tions, by the very fact of being closer to us, provide us with greater opportuni-
ties for acquiring knowledge. And even from this type of research, pleasure can
be derived (645a811):
For though there are animals which have no attractiveness for the senses,
yet for the eye of science, for the student who is naturally of a philosophic
spirit and can discern the causes of things, Nature which fashioned them
provides joys which cannot be measured.96
93 More than any other consideration, what seems significant to me is the sentence: So
(...) it was only after Philitas, the poet and scholar, that the true scholar came into being
(Pfeiffer [1968] 9293). Rather fascinating, but also somewhat beyond its best before
date, is this almost teleological commitment to identifying the true scholar.
94 Montanari [1999] 2930; Montanari [2000b].
95 Montanari [2012d].
96 Transl. Peck [1937] 99.
Philological Observations and Approaches 1037
towards the great, the beautiful and the important, but is non-evaluative, in the
sense that attention is denied to no-one and nothing. An overall assessment of
the influence of the Peripatos on Alexandrian scholarship97 was delineated by
N. J. Richardson in a study published in 1994. Strabo 13.1.54 (608609) relates, as
a conclusion to the story on the Aristotelian writings, that Aristotle taught the
kings of Egypt the arrangement of a library (
).98 Clearly, the teaching was not imparted directly, given
that Aristotle died long before the Ptolemies came to power in Egypt. However,
one may discern in the testament of Aristotle a hint of similarity between the
organization of the Lyceum and that of the Mouseion. The collection of books
and the corresponding system of cataloguing may have acted as a model for
the library in Alexandria; materially, the connecting link may have been the
Peripatetic Demetrius of Phalerum, who moved from Athens to Egypt.
Even clearer links between the Peripatetic school and Alexandrian philol-
ogy can be perceived in the field of antiquarian and documentary studies.
Aristotle produced erudite works on Non-Greek Customs ( ),
on the constitutions () of 158 Greek cities, on the chronographies of
winners of the Pythian and Olympic games, as well as on the victors of drama
competitions. These writings were certainly used by the Alexandrians as source
material: by Eratosthenes in his Chronographiai and in his Olympionikai,
by Callimachus in his Table and List of Dramatic Poets and in the Aitia, by
Aristophanes of Byzantium in his hypotheseis. To this should be added that
Aristotle was the author of a work which would later become known by the
name Homeric Problems ( ).99
What this suggests is that if there was a dependency between Alexandrian
scholarship and Aristotle, there is also likely to have been a relation between
Alexandrian aesthetics and Aristotelian aesthetics. Besides, the persistence
of Aristotelian elements in the tradition of Homeric studies (and this tradi-
tion has an uncontrovertible Alexandrian base) is an acknowledged fact.
See e.g. sch. Od. 69a, which indicates the recognition of Odysseus as being
a recognition with , making use of categories that were found in
the Aristotelian Poetics, in particular chapters 10, 11 and 24.100 Note, though,
that it is precisely in this regard that Pfeiffer raised objections,101 arguing that
Aristotle loves that which is complete and unitary whereas the Alexandrians
cultivate that which is small, episodic, refined. But Aristotelian admiration for
Homer does not imply that he believed it was possible for the Homeric man-
ner to be replicated in the present. As pointed out by Montanari, Aristotle and
the Peripatetics were the premise, the ferment of Alexandrian scholarship.102
Moreover, the method of posing questions and showing solutions, as docu-
mented in Poetics (see the beginning of ch. 25 ),
has been construed by some as a prototype of the Alexandrian commentary.103
Establishing whether a given person was or was not the first to have said a
certain thing gave rise, starting from the 4th century BC, to a vast protoheu-
rological production. Theophrastus composed a work On Discoveries, in two
books (D. L. 5.47); Heraclides Ponticus also was named as the author of a book
with the same title (D. L. 5.88). Although both these texts are now lost, in antiq-
uity they constituted a valuable source of information for erudites of the impe-
rial age such as Favorinus of Arleswhose Miscellaneous History (
) certainly contained a heurematic section105and even for Diogenes
Laertius, whose interest in can be vividly perceived through-
out his production.106
Another remarkable aspect is that those who hunted for
dealt not only with doxai, laws, customs and institutions, but also with quite
down-to-earth questions. Consider for instance the following passage from
Diogenes Laertius (which, however, is unfortunately far from clear) on the
vicissitudes of Heraclitus book (9.1112): no less than a veritable example of
berlieferungsgeschichte:
The story told by Ariston of Socrates, and his remarks when he came upon
the book of Heraclitus, which Euripides brought him, I have mentioned
in my Life of Socrates (2.22). However, Seleucus the grammarian says that
a certain Croton related in his book called The Diver that the said work of
Heraclitus was first brought into Greece by one Crates, who further said
it required a Delian diver not to be drowned in it.107
was that Cleanthes and Chrysippus had already drawn some citations from the
work, thereby recognizing it as authentic.
It has to be admitted, however, that misgivings as to the authorship of
works, including some of the important works, were not always unfounded.
The very structure and manner of working within the philosophical schools lay
at the root of these uncertainties. In examining the various corporaPlatonic,
Aristotelian, Hippocratic, etc.one finds both authentic and spurious works,
but also cases where it is hard to tell whether a work is authentic or spuri-
ous. Furthermore, if a work is indeed spurious, this may not necessarily be the
result of deliberate falsification. K. Gaiser has claimed that when Platos pupils
moved to settle in the Troas and took up residence under the protection of the
tyrant Hermias, a Philosophenkreis was gradually built up where research was
carried out in common, and then signed jointly by Aristotle and Theophrastus,
and presumably also by other pupils who enjoyed a certain prestige.109 And the
habit may have continued later as well, after the foundation of the Peripatos.
Aristotle and Theophrastus may have consciously sat down to divide up the
study of biology between themselves, with one devoting himself to zoology
and the other to botany.110
A few examples may offer insight into this manner of working. The last
six chapters of the Categories (1015) are classed as Aristotelian, but perhaps
they are not, as already suggested by Andronicus. It is conceivable that some-
one composed them to create a link between Categories and Topics,111 a step
Aristotle himself perhaps did not take but would have done, if he had had the
means and the time.
Aristotle wrote a treatise on physiognomy, and in the corpus Aristotelicum
there does indeed exist a monograph that goes precisely by the name of On
Physiognomy (D. L. 5.25 ). But the work preserved in the cor-
pus is not the same as that mentioned in D. L. 5.25. What we actually have is
a composite product, not attributable to Aristotle. It consists of two distinct
parts: the second part is a rough and ready composition, simplistic, clearly of
a late date. The first part, on the other hand, is written by someone who knew
how to put forward arguments and conduct a line of reasoning in Aristotelian
109 Gaiser [1985]. See in particular the chapter entitled Die Zusammenarbeit zwischen
Aristoteles und Theophrast (8789). The first fruits of this collaboration (which may
already have begun, Gaiser conjectures, within Platos Academy) are said to have been
the treatises On Fire and Meteorology IV.
110 Gaiser [1985] 89.
111 Moraux [1974] 271272.
Philological Observations and Approaches 1041
terms, and this section may genuinely reflect the true thought of Aristotle.112 It
cannot be ruled out that the author was a pupil, quite unaware that his text
could be taken as a forgery.113
Whenever a scholar found himself faced with the accusation that an impor-
tant work of his school was false, or when a member of his school was lam-
basted for producing an indecent work or making an offensive statement,
there were two possible lines of defense: blame either a disreputable saboteur
() or the scribe (). The genetic core of the Textkritik known
as Lachmannian resides precisely in this dilemma: trying to ascertain what is
authentic and to free it from the layers of false encrustations.
The school that made the greatest contribution to formal philology is
that of the Epicureans. This is somewhat ironic, given the scorn this school
displayed towards erudition.114 But two aspects inevitably induced the
Epicureans to turn to philology: firstly, it was a school unius viri, and secondly,
its message concentrated, more so than in other cases, on basic principles. For
Epicureanism, ataraxia was of paramount importance and, according to its
teachings, it consisted of a limited number of intangible certainties, which
could often be expressed in the form of maxims. Such certainties represented
the raison dtre of the school itself and were to be defended at all costs: casting
doubt on them would be a ruinous perspective.115 Unfortunately, Epicurus was
hardly a very readable writer.116 He himself was aware of this, and sought to
remedy the defect by composing epitomes in the form of letters which had
or aimed to havea more widely understandable character.117 Dissemination
of his thought in more readable forms continued after his death (Philonides,
Artemon, etc.).
But a commentary needs to be founded on certainty of the text, a certainty
that was often in abeyance. The schools had specialist libraries, which held
more than one copy of a given text, necessarily different from one another,
and it was often not a question of purely minor differences. For instance,
the Herculaneum library undoubtedly held several exemplars of Peri physeos.118
Its most ancient nucleus of works may have been brought to Herculaneum
by Philodemus, and may have consisted of exemplars from the 3rd2nd
century BC, i.e. exemplars very close to the library Epicurus bequeathed to
Hermarchus (D. L. 10.21). But there are known to have been other copies in
circulation as well.
Eudemus, who was the head of a school in Rhodes, sent a letter to
Theophrastus, in Athens, to enquire whether a passage from Theophrastus
own version of Physics was identical to the copy he himself possessed.119 Events
of this kind are likely to have been quite common, although they probably
involved rather more important questions than establishing a or a or an
. The special aspects of ancient publication should also be borne in mind.120
Once a stage A of a certain text had been reached, the author would publish
it, in other words, he would allow copies of it to be made.121 But he would then
continue to work on the text up to phase B, then C, etc.; the exemplars cop-
ied during the phases A, B, C, etc. certainly diverged from one another, and
undoubtedly became hybridized. It is this process of hybridization that was
ultimately responsible for the survival of an authors variants, which were diffi-
cult to identify yet without a doubt did exist, and they continue to existcam-
ouflaged in one way or anotherat the present time.122
PHerc. 1012. The title of the work contained in PHerc. 1012 is unknown,124 but
we do know that it was a defense of the writings of Epicurus against incon-
gruities and contradictions that some unnamed adversaries purported to have
spotted. Demetrius availed himself of two possible procedures: one was to
adduce a comparison with other Epicurean passages (the method of Epicurum
ex Epicuro , explaining Epicurus by Epicurus himself), thereby
demonstrating that his opponents were wrong, due to ignorance or malice,
while the other was to maintain that the masters sentences had been distorted
by careless scribes. In the latter case, Demetrius made extensive use of specific
philological terminology, e.g. col. 25 ll. 12125 ] []
; col. 25 ll. 34 ; col. 25 l. 7
; col. 21 ll. 45 ; col. 44
ll. 79 ; col. 50 ll. 34
etc.
Some of Demetrius discussions can be almost entirely reconstructed. Some
parts of coll. 3132 are given here below:
]| [] [] | []
|[] |5 [] |[] []
|[][ ]. [ ] | [] .
(...)
| | , |
| | .
(i.e. ) ,
,
.
hese (i.e. the Carians) were led by captains twain, Amphimachus and
NastesNastes and Amphimachus, the glorious children of Nomion.
And he came to the war all decked with gold, like a girl, fool that he
was etc.126
In antiquity, scholars wondered whether the one who went to war covered in
gold was Amphimachus, as the syntax would appear to suggest, or Nastes, as
Simonides believed (fr. 60 Page): see sch. A Il. 872a
, , the sentence
refers to Amphimachus, but according to Simonides it should be
construed as referring to Nastes.
In col. 32 there was certainly a discussion of what is now fr. 68 Usener,
taken from Epicurus On the End ( ) and transmitted by Plutarch in
the treatise That Epicurus Actually Makes a Pleasant Life Impossible 1089d in
the form:
The stable and settled condition of the flesh and the trustworthy expec-
tation of this condition contain (...) the highest and the most assured
delight for men who are able to reflect.127
| []. []|
[]|5 [|] | [] { }
|[] [] |10 [] [] |[ .
]| |. |
, |5 | ,
|. | [] []| []
|10 []|, [ ] [ ]|
[] [] []| .
]| []|
, [] []| |5 ,
| [] , | [] |, []
[] , | [] [ ] |10[][. ]
130 Likewise, in the tenth Principal Doctrine one also reads (D. L. 10.142):
, , For they would be filling themselves
full with pleasures from every source and never have pain of body or mind, which is the
evil of life (transl. Bailey [1926] 97).
131 Transl. Bailey [1926] 95.
132 Provided that [.] is to be integrated (as is indeed the preferred
reading now) rather than .
133 On the value of (ll. 34), see Roselli [1990] 128, according to whom the term does
not mean to vary in the sense of the variant, as Puglia maintains on the basis of LSJ, but
rather to be the object of discussion.
Philological Observations and Approaches 1047
|[ ][ |, ] | [
] | [
as a pretext, because naive and untrained readers could be deceived. All read-
ers should be alerted to the fact that despite possible differences in the man-
ner of setting out doctrines, the doctrines themselves remain unchanged, and
therefore there is no need to become agitated (). If the reader bases
his interpretation on the meaning, rather than simply on the words, he will not
be led astray. Thus in col. 69 Demetrius writes:
] | | |
|5 | , |
|, | |10[]
| [] .
|[] [] | []
| |10 [] [] | [] .
| [] |[] [][] |[],
|15[] [] | [], |[]
|, | [] .
The translators have seen the final words as referring to the opponents
of the Epicurean school. But in actual fact this could also be a reassurance
addressed to the followers, to our people. And the same holds true for col. 69
of PHerc. 1012, where could mean shows that those
who in this manner do so pointlessly, mistakenly.
2.4 Two Philologists Greater than Their Time: Panaetius and Galen
Among the Stoic texts that have come down to us, there is nothing compa-
rable to PHerc. 1012, but Stoicism did have among its exponents a highly com-
petent Textkritiker: namely Panaetius of Rhodes. We have information on his
linguistic-grammatical observations (fr. 155)139 and on antiquarian questions
dates to the 1st century BC.146 How these commentaries were organized is not
known, but they focused on the same interests as the other cases seen so far:
language and identification of the correct form of the text.147
But from the point of view that concerns us here, the commentaries that
were truly important, in that they founded a precise literary genre, are those on
Aristotle. They begin from the moment when the exoteric works were brought
to light, and therefore considerably later than the floruit of the Alexandrian
school, of which they inherited the methods and goals.
The story of the reappearance of Aristotles exoteric books,148 recounted
by Strabo,149 partly has the air of a fairytale, but is partly realistic. They are
said to have disappeared from the age of Theophrastus to the age of Sulla. A
particularly controversial figure in this question is that of Apellicon of Teos,150
who, having come into possession of the library of Aristotle and Theophrastus,
worked on some of the books that had deteriorated, supplementing them
wrongly and editing them with a great quantity of blunders (
, , Strab. 13.1.54).
Andronicus, the eleventh successor of Aristotle and first commentator of
Aristotles recovered works, is a key figure, if for no other reason than that he
was the one who materially gave the corpus Aristotelicum roughly the struc-
ture it still has now.151 His activity was not merely redactional and ecdotic but
also philological stricto sensu. Among the works he declared to be inauthentic,
mention should be made of On Interpretation (see Alexander of Aphrodisias,
On Aristotles Prior Analytics 160.32161.1 Wallies) and the final part of
Categories (see Simplicius, On Aristotles Categories 159.3132 Kalbfleisch).152
His reason for issuing this judgmentwhich was also a typical manner of
on the Silloi. It is probably from this work that Diogenes Laertius took the citations from
Timon: see Mejer [1978] 30 n. 61.
146 Sedley [1997] 112. But it is possible that the range of time was more extensive: 1st century
BC1st century AD (Bastianini-Sedley [1995] 515).
147 Sedley [1997] 116, 124.
148 Cf. Montana in this volume.
149 Cf. also Plut. Sull. 26. For a recent overview on this issue, see Primavesi [2007].
150 Cf. e.g. Moraux [1973] 3031.
151 Moraux [1986] 131.
152 What Andronicus declared to be spurious has nevertheless come down to us. This does not
demonstrate that there were rival editions, nor that Andronicus judgment was ignored,
but rather that the (exceptionally important) principle adopted by the Alexandrians
had already gained widespread acceptance, namely, the principle that a judgment of
inauthenticity did not imply the consequent physical suppression of what was held to be
inauthentic.
1052 Lapini
Erudite activity did not involve only one genre. A distinction should be made
between the epitome, the thematic essay on specific points, paraphrase, and
the running commentary, though genre distinctions are naturally valid only up
to a certain point155 (in effect it would be helpful to make use of the definition
proposed by F. Montanari, text on a text or text about a text).156 But let us
examine the running commentary, which was the predominant form. Its char-
acteristic feature is the chapter-by-chapter analysis, or sentence by sentence,
sometimes word by word. Strictly speaking, it does not presuppose a paral-
lel text on the facing page, because the commentary practically rewrites the
whole text from the beginning to the end. Effectively, by uniting the lemmata,
one obtains an exemplar insertable into a stemma.
Scribal errors often impair the exact correspondence between the lemma
and the commentary. For instance, the scribe may make a mistake while tran-
scribing the lemma, in which case the explanation will refer to a reading that
no longer exists. Some reader may then try to restore coherence, but there is
no guarantee that this will be done correctly: if the procedure is not carried out
properly, the reader is likely to adapt the (right) explanation to the (corrupted)
lemma, with the result of corrupting the explanation as well. Or, vice versa, an
error may occur in the explanation, resulting in an inappropriate correction
of the lemma. At other times the error may lie in the source, as documented
by the following small fragment from Sextus, Against the Mathematicians 7.111,
the single witness for the greater part of the proem of Parmenides Peri physeos.
153 Some of the causes at the root of the judgments on falsity are indicated by Mller [1969];
often these were causes we would not imagine: see Barnes [1992] 267268.
154 Transl. Blank [1996] 15.
155 For a typology of philological writings, see Dubischar in this volume.
156 Montanari [2011a] 15.
Philological Observations and Approaches 1053
of mind.162 Between the end of the 2nd and the beginning of the 1st century BC,
Pythagoreanism and Skepticism had lost much of their impetus, or were only
just beginning to experience a reawakening of interest. Stoicism, which had
come under the influence of the Roman world, had lost its sheen and liveli-
ness in the realm of theory. As far as Epicureanism is concerned, it persisted
in its vocation of faithfulness to the Masters thought. If Sedleys basic theory
is right, as I believe it is, then it is not incorrect to say that Lucretius set down
Epicurean doctrine in verse without introducing any noteworthy innovation
at all.163 During the 2nd1st centuries BC and 1st AD the tendency of philoso-
phers to look back to the past had already become established. This is the ten-
dency Seneca was challenging when he wrote philosophia philologia facta est
(Ep. Lucil. 108.23), remarking that it was by now an entrenched habit for think-
ers to draw up erudite commentaries on the authors of the past rather than
putting these authors precepts into practice.
Up to the 3rd century AD, Aristotelian commentaries were characterized
by an absolute veneration for the Master.164 The basic idea was that the truth
had already been discoveredby Aristotleand that a good follower of the
Master should bring it back into the limelight, confute misguided interpreta-
tions of his words and clarify his vision for the benefit of those who are (or
pretend to be) incapable of understanding it. And in effect, even in this first
period the commentaries were well constructed, probing into the topics in
considerable depth. So why did the production of commentaries on Aristotle
continue for centuries? One reason is that the commentaries concerned the
exoteric works, which were difficult, technical, and in need of exegesis. But
above all, the commentaries continued because from a certain moment
onwards they became a form of philosophical expression: being a philosopher
began to coincide with the art of being a commentator.165 In this period the
Peripatetics focused almost entirely on Aristotle; Adrastus of Aphrodisias is
the only Aristotelian who wrote commentaries on a figure other than Aristotle
(Plato). After Epicureanismwhich had a distinct history of its ownthe
Peripatos was the most autistic school. In addition to investigating and study-
ing the truth, scholars of the Peripatos busied themselves with cataloguing,
putting in the proper order, perfecting and justifying the work of Andronicus.166
The commentaries were founded on the idea that the whole of a text must
be examined, in all its aspects. There was a precise order of events to be fol-
lowed in the investigation: the enquiry started out from the author (the prin-
ciple of ) and then extended the range of vision to include a glance
at other authors. As far as Aristotle was concerned, as time went on it would be
above all Simplicius who would feel that it was important to quote actual pas-
sages from the Presocratic philosophers mentioned or alluded to by Aristotle.167
Of course, the commentators of Aristotle often came across variants. The
older a reading was, the more it was likely to be reliableso they reasoned,
on the assumption that the older commentators had respect for the text
whereas the newer generations had no scruples about altering it. In contrast to
Galen, the Aristotelian commentators did not specifically concern themselves
with issues of theoretical text criticism,168 but they were well aware of such
aspects; particular attention to these problems was shown by Alexander of
Aphrodisias, who cites commentaries and variants, and hypothesizes lacunae
and translocations.
During the imperial age the running commentary gradually became scler-
otized into a single fixed form, with a precise subdivision into parts. One of
the elements always present was the prolegomena, which addressed a series
of preliminary questions such as authenticity, relations with the rest of the
corpus, etc. This repetitiveness can be explained in the light of scholastic
teaching. The teacher, who has already composed a commentary on Aristotle,
utilizes his own commentary when holding lessons at school. His pupils take
notes. Over time, a pupil then himself becomes a teacher, and in order to
comment on Aristotle, he willquite naturallydraw on his old teachers
notes, integrating them with his own observations, but without taking care
to distinguish the other mans work from his own additions, since such a dis-
tinction was regarded as being of no significance from the point of view of
the user. The composite commentaries gradually acquired a typical shape,
166 On this aspect of Aristotelian exegesis belonging to the first period, see e.g. Montanari
[2006a] 1112.
167 See Baltussen [2002], esp. pp. 175 and 182.
168 With regard to Galen and his manner of dealing with Textkritik on the theoretical level,
the work of Manetti-Roselli [1994], passim, remains fundamental.
1056 Lapini
169 On this aspect, which is in any case well known, see e.g. Fazzo [2004] 56.
170 See Rashed [2007] 42. A new commentary on Categories (by Porphyry?) has been
identified in the Archimedes Palimpsest: cf. Chiaradonna-Rashed-Sedley [2013].
171 See e.g. Donini [1994] 5040.
172 Fontaine [19832] 54.
chapter 3
Mythography
Claudio Meliad
What was mythography in ancient times? The answer to this question can-
not be easily deduced from the occurrences of terms such as mythographers
and mythography (understood as a literary genre). is used for the
first time by Polybius, who in The Histories 4.40.2 classifies the mythographers
together with poets, as witnesses on whom to base knowledge of the world;
in this context the mythographers are probably to be identified with the first
logographers, who were authors of Genealogies.1 This association also occurs
in later writers and in some cases it is the poets themselves who are called
. In this sense an excerptum of Diodorus Siculus handed down
by Eusebius of Caesarea is characteristic (Praep. evang. 2.2.54 = Diod. Sic. 6.,
fr. 1): among the historians () Euhemerus of Messene is mentioned,
while poets such as Homer, Hesiod and Orpheus are considered as mythog-
raphers. As regards the latter, Philon (De spec. leg. 1.28) states in the same way
that in order to charm the readers of their works they had adapted the false-
hood to the melodies, rhythms and metres (
,
).
In a more general perspective, the mythographer () was he who
narrated a myth (), either in prose or in poetry. In this sense the term
is used by Strabo (4.1.7) with reference to Aeschylus who had been attacked by
the philosopher-scientist Posidonius because, in his Prometheus Unbound, he
had tried to explain the origins of the Stony Plain, situated between Massilia
and the mouth of the Rhodanus river, narrating that Heracles had found him-
self without arrows and Zeus, in order to help him in the fight against the
Ligurians, had made stones rain down from a cloud, which the hero of Tiryns
had then used to beat his enemies.
The first narrator of myths in prose was Hecataeus of Miletus (550480 BC),2
an author of Genealogies, which the sources also called Histories or 3
(Heroic Tale), as well as of a work of a geographical nature, the Periegesis, writ-
ten at the end of the sixth century accompanied by a sort of map, the
, in which a description of the lands known until then appeared.
His activity is usually seen from two points of view. One approach sees him
as the inventor of genealogical chronography and the rationalistic exegesis of
mythical traditions;4 the other denies him the merit of being the father of his-
torical research and considers him merely as a continuer of the Hesiodic tradi-
tion: thus in this interpretation the only novelty he introduced would be the
transformation into prose of material dealt with by poets.5 The main features
of his Genealogies can however be drawn from the opening phrase of the work
preserved for us by Ps.-Demetrius (Eloc. 12 = Hec. fr. 1 Fowler):
,
, , .
What emerges clearly from this proem is the embarrassment of the historian
in the face of the contradictory nature of innumerable and ridiculous stories.
If previously it was the authority of the poet inspired by the Muses which guar-
anteed the veracity of the song, for Hecataeus instead the only yardstick is the
authors opinion. In fact he does not limit himself to recording the traditions
he encounters but makes a krisis of his sources so as to be able to create inter-
pretations that respond to a precise criterion of likelihood. Proof of this is, for
2 Bertelli [2001].
3 From a Heroologia of Anaximander of Miletus (to be identified with the historian of the first
part of the fourth century BC mentioned by D. L. 2.2 and by Suidas s.v. ) only
one certain fragment remains: Ath. (Deipn. 11.99 p. 498a = Anax. fr. 1 Fowler) informs us that
Anaximander had used, as did Hesiod before him in the Melampodia (frr. 271272 M.-W.), the
form instead of ; the fragment is also important as it states that in the view
of Anaximander, Pterelaus was the son of Teleboas, in his turn the son of Poseidon, while
Pterelaus is usually given as the father of Teleboas. Cf. Schubart [1832], pp. 6263.
4 So for example Jacoby [1912a] 2667ff.
5 Pearson [1939] 105ff.
Mythography 1059
instance, his explanation about the real nature of Cerberus the dog of Hades
of Cape Taenarum (Paus. 3.25.4 = Hec. fr. 27 Fowler):
[...]
.
[...]. ,
, ,
.
Among the criteria to establish the veracity of a story, Hecataeus therefore paid
attention to the possible linguistic misunderstandings caused for example by
homonymy. This certainly does not make him a rationalist tout court; one has
the impression that in his Genealogies he has attempted to demythicize some
elements in the narration that provoked particular surprise and incredulity,
with the aim of strengthening the reliability of the tale. On the whole, he does
not seem to have had doubts about the traditional mythological system.
Acusilaus of Argos,6 who lived in the sixth century BC or more probably in
the first half of the fifth, also wrote Genealogies. Suidas s.v. (test. 1
Fowler) defines him as the oldest of the investigators ( )
and reports a legend according to which he wrote his work on the basis of the
text of some bronze tablets, found by Cabas, his father, digging somewhere in
his house. From Welcker7 onwards this tradition has been considered a late
invention, but, even though it was Acusilaus himself who stated this, it could
represent an interesting parallel for the use that Herodotus would make of
Theban inscriptions of the archaic period, which he had interpreted with ref-
erence to the breed of the Labdacids.
6 Mazzarino [1966] 6070; Dowden [1992] 30; Toye [1995]; Calame [2004]; Pmias [2008] 166
169; Fontana [2012].
7 Welcker [1844] 444.
1060 Meliad
During the writing of his Genealogies, in at least three books, he had proba-
bly used material found in the epic poems written before his time: he followed
the Phoronis making Phoroneus, the first man, the father of Niobe (mother of
Argos and Pelasgus) and of Sparton (father of Mykeneus), thereby foreshadow-
ing the conflict between Mykene and Argos that led to the destruction of the
former in 468 BC (fr. 24 Fowler). Because of his dependence on poetic sources,
in particular on Hesiod, he had been accused of plagiarism for having limited
himself to transforming into prose what his predecessors had expressed in
poetry (Clem. Al. Strom. 6.26.8 = test. 5 Fowler); however, this opinion is not
completely trustworthy if we consider what can be deduced from the frag-
ments of Acusilaus. In the Bibliotheca attributed to Apollodorus (2.1.1), it is
stated that according to Acusilaus (fr. 25 Fowler), Pelasgus was the son of Zeus
and Niobe, while Hesiod (fr. 160 M.-W.) had defined him as (born
from the earth). Similarly, while according to the poet of Ascra (fr. 131 M.-W.),
the daughters of Proetus had gone mad for not having accepted the Dionysiac
rites, Acusilaus had attributed the reason of their folly to the fact that they had
not honored a statue of Hera (fr. 28 Fowler).8 All this confirms what was stated
by Flavius Joseph (Apion. 16 = Acus. test. 6 Fowler):
,
.
In the Genealogies there were also alternative versions about the fleece cap-
tured by the Argonauts, which in reality was not golden but turned purple by
the sea (fr. 37 Fowler), and as regards the Trojan war, provoked by Aphrodite to
bring about the defeat of the breed of Priam and to favour the descendents of
Anchises (fr. 39 Fowler).
Probably of the same period as Acusilaus was Pherecydes of Athens.9 Despite
the condition of information about him in Suidas Lexicon, in which a certain
confusion with two other homonymous writers of Syros and of Leros occurs,
it is possible, with a high degree of certainty, to attribute to him a genealogical
work, on the basis of the evidence of Diogenes Laertius (1.119) and of Dionysius
,
. ,
,
, .
, ,
<>
, ,
. ,
.
Among the constant elements of the Pherecydean tales there emerges, accord-
ing to Dolcetti,14 the attempt to blend different traditions, harmonizing them
in order to present a consistent and exhaustive treatment, sometimes almost
rationalistic. Fowler, on the other hand, argues that Pherecydes seems to have
given his genealogies straight, without qualification, variants, or anxiety about
truthfulness.15
An author of Genealogies in three books and of a (On discov-
eries) was Simonides of Ceus, writing in the second half of the fifth century BC,
perhaps the grandson of the more famous lyrical poet. Only two fragments
remain of the first work: they deal with the two daughters of Itonus, Athena
and Iodama, who was killed by her sister (Sim. fr. 1 Fowler),16 and with the
genealogy of Ancaeus, the son of Poseidon and of Astypalaea.17
Of the same period as Thucydides, but older, was Hellanicus,18 probably a
native of Mitylene. Tradition attributes at least 23 works to him, in prose and
in verse according to Suidas s.v. , classifiable in three main groups:
19 According to Mller [2001] 250 in the Deukalioneia, Phoronis, Asopis, and Atlantis
he reduced the mass of mythological tales and genealogies to just four ancestors. He
managed to tell the stories of those four lineages in a parallel and synchronistic manner,
leading to the generation of the Trojan war, which he described in the Troika.
20 Such a large number of titles might be due to the fact that ancient authors may have
referred to the same work when citing a subtitle or title of a section. For example, Pearson
[1939] 170 suggests dividing the Phoronis into three parts (or books), and identifying the
first, concerning the descendents of Agenor and the Theban saga, with the Boiotika, the
second, connected to the race of Iasus and Heracles, with the Argolika, the third, centering
on the descendents of Pelasgus, with the Thessalika.
21 Pearson [1939] 176.
22 As Ambaglio [2005] 137 conjectures.
23 Kullmer [1902].
1064 Meliad
the race of Aeolus. The second book might have contained the narration of the
propagation of the Hellenic tribes in Asia.
The other mythographic works (the Atlantis, the Asopis and the Troika)
are connected to the Trojan war. In Book 1 of the Atlantis a Homeric scholion
(Il. 18.486) places the catalogue of the divine lovers of six of the seven daugh-
ters of Atlas and of the children of each couple: Taygete and Zeus gave birth
to Lacedaemon, Maya and Zeus to Hermes, Electra and Zeus to Dardanus,24
Alcyone and Poseidon to Hyrieus, Sterope and Ares to Oenomaus, Celaeno and
Poseidon to Lycus.25 Fr. 21 lists the names of the children of Niobe, who had
married Amphion, a descendent of Alcyone and Poseidon, and was the daugh-
ter of Tantalus and therefore sister of Pelops. The latter is the protagonist of
fr. 157, dealing with his relationship with Hippodamia, daughter of Oenomaus
(perhaps the grandson of that Oenomaus who was born from the union of
Sterope and Ares), and the curses he sent down on his children, Atreus and
Thyestes. Almost nothing can be said about Asopis. The only explicit reference
to this work is in a passage of The life of Thucydides by Marcellinus (24 = fr. 22),
from which we learn that, like Pherecydes, Hellanicus likewise considered
Miltiades as a descendent of Aeacus. The Troika were in at least two books: the
first had an exclusively genealogical character, while the second was devoted
to the story of the events of the Trojan war. The work, in which the author tried,
among other things, to clarify obscure elements in the Homeric poems, some-
times giving an interpretation of a rationalising nature, might have contained
references to the wanderings of Aeneas and Odysseus.
A genealogical treatment of the Trojan Saga is suggested by the title of the
(On the sons and grand-
sons of those who fought against Troy) of Damastes of Sigeus, pupil or master of
Hellanicus.26 Nothing remains of this work in two books, mentioned by Suidas
s.v. (Dam. test. 1 Fowler) and it can possibly be identified with the
(Genealogy of
the Greeks and of the Barbarians who fought at Troy) of Polus of Acragas, about
which Suidas s.v. (test. 1 Fowler) notes that somebody attributed the
work to Damastes ( ).
24 Fr. 23 attributes information about Dardanus and Electra to the first book of the Troika,
therefore Sturz [1826] 103 proposed considering this work as a section of the Atlantis.
Instead Pearson [1939] 181 believes that Atlantis and Asopis were parts of the Troika.
25 A similar treatment is also present in P.Oxy. 8.1084, attributed by Hunt to the Atlantis.
Contra Pearson [1939] 177178.
26 On the chronological relation between the two authors, see Gallo [2005].
Mythography 1065
,
.
Similarly, his vision of the conquest of the apples of the garden of the Hesperides
is very eloquent (fr. 14 Fowler = Io. Antioch. Archaeol. I fr. 6.2 Roberto). In fact
according to Herodorus, Heracles had to kill the snake of evil passions (
) with the club of philosophy (
), wearing a lion skin to indicate the nobility of
the spirit ( ). Moreover the apples
symbolized three virtues, abstention from anger, greed and hedonism (
, , ).
[]
[ cod.] , ,
. ,
,
.
, , ,
, , ,
. ,
.
Asclepiades in the second book [of the Tragodumena] speaks thus about
these characters: it is said that Thamyris was of an extraordinary beauty:
his right eye was blue, while the left one was black and his singing was
different from all the others. When the Muses reached Thrace, Thamyris
had asked to join them, saying that for the Thracians it was usual for one
man to couple with many women. The Muses, on hearing this, proposed
him a song contest: if they won they would do what they wanted with him,
but if Thamyris won, he could take as many women as he wanted. After
arranging this, the Muses won and blinded him.
The first work about myths on the stage is however recognizable in the
(On the Myths of Aeschylus) by Glaucus of Rhegium (second
half of the fifth century BC)39 of which only two fragments survive. In the first
(Glauc. fr. 1 Bagordo), handed down by the hypothesis of the Persae, Glaucus is
cited as a witness of the fact that the subject of the tragedy had been taken from
the Phoenissae of Phrynicus; very probably what is reported by a Euripidean
scholion (Hec. 41 = Glauc. fr. 2 Bagordo) can also be attributed to the same work;
according to the scholion, while Ibycus and Euripides had stated that Polyxena
had been killed by Neoptolemus, Glaucus writes that for the author of the
Cypria, Diomedes and Odysseus had killed her. The
(On the <Myths> of Euripides and Sophocles) by Heraclides Ponticus
(fourth century BC) must have been complementary to the text of Glaucus; this
work probably narrated the plots of the tragedies by the other two tragedians.40
39 Lanata [1963] 278279; Huxley [1968]; Bagordo [1998] 1415 and 137138; Caroli [2006] 9;
Ucciardello [2007a].
40 Cf. Hiller [1886] 428, Wehrli [1953] 123, Bagordo [1998] 3031 and Ippolito [2009]. For the
collection of (Plots of the Myths narrated
by Euripides and Sophocles), apparently attributed by Sextus Empiricus to a pupil of
Aristotle, Dicaearchus of Messene, and for all the problems concerning the identification
Mythography 1069
, , . .
,
.
of this work with the hypotheseis handed down by some papyri, see Cannat Fera [2002],
cf. also Montana and Dickey in this volume.
41 Bagordo [1998] 33 and 155156; Caroli [2006] 10.
42 Schwartz [1901]; van Looy [1970]; Bagordo [1998] 35 and 118; Caroli [2006] 11. The correct
name is Demaretes for Wendel [1931].
43 Bagordo [1998] 7071 and 168; Cameron [2004] 59; Caroli [2006] 10.
44 Welcker [1865] 70ff.; Bethe [1887]; Rusten [1982b]; Stephens [2003] 3943; Ippolito [2006a].
Lehnus [1993] had proposed to identify Dionysius Skytobrachion with one of the Dionysioi
listed among the Telchines in the Scholia Florentina to Callimachus fr. 1,1 (ll. 38) Pf. The
correctness of this hypothesis has been recently demonstrated by Bastianini [2006], who
after revising the papyrus that transmits the scholia was able to rectify the transcription
of the previous editors, reaching the conclusion that at lines 34 it is possible to read
] [], |[].
45 is certainly an error: the works of Dionysius, as can be seen from P.Hibeh 2.186
and P.Oxy. 37.2812 which transmit some fragments of them, were in prose.
46 Lehnus [1993] 2728 insightfully proposes the hypothesis that this may be Parmenon, a
native of Byzantium, but working in Alexandria, a contemporary of Callimachus. We have
an iambic fragment (fr. 2 Diehl) by this Parmenon, in which a clear relationship with the
1070 Meliad
The title of the first work can be explained in the light of a comparison with
Diodorus 3.71.34 (Dion. fr. 10 Rusten), where it can be read that Athena and
the Libyan Amazones joined Dionysus in the struggle against the Titans. The
setting of the fight must have been Libya and, although the actual title of the
work is unknown, we can imagine that it was something like Libyan Stories.
The Argonauts are known through the citations in the scholia to Apollonius
Rhodius also as (Argonautic Stories).
Another entry of Suidas may be of help in reconstructing the production of
Dionysius:
, , . ,
, , , ,
.
Excepting the Facts after Darius in five books, the Periegesis of the inhab-
ited earth and the Persian Stories, works of a late Archaic historian, native of
Miletus, the is without doubt to be attributed to Dionysius
the Cyclographer, son of Musonius, while the and the
to Skytobrachion. For the first work a correspondence can be found in
Diodorus who attributes to the Dionysius in question a text about the Trojan
War (3.66.6 ), and the are to be
identified with the of the Suidas entry on Dionysius
of Mitylene.
The scholia to Apollonius Rhodius sometimes define him as Mitylenean
and sometimes as Milesian, but a comparison with Diodorus Siculus, who is
one of the primary sources of our knowledge of this grammarian, allows both
ethnics to be referred to the same character. Mller made a suggestion that had
a certain popularity: Dionysius of Mitylene, who in his Libyan Stories seems to
have introduced false sources speaking of the and of Dionysus, and
who was accused by Artemon of Cassandreia of falsifying the Lydian Stories
(), a work transmitted under the name of Xanthus of Lydia (test. 4 R.),47
beginning of the first Iamb of Callimachus can be recognized. Cf. Gerhard [1909] 211 and
Maas [1949].
47 See Rusten [1982b] 8284.
Mythography 1071
may have invented Dionysius of Miletos as his source, to give authority to his
narratives.48 According to Rusten, Welckers thesis is more probable; in his
opinion, one of the epithets and was the result of an
error, caused perhaps by its belated insertion by a scribe.49 As the historian
Dionysius of Miletus, author of Persian Stories, who lived between the sixth
and fifth century BC, could not have had any link with the Argonauts, very
probably the correct ethnic must be Mitylenean.50
The approach adopted in the treatment of the mythological sagas, subjects
of the two works we know best, the Libyan Stories and the Argonauts, is of a
rationalistic nature, based on the ideal of the , so that mythogra-
phers attempted to explain certain extraordinary features of the myth, justify-
ing them as mere misunderstandings of far more normal facts. In this sense,
Dionysius re-elaboration of the Argonaut myth is exemplary. Tradition had it
that Aeetes, king of Colchis, had received an oracle who informed him that he
would die when some foreigners had succeeded in stealing the golden fleece
of a ram () kept in a sanctuary. According to the legend, the place was
guarded by fire-breathing bulls (), and the fleece by a never-sleeping
serpent (). Dionysius notes that (Dracon) was really the name
of the man who guarded the sacred enclosure, while the were nothing
but the guards from the Chersonesus Taurica. Moreover, the skin preserved
inside the temple had an origin different from the fantasy reconstructions of
the myth: while Phrixus was a prisoner in Colchis together with his paeda-
gogus, the king of the Scythians, brother of Aeetes, the fell in love with the
young man and received him as a gift from Aeetes. The paedagogus, whose
name was Krios (), was sacrificed to the gods and his skin was hung in the
sanctuary. Characteristic is also the role played by Heracles within Dionysius
narrative: if it is up to Iason to organize the expedition and build the ship, his
role as captain seems to be reduced in favour of the predominant presence
of the hero of Tiryns, whose responsibility for the civilization of distant and
savage peoples the mythographer strongly underlines. This probably underlies
the characterization of the inhabitants of Colchis as cruel barbarians (fr. 14 R. =
Diod. 4.40.4):
48 Mller [1848] 6. And also Schwartz [1905] 932 and Jacoby [1957] 510.
49 Welcker [1865] 80.
50 So Rusten [1982b] 7276.
1072 Meliad
The Pontos, inhabited at that time along its coasts by barbarian and
extremely savage peoples, was called Axenos (Unhospitable), because
the people of the region killed the foreigners who landed there.
As he was the object of admiration for his courage and commanding skill,
he rapidly put together a very strong army and visited all inhabited lands,
benefitting mankind.
51 This naturally meant celebrating at the same time all the Ptolemaic race who boasted
origins from Heracles, precisely through Alexander. See Meliad [2004a].
52 Heyne [1783] 980982; Meliad [2005a] with further bibliography.
53 It cannot be excluded that in 10 books is to be identified with the
in 7 books, even though the fragments do not allow for a certain conclusion.
54 Bethe [1891].
Mythography 1073
....
. ,
. ,
, .
,
57 Pfeiffer [1968] 253266; Fraser [1972, I] 471 and 538539; Habicht [1997] 119121; Pontani
[2005b] 54.
Mythography 1075
, ,
.
Several extracts have been read in the twelve books by the soph-
ist Sopater...the first, then, deals with Greek theology expressed in
the myths, and it is a collection from the third book of On the Gods of
Apollodorus. Apollodorus came from Athens, and from the point of view
of his skills, he was a philologist and man of letters. (Sopater), however,
did not only summarize the third book, but also the fourth, the fifth and
the ninth books and again, the first, the twelfth, the fifteenth and six-
teenth, as far as the twenty-fourth. And this sylloge embraced both the
invented myths about the gods, and what has been said in historical-
legendary narratives, both regarding the heroes of their tradition and the
Dioscuri, and also about what is in the Hades and other similar matters.
In his work Apollodorus seems to start from the interpretation and etymol-
ogy of the names of the gods, in order to show, by resorting to the most varied
literary testimonies, the link between their functions and the epithets.58 It is
not clear whether, in this respect, was influenced by Stoic doctrine;
even though the recourse to etymology for analyzing the names of the gods
may have been inspired by the writings of the philosophers of the Stoa, this is
not sufficient to prove that Apollodorus subscribed to those theories. In any
case it is certain that he derived the names of the divinities not from toponyms
connected with cults ( ), but
... , from the active faculties of the soul, or from
qualities of the body (FGrHist 244 F 353.11). For instance, he maintained that
the epithet of Apollon had no connection with the island of Delos, but
was due to the fact that the god made everything visible (FGrHist 244 F 354
).59 The explanation of why the sun was also
called is likewise significant: according to Apollodorus (FGrHist 244 F 95 =
Macrob. Sat. 1.17.19), this was attributable to the fact that it races round the uni-
verse ( , quod sol per orbem impetu fertur).
58 Parsons [1993] 167. Etymological explanations of the names of the Greek gods are
contained in (Introduction to
the Traditions of Greek Theology) by L. Annaeus Cornutus, who lived in the second half of
the first century A.D. and was the teacher of Persius and Lucanus.
59 A summary of the section of centred on Apollo is transmitted by the second
column (ll. 136) of P.Oxy. 37.2812, identified by the editor princeps E. Lobel with a
commentary on a tragedy. See Rusten [1982b] 30ff.
1076 Meliad
[] []
[] .
Apollodorus idea that the Meropis was a post-Homeric work is certainly cor-
rect, but even today it is debated whether it should be dated to the sixth cen-
tury BC (as seems more probable) or even to the Hellenistic period.
The fame of this work On the Gods and of its author were such that
Apollodorus was even credited with the authorship of an anonymous myth-
ological handbook entitled Bibliotheca (). While the name of the
author is destined to remain unknown, some attempts were made in the past
to understand to what period the composition dated back.66 A terminus post
quem for the dating is clearly identifiable in the citation of Castor of Rhodes
(first century BC) author of Chronica, in Bibliotheca 2.1.3; thus C. Robert
in 1873 proposed to date the work to the first half of second century AD
(M. van der Valk thought the first century AD was more likely)67 and to con-
sider it an instrument for use in schools.68 The mainly linguistic criterion used
by Carrire and Massonie,69 who believed that the handbook was addressed to
a cultivated public in the period of the second sophistic, led them to identify a
language with an imperial tone, datable between the first and the third century
AD, with numerous lexical and semantic contacts with Plutarch, Lucian and
the scholiographic tradition; accordingly, they dated the work to the age of the
Severi, between the end of the second and beginning of the third century AD,
probably around 200 AD, also keeping in mind that Philostratus in his Imagines
of 175 AD seems to know the Bibliotheca.
In his summary of the Greek myths, the author of the handbook resorts
to numerous literary sources. As well as Homer and Hesiods Theogony, he
certainly knew the Catalogue of the women, which he seems to have used as a
basis for the genealogical structure of his work.70 He knew the contents of the
66 The attribution of the work to Apollodorus seems implicitly testified in the subscriptiones
of some historiae present in Homeric scholia (ad Il. 1.195, 2.103, 1.42, 2.494), in which at the
foot of the treatment of the myths present also in the Bibliotheca we can read
, , . On the other hand, this may instead be a reference to the monumental
by Apollodorus of Athens. Cf. Diller [1935] 297301.
67 Van der Valk [1958].
68 Some scholars were convinced that it was correct to credit Apollodorus of Athens with
the Bibliotheca, which might be an epitome of his : Haeniche [1875] (according
to whom it could be Sopaters epitome) and Lehrs [1878] (who moreover hypothesizes
that the reference to Castor of Rhodes is the result of an interpolation).
69 Carrire-Massonie [1991] 912.
70 See among others Drger [1997] 11, 36107.
1078 Meliad
Epic Cycle, even though we cannot tell whether it was from a direct reading of
the various poems or thanks to mythological summaries. As regards archaic
poetry he also cites Orphic texts, the Geryoneis and the Palinodia by Stesichorus,
Telesilla. Furthermore, the mythographers of the fifth century BC are present
in the Bibliotheca (Acusilaus, Pherecydes, Herodorus and Hellanicus; knowl-
edge of Philochorus cannot be excluded) and authors less well-known to us
(Amelesagoras, Demaratus, Philocrates). Ps.-Apollodorus mentions the three
tragedians (in particular Euripides) and also Callimachus, Apollonius Rhodius,
Asclepiades and Dionysius of Mitylene.
Photius (858 A.D.) attributes the handbook to the grammarian of Athens
and summarizes its purposes perfectly (Bibl. cod. 186):
,
,
,
,
,
.
In the text of Photius an epigram follows which may have opened the
Bibliotheca originally, but its authenticity is debatable:71
, ,
,
, ,
.
Draw your knowledge of the past from me and read the ancient tales of
learned lore. Look neither at the page of Homer, nor of elegy, nor tragic
muse, nor lyric strain. Seek not the vaunted verse of the cycle; but look in
me and you will find in me all that the world contains.
72 Wagner [1891].
73 Papadopoulos-Kerameus [1891].
74 Mastronarde [1994] 3138; Lloyd-Jones [2002].
75 Welcker [1865] 91ff.
1080 Meliad
Dicturumne me putatis ea, quae uulgo nota sunt, quod Theocritum sibi
fecerit pastoralis operis auctorem, ruralis Hesiodum, et quod in ipsis
Georgicis tempestatis serenitatisque signa de Arati Phaenomenis traxerit,
uel quod euersionem Troiae cum Sinone suo et equo ligneo ceterisque omni-
bus, quae librum secundum faciunt, a Pisandro ad uerbum paene tran-
scripserit, qui inter Graecos poetas eminet opere, quod a nuptiis Iouis et
Iunonis incipiens uniuersas historias, quae mediis omnibus saeculis usque
ad aetatem ipsius Pisandri contigerunt, in unam seriem coactas redegerit.
You are perhaps thinking that I shall speak of things that are common
knowledge: for example, that in his pastoral poetry Vergil took Theocritus
for his model, and in his work on husbandry, Hesiod; and that in the
Georgics he drew on the Phaenomena of Aratus for the signs of bad and
good weather; or that he copied his account of the overthrow of Troy,
with the tales of Sinon and the wooden horse and all the rest that goes
to make up the second book of his Aeneid, almost word for word from
Peisandros, a writer eminent among the poets of Greece for a work
which, beginning with the marriage of Jupiter and Juno, has brought
within the compass of a single sequence of events all the history of the
world through the intervening ages down to its authors own day. (Transl.
by Cameron [2004] 257)
There are, as we have seen, two Greek poets called Pisander, one native of
Kamyros, the author in the archaic period of an epos on Heracles, and one from
, ,
.
, ,
. <> ,
81 Lightfoot [1999].
82 Meineke [1843] 256.
1082 Meliad
,
, , , .
Thus the official purpose of the collection was that of serving Cornelius Gallus
as a background to poetic compositions.
Some elements present in the Erotika pathemata can be classified as typi-
cally Hellenistic: loves with a tragic epilogue, morbid passions, an interest in
foundation myths. Moreover, the narratives seem to contain moralistic ele-
ments and in the main the author avoids divine intervention and recourse to
adynata (only three metamorphoseis: Daphne, Harpalyce and Byblis can be
found), but this does not imply that he was a euhemeristic mythographer. As
J. Lightfoot underlines, some plots repeat well-known myths: the ventures of
Lyrcus (Amat. narr. 1), which center on an oracle received at Didyma, seem to
resemble those of Ion and Aegeus associated with the Delphic oracle; the story
of Leucippus (Amat. narr. 5) has much in common with that of Althaemenes
and his sister; the very famous myth of Oenomaus and Hippodamia is repeated
in the narratio about Sithon and Pallene (Amat. narr. 6) and the same happens
for Leucone and Cyanippus (Amat. narr. 10) who are modelled on Cephalus
and Procris.
Even though in the dedicatory epistle Parthenius states expressly that he
took his stories from previous poems (but he probably also consulted prose
works, as can probably be suggested for the ventures of Oenone, Paris and
Corythus, which may depend on Hegesianax),83 he rarely mentions his sources
during his discussion. Speaking of Byblis (Amat. narr. 11) he quotes ten lines of
Nicaenetus, a long passage of his Apollon; he adds twenty-one lines from the
Foundation of Lesbos in the story of Peisidice (Amat. narr. 21) and three lines
from Nicander testifying a variant of the story of Corythus (Amat. narr. 34).
The possible sources of Parthenius are however explained by some brief anno-
tations, commonly known by the name of manchettes, inserted in the upper or
lower margin of the only manuscript that hands down the Erotika pathemata
and the Metamorphoses by Antoninus Liberalis (Palatinus Heidelbergensis
gr. 389ninth century AD), corresponding with most of the narrationes in
the two collections. Modern scholars mostly reject the hypothesis that these
indications date back to the authors (Sellheim for example thought so)84 and
regard it as probable that these brief notes refer to works in which the various
stories were presumably present. We do not know when the manchettes were
compiled but, judging from the authors cited, a dating for the middle of the
third century AD has been proposed, even though an earlier date cannot be
excluded. A. Cameron85 has recently brought back to favour their attribution
to the author; he argues that they must originally have been contained in a
bibliography placed at the beginning of the roll containing the work, a sort
of index in which were specified the titles of the various chapters followed by
their sources. According to this reconstruction, we could have had the begin-
ning of an imaginary roll of the Erotika pathemata:
() .
.
() . .
() . .
() .
.
() . . And so on.
It may have been a copyist who transferred the information into the margins
when he copied the two works from the roll to the codex.
The same solution is also proposed by Cameron for the manchettes which
accompany the work of Antoninus Liberalis.86 We know nothing about this
authors life: on a linguistic basis he has been tentatively placed in the sec-
ond century AD,87 while the gentilicium Antoninus would seem to lead to a
placement in the third AD.88 We have a collection of forty-one narratives of
metamorphoses handed down under his name: they are metamorphoses of
84 Sellheim [1930].
85 Cameron [2004] 106116, 321327.
86 Papathomopoulos [1968]; Celoria [1992].
87 Knaack [1890] 39; Blum [1892] 2627.
88 Bcheler ap. Oder [1886] 56 n. 1.
1084 Meliad
,
.
A little book was read, the Narratives of Konon. He dedicates this short
work to King Archelaos Philopator, and it consists of fifty narrations gath-
ered from many ancient sources. (Transl. by Brown [2002] 47).
Rhodius (1.1165 = fr. 2), by Servius (Ad Aen. 7.738 = fr. 3), and by Flavius Joseph
(Apion. 1.216 = fr. 4). The Narrationes, on the grounds of what can be deduced
from Photius summaries, were a collection of fifty tales on mythological
subjects, which often give new versions of well-known myths. In some cases
(Narr. 1 about Midas, 37 about Cadmus, 40 about Cepheus and Andromeda)
Conon produces stories supplying a rationalistic reading, in order to eliminate
the elements that violated the principle of veracity.
That the were little read is demonstrated by the total lack of men-
tion in ancient sources except for Photius, even though some papyrus frag-
ments (P.Oxy. 53.3648), published in 1984, which hand down stories 46 and 47
in a more extended version, testify to a certain circulation in Egypt in the sec-
ond century AD. The aims of the work cannot be clarified with certainty. As
regards Conons style, Photius writes:
,
,
.
1) foundation myths (): 2, 3, 4, 8, 12, 13, 14, 19, 21, 28, 29, 36, 37, 41, 46,
47, 48;
2) aetiological myths: 6, 11, 15, 17, 19, 20, 24, 30, 33, 35, 44, 45, 49;
3) love vicissitudes: tragic and unhappy loves already recounted by
Parthenius (2, 10, 23) and homosexual loves (16);
4) paradoxographical stories: 5, 22, 43;
5) paroemiographical myths: 28 and 34;
In the surviving text the sources used by the mythographer are never speci-
fied, but this does not mean that Conon had been deliberately reticent, as
suggested until now by scholars;92 on the contrary the fact that Photius says
that the (narrations) had been taken from various ancient authors
( ), means instead that the indications of the
sources were probably present in the original version of the work and that the
Patriarch had omitted them in the process of abridgement.93 This work had
led him initially to transcribe the first three stories almost completely, later
deciding to summarise further the contents. At the end of the third Narratio
he notes:
,
;
or part of a line, sometimes in ekthesis) which gives a link with the Homeric
poems; this is followed by the mythological narrative, centred on the name in
the lemma; at the end almost always a subscriptio is inserted stating the source
from which the story is taken, usually in the form , ()
or . See for example the narrative about Il. 3. 151:
, ,
, .
.
. .
similar to the crickets: Hemera fell in love with Tithonus, son of Laomedon
and brother of Priam, and with him she generated a son, Memnon. As
Tithonus had lived a long life, the goddess transformed him into a cricket.
For this reason the poet compares the chiefs [of the Trojans], his rela-
tives, to crickets. Hellanicus recounts this.
, ,
, .
That Odysseus, because he had big ears (), was first called ; but
he says also that, one rainy day, his pregnant mother, as she could not
hold out, gave birth by the side of the road (), and that is why he was
then called Odysseus.
99 Montanari [1995c].
100 Cameron [2004] 104106.
101 Tomberg [1968]; Pagani [2005a] n. 5; Pontani [2005b] 7273; Pagani [2006a]; Hose [2008].
Mythography 1089
102 The first to propose this was Hercher [18551856] 276 and 282.
103 Cf. Chatzis [1914].
chapter 4
1 Foreword
2 Historians as the Object of Philological Enquiry and Rhetorical Analysis
3 Genealogists and Geographers
4 The Great Historiography of the Classical Age
5 Hellenistic Historiography and Geography
6 Brief Afterword on Latin Historiography and Antiquarianism
1 Foreword
(
),
.
It is clear, therefore, that for legislation books of travel are useful, since
they help us to understand the laws of other nations, and for political
debates historical works. All these things, however, belong to Politics and
not to Rhetoric. (Transl. J. H. Freese)
line will hold between an alleged pre-history of philology, with ethical con-
cerns to the fore, and an age of the philologists, in which the interest shifted
wholly to the questions of editing and interpreting the texts. What had been
displayed as the criterion as early as Hdt. 2. 116. 1, also remained the
foundation for the textual criticism and exegesis of the Alexandrian scholars.
5 On the use of historians as sources of information for orators, who certainly did not enjoy
a robust foundation in philology, see Nicolai [2007]. The authors of commentaries on the
orators greatly relied upon the historians, as the commentary to Demosthenes by Didymus
shows.
6 On the problems raised by anonymous epitomes and other issues regarding fragments by the
philologists or otherwise of philological interest, see Montanari [1997o] (on the epitomes in
particular, see p. 283).
7 See Nicolai [1992] 272f., cf. also Montana and Dickey in this volume.
8 Thus Pfeiffer [1968] 349f.; on the provenance of the scholia on Thucydides, see Luschnat
[1954].
Historiography, Ethnography, Geography 1093
History also can nourish the orator with its rich, sweet sap. But we should
read it too in the knowledge that many of its excellences are to be avoided
by the orator. History is very close to the poets. In a sense it is a prose
poem, and it is written to tell a story, not to prove a point. Moreover, it
is wholly designed not for practical effect and present conflicts, but to
preserve a memory for future generations and for the glory of its authors
Interest in the exegesis of poetic texts, particularly epic poetry, originated with
the early prose authors, especially those whose subject matter was mythology
and genealogy.14 All of them, though in varying degrees, may be regarded as
exegetes of epic, which they would use for information, but would also com-
ment on and criticize, at times comparing versions of the same narrative
episode.15 The term to designate these prose writers indicates they
were authors of to be recited in public ().16 We cannot follow
Dionysius of Halicarnassus in his inclination to regard them as a homogeneous
group, mainly on stylistic grounds (Thuc. 5). There are marked differences
among them regarding purposes and methods of composition, and, not least,
the use they made of epic poetry. I will offer a limited number of examples,
mostly from Hecataeus, to illustrate different typologies of usage and exegesis
of epic.
Very little is known of Pherecydes of Syros:17 although his theogony is clearly
independent of both Homer and Hesiod, the surviving fragments are insuffi-
cient to establish whether he wrote allegorical interpretations of epic poetry,
as some have surmised.18 The words of Celsus reported in Origen C. Cels. 6. 42
(= fr. 83 Schibli) could be seen to provide an indication:
,
.
numerous and unreliable, and which he unfavourably contrasted with his own
narrative (FGrHist 1 F 1, ap. Ps.-Demetr. Eloc. 12 .
, . ,
, , Hecataeus of Miletus thus relates. I write these things
as they seem to me to be true. For the tales told by the Greeks are, as it appears
to me, many and absurd [Transl. W. Rhys Roberts]), makes it clear that he
strove for a dependable reconstruction of the heritage of Greek traditions,
emendated of the most controversial and outrageous elements. Hecataeus
never challenged the view that the gods interacted with human beings, as
Hdt. 2. 143 clearly attests, but he did want to consolidate the traditions upon
which the hegemonic claims of the Greek aristocracy were grounded.
The fragments of the Genealogies refer to the most important , the
ones most widely celebrated in epics: Danads, Deucalionides, Argonauts,
Heracleidae, Labdacids. The clearest illustration of the use of epos as a docu-
mentary source is F 19, ap. sch. Eur. Or. 872:
,
.
, , <>, [fr. 127 Merkelbach-West]
, , , .
As fame widely has it, Egypt never came to Argos. Among others there is
the account of Hecataeus, who writes: Egypt did not come to Argos, but
his sons did, whom Hesiod said to be fifty in number, but I say were not
even twenty.
Hecataeus set himself on a level with Hesiod, claiming for himself the author-
ity to establish () the number of the sons of Egypt, and to amend the
figure given by Hesiod. The scholiasts reference to relates to the
corpus of poetry (mainly epic and tragic) and mythography which was avail-
able to him. F 18, ap. sch. Ap. Rhod. 4. 259 refers to the journey of the Argonauts
along the Phasis: as far as can be established, on this point the opinions of
Hecataeus and Hesiod (fr. 241 Merkelbach-West, ap. sch. Ap. Rhod. 4. 259) must
have differed only in part, and the problem was represented by the course
of the Phasis, which Hecataeus believed to flow into the Ocean. In F 27, ap.
Paus. 3. 25. 4 (cf. sch. Antimach., ap. PCairo 65741, col. II 26ff., p. 83 Wyss)
Hecataeus examines the vulgate, of Homeric descent, regarding Cerberus, pos-
sibly following a Hesiodean version according to which the keeper of Hades
was actually a dreadful serpent.22 An interesting case, in which the epic poets
are not explicitly referred to, is F 15, ap. Ath. 2. 35 ab, with its account of the
genealogy of Oeneus () and of the ancient name for the vine (), and
in which the evocation of the must correspond to the epic
poets (cf. Hes. Op. 572 and [Sc.] 292).23
As a criterion for the study of the geographical work of Hecataeus when
direct references to epic sources are absent, it is viable to cross-reference the
disappearance of the site for which we are given a toponym with its survival
in the epics.24 Some instances are the toponyms recorded in the Catalogue
of the Ships, and the Trojan Catalogue: Kynos (F 131), Olizon (F 135), Enete,25
Alazia (tied to the Alazones: F 127), (F 239). Thus, F 308, ap.
Aristid. 36. 108, II 297 K. ( ,
,
, Canobus was the name of Menelaus
steersman, according to the account of Hecataeus logographer and to com-
munis opinio. When he died, the place we speak of was named after him)
names Canobus, steersman of the ship of Menelaus, whose name was given
to the site of his death; and when the name of Hecataeus is associated with
, this probably ought to be taken to indicate the epic tradition
(Nostoi) on the one hand, and the exegesis of the epics, and mythography on
the other.
No less than Hecataeus and the other logographers, Hellanicus of Lesbos26
found epic poetry to provide a frame of reference, as a passage in the Contra
Apionem of Flavius Josephus (1. 16 = FGrHist 4 T 18) clearly attests:
(FGrHist 2 T 6) ,
,
.
Apart from the sequence of the corrections, what stands out is that Hesiod
should be placed at the start of the chain. In actual fact, F 94, ap. sch. Eur.
Rhes. 29 ( ** , Hesiod says
that [Sarpedon] was born of Europa. Likewise Hellanicus) testifies to a con-
vergence of Hellanicus with Hesiod; but then, on the other hand, F 95, ap. sch.
Ap. Rhod. 2. 178 ( , .
[fr. 138 Merkelbach-West], , [Phineus]
is the son of Agenor, as Hellanicus records. Otherwise, as Hesiod maintains,
he was the son of Phoenix, in turn born of Agenor, and of Cassiopea) shows a
divergence. Equally, divergences from Homer are to be found in F 141 and F 144.
In F 141, ap. sch. Hom. Il. 24. 495, it is said that Priam had 56 sons and daugh-
ters, and not 50, which makes it likeky Hellanicus was also drawing on the epic
cycle, alongside the Iliad;27 the argument in F 144, ap. Strab. 10. 2. 14, instead,
largely relies on the Catalogue of the Ships (spec. Hom. Il. 2. 631ff.), although
some of its elements probably originated with Hellanicus, and filtered to Strabo
through Apollodorus of Athens. Hellanicus also studied the biographies of the
more archaic poets, and traced back the genealogies of Homer and Hesiod to
Orpheus (F 5).28
An author we frequently find in connection with Hellanicus, of whom Suda
claims he was the pupil (FGrHist 5 T 1), is Damastes of Sigeum,29 who pro-
duced, among other things, a work in two books Ancestors of Those who Fought
at Troy ( ), a Catalogue
of the Peoples and Cities ( ) and one On Poets and
Sophists ( ). No fragment is extant of the genealogi-
cal work; of the geographical tract there is only a surviving fragment on the
northernmost peoples, including the Hyperboreans (F 1, ap. Steph. Byz. s.v.
, p. 650 M.). Of the fragments bearing no title or heading, F 3, ap.
Dion. Hal. Ant. Rom. 1. 72. 2, deals with Aeneas and the founding of Rome; F 7,
ap. Plut. Vit. Cam. 19, establishes a date for the capture of Troy; F 9, ap. Strab. 1.
3. 1, the extent of the Troad; F 11, ap. Vit. Hom. Rom. p. 30. 24 Wil., records the
place of birth and genealogy of Homer.
There is not enough room to attend to other great genealogists, such as
Acusilaus of Argos and Pherecydes of Athens;30 nor to touch upon the corpora
of the sophists who were most interested in the , i.e. the most remote
history, such as Hippias of Elis, or to human progress, such as Protagoras of
Homer, since this poem states that Paris arrived at Troy with Helen on the third
day due to favourable winds and sea condition, whereas the passage from the
Iliad clearly has Paris sailing across the Mediterranean before he reaches Troy
(2. 117). This section of Homeric exegesis, the conclusion of which is explicitly
marked (2. 117 , enough, then, of
Homer and the Cyprian poems), offers the clearest illustration of Herodotus
use of the principle of (2. 116. 1 ). Herodotus wished to question
the attribution to Homer of the poems of the epic cycle, bringing to light the
inconsistencies with the Iliad and Odyssey accounts, and thus gave proof of his
deep familiarity with the entire corpus of texts attributed to Homer (2. 116. 2
, and nowhere else does he return to the story
[Transl. A. D. Godley]). The digression on Homer indicates that Herodotus
must have been writing for readers who regarded the matter worthy of inter-
est, and who must also have been able, though not uniformly, to follow his line
of argument.
In terms of authorial ascription, Herodotus (4. 32) also doubted the Homeric
attribution of the Epigons, this being a poem from the Theban cycle which was
also ascribed to one Antimachus of Teos. The passage swiftly goes over the few
mentions made of the Hyperboreans, and quotes the names of Hesiod (fr. 150
Merkelbach-West) and, precisely, of the Epigons.
The two passages cited above should confirm the view that the Iliad and
Odyssey were the only poems to be regarded as undoubtedly Homeric, and as
such were rated as more reliable informants than the remainder of the epic
cycle. What is particularly significant in this connection is the determination
of a time-line for Homer and Hesiod in Hdt. 2. 53. Herodotus established that
both poets had been active four hundred years before his time, and credits them
with the creation of a theogony for the Greeks, specifying that they bestowed
on their several gods their , and individuated their and
(2. 53. 1). Some poets who were assumed to pre-date Homer and Hesiod should,
according to Herodotus, be regarded as posterior. Herodotus ascribes to him-
self the authority with regard to all information concerning either poet (2. 53.
3 ,
, . ,
, But those poets who are said to
be older than Hesiod and Homer were, to my thinking, of later birth. The ear-
lier part of all this is what the priestesses of Dodona tell; the later, that which
concerns Hesiod and Homer, is what I myself say [Transl. A. D. Godley]).
Herodotus did therefore rely on the testimony of the poets, Homer espe-
cially, but, unlike Hecataeus and the other genealogists, did not set them as the
starting point for his investigation. He moved, rather, from his own empirical
1102 Nicolai
enquiry and, based on the data so obtained, sought confirmation in the work
of the poets, which he subjected to critical scrutiny.38 An instance is 4. 29,
where Od. 4. 85 is cited in order to corroborate, contrastively, the observation
that hornless oxen live in the cold climes:
.
(Hom.
Od. 4. 85)
,
, .
.
And in my opinion it is for this reason that the hornless kind of cattle
grow no horns in Scythia. A verse of Homer in the Odyssey attests to my
opinion: Libya, the land where lambs are born with horns on their fore-
heads, in which it is correctly observed that in hot countries the horns
grow quickly, whereas in very cold countries beasts hardly grow horns, or
not at all. [Transl. A. D. Godley]
.
,
.
In 3. 115. 2, the name of Eridanus, Greek and not barbaric, is shown to betray its
origins as the making of some poet (
, , The
very name Eridanus betrays itself as not a foreign but a Greek name, invented
by some poet [Transl. A. D. Godley]).
In 6. 52. 1 ( ., The
Lacedaemonians saybut no poet agrees etc. [Transl. A. D. Godley]), a tradi-
tional Spartan narrative is compared with its version in the poets, with which
it does not tally. Herodotus compares three distinct traditions (the Spartan,
the common Greek, and Persian) regarding the origins of Spartan royalty.39 In
recounting the common Greek tradition, Herodotus dwells upon the limita-
tions that impinge upon genealogies, remarking how he could go no further
than Perseus, son of Danae (6. 53. 1f.):
,
,
, ,
. , ,
,
.
The Lacedaemonians are the only Greeks who tell this story. But in what
I write I follow the Greek report, and hold that the Greeks correctly
recount these kings of the Dorians as far back as Perseus son of Danae
they make no mention of the godand prove these kings to be Greek; for
by that time they had come to be classified as Greeks. I said as far back
as Perseus, and I took the matter no further than that, because no one is
named as the mortal father of Perseus, as Amphitryon is named father of
Heracles. So I used correct reasoning when I said that the Greek record is
correct as far back as Perseus. [Transl. A. D. Godley]
Herodotus implicitly identified the common Greek tradition with that of the
poets, but did not ratify it as far as acknowledging that the father of Perseus
is Zeus, who was united to Danae in the guise of a rainfall of gold. The term
of comparison appears in the renowned passage from the second book of the
Histories, where Herodotus introduces Hecataeus, who recites his own geneal-
ogy before the priests of Thebes as far back as a god (2. 143. 1 ,
39 The method is employed by Herodotus in the proemium (Hdt. 1. 15), for instance, where
he discusses the abduction of women (Io, Medea, Helen), which supposedly gave rise to
hostilities between the Greeks and Barbarians.
1104 Nicolai
in the sixteenth generation), while the priests point him to the 345 statues of
their own human forebears ,
for they would not be persuaded by him that a man could be descended from
a god (2. 143. 4) (Transl. A. D. Godley). The emendation in Herodotus is thus
far more radical than Hecataeus, of whom Mazzarino remarked that he could
not cast aside, for this Egyptian encounter, his conviction that the gods had
had dealings with humans, in Greece, up to a date that would approximately
coincide with the year 1100 BC.40 The poetic tradition was thus seen as the
originator for names and myths which later were to become part of the com-
mon tradition: however, albeit implicitly, the poets were also held accountable
for the lapses in verisimilitude of the tradition.
Thucydides Archaeology (Thuc. 1. 219) is a piece of demonstrative dis-
course with the aim of establishing the greatness of the Peloponnesian War,
which Thucydides is about to relate, over all former wars.41 A framework of
socio-economic dynamics, rooted in the development of non-nomadic com-
munities, in their accumulation of wealth, and their efforts to defend it is
employed to relate the more remote history. Thucydides makes use of Homer
as witness, at times quoting him directly. In 1. 2. 2, the general frame of what
we would call pre-historic Greece is clearly moulded on the description of the
social organisation of the Cyclops in book IX of the Odyssey:
,
,
,
,
, ,
.42
There was no commerce, and they could not safely hold intercourse with
one another either by land or sea. The several tribes cultivated their own
soil just enough to obtain a maintenance from it. But they had no accumu-
lations of wealth, and did not plant the ground; for, being without walls,
they were never sure that an invader might not come and despoil them.
Living in this manner and knowing that they could anywhere obtain a
bare subsistence, they were always ready to migrate; so that they had nei-
ther great cities nor any considerable resources. [Transl. B. Jowett]
Contrast, in particular, the passage from Thucydides above with Hom. Od. 9.
105129, esp. 108, 123 (the absence of agriculture among the Cyclops), 125129
(the absence of ships and sea-faring skills among the Cyclops). Other infor-
mation from the same passage in the Odyssey was thoroughly revised in
Thucydides adaptation. For instance, the remark on the absence of assem-
blies and civic institutions (Hom. Od. 9. 112
) and of the cave-dwelling of the Cyclops translates with Thucydides
into the notion that there were no major powerful cities in pre-historic times.
A hint of elaboration upon the Homeric source is probably to be seen in the
passage of the Archaeology of Thucydides in which it is reported that the cul-
tivation of the land in more ancient times was limited to a level of subsistence
(Thuc. 1. 2. 2 ), and should also be
set in relation to Hom. Od. 9. 114f. ( / ,
, and each one is lawgiver to his children and his wives,
and they reck nothing one of another [Transl. A. T. Murray]), where it is held
that the Cyclops only look after their closest kin, with no regard for the others.
It may thus be maintained that both the treatment reserved by the Cyclops
for their guests, and their weak social bonds (Hom. Od. 9. 401412), the same
which, combined, ensured the outcome of Odysseus verbal trick, might have
contributed to the picture painted by Thucydides. Ex post confirmation may
be derived from Pl. Leg. 680b 3ff., which, within a like context, explicitly cites
Hom. Od. 9. 112115.
In Thuc. 1. 3. 3, the authority of Homer is called upon to prove the name
Hellenes is subsequent to the Trojan war;43 in 1. 5. 2, the ancient poets serve
as evidence upon the problem of piracy, and reference is clearly made to such
testimonies as, e.g., Hom. Od. 3. 71 ss., 9. 252ff.
The much-debated Thucydidean passage on the founding of cities (Thuc. 1. 7)
,
.
,
43 See Vannicelli [1989] esp. 37ff. and 45ff., according to whom Thucydides has in mind Il. 2.
681685 particularly.
1106 Nicolai
(
), .
In later times, when navigation had become general and wealth was
beginning to accumulate, cities were built upon the sea-shore and forti-
fied; peninsulas too were occupied and walled-off with a view to com-
merce and defence against the neighboring tribes. But the older towns
both in the islands and on the continent, in order to protect themselves
against the piracy which so long prevailed, were built inland; and there
they remain to this day [Transl. B. Jowett]
may perhaps be clarified in the light of its Homeric hypo-text (Il. 20. 215218):44
,
,
,
.
The position of Troy enabled the city to control a stretch of coastline (thus
Strab. 13. 1. 7, who bases himself on Homer) and the straits, whereas old
Dardania, perched upon the slopes of Ida, had no outlets for economic and
military expansion. Over the fifth and fourth centuries, the frequency of direct
or indirect references to Il. 20. 215ff. (aside from Thucydides: Hellanic. FGrHist 4
F 25a; Pl. Leg. 681e 15) suggests the Homeric source had been commented
upon prior to Thucydides, not only for its documentary evidence on antiquity,
but also because it may have entered the speculative debate on the develop-
ment of human civilization. The topos of city-foundation as a fundamental
stage towards progress appears in Protagoras (Pl. Prt. 322a 8f., which possibly
derives from the ) and Hippias (Pl. Hp. mai. 285d
6e 2). The words Plato has Hippias pronounce point to some correlation
between the genealogies of humans and of heroes, and the foundation of cit-
ies; meaning that the foundation of the oldest cities was the work of heroes,
most of whom were the heroes of epic poetry.
,
, .
.
( )
, .
.
Of the chiefs who came to Troy, he, if the witness of Homer be accepted,
brought the greatest number of ships himself, besides supplying the
Arcadians with them. In the Handing down of the Sceptre he is described
as The king of many islands, and of all Argos. But, living on the main-
land, he could not have ruled over any except the adjacent islands (which
would not be many) unless he had possessed a considerable navy. From
this expedition we must form our conjectures about the character of still
earlier times. [Transl. B. Jowett]
Even the notorious argument upon the archaeology of the future is set in
between two statements concerning the numbers of the Achaean expedition
on Troy and the reliability of the testimony of the poets (1. 10. 1 and 1. 10. 3
respectively):
,
,
.
When it is said that Mycenae was but a small place, or that any other
city which existed in those days is inconsiderable in our own, this
1108 Nicolai
argument will hardly prove that the expedition was not as great as the
poets relate and as is commonly imagined. [Transl. B. Jowett]
,
,
, ,
, ,
.
The bent of the argument in Thucydides is to prove that the scale of the Trojan
war was comparatively small, when set against the Spartan-Athenian war he
narratesalthough the former had, in fact, been the greatest of all previous
conflicts. For the same quantitative ends, Thucydides examines the figures
in the Catalogue of the Ships, estimating the number of participants in the
Achean expedition on the basis of the men embarked on the Boeotian ships
and on the ships of Philoctetes. The exegesis of the Catalogue in Thucydides is
a display of remarkable accuracy (1. 10. 4f.):
, , , ,
. .
, .
.
,
,
, .
, .
For it numbered, as he tells us, twelve hundred ships, those of the Boeotians
carrying one hundred and twenty men each, those of Philoctetes fifty;
and by these numbers he may be presumed to indicate the largest and
Historiography, Ethnography, Geography 1109
the smallest ships; else why in the catalogue is nothing said about the
size of any others? That the crews were all fighting men as well as rowers
he clearly implies when speaking of the ships of Philoctetes; for he tells
us that all the oarsmen were likewise archers. And it is not to be sup-
posed that many who were not sailors would accompany the expedition,
except the kings and principal officers; for the troops had to cross the
sea, bringing with them the materials of war, in vessels without decks,
built after the old piratical fashion. Now if we take a mean between the
crews, the invading forces will appear not to have been very numerous
when we remember that they were drawn from the whole of Hellas.
[Transl. B. Jowett]
The information regarding the problem of supplying the Achaean army with
food, and the need to cultivate the lands of Chersonesus and also practice
piracy (1. 11) is derived from the combination of the Iliad with other sources,
most likely the Trojan cycle, and provides an explanation for the ten-year dura-
tion of the war. Thucydides also mentions the Achaean wall (1. 11. 1), a greatly
debated issue in the exegeses of antiquity (and one surviving into modern
times).45 Once again the drift of the argument is to confirm the small scale of
that war, when measured both against the data, and against its representation
in the poets (1. 11. 2
, ,
, Poverty
was the real reason why the achievements of former ages were insignificant,
and why the Trojan War, the most celebrated of them all, when brought to the
test of facts, falls short of its fame and of the prevailing traditions to which the
poets have given authority [Transl. B. Jowett]). The epic cycle is also the source
for information regarding the perilous homeward voyages of the Achaeans
from Troy (1. 12. 2), whereas the calculations concerning Boeotian colonisation
may be relying on the exegetical work of the logographers, who had most likely
taken an interest in the toponym Arne (Il. 2. 507; Thuc. 1. 12. 3).46
Chapter 21 is a synthetic presentation of the methodology for the investiga-
tion in the Archaeology, and of its results: in this chapter, the current war is
by far the greatest of all former conflicts, although the available evidence will
only serve for quantitative estimates. Here too, the value of the poets as docu-
mentary sources is played down, and that of the logographers as well (1. 21. 1):
,
,
,
,
.
Yet any one who upon the grounds which I have given arrives at some
such conclusion as my own about those ancient times, would not be far
wrong. He must not be misled by the exaggerated fancies of the poets, or
by the tales of chroniclers who seek to please the ear rather than to speak
the truth. Their accounts cannot be tested by him; and most of the facts
in the lapse of ages have passed into the region of romance. At such a dis-
tance of time he must make up his mind to be satisfied with conclusions
resting upon the clearest evidence which can be had. [Transl. B. Jowett]
The fact that this formulation comes after the section bemoaning the excessive
credulity among the sayings of those who were our predecessors (
), illustrated by two mistakes in Herodotus (who is not named,
however) (1. 20), demonstrates that Thucydides is at once measuring himself
against Homer and epic poetry, on the one hand, and also, on the other, against
both Herodotus and the logographers.
Outside the Archaeology, Homer is named only twice, in 2. 41. 4 and 4. 104,
where two passages from the Hymn to Apollo are cited. In the latter instance,
Thucydides borrows from the Homeric testimony in order to demonstrate the
importance and great antiquity of the celebrations at Delos, in a pattern that
is modelled upon the Archaeology in the first book. Epic is also relied upon in
4. 24. 5, in a discussion of the strait of Messina,47 and in a further Archaeology,
namely the Sicilian Archaeology (6. 2. 1), where the poets provide the informa-
tion on the most ancient inhabitants of the island: Cyclops and Laestrygonians.
As in the first book, Thucydides draws from the poets only a modicum for his
discussion, deliberately refraining from venturing into speculation that must
have been current in his own time. Besides, the localisation of the Cyclops and
Laestrygonians in Sicily is not to be found in the Homeric epics, and is a deriva-
tion from successive exegesis, possibly starting as early as Hesiod (fr. 150. 2526
Merkelbach-West; cf. Strab. 1. 2. 14).
When comparing the attitudes of Herodotus and Thucydides towards the
poetic tradition, particularly the Homeric tradition, it is interesting to note that
the Herodotean critical approach seeks to pinpoint whether poets may serve
as informants; Thucydides, on the other hand, neatly discriminates between
recent and ancient history, and will only allow the authority of the poets in
the investigation of the latter, whilst marking the limitations of the validity
of their testimony. The Archaeology of the first book may thus be read as a
virtuoso piece of epideictic writing, proving the incomparable greatness of the
Peloponnesian war; it is also, however, a discourse on method, dealing with
the tools and sources that will allow past events to be reconstructed correctly.
The relation of Ephorus of Cyme with Homer, whom he considered a fellow
native of Cyme, must have been particularly strong. In FGrHist 70 F 1, ap. [Plut.]
Vit. Hom. 1. 2, Ephorus discusses the genealogies of Homer and Hesiod. In F 9,
ap. Harp. 244 Keaney, in the context of a statement on methodology, we are
offered an appreciation of the wealth of narrative detail that is of the essence
in epic and myth:
[...]
,
,
.
On the fact which occurred in our own time [...] we have deemed to be
wholly worthy of credit those authors who related most precisely; but
upon ancient events, those whose narratives were equally precise we
have regarded unreliable entirely, assuming it to be most contrary to rea-
son that they should remember at such a removal in time neither all of
the facts, nor most of the words.
The ancient facts of which Ephorus speaks match chronologically with the
Trojan war, and belong to that heroic age which only the words of the poets
made accessible. Just like the genealogists, Ephorus refers to the entire epic
corpus: in F 15, ap. Steph. Byz. s.v. , p. 240 M. he discusses Egymius, king
of the Dorians and the subject of a poem which in antiquity had been attrib-
uted to Hesiod, though the attribution is doubtful.
Ephorus approach to , and to more in general apparently
contradicts the time constraints determined by the upper limit of the home-
coming of the Heracleidae.48 F 11, ap. Ath. III p. 105d, takes us to the survivors
48 On the issue of delimiting the spatium historicum in Ephorus, see Parmeggiani [1999], who
excludes there are rigid boundaries and highlights the use in Ephorus of the investigation
techniques employed by Thucydides in the Archaeology. See also Breglia Pulci Doria
1112 Nicolai
in the deluge of Deucalione; F 14, ap. sch. Ap. Rhod. 1. 1168, takes us to Heracles;
F 23, ap. Sud. 1168 Adler, and F 34, ap. Ael. Theon Progymn. 2 p. 95. 27 Sp., take
us to Pirithous and Theseus; in F 24, ap. Steph. Byz. s.v. , p. 33 M. Diante
is discussed; in F 122, the Curetes. This latter fragment combines a statement
that may be compared to the Archaeology of Thucydides (1. 2. 6: whether the
Athenians are autochthones) with the documentary use of inscriptions (see
discussion below). F 147, ap. Strab. 10. 4. 8, contains a version of the myth of
Mynosses which is, let us say, more acceptableMynosses being represented
as following the example of a notably righteous man, who was also his brothers
namesake, being called Rhadamanthys. The nine years spent by Mynosses in
Zeus cave would explain Od. 19. 178. F 31ab, ap. Ael. Theon Progymn. 2 p. 95. 8
Sp. + Strab. 9. 3. 11f., offers diverging evaluations, within a limited time-frame,
of the treatment reserved by Ephorus to mythical subject matter. The rhetori-
cian Theon interprets the Ephorean version of the myths of Tityos and Python
as models of of myth (F 31a, ap. Ael. Theon Progymn. 2 p. 95 Sp.).
Strabo, instead, is keen on distinguishing the domains of history and myth
and condemns Ephorus for having betrayed his promise, and for pointlessly
humanizing mythical characters (F 31b, ap. Strab. 9. 3. 1112 ).49 Ephorus man-
ner of proceeding has a precedent at least in Hecataeus (cf. FGrHist 1 F 27).
The question that might be asked, rather, is whether, in the fourth century,
Ephorus could have been in a position to disregard myths, as Strabo would
have it. We must answer in the negative. Ephorus at least still depended on
Apollo as founder of the Pythian oracle and had no access to an abstract, tran-
scendent conception of the divine:
[2001] 154162, who documents the interest in Ephorus towards the pre-history of the
different sites.
49 For an analysis of the fragment in its context in Strabo, see Parmeggiani [2001].
Historiography, Ethnography, Geography 1113
I would not rule out that Strabo conflated the proemial remarks of Ephorus
concerning veritas and the space to be assigned to traditional narratives (along
the lines of Thuc. 1. 22. 4) with the preamble to the story narrated in book four
of his work.50 The key to the passage lies in the definition of the Delphic ora-
cle: , which is the most truthful of all. If Ephorus
had undermined the Apollonian lineage of the oracle, he would not only have
disqualified that particular myth: the framework for the founding of the cit-
ies would have caved in on itself, since most acts of foundation were, as we
know, inspired by the oracles pronouncements. The public demanded to have
the , to use the Thucydidean term, because it needed it. Moreover, we
should not disregard the renewed importance of Delphic amphictyony in the
fourth century.51 Riccardo Vattuone has noted that in F 31b Ephorus studies a
remote tradition and seeks to verify it by means of contemporary data, such as
the Pythian procession.52 Comparable use of the same method is made within
50 On this point, see Vattuone [1998] 193: La soluzione pi plausibile entro largomentazione
di Strabone che Eforo abbia biasimato i allinizio dellopera (
) in un discorso pi ampio su , e che poi sia ritornato sullargomento
allinizio del IV libro, a proposito di un tema che si prestava in maniera particolare al
rifiuto di , vale a dire il racconto delle vicende del santuario
delfico.
51 On this point, see Spreca [2007] 196, in the context of a work which emphasizes the
relationship between Athens and the civilizing impulse of the god Apollo. See also
Avagianou [1998] 136, according to whom the Ephoran Apollo functions like the
Isocratean Theseus and fulfills the panhellenic point of view of the historian, standing for
humanity and virtue.
52 Vattuone [1998] 192194. On the presumptive method of Ephorus, see Parmeggiani
[2001] 190 who suggests the definition autopsia archeologica. On the fragment, see also
Parmeggiani [1999] esp. 120f. and Breglia Pulci Doria [2001] 153f.
1114 Nicolai
the like context of Isocrates Panegyric (28), in the discussion on the gifts of
Demeter, which is fundamentally a discussion of traditional lore, of its cred-
ibility and value.
Ephorus appears to be placed firmly within a historical-geographic tradi-
tion which had begun with the early exegetes of epic and had been continued
in historical writings as well as in the of the grammarians.53
The use of poetic texts as (F 122, ap. Strab. 10. 3. 24) finds a parallel in
the Thucydidean discussion upon Delos (Thuc. 3. 104). It is a fragment we owe
to Strab. 10. 3. 24, which ends with a citation that Strabo reads as a climax of
proud assertion, sealing the successfulness of the argument:
,
,
.
His interest in the myths transmitted by the poets surfaces in F 3, ap. Ath. VI
p. 231c, where he accounts for the fame of the necklace of Eriphyle in terms of
the rarity of gold among the Greeks at that ancient time. This account is analo-
gous to the discussion in Thuc. 1. 5. 2 of the testimonies in Homer regarding
piracy. Anaximenes interest in grammar finds confirmation in the work of his
pupil Timolaus of Larissa in Macedonia, who reportedly added in his
one verse of his own for each verse in the Iliad (T 20, ap. Sud. 626 Adler). This
operation required a great knowledge of epic poetry, equal or possibly greater
to that of the later composers of centos.
Callisthenes of Olynthus, a nephew and pupil of Aristotle, also approached
the text of Homer in terms of textual criticism. Strab. 13. 1. 27 (= FGrHist 124 T
10) records that Alexander the Great, with Callisthenes and Anaxarchus, would
read and annotate with critical symbols () a copy of the Iliad,
dubbed . Strabo, with a display of great propriety in the use of
philological terminology, acknowledges that Alexander and his companions
were engaged in a practice that was not wholly dissimilar from that of the
Alexandrian philologists.56
In F 10, ap. sch. Eur. Hec. 910, from the second book of the Hellenika,
Callisthenes expresses his view on the date of the capture of Troy, developing
an argument of astronomy based on a line from the Little Iliad (9 Bernab).57
Callisthenes contrasts his own view with the opinion of
though the label is somewhat broad, and could refer to anyone researching
any field, including investigations in the past. Recourse to the natural sci-
ences in the study of epic poetry stems, on the one hand, from the scientific
mindset of the Peripatetic school while, on the other, it leads back to epic as
the starting point for the exposition and elucidation of scientific doctrines.
Commentaries on epic allowed the ancient schools to include a number of
subjects which would otherwise have been excluded, such as astronomy, geog-
raphy, the natural sciences. The Trojan war was perhaps also treated in F 1,
ap. Ath. XIII p. 560bc, from the writing On the Holy War, which possibly con-
ducted a comparison of the Third Holy War with both the First Holy War and
the Trojan war.58
56 See Nicolai [20052006] 59f., also with regard to the position of Pfeiffer, who denies the
existence of an Aristotelian edition of Homer.
57 See Prandi [1985] 61f.
58 Thus Prandi [1985] 6668; see also Nicolai [2006] 712f.
1116 Nicolai
,
,
.
Another fragment forms part of the same context (I A 12 Berger = 6 Roller, ap.
Strab. 1. 2. 12), in which it is said:
,
.
,
,
.
,
,
Is it fitting for Hesiod not to talk nonsense and to follow prevailing opin-
ions, yet for Homer to shout forth everything that comes to this untimely
tongue? (Transl. D. W. Roller)
, , .
Therefore, it appears that Homer did not speak without meaning, but
was stating a truth allegorically when he called Aeolus steward of the
winds. (Transl. E. S. Shuckburgh)
61 On the discussion surrounding this passage in the Odyssey, cf. Cratet. fr. 37 Broggiato
(with the annotations in Broggiato [2001] 200203).
62 See Nicolai [1999] especially for the critique of the speeches of Timaeus.
Historiography, Ethnography, Geography 1119
Ephorus for a mistake in stating the timeline for the kingdom of Dionysius the
Elder. Polybius remarks (ibid.):
,
.
,
.
*** ,
.
For surely no one could say that the mistake here was the authors, but it
is obviously the scribes. Either Ephorus must have surpassed Coroebus
and Margites in stupidity if he could not reckon that forty-two added to
twenty-three make sixty-five, or as nobody would believe this of Ephorus,
the mistake is evidently due to the scribe. No one, however, could approve
of Timaeus love of cavilling and fault-finding. (Transl. W. R. Paton)
In the course of the diatribe against Timaeus, Polybius stresses that Timaeus
has a reputation for historical accuracyonly to then remark on his omissions
and accuse him of wilful untruths (12. 10. 4):
***
, .
And yet Timaeus special boast, the thing in which he outvies other
authors and which is the main cause of the reputation he enjoys, is, as
I suppose we all know, his display of accuracy in the matter of dates
and public records, and the care he devotes to such matters. (Transl.
W. R. Paton)
,
, ,
.
1120 Nicolai
.
.
For this is the author who compares the dates of the ephors with those
of the kings in Lacedaemon from the earliest times, and the lists of the
Athenian archons and priestesses of Hera at Argos with those of the vic-
tors at Olympia, and who convicts cities of inaccuracy in these records,
there being a difference of three months. Yes, and it is Timaeus who dis-
covered the inscriptions at the back of buildings and list of proxeni on
the jambs of temples. We cannot then believe that he would have missed
any such thing had it existed, or omitted to mention it had he found it,
nor can we in any way excuse his mendacity. (Transl. W. R. Paton)
Beyond the interest in the different outlooks on the work of the historian that
are upheld by Timaeus and Polybius, it is worth pointing out that by the time
of Polybius, the practice of gathering documentation, especially epigraphic,
and analysing it had become established.63 The very critique of Timaeus
shows with great clarity that the works of the predecessors were carefully anat-
omised, in a search for contradictory or fallible statements. In this regard, the
historian operated much like the judicial orator, whose task it was to bring out
the non-conformities of the opponents with . Familiarising with
the work of the predecessors and comparing accounts is, according to Polybius
(12. 25e. 1), the first stage in pragmatic history. Polybius, however, deemed
scholarly knowledge to be insufficient, needing the support of first-hand mili-
tary and political experience, and calling for personal acquaintance with the
territories (12. 25g. 125i. 2). What Polybius (12. 25i. 1) does concede to Homer,
instead, is profound knowledge of lifeonce more marking his variance from
Eratosthenes on the subject of geography in Homer (see above).
Dionysius of Halicarnassus chose ancient Roman history as his object of
study, ending with the first Punic war. As a matter of course, the investigation
of his topic forced him to rely heavily on previous historiography, to compare
the several versions of obscure and remote events, and to turn occasionally to
poetic sources. Dionysius traces a brief status quaestionis, listing the histori-
ans who had covered ancient Roman history (1. 6. 1f.): the picture he gives the
reader is one of great familiarity with the Greek tradition of historiography,
63 We must at least recall the gathering of epigraphs in Polemon of Ilion (first half of the
second century BC). On the use of documents in ancient historiography, see Biraschi-
Desideri-Roda-Zecchini [2003].
Historiography, Ethnography, Geography 1121
as well as with the Latin authors who had written in Greek. I shall give one
example of the exegetical method in Dionysius from chapter fifty-three of the
first book, on the arrival of Aeneas in Italy. Dionysius uses two monuments
as ; then he justifies his digression () as necessary and as
motivated by the divergence of opinion among the . The notion that
Aeneas should have returned to Troy after leading his people into Italy derived,
in his opinion, from a misinterpretation of Il. 20.307308 (1. 53. 4):
,
. ,
.
,
.
.
.
The first country on this seaboard is the Troad, the fame of which,
although it is left in ruins and in desolation, nevertheless prompts in writ-
ers no ordinary prolixity. With this fact in view, I should ask the pardon
of my readers and appeal to them not to fasten the blame for the length
of my discussion upon me rather than upon those who strongly yearn for
knowledge of the things that are famous and ancient. And my discussion
is further prolonged by the number of the peoples who have colonized
the country, both Greeks and barbarians, and by the historians, who do
not write the same things on the same subjects, nor always clearly either;
among the first of these is Homer, who leaves us to guess about most
things. And it is necessary for me to arbitrate between his statements and
those of the others, after I shall first have described in a summary way the
nature of the region in question. (Transl. H. L. Jones)
name of Messenia ever existed, but only a Messenian people.68 The argument
closely resembles that in Thucydides 1. 3 on the subject of the designation
Hellenes. In 4. 30. 14. 31. 1 a discussion of historical geography appears, based
on references to the Iliad and to the Hymn to Demeter. Pausanias makes docu-
mentary use of epic, with a keen exegesis of geographical and genealogical
issues in particular. The example of Messenia is of interest, in that the loss of
political autonomy prevented a local historiography from being established in
Messenia prior to 370 BC: the understanding of Messenic identity therefore
had to be derived from epic poetry, for accounts of remotest history, and from
oral traditions for the Messenic wars and more recent events. In this perspec-
tive, we find epic poetry being used as a matrix for new epic compositions
(Rhianus of Bene) and as providing the grounds for erudite discussions, such
as those in Pausanias.
68 See Musti-Torelli [1991] 204f., with valuable insights on the role of Andania and the
reasons for certain decisions in Pausanias.
69 See Nicolai [2001].
70 See Della Corte [1981] 149216, in particular p. 166.
Historiography, Ethnography, Geography 1125
although Greek and Latin historiography did gradually incorporate the study of
proper documents, it has to be acknowledged that the integration of narrative
and antiquarian historiography was never fully accomplished by the ancients,
and that only with the accomplishment of that step do we have, in the defini-
tion of Arnaldo Momigliano,71 the turning point for modern historiography.
71 Momigliano [1950].
chapter 5
It has been remarked that the adoption of writing by the ancient Greek physi-
cians brought about a profound innovation in the history of medicine.1 Thus
although apparently ancient doctors did not need to use writing in their craft,
written medical texts are already attested in the mid 5th century as part of a
consolidated activity of the techne.2 For instance, this was the period (the sec-
ond part of the 5th century BC) that saw the creation of some important medi-
cal texts attributed to Hippocrates, under whose name roughly sixty writings
have come down to us. They are the core of the present essay. But on reading
through these materials it immediately becomes evident that even the most
ancient are by no means the first experiments in writing: the authority of the
so called Knidiai gnomai (Cnidian Sentences) reveals a firm tradition, which no
physician could afford to disregard, as is made clear by the polemics against
this work in the Hippocratic treatise Regimen in acute diseases. The text, pro-
duced by a collective writing community ( ), is assumed to be
known to its readers; moreover, it has been also revised or re-edited3 to cor-
rect mistakes and make improvements. What the author of Regimen in acute
diseases attacks is a written tradition4 and he seeks to counterbalance its influ-
ence by using the same kind of medium. And the opinion of Epidemics III The
power to study correctly what has been written I consider to be an important
part of the art of medicine5 shows that the medical profession was exten-
sively literate.6 Furthermore, rewriting for a different purpose, with additions
or adaptations, seems to have been a widespread feature in ancient medi-
cal texts, as was pointed out in the 1970s by J. Jouanna and H. Grensemann7
with regard to the so-called nosological and gynecological treatises of the
Corpus Hippocraticum.
It should also be borne in mind that while some physicians began either to
write lists of diseases for internal use or to produce pamphlets promoting their
craft (for ex. On the art), or addressed a lay public in order to disseminate basic
knowledge on healing (for ex. On affections), they also used private notebooks
12 For the production, circulation and reutilization of technical texts, Perilli [2006] and
[2009], Langholf [2004]; on Antiphanes, Langholf [1986] 1721. The presumed allusion
to Oath in Aristophanes Thesm. 272274, proposed by H. S. Jones, is not plausible: see
Anastassiou-Irmer [1997] I, 289 with bibliography.
13 See Aristotles remark that medical writings are useful only for lay people in Eth. Nic.
X 10, 1181b 2 ff. For the secondary effect of nevertheless addressing a larger audience in
producing written texts, Althoff [1993] 221223.
14 Perilli [2006] and [2009].
15 Marganne [2004b]; the most recent printed list in Andorlini [1993] 458562, and for a
continuous up-dating see http://promethee.philo.ulg.ac.be/cedopal.
1130 Manetti
rolls re-utilized by writing on the reverse side, there are also numerous cases of
carefully written books, showing a professional hand.16
The 4th century tradition of medicine soon became centered around the name
of Hippocrates, who will be taken as the reference point in addressing the phi-
lology of medical texts.17 The so-called Hippocratic question, which galva-
nized the attention of generations of scholars in the attempt to identify the
real Hippocrates, has become somewhat outdated and has given way to differ-
ent trends of investigation, but it needs to be addressed at least as part of the
quest to delineate the history of the Corpus Hippocraticum, and, if possible, to
highlight the stages of a body that was growing over time.18 Pre-Alexandrian
evidence is limited to a famousand much debatedpassage from Plato, in
which the name of Hippocrates is brought up, as well as a mention of Aristotle
and a medical doxography produced in the Peripatos, all of which have given
rise to a never-ending bibliography.19
Several preliminary remarks will be helpful here. Since we will be deal-
ing essentially with the reception of Hippocrates, it should be noted that his
reception can be understood in several different meanings: 1) a reference to
Hippocratic doctrines, 2) a reference to Hippocratic writings, 3) philological-
exegetic interest (collection and cataloguing of texts, text criticism enquiring
into the issue of authenticity, glossaries, editions, commentaries), 4) doxog-
raphy, 5) genealogy/biography.20 In this paper the focus of investigation will
mainly concern the last three points, which are interrelated and include the
definition of a Corpus of works attributed to Hippocrates and the debate on
their authenticity, the first signs of interest and attention to the texts in the
form of lexicons and editions, the extensive development of works devoted
to interpretation of the Hippocratic texts and the historical interest in the life
and works of Hippocrates within the context of the history of the discipline.
16 For ex. the most ancient exemplar of a Hippocratic work, Epidemics II, 1st c. BC, PSchyen
inv. MS 2634/3 + PPrinc inv. AM 15960A = CPF [2008] 137143, or a late exemplar of Gal.
De comp. medic. per genera, PAnt 186 (5th c. AD) = CPF [2008] 1043.
17 On the scanty traces of philological and exegetic activity devoted to other medical
authorities, see infra 4.
18 Roselli [2000].
19 For an overview see Lloyd [1975], Smith [1979] 3444; Jouanna [1992] 85105.
20 Kudlien [1989] 355.
Medicine and Exegesis 1131
21 On the importance of local traditions of Cos and Cnidos, and on the confirmation that
some biographic texts of the Corpus, considered to be of a late date, have received by
epigraphic testimonies, see Jouanna [1992] mainly 4384.
22 Pherekydes FGrHist 3 F59 = Sor. Vit. Hipp. 1 (CMG IV, p. 175, 37).
23 Pl. Prt. 311bc, Phdr. 270c; Arist. Pol. VII 4 (1326a 15). I do not intend to address the
interpretive problems connected with these passages.
24 F67 Lenfant = Gal. In Hipp. Art. IV 40 (XVIIIA 731, 69 K.), see Jouanna [1992] 92.
25 Jouanna [1992] 592 n. 14; Lenfant [2004] clixclxi.
26 T162 van der Eijk [20002001] I, 262, and commentary, II, 302. Galens commentary on
Airs, waters, places quotes a passage of Hygieina (or of On catarrhs, see fr. 137 van der Eijk)
by Diocles, where he discusses the meaning of the so called (Ar. II 78, 3 L. etc.).
kedma is attested only in the Hippocratic Corpus and in Aretaeus, see Anastassiou-Irmer
[2001] II 1, 48 n. 1. Inference of Diocles direct connection to Aph. is possible in frgs. 55a,
55b van der Eijk.
27 For the relationship between the Problemata and the Corpus Hippocraticum (especially
with Aphorisms, Airs waters places and Epidemics) see Bertier [1989], Jouanna [1996].
28 This has to do with a Hippocratic quotation in Hist. an. 512b 12 ff. where Aristotle quotes
a passage of On the nature of man (Nat. hom. 11, CMG, I 1, 3, pp. 192, 15196, 5) but ascribes
it to Polybus (who in the biographical tradition is Hippocrates son-in-law and pupil). In
general I do not approach the problem of the traces of Hippocratic treatises in Aristotle:
for a summary see Kudlien [1989] and most recently on the issue Oser-Grote [2004].
1132 Manetti
in PBritLibr inv. 137, the so called Anonymus Londiniensis.29 Upon its first
appearance the Hippocrates of the Anonymus, clearly based on the treatise
On Breaths, ill accorded with the image that had taken shape in the course of
nineteenth-century scholarship. I will not go into the history of the question
in detail; suffice it to note that according to the doxography, both On breaths
and Aphorisms were regarded in the Peripatos as belonging to the Hippocratic
context.30 That the treatise On breaths, a work bearing the hallmark of a decid-
edly rhetorical approach, was very well known (although its Hippocratic
authorship cannot be demonstrated) emerges from the above cited fragment
of Antiphanes comedy The physician and also from a Callimachean allusion
identified by V. Langholf.31 Finally, Callimachus takes us to Alexandria, which
is not only the turning point of the history of the reception of Hippocrates but
also the place where a philological interest in medical texts blossomed in its
most concentrated form.
Two preliminary problems must at least be raised: 1) from where and how
did medical texts, whether connected to the figure of Hippocrates or not, arrive
in Alexandria? 2) is it possible to identify a certain number of works, attributed
to him in the most ancient period, which may have constituted the starting
point of a Corpus?
An erudite culture with literary and dialectological interests had
already developed on Cos in the 4th3rd century, with Philitas of Cos
the author of the 32who was Ptolemy IIs tutor and
also the teacher of Zenodotus, the first chief librarian of the Museum.
Therefore it is hardly a coincidence that the earliest writer to take an inter-
est in the language of Hippocrates, dedicating a glossary to him, was pre-
cisely a grammarian, Xenocritus of Cos, probably a contemporary of Philitas.33
34 Anth. Pal. XII 150: see also Marasco [1996] 450451. On the decree of Cos for Kephisophon,
see Samama [2003] n. 132, 240243. In the 3rd and 2nd century the island of Cos was
an exporter of physicians towards other cities, which issued honorary decrees for
them: these decrees stated they had behaved in a manner that was worthy () of the
homeland (see Samama [2003] 22555, 26668, nn. 56, 126, 136): a confirmation of the
prestige enjoyed by the local medical tradition.
35 A symptom of this prestige, which continued the biographic tradition of the relations of
Hippocrates the Great with Perdiccas (see infra), can also be seen in the presence of a
Hippocrates (IV) of Cos as the physician of Roxane (the widow of Alexander the Great);
this physician was murdered upon the instigation of Cassandrus: Suidas s.v.
( 567) and Jouanna [1992] 6971.
36 Soranus, Vit. Hipp. 4 (CMG IV p. 175, 9 ff.).
37 They are similar to those represented, for ex., in the pseudohippocratic treatise Presbeutikos
7, in which Hippocrates journeys are depicted in a eulogistic manner, confirming the
renown and philanthropy of Hippocrates. An ancient genealogical tradition that linked
the Asclepiadae of Cos to Podalirius is testified as early as Theopompus (FGrHist 115 F103,
14), in the 4th century BC.
38 Fraser [1972] 346 admits that there may be some elements of truth in the hypothesis that
the Corpus was perhaps the core of a library of a physician of Cos.
1134 Manetti
39 Montana 2.2 and Lapini 2.1 in this volume; Irigoin [1994] 5053.
40 Ath. 1.3ab; Strab. XIII 1.54 C609, Plut. Vit. Sull. 26.12.
41 See Blum [1991] 53; cf. also Montana 2.2 and Lapini 2.5 in this volume.
42 For a sketch of the different positions, see Blum [1991] 5758; for the conciliatory
interpretation, Irigoin [1994] 5253.
43 Skepticism regarding the disappearance of the Aristotelian works has been espressed
by Barnes [1997] in particular 1216, where he underlines the attested knowledge of
Aristotelian texts at Alexandria; a critic of Strabos version in Lindsay [1997] 29098.
Medicine and Exegesis 1135
Blum [1991] 6164 argues that the Ptolemies did indeed buy Neleus library, and this
position is then followed by Nagy [1998] 205206.
44 Irigoin [1994] 53: Apparemment, ces traits ne se trouvaient pas au Muse lorsque
Callimaque tablit ses Pinakes, mais il est probable que des copies en ont t acquises
plus tard.
45 Wendel [1949a] 7173; Blum [1991] 151. Pfeiffer [1949] 344 (fr. 429), attributes the mention
of Eudoxus of Cnidos to his tutor Philistion of Locri, in the class of physicians. Wellmann
[1929] 17, goes so far as to suggest that the edition of Hippocrates by Bacchius of Tanagra
was based on the list provided by the Pinakes of Callimachus, but there is no solid
evidence for this.
46 , see Gal. In Hipp. Art. XVIIIA 379, 614 K.: dating these catalogues
is difficult, but Galens formulation is analogous to that used by Dionysius of Halicarnassus
for the lists of rhetorical authors in a passage of Ep. Amm. 4 (= fr. 432 Pfeiffer),
, which, although anonymous, is nevertheless identified
as a Callimachean citation, connected with that of fr. 447. Blum [1991] 150, takes it to be
a broader citation, which includes Callimachus and the lists drawn up by the Pergamon
grammarians. Certainly, it cannot be ruled out that the Hippocratic works may have
entered the lists somewhat later, for example in the work by Aristophanes of Byzantium,
devoted to correcting and completing Callimachus Pinakes (frr. 368369 Slater). On the
other hand, Bacchius Hippocratic Lexicon (see infra) does suggest the possibility of an
early presence of Hippocratic works, already in the Callimachean work.
1136 Manetti
...and (they say) the Ptolemy who was then king of Egypt became so
greedy for books that he ordered that the books of everyone who arrived
by ship be brought to him. After he had them copied on new paper he
gave the copies to the owners of the books that had been brought to
him on the debarkation, and deposited the confiscated books in the
library with the inscription Of those from the ships...The kings agents
inscribed the names of all the travellers on the books that were put in
storage, because they did not take the books straight to the library, but
stored them in houses in heaps.47
47 Gal. In Hipp. Epid. III, CMG V 10.2,1 p. 79, 815, 1922. This procedure (namely, having a
copy made of the procured texts and then giving the copies instead of the originals back
to the owners) is linked to the immediately following story of asking Athens to lend the
official copy of the three tragedians (p. 79, 23 ff.); cf. also Montana 2.3 in this volume.
48 According to Smith [1979] 201 the medical books were collected in haphazard fashion,
though in my view this does not necessarily exclude the possibility that there may also
have been a nucleus of books deriving from a library repository, whether of Cos or of
some other origin: cf. Smith [1990] 9.
49 The opinion is not new, see Wellmann [1929] 1621, Edelstein [1935] 13101312, Smith
[1979] 199 ff., Nutton [1975] 315. But, far from imagining a collection that immediately
gave rise to a canonical Alexandrian edition, already coinciding with that of Erotianus,
I believe there are likely to have been groups of works of diverse origin.
Medicine and Exegesis 1137
to the same author, until the moment when Bacchius of Tanagra (275200 BC),
a pupil of Herophilus, decided to compose a Hippocratic lexicon (Hippocrates
Lexeis). As pointed out by von Staden, the work drawn up by Bacchius50 is
the oldest example of a lexicon dedicated to a specific author.51 The extant
fragments allow identification of about 18 works, considered to be clearly
Hippocratic: Prognostic, On the sacred disease, On joints, Instruments of reduc-
tion, Epidemics I, II, III, V and VI, Prorrhetic I, In the surgery, On wounds in the
head, On fractures, Regimen in acute diseases, On the nature of bones, On places
in humans, On diseases I, On the use of liquids, Aphorisms.52 Von Stadens remark
that the so-called Coan treatises dominating in this group confirm the lively
contact between Herophilus school, the Ptolemaic court and Cos is plausible,
although one should not presume the distinction of any particular school in
the term Coan.53 This list should be extended by addition of On the nature of
child, if credit can be given to the Vita Bruxellensis of Hippocrates, according to
which Bacchius reported that Hippocrates ordered On the nature of child after
the Aphorisms.54 Bacchius possibly also suggested that In the surgery should
be read before all of the other works by Hippocrates.55 Thus Bacchius, writing
roughly in the middle of the 3rd century BC, presupposes both an already large
list of tests (formed of at least 19 works) attributed to Hippocrates and a discus-
sion concerning the proper order in which they were to be read.
While the number and titles of the progressively growing collection of
Hippocrates works are known from the lists in Erotianus Lexicon (1st c. AD)
and other sources,56 a material order of the collection, that is to say, a sequence
of writings, current at the time and concretely attested in ancient rolls, can-
not be identified. Traces that could provide evidence of an accepted order in
the papyri are rare and late.57 Only PAnt I 28,58 a parchment codex of the 5th
century AD, presents the sequence PrognosticAphorisms, and PAnt 184,59 a
papyrus codex of the 6th century AD, On superfetationOn womens diseases.
In both cases the sequence is not attested in the medieval manuscript tradi-
tion. But the medieval manuscripts do preserve traces of ancient sequences
(dating from the phase when the Hippocratic texts were still transmitted on
papyrus rolls) in the so called reclamantes (catch-words), short passages that
were copied at the end of some works (without any separative marks) and
coincided with the incipit of the first text in the next roll.60 In general they
seem to bear witness to different sequenceswhich are also in contradiction
with one anotheras compared to the order that has actually come down to
us through the manuscripts themselves. Thus the individual Hippocratic texts
seem to have been gathered together and then have undergone a process of
separation, followed by redistribution in different sequences over time, from
an ancient phase up to the 5th6th century AD.61 What interests us here is
the one and only case that can with certainty be dated earlier than Galen
(2th century AD): at the end of On the nature of man, all the medieval mss. have
the text of On diseases II 12.62 Furthermore, the ancient Latin translation (5th
6th c. AD), but above all Galen himself, in his commentary on the treatise, both
contain this passage. In fact, Galen considers it to be an integral part of the
workeven though open to doubt63yet without realising its nature. Now,
Jouanna has demonstrated that the treatise On diseases II, as we know it in its
present form, took shape through the juxtaposition of two distinct treatises,
one going from chap. 1 to chap. 11 (Morb. II B) and the other going from chap. 12
to the end (Morb. II A), which is older than the previous part. Therefore the
passage preserved in the medieval mss. dates back to a very ancient phase, in
which the treatise Morb. II A was still an independent work and was located at
the beginning of a roll that followed the one ending with On the nature of man,
57 The fragment of an Oxyrhyncus roll testifying to the sequence Alim. > Liqu., currently in
press, edited by David Leith (32 4B.3/K(46)b), is thought to date from the 2nd century AD.
58 CPF [2008] 7782, 174176.
59 CPF [2008] 113125, 130133.
60 On the subsequent material, see Jouanna [1977].
61 For the late origin (10th century) of a large collection of Hippocratic treatises from smaller
groups of works, see Irigoin [1975]. Edelstein [1935] 1311, explicitly says muss es schon im
Hellenismus verschiedene Corpora Hippocratica gegeben haben.
62 Nat. hom. 23, CMG I 1, 3, p. 220, 17 = Morb. II A 1 (12).
63 In Hipp. Nat. hom., CMG V 9.1, p. 113,118: Galen generally ascribes the doubt on the
authenticity of this passage to some writers, who, however, do not recognize the origin of
the text at all.
Medicine and Exegesis 1139
before the chapters of Morb. II B were joined with the above-stated treatise.
This merger took place at a time long before Galen, who had no knowledge of
what had happened. It probably came about because at some point, during
a copying procedure, Morb. II B must have come to occupy a position at the
end of a roll, while the next roll began with Morb. II A, and the homogeneous-
ness of their content resulted in their merger into a single work. This phenom-
enon also highlights two aspects of the tradition, both of which are of major
importance in the history of Hippocratic exegesis, influencing the structure of
Hippocratic works, even prior to their arrival in Alexandria: a) the possibility
of the merger and/or separation of blocks of text,64 b) the probable absence
of specific titles, as well as of author, in groups of treatises or, in contrast, the
great variability of the titles.65 However, while mergers and separations may
have been partly due to mere chance, as soon as reflection on the texts attrib-
uted to Hippocrates began to play a more significant role it certainly became
vital to impose some ordering and cataloguing system (in the Pinakes?), as a
means of dealing with texts that had similar content or were already known
by similar titles, or had no title at all.66 We have only very late traces of debate
on the composition of certain works or on the variables influencing the titles,
but it should be kept in mind that from the very beginning of their reception in
Alexandria such problems must have been a feature of the medical texts that
became incorporated in the Corpus Hippocraticum.67
64 For instance, one needs only mention the insertion, within the block of Hippocratic
gynaecological treatises On womens diseases III, of a piece on womens diseases, partially
conserved, by the so-called Author C (who coincides with the author of On generation
On the nature of childOn diseases IV), while On generation and On the nature of child, in
turn, constitute what was originally a unitary work which became divided in the medieval
tradition): cf. Grensemann [1975] and [1982], Lonie [1981] 4353. In general cf., again,
Langholf [2004].
65 Jouanna [1997] 6073.
66 On the activity of reordering or attribution of new titles in Callimachus Pinakes, see Blum
[1991] 156. As an example of the variety of titles, see Regimen in acute diseases, as in the
testimony of Galen and Athenaeus (Anastassiou Irmer [1997] I, 13; [2001] II 1, 1) or those
of Airs waters places (ibidem I, 23).
67 The Vita Hippocratis secundum Soranum 13 (CMG IV, p. 177) briefly mentions the problems
relating to the debate on the authenticity of the Hippocratic works; in Galen one finds
numerous mentions of the criteria he adopted in analysis of the problem, and it is likely
that he derived them from the previous exegetic tradition: see for ex. In Hipp. Art. quoted
supra n. 46, and In Hipp. Nat. Hom. CMG V 9.1, pp. 55, 57, where Galen attributes to the
Ptolemaic voracity for books the start of faking literary works. On the authenticity criteria,
see for ex. Manetti Roselli [1994] 15671568.
1140 Manetti
The history of the Museum and its library has already been outlined in this
volume (see the contribution by F. Montana). As far as medicine is concerned,
it must be underlined, as I suggested earlier, that knowledge of the medical
(perhaps specifically Hippocratic) tradition was present at the court of the
Ptolemies, and that it enjoyed considerable prestige. On the other hand, there
is no evidence indicating that the first and prominent physician in Alexandria
in the 3rd century BC, Herophilus of Calchedon, was ever a member of
the Museum or in general that any medical research was ever conducted at
the Museum.70 But the fervent intellectual climate of the new capital stim-
68 The citations by Demetrius of Laconia (15075 BC) of three Hippocratic works, Epidemics
VI, Prognostic and Prorrhetic (the last explicitly attributed to Hippocrates), merely confirm
that these treatises were stably associated with Hippocrates (even outside of Alexandria:
it appears that Demetrius lived partly in Miletus and partly in Athens): they were already
present in the group of Bacchius of Tanagra, cf. Roselli [1988].
69 See infra 4.1.1.
70 A certain Chrysermus, in an inscription of the 2nd century BC from Delos, is described as
Superintendent of physicians, Administrator of the Museum, but he was probably not a
Medicine and Exegesis 1141
ulated and may well have offered at least an indirect patronage of the arts.71
Herophilus (320250 BC), probably educated in Cos by Praxagoras, exploited
the possibilities of a frontier environment, gained exceptional progress in
anatomical knowledge and founded a new school.72 The milieu in which such a
vast community of poets, intellectuals and scientists were active proved favour-
able to a fertile exchange of ideas among disciplines: it is by no means surpris-
ing that the poet Callimachus alluded to the anatomy of the eye discovered
by Herophilus, nor is there anything remarkable in the fact that Callimachus
was aware of and referred to the Hippocratic treatise On breaths,73 but it is
perhaps worth emphasizing Herophilus knowledge of dialectics (as emerges
from the anecdote of his dialogue with Diodorus Cronus),74 or his application
of geometric theorems to medicine.75 It was no less than a genuine osmosis
among disciplines, which was made possible if not by the Museum itself or by
the royal patronage, certainly by the common presence of so many artists and
scientists within the restricted circles of the Greek lite of Alexandria, and this
aspect must be borne in mind when evaluating the first stirrings of a philology
applied to medical texts. One should not underestimate the detail provided by
Erotianus (1st century AD) in his Hippocratic Lexicon, where he states that the
first to take an interest in glosses of Hippocrates was a grammarian, who in
other respects was quite unknown, Xenocritus of Cos. His dating is uncertain,
physician and, most of all, was not the Herophilean Chrysermus (1st century BC), see von
Staden [1989] 523528.
71 Apart from the information that the Ptolemies permitted Herophilus to dissect human
cadavers or even live condemned prisoners (on which see von Staden [1989] 139153),
it is worth noting that Andreas, one of Herophilus first pupils, was present at court as
the personal physician of Ptolemy IV Philopator; note also the association of Dioscurides
Phacas with Ptolemy XII Auletes and Ptolemy XIII (von Staden [1989] 519522). For the
connection of Apollonius of Citium with Ptolemy XII Auletes, see Fraser [1972] I 312; von
Staden [1989] 455456.
72 A tradition of which we have evidence from Polybius (XII 25d 26 = Herophilus T56 von
Staden) believed Callimachus (275205 BC) to have been a co-founder of the rationalist
school, and speaks of Herophileans and Callimacheans as parallel groupings forming part
of the same school. He was probably a contemporary of Bacchius of Tanagra and Philinus
of Cos, that is, the first generation pupils.
73 Callimachus Hymn. Artem. 53, cp. Herophilus T87T89 von Staden: Oppermann [1925].
For the Hippocratic On breaths, see supra n. 12 and Langholf [1986] in particular 517. For
scientific echoes in Callimachus, see also Most [1981] and White [1986].
74 T15 von Staden.
75 See T236 von Staden and Manetti [2011b]: von Staden [1998a] has already underlined the
Herophilean Andreas exploitation of the new mechanical technology in the construction
of an instrument for reducing dislocations of larger joints.
1142 Manetti
but what is striking is that Xenocritus shared a common homeland and inter-
ests with Philitas and the latters Ataktoi Glossai. Once again, the link between
Cos and the development of medicine in Alexandria seems significant. The
interest in the Hippocratic texts shown by grammarians, i.e. an interest aris-
ing from the general point of view of language, would subsequently continue
even beyond Alexandria, and would influence the early stages of Hippocratic
lexicography, which seems to have flourished in an atmosphere shared with
Alexandrian philology.76
In Alexandria two medical schools opposed each other throughout the
Hellenistic period: the Herophilean school, defined as rational or dogmatic,
and the Empirical school.77 Both derive historically from Herophilus. As
demonstrated by von Staden, Herophilus pupils never took up a staunchly
dogmatic position aligned with the doctrines of their master: rather, they
maintained a considerable sphere of freedom of research and criticism,78 so
much so that it was in fact a first generation pupil of Herophilus, Philinus of
Cos, who made a break with the original group and founded another school,
which he called Empirical. This new school, polemicising with the doctrines
of the master, proclaimed that all speculation on causes and all anatomical
research was pointless. The rivalry between these two schools exerted a cer-
tain effect on their history, at least with regard to the Herophileans, who soon
turned their backs on anatomical research in favour of greater attention to
pharmacology, a theme certainly cherished by the Empiricists. However, the
Empiricists, supporters of an empiricism that rejected all rational hypotheses,
developed a conception of experience that was historicized in the concept of
, which made it possible to accept as valid the texts of earlier physicians,
with the argument that their writings bore witness to observation of real data
of the past.79 This accounts for the Empiricists interest in reading and working
76 See von Staden [1989] 454 and infra 3.1. On the osmosis among disciplines, see also,
in this volume, Montana 2.4, who cites a work on wounds in the Homeric poems,
composed by Aristarchus pupil Ptolemy Epithetes. On historians-philologists and the
interpenetration of these interests within the same person, see also Montana [2009c]
175181.
77 Here I will leave aside both Erasistratus, whose activity in Alexandria rather than Antioch
is still the object of debate, and also his followers, who, however, do not seem to have
played a role in relation to Hippocrates: see von Staden [1989] 4648, 142 n. 7 with the
bibliography.
78 See in particular von Staden [1989] 445462.
79 For the three ways in which experience could be gathered (teresis, metabasis tou homoiou,
historia), see the brief sketch in Vallance [2000] 106107. The reference text is of course
Deichgrber [1930].
Medicine and Exegesis 1143
on the Hippocratic texts: from the very beginning they were eager to dip into
this veritable treasure chest of medical experience of the past. For instance,
Philinus is said to have written a Glossary of Hippocrates, but he was equally
quick to respond to an earlier work composed by a former fellow student of
his, who had subsequently become his rival, namely Bacchius of Tanagra, of
Herophilean persuasion (and prior to Bacchius another pupil of Herophilus,
Callimachus, had also entered the fray). Thus within a very short space of
time three physicians who had received their training in the same environ-
ment, but had subsequently parted ways on account of divergent professional
choices, began to test their mettle by undertaking the same kind of work on
Hippocratic texts. The Empiricists certainly used Hippocrates to build up their
image and they emphasized the Empirical character of these studies, but ideo-
logical interest was not, it seems, the prime mover underlying the birth of the
study of Hippocratic texts, since by right that title belongs to the Herophilean
school, which did not claim to have any descendance from Hippocrates, despite
remaining in some respects within its frame of reference.80 The Herophilean
school placed great emphasis on high literacy, viewed as one of the benefits
of the shared cultural climate, which brought to the fore the ancient tradition
of every discipline. And medicine, in particular, had long been based on the
transmission of a written tradition, as noted above.
One of the common-place remarks, in this context, is that Bacchius, in
writing his Hippocrates Lexeis, derived some of the material from the Lexeis
composed by Aristophanes of Byzantium.81 Given that the two scholars were
contemporaries and belonged to environments that were certainly close to
each other, the assumption is in itself quite plausible, though it is based only
on the preface to Galens Glossary. Galen cites Bacchius in the polemic with
Dioscorides Lexicon, which had also provided explanations for common and
well known words: 1) Galen argues that, in contrast with Dioscorides, Bacchius
had concerned himself only with glosses, in the specific sense of difficult
words, yet the very title of Bacchius work, Lexeis, as well as the reconstruct-
ible fragments, clearly show that this is not accurate; 2) additionally, Galen
asserts that the grammarian Aristarchus (referred to in this manner in the
manuscripts), as they say, had collected a large quantity of examples for
Bacchius. Kleins conjecture Aristophanes, universally accepted, corrects the
obvious anachronism in Galens text, but as suggested by Perilli,82 it may not
83 Besides, the only other time that Galen cites the philologist Aristarchus (In Hipp. Nat.
hom. CMG V 9,1, p. 58, 79), is for the purpose of attributing to him the use of the obelos as
a sign of expunction in his Homeric text (which would be imitated by Dioscorides in his
edition of Hippocrates), although this sign is by no means a characteristic of Aristarchus
alone.
84 Von Staden [1992] 553559.
85 See also the accusation launched against him by Apollonius of Citium, who argued that
Bacchius was more interested in the linguistic elements than in the medical content
(infra, 1151).
86 Erotian. p. 5, 1419 N.:
(codd., Meineke edd.)
,
. In
the text a transposition is performed: Schmidt [1854] 24 moved , which
in the codices is transcribed after , to the start of the sentence. The correction
introduced by Meineke, generally accepted, presupposes a previous lacuna
which allows the possibility that may refer to some character other than Hippocrates,
conceivably a glossographer whom Euphorion may have used as a source (Klein [1865]
32,8). In actual fact the text of the mss. can be accepted, admitting that Euphorion may
well have read (on the meaning, cf. Callim. Anth. Pal. VII 471, 4) Hippocrates: the defense
of the transmitted text is already in Schmidt [1854] ibidem (but afterwards he changed his
mind, see Kleins apparatus criticus ad loc.), but it is not necessary to agree with him in
taking the verb in the technical sense of praelectio.
Medicine and Exegesis 1145
Since Euphorion87 had been active in Athens, before the time when he arrived
in Antioch and became chief librarian of the royal library, it has to be pre-
sumed that knowledge of Hippocratic texts (whatever they were) had con-
tinued to flourish in Athens in parallel with the development in Alexandria.88
Erotianus derives this piece of information from later lexicographers, such as
Aristeas and Aristocles,89 and no conclusion can be drawn concerning the lit-
erary form in which Euphorion expressed his interest. Erotianus list (which
ends with Didymus of Alexandria) also includes Aristarchus who is not known
from any other source to have been interested in Hippocrates: it has been sug-
gested that Aristarchus is a mistake and that the reading should have been
Aristophanes,90 as in the suspected error in Galens Glossary, and indeed this
is the assumption generally made. But Erotianus text, unlike Galen, does not
link Aristarchus to Bacchius and in its concise (epitomized?) formulation it
does not specify the concrete nature of the exegesis of Aristarchus. Although a
cautionary approach is essential, it should be recalled that even if no ancient
source ever cited the circumstance,91 a papyrus has revealed that Aristarchus
wrote a commentary on Herodotus, that is to say, on an Ionian text, just as the
Hippocratic texts were Ionian. However, nothing is definable as regards the
form in which Aristarchus may have devoted attention to Hippocrates.92
Erotianus did undoubtedly emphasize the extent to which grammarians
contributed to study of the language of Hippocrates, because he was inter-
ested in highlighting Hippocrates nature as an ancient writer, on a par with
87 Frr. 175176 in Gronigen [1977] 228229: according to van Groningen Euphorion may
have corrected, summarised or criticised the lexicon of Bacchius, his contemporary, and
this activity could be seen as a symptom of the rivalry between Antioch and Alexandria.
However, this does not seem credible, considering also that Euphorion probably became
the librarian at Antioch at an advanced age. Furthermore the hypothesis seems to follow
the conventional vision of the rivalry among libraries, which is not supported by other
evidence. van Gronigen points out that in fr. 157 Euphorion can be seen to use technical
medical language.
88 Study of the language of Hippocrates other than in Alexandria can also be traced in the
Glosses of Nicander (2nd century BC), whom some believe to have had close ties to the
Pergamon court, and additionally in Demetrius Lacon (2nd1st century BC), cf. infra in
this paragraph.
89 Corradi [2007]; Pagani [20112].
90 Cohn [1895a] 873 and 1002; von Staden [1992] 566567.
91 Pfeiffer [1968] 224225; Montana 2.4 and Dickey 2 in this volume.
92 Klein [1865] 32,11 assumed there was a lacuna after Aristarchus and argued that Aristarchus
did not write a Lexicon, but a more general treatment (p. XXXVII): the conjecture
Aristophanes offered a widely accepted reasonable solution: cf. Strecker [1891] 263. See
also Ihm [2002a] 6869 (with bibliography).
1146 Manetti
others such as Democritus, Thucydides and Herodotus.93 But even so, the
picture that emerges appears to me to confirm that from the very start the
language of Hippocrates had a place in the studies of the Alexandrian gram-
marians. Indeed, it was the object of attention continuously throughout the
Hellenistic period, both in medical and grammatical circles94 and also among
philosophers, if one is to believe the identification of the Demetrius cited by
Erotianus, in the preface95 and in the Lexicon,96 with the Epicurean Demetrius
Laco. It may be somewhat rash to attribute a Glossary to Demetrius, but it is
by now widely known that he had a good knowledge of the Hippocratic texts.97
Conceivably, the awareness that medical texts acted as a form of storehouse
of special or dialectal expressions may have been one of the reasons for the
interest displayed by philologists: Philitas, for example, had glossed dialect
words and Zenodotus had likewise concerned himself with Ethnikai Lexeis,98
as did Callimachus and Aristophanes subsequently.99 Interest in dialect phe-
nomena is a natural by-product of the conception of gloss formulated by
Aristotle, word not in common use, which allows the possibility that one and
the same word may be a gloss in a given author but a commonly used word in
100 Montanari [2012a]. The case cited above in n. 94, of the grammarian Thoas, is held to
show that Aristotle made use of the work on the Phrygian dialect of Thoas to explain
, a word attested, as far as we know, only in the gynaecological treatises of the
Corpus Hippocraticum and in Aretaeus.
101 Cassio [1993b] 7081.
102 The material is conventionally subdivided between editions and lexicons, hypomnemata
and syngrammata, biography and doxography. This subdivision has the aim of making it
possible to give a single treatment of some clusters of problems that are linked to specific
Hippocratic passages, which have over time become loci classici of debate, but with a
warning that there will inevitably be many cross-references. For a general survey of the
typology of philological writings, see Dubischar, in this volume.
103 Von Staden [1989] 7475, while Smith [1979] 191193, remains skeptical.
104 T262265 von Staden.
105 T270 von Staden, with Cobets correction , attributes
to Bacchius rather than to Herophilus (as in the mss.) the composition of a Lexicon.
A similar correction by Cobet in T34 restores the attribution to Bacchius of a comment on
Aphorisms. T 269 shows that Bacchius very likely endorsed an explanation of Herophilus
that considered and to be synonyms (once again a comparison among
terms), but without direct relation to the Hippocratic text. As regards the word of
Hipp. Epid. VI 1.4 (frr. 267a and b von Staden), the role of Herophilus, adduced by Zeuxis
1148 Manetti
the Empiricist (infra) as a witness in his commentary ad loc., by no means implies that
Herophilus offered lexical reflections on Hippocrates text.
106 Gal. In Hipp. Epid. III, CMG V 10.2.1 p. 87, 212 = fr. 7 von Staden.
107 See infra 4.1.1.
108 Namely, the Collectio Alexandrina, edited by Bacchius: Wellmann [1931] 2, 6, 8.
109 Montanari [2011b] 115, who records all his previous articles, and Montanari 1, in this
volume. See also Montana 2.4, in this volume.
Medicine and Exegesis 1149
from Epid. III.110 More concrete indications can be drawn from a rapid over-
view of the papyrus documentation pertaining to Hippocrates. Unfortunately,
this documentation dates mainly from the Roman era, but it can provide some
examples of various redactional practices. Naturally, one finds first-hand and
second-hand corrections, which were the accompaniment of the production
of copies in any scriptorium,111 whereas variants112 or small marginal titles113
are more rarely encountered. In the more carefully produced exemplars there
is evidence of reading signs and punctuation marks, such as dipl, dipl obelis-
mene, paragraphos, ektheseis and blank spaces.114 In the most ancient exem-
plar to have survived, a fragment of a roll of Epidemics II (1st century BC), no
marginal diacritic signs can be seen, on account of the lacunae, but it can be
observed that the scribe constantly uses on-line spaces to divide text units,
which in various cases do not correspond to the punctuation preserved in the
medieval codexes.115 This text, despite its reduction to essentials, represents an
ekdosis, because the segmentation (i.e. the syntactic interpretation) of com-
posite and elliptical texts like Epidemics (in particular II and VI) or Aphorisms
was one of the main critical problems: thus the blank space may itself consti-
tute a critical intervention.116
As noted earlier, the grammarian Xenocritus of Cos was the first, accord-
ing to the Empiricist Heraclides of Tarentum, to attempt an explanation of
Hippocrates words. It is perhaps no coincidence, in the perspective of our
110 The fact that one finds a mention, in the passage, in a coordinated sequence, of the royal
library, the books from the ships and the ekdosis of Bacchius, allows the surmisal that
there may have been a copy of Bacchius ekdosis in the Museum library. I regard as rather
implausible the hypothesis advanced by Mansfeld [1994] 201, who argues that ekdosis
should be taken here as meaning interpretation, which would certainly be appropriate
in the context of Galens exposition, and support the ascription of a commentary to
Bacchius.
111 Corrections by a second hand, for ex. in PDubl 1 (Epidemics VII, 1st2nd c. AD); twofold
redaction in POxy 1184 (Hipp. Epist., 1st c. AD), CPF [2008] 150157.
112 PRyl 56 (Hipp. Acut., II sec. in.) in CPF [2008] 134137, twofold redaction in POxy 1184
(Epist., 1st century AD).
113 PAnt I 28 (5th century AD), which contains the end of Prognostic and the beginning of
Aphorisms: CPF [2008] 80.
114 A diple at the beginning of the clinical history PSI 116 (Epid. III, end of 3rd century AD),
CPF [2008] 144148; paragraphoi and ekthesis in PBerol 21137v+6934v (Epist., end of 2nd
century AD), CPF [2008] 162167; diple obelismene and paragraphos in POxy LXXIV 4969
(Art., 2nd3rd century AD).
115 PSchyen inv. MS 2634/3 + PPrinc inv. AM 15960A, CPF [2008] 137143.
116 Hanson [1997] 310314 and Hanson in CPF [2008], quoted in n. 115; Montanari [1997o]
279280, and in this volume.
1150 Manetti
observations, that the only mention of Xenocritus of Cos concerns the mean-
ing of in Hippocrates Prognosticon,117 where he accounts for
this meaning by invoking the Ionian usage current at the time. In contrast,
Callimachus of Herophilus House (last half of the 3rd c. BC)118 is quoted
by the Empiricist Apollonius of Citium as the the first to explain the difficult
words in Hippocrates.119 Thus the reconstruction of the history of Hippocrates
glossaries points to competition between physicians and grammarians.
Bacchius of Tanagra (275200 BC),120 a pupil of Herophilus, composed the
Lexeis of Hippocrates, the earliest author-specific lexicon, which was divided
into three books or collections and appears also to have contained identifica-
tion of the Hippocratic works. Each collection was not based on a grouping
of Hippocratean treatises, as was done by later lexicographers like Erotianus,
because one finds glosses drawn from the same text in more than one book
(were there thematic criteria, perhaps?). As stated above, at least nineteen
works known by Bacchius can be identified. Bacchius is the chief source of
Erotianus Lexicon, our main source, and is used directly. The lexicon cer-
tainly provided an aid to reading and comprehension, organizing the lemmata
according to the sequence of the text. Bacchius interest focused not only on
difficult words but also on common words used in a special manner (following
the Aristotelian approach of the glossai). The presence of quotations adduced
for purposes of comparison mainly involved Homer and Euripides, but also
Aeschylus and some local dialects, such as that of the population of Rhodes,
and of the Eleans and the Thymbrians; thus he did not restrict his investiga-
tions to poetic sources. In the glossaries, debate naturally focused on variants
or put forward evidence for variants,121 and some of Bacchius readings can be
identified in Instruments of Reduction, Wounds in the Head and Prorrhetic, but
this is not sufficient to demonstrate that he also drew up an edition of the text.
117 Prog. 20 (II 170, 15 L.) in Erotian. 12, 610 N.); see Fuhrman [1983] 1533.
118 On the role of Callimachus in Herophilus school, see supra, n. 72, and von Staden [1989]
480483. He is quoted only once by Erotianus for in On sacred Disease, fr. 33, 108, 17 N.
119 Apollonius wrote a Hippocratic Lexicon in polemics with Heraclides.
120 Von Staden [1989] 484500; von Staden [1992].
121 From Erotianus we glean the following information: lemma Erotian. 28, 10 N.:
Lysimachus and Bacchius write . Lemma Erotian. 46, 19: Ischomachus and
Kydias of Mylasa write wrongly (47, 13). Lemmata and (47,
7 and 48, 15): note, also, the alternative spelling - -. Bacchius (57, 68) writes
instead of ; Bacchius (57, 22) writes instead of . Variants
are also found in the lemma (85, 25); (90, 820); (93, 1519):
(102, 1921).
Medicine and Exegesis 1151
that Glaucias also wrote commentaries on the Hippocratic texts, of which only
scanty traces remain (frr. 350, 354, 356 D. on Epid. VI): they may have been
overshadowed by the enormous exegetic activity of his contemporaries and
successors Zeuxis and Heraclides of Tarentum. Little information can be found
that could characterize his Lexicon, although, it is noteworthy that in order to
explain in Aph. VII 43 (Erotian. 15, 21 N.) he resorts to a traditional
embryological theory.126
Cited among subsequent physicians is Lysimachus of Cos127 (end of 2nd
1st century BC), who wrote a glossary composed of twenty books and two works
composed of three books against Kydias the Herophilean128 and a Demetrius
(of Laconia, the Epicurean),129 which presumably were likewise devoted to dis-
cussion of the language of Hippocrates.
The polemic flared up again in the 1st century BC, with the three books
by the Empiricist Heraclides of Tarentum against Bacchius, but his work was
attacked in no fewer than 18 books by Apollonius of Citium, himself also an
Empiricist. Thus lexicography represented an important sector in the rivalry
between the Herophilean and Empirical school, but the polemic also raged
within the schools themselves.
In line with his polemical intent, Heraclides of Tarentum (1st century BC)
is often cited in Erotianus, in opposition to Bacchius.130 Heraclides testifies
that in his own day the work On Art, on which we have no further informa-
tion in the Hellenistic era, had already become part of the group of works
attributed to Hippocrates. Heraclides makes use of literary parallels and pro-
poses etymologies,131 but the relatively few exegetic notes Erotianus cites
could also derive from the commentaries: there is at least one case in which
immediate and simplest level of exegesis, and therefore it is the most constant element,
although it is naturally the least documented for the authors of Hellenistic commentaries.
This is also the reason why the text preserved in PTebt 897 = PBingen 1 (end 3rd century
BC), which Marie-Hlne Marganne considers to be either another version ofor a
commentary onRegimen II 49 (most recently CPF [2008] 228233) is difficult to classify.
Since we have no information on any ancient commentary on the treatise, the text could
be a paraphrase of the passage in another context, cf. Anastassiou-Irmer [1997] I, 460.
139 The term is used as a conscious anachronism because, as regards the secondary literature
of antiquity, there certainly never existed a classification of genres properly speaking,
analogous to that existing for other literary forms, see Sluiter [2000a]; however, the
distinction among these forms is commonly adopted in the critical literature. See also
Dubischar 1, this volume.
140 See infra 4.1.2.
141 Smith [1979] 193 considers it unlikely in the extreme that Herophilus wrote a book against
Hippocrates Prognostic and suggests that the discussion about prognosis and prorrhesis
seems appropriate to his book Against Common Opinions.
Medicine and Exegesis 1155
authors, these are not to be considered tout court as evidence of the existence
of a written hypomnema. Information such as Galens concerning Serapion of
Alexandrias comment on Epid. VI 7.2142 could also be the trace of oral inter-
pretation in class,143 preserved in the series of commentaries in the same
Empiricist tradition. This aspect of oral exegetical practice in school must not
be overlooked when considering the history of Hippocrates reception: indeed,
it constitutes the premise of interest in a thorough critical examination of the
language and doctrine of Hippocrates. More than once Galen states that he
had written his commentaries upon the request of or pupils, who had
listened to his lectures and urged him to commit them to writing: thus the writ-
ten version of commentaries would arise after prolonged practical experience
of oral teaching.144
The practice of writing commentaries seems to have been more pervasive
in the Empiricist rather than the Herophilean school, as suggested by the fact
that Galen mentions the four commentators who (in his judgment or, possibly,
as far as he knew) were the most ancient. Among these, the Empiricists Zeuxis
and Heraclides of Tarentum wrote commentaries on the complete range of
works by Hippocrates, whereas the Herophilean Bacchius and the atomist
Asclepiades wrote commentaries only on the more difficult works.145 On the
other hand, this piece of information may be biased as a result of the particular
perspective adopted by Galen, who seems to show a certain degree of depen-
dence on Empiricist exegetic tradition146 and, above all, it does not grant us
any insight into the extent of the material which, during the period of time
spanning the transition from Bacchius to Asclepiades, was subsumed under
the heading the complete range.147 But the central role of the in the
enormous amount of Hippocratic exegesis that has been lost, and that great
caution should be exercised before making any generalization.
Bacchius was the first in exegetic activity as well. We know from Galens
observations that he wrote commentaries on Aphorisms, Epidemics VI, and In
the surgery;156 Galen asserts that he dealt only with the difficult Hippocratic
works,157 but what exactly was meant by this statement remains unclear. It
is equally difficult to gain a clear idea of the approach Bacchius adopted in
his commentaries; however, the mention of variants and the observations on
the sequence of the Hippocratic treatises, together with his considerable lin-
guistic and literary refinement, as can be gathered from the fragments of the
Lexeisa circumstance suggesting he set the Hippocratic text in a broader
comparative contextall point to a base in common with Alexandrian philol-
ogy of the time.158
Bacchius contemporary, Zeno the Herophilean (3rd2nd century BC),
sparked a famous controversy on Epidemics III which raged in the wrangles
between Herophileans and Empiricists right up to the 1st century BC.159 He
may in fact have written a continuous commentary on the treatise, as he did
with regard to other works,160 but the information given by Galen indicates
that he devoted an extensive syngramma to the so-called question of the
.161 This concerned the provenance and meaning of certain marks
present in some Alexandrian copies of Epidemics III: these symbols, mainly
Greek letter symbols, or combinations of them, were written in clusters of four
although cited in both cases for a word that can be traced to Epid. IV, could also be
the author of a glossary. It would be a further unwarranted generalisation if one were
to attribute a specific action of exegesis to figures cited for the interpretation of an
individual Hippocratic lemma, such as Erasitratus with regard to (23, 824, 10 N.
= fr. 72 Garofalo) who, even though he glossed the word, may have done so in a medical
work of his, like Diocles on kedma, cf. supra n. 26.
156 See fr. 8, 9, 10, 11 von Staden: the case of Epid. II (fr. 11) is somewhat doubtful, because the
text seems to allude to his glossographic activity rather than that as a commentator. For
a discussion of the supposed commentaries on Epid. II and III see von Staden [2006]
17 ff. For the commentary on Aphorisms (Gal. In Hipp. Aph., XVIIIA 186, 11187, 4) see von
Staden [1989] 7476.
157 Fr. 8 von Staden (Gal. In Hipp. Off. XVIIIB 631, 17 K).
158 For more detailed features of Bacchius exegesis, see von Staden [2006] 1727. I doubt that
it is possible to infer from Bacchiususe of literary parallels that he did not use consistently
the critical principle Homerum ex Homero (ibidem, 26).
159 For a synthesis of the question, see von Staden [1989] 501503.
160 Zenon conceivably composed a commentary on On places in Humans, cf. supra n. 151, and
On joints (Erotian. 23, 10 N. = fr. 7 von Staden).
161 On the form of the text, see Gal. In Hipp. Epid. III, CMG V 10.2.1, p. 86, 2022.
1158 Manetti
or five at the end of individual case histories. Evidently Zeno found them in
the copies at his disposal (apparently starting from the seventh case history)162
and attributed them to Hippocrates himself. Zenos opinion was immediately
attacked by the Empiricist Apollonius the Elder of Antiochia (fl. 175 BC), who
wrote a syngramma on the subject. Zeno continued the dispute with another
pamphlet, but the question was again treated polemically by the Empiricist
Apollonius Byblas (the Bookworm, fl. 150 BC), who elaborately refuted
Zenos defense of the symbols authenticity. The great Heraclides of Tarentum
(1st century BC) also entered into the fray, but it was only the Herophilean
Heraclides of Erythrae (1st century BC) who succeeded in putting an end to the
story, agreeing with the Empiricists on the spurious nature of the .
Zeno interpreted the sequence of signs as a code for summarizing the essen-
tial data of the clinical history; the subsequent controversy focused not only
on the question of whether such signs were authentic or not but also on their
interpretation in the individual passages (with the presence of variants and
attempts at correction). His Empiricist adversaries rejected the attribution of
authorship of the signs to Hippocrates himself, but recognized that they must
have been arranged according to some sort of system, although the Empiricists
did not agree with the readings proposed by Zeno, whom they accused of
falsifying the text to confirm his own interpretation. To confute Zenos argu-
ment, the Empiricist Zeuxis163 possibly introduced (or he may simply, in his
commentary, have made a brief mention of the reported episode) the story
of a man named Mnemon from Side in Pamphilia, who borrowed a copy of
Epidemics III from the great Alexandrian library and returned it after adding
the symbols written in similar letters and ink, or, alternatively, he may have
related that a personal copy belonging to Mnemon (with the symbols) was
one of the books to have arrived at the Museum Library bearing the inscrip-
tion Of those from the ships by the redactor Mnemon of Side, or conceivably
162 Gal. In Hipp. Epid. III, CMG V 10.2.1, p. 46, 1924, says that the ancient manuscripts (and
the ancient commentators) had knowledge of the signs starting from the seventh clinical
history, whereas more recent copies (and the edition of Dioscurides, see infra 4.1.1)
mentioned them from the first onwards; with regard to the latter, Galen does indeed
begin to speak of these signs, which, however, are immediately declared to be spurious
and not present in all the manuscripts (CMG V 10.2.1, p. 27, 128, 28). In the medieval
tradition (represented only by Codex V), the signs are present in all cases. Among the
modern interpretations, Fraser [1972] I 325326, compares the signs to those used in
the Homeric editions, but without any foundation.
163 Galen declares that he derived this information from Zeuxis commentary, which at
that time was still available to him, although it was extremely rare (CMG V 10.2.1, p. 78,
2979, 3).
Medicine and Exegesis 1159
marked only with Mnemons name.164 What can be inferred from this story is
that in the 2nd century165 it was quite normal to enquire into the authenticity
of parts of the Hippocratic text, and that consultation and comparison of cop-
ies held in the Royal Library and accessible to external scholars was a routine
practice. Clearly, there was general awareness of the diverse provenance of the
copies, and attention was paid to details and documentary evidence regard-
ing their redaction, e.g. ownership notes or subscriptions, if any such marks
were visible. Although no official link between the Museum and medicine is
documented, everything suggests that the cultural environment was open to
exchange and circulation of ideas. Moreover, Apollonius Byblas confutation of
a specific passage of Zenos text,166 adducing the argument that no manuscript
of the Royal Library nor any manuscript of From the ships, nor even Bacchius
ekdosis, contained the symbols as Zeno had written them, reveals the same
method of collating copies, and this in turn presupposes the accessibility both
of manuscripts held in the library and distinct from those of From the ships,
and also of the authoritative ancient edition of Bacchius.
Although Zeuxis had composed commentaries on all the Hippocratic
works, the evidence derivable from the sources merely shows that he inter-
preted six works: On places in humans, Aphorisms, Prorrhetic I, Epidemics II,
Epidemics III and Epidemics VI. It is worth highlighting the presence of On
places in humans,167 which since the very beginning had been included among
Hippocratic works and attracted the attention not only of the Herophileans
but also of the Empiricists, whereas it later disappeared from Galens hori-
zon. In addition to concerning himself with the interpretation of words
or expressions,168 Zeuxis also addressed textual questions.169 For instance,
when commenting on Epid. VI 2.22170 he criticized Glaucias for inserting
negations into the Hippocratic text and thereby trying to make it consistent
with the doctrine; Zeuxis likewise appears to have been inclined to preserve
the transmitted text in the case of another reading he regarded as incorrect in
164 This question is dated to the reign of Ptolemy III Euergetes I (246221 BC).
165 Deichgrbers dating of Zeuxis around the end of the 1st century BC (Deichgrber [1965]
209, 263) has been abandoned in favour of a dating in the first half of the 2nd century: see
Fraser [1972] 32527, Manetti-Roselli [1994] 1594, von Staden [2006] 30.
166 Gal. In Hipp. Epid. Iii, CMG V 10.2.1, p. 87, 112, cf. supra 3.1.
167 Lemma , Erotian. 51, 110 N.
168 He addresses problems of accentuation (i.e. of syntactic interpretation) in fr. 358
Deichgrber (= Gal. In Hipp. Epid. VI, CMG VI 2.2.2, p. 217, 13 ff.).
169 He carried out expunctions of words and sentences: Gal. In Hipp. Epid. VI, CMG V 10.2.2,
p. 219, 1920, and ibidem p. 251, 11 ff. = fr. 359 D.
170 Gal. In Hipp. Epid. VI, CMG V 10.2.2, p. 114, 19 = fr. 354 Deichgrber.
1160 Manetti
177 Fr. 356 Deichgrber = In Hipp. Epid. VI, CMG V 10.2.2, p. 174, 20 ff. von Staden [2006] 4445,
points out that ideological appropriations of the Empiricists in the texts of Hippocrates
are, from an overall perspective, fairly rare.
178 Gal. In Hipp. Epid. VI, CMG V 10.2.2, p. 451, 40452, 2 = fr. 361 Deichgrber, where Zeuxis
seems to accept the reading and interpretation of Glaucias.
179 Von Staden [2006] 41, 45.
180 Gal. In Hipp. Epid. VI, CMG V 10.2.2, p. 3, 810.
181 T 12 Guardasole (Gal. In Hipp. Art. XVIIA 735, 1014 K.) = F 43 (see also T 6).
182 Gal. In Hipp. Epid. III, CMG V 10.2.1 p. 87, 1314. Galen had direct knowledge of Heraclides
of Tarentums commentary on Epid. II and perhaps also on other treatises (Epid. III):
Manetti-Roselli [1994] 15941600.
183 Together with Zeuxis and Glaucias, see the passage cited in n. 177.
184 Doubts can be expressed with regard to the information concerning his commentaries
on De humoribus and De alimento, deriving from the pseudo-Galenic Renaissance
commentaries, see T 32 and F 96, F 97 Guardasole.
185 Guardasole [1997] with previous bibliography.
186 Mention should be made of the dietetic work Symposium, in dialogic form.
1162 Manetti
187 In addition to the commentary ad loc. Galen mentions it in the proem to the commentary
on Epid. VI (F 83 Guardasole = In Hipp. Epid. VI, CMG V 10.2.2, p. 4, 415), as a guide to his
behaviour as a philologist: preserve the ancient reading as far as possible, and interpret it,
but where this does not prove possible, then a plausible correction should be performed,
like that of Heraclides.
188 F 82 Guardasole = In Hipp. Epid. Ii, CMG V 10.1, p. 231, 25233, 1; 233, 442.
189 For the probable literary parallels of the original version of Heraclides, see Manetti-
Roselli [1994] 1598.
190 Heraclides exercised also his own judgment in segmenting the text, i.e. in determining
its syntactic structure: F 80 = In Hipp. Epid. II, CMG V 10.1, p. 220, 34221, 7; F 81 = In Hipp.
Epid. II, CMG V 10.1, p. 222, 3033.
Medicine and Exegesis 1163
of Erythrae to be built up, apart from the fact that he belonged to the ancient
authorities.195
While all the above mentioned characters had an ideologically neutral
or generally favorable attitude towards Hippocrates, a special case is that of
Asclepiades of Bithynia (2nd century BC),196 a physician active above all in
Rome, who championed a corpuscular theory of matter, in open opposition
to the Hippocratic tradition of humoral physiopathology. He wrote at least a
commentary in 2 books on Hippocrates Aphorisms and a commentary on In
the surgery.197 We now know that he also wrote a commentary on Epidemics I.198
Here we find for the first time a work dealing with medical exegesis that does
not depend on the Alexandrian milieu and is an external exegetical endeavour,
because Asclepiades rejected many of the Hippocratic theories. Therefore it
is unfortunate that little can be deduced from Galens citations, which concern
only the commentary on In the surgery: possibly due to a form of censorship,
the ideological elements that were probably present in Asclepiades commen-
taries were sidelined and the only elements to emerge are some philological
characteristics, which concern the interpretation of individual words or the
attestation of variants.199
speech the Embassy, dated to the 4th3rd c. BC,201 shows some essential ele-
ments of family genealogy and may be seen partly as the outcome of local tra-
ditions, as indeed is suggested by the most ancient pseudo-epigraphic texts
of the Corpus Hippocraticum. Although these texts are difficult to date, they
seem to go back to traditions on Cos, with close family ties to Thessaly. They
seem to show no knowledge of the existence of a Corpus of works attributed
to Hippocrates, and thus were composed, according to Smith, before the col-
lection was put together in Alexandria.202 Subsequently, in Alexandria it was
Andreas, Herophilus pupil, who composed a treatise On medical genealogy in
which he argued that Hippocrates had left Cos because he had allegedly burnt
the repository of books in Cnidos;203 a detail of this kind presupposes the exis-
tence of a biography already properly structured and included within a broad
genealogical-biographic context. The allegation, defined as malicious by the
biography attributed to Soranus, sprang from a polemical motive, but it may
also have had some relation to the contemporary debate on the authenticity of
the works of Hippocrates.204 Eratosthenes is likewise said to have addressed
the genealogy of Hippocrates, probably in his great chronological work, and he
apparently polemicized with Andreas, accusing him of plagiary.205 The accusa-
tion underlines once again the intensity of exchange of ideas among intellec-
tuals of different disciplines206 in 3rd c. BC Alexandria. Additionally, a Soranus
of Cos (otherwise unknown), probably through Ischomachus citation (see
infra in this paragraph), is said to have checked all the archives of the island in
search of precise details on the chronology of Hippocrates life.
One rather remarkable composition, probably more of an ethical-bio-
graphic than doxographic character,207 is Bacchius Memoirs on Herophilus
and the members of his House.208 The only item of information taken from this
work comes to us from Galen, who derives it from the commentary by Zeuxis
on Epid. VI. Galen speaks disapprovingly of several episodes relating to the
physician Callianax, whose response to the words I am going to die uttered
to him by a patient consisted in reciting the tragic line unless you were gener-
ated by Latona mother of fine children (TGF adesp. 178), and who responded
to the same words of another patient by citing the Homeric line death was the
fate of Patroclus as well, who was far better than you (Il. XXI 107). Zeuxis had
gleaned this piece of information from the work of Bacchius: Amneris Roselli
has recently demonstrated that the original character of the anecdote was far
from having the negative connotation portrayed by Zeuxis and Galen,209 and
that it invoked the topos of the contemptus mortis.210 In actual fact, once the
anecdote is cleansed of the deforming filter of the Empirical source, it seems
to indicate a link with the ethical-philosophical tradition of biographies like
that of Diogenes the Cynic,211 a forerunner of the Chreiai genre, which had
the aim of illustrating the coherence between the life and the doctrine of the
character in question by narrating episodes of a persons life but, above all,
highly revealing sentences uttered by the character. The Chreiai arose from
the model of Xenophons Apomnemoneumata, and it is no coincidence that the
latter work bears a title similar to that of Bacchius. It is thus a bio-ethical text,
internal to the school, highlighting the bond linking the Herophilean milieu
to the contemporary philosophical culture and interpreting in its own manner
the identity of and (parallel to that between philosophical doctrine
209 Roselli [2009] 6974, who believes that the analogous anecdote of a rough answer by
Herophilus to the philosopher Diodorus Cronus (T15, von Staden) may also derive from
the same work by Bacchius. This seems probable if one considers that Sextus (Pyrr. 2.245)
also quotes it as a lovely reminiscence () of Herophilus. Fraser [1972] IIa
533 (n. 204), remarks The third century work of Bacchius...seems to have been a memoir
of his own teacher, and not a historical work in the same sense (scil. like Heraclides work
on the Empirical school).
210 The aim of these quips of preparation for death, which can be seen as belonging to the
tradition of spiritual exercises, could also reveal a , like that attributed to
Antiphon (87 A 35 DK: note that the source, Plutarch, speaks of an application of this
techne to the sick (see Pendrick [2001] 3031, 95, 241), based on different philosophical
premises compared to the consolatory ethos that was to gain great popularity in the
Roman era.
211 Fragments of a biography of Diogenes are attested in a papyrus, PVindob G 29946, dated
to the 3rd century BC, and they attest to its circulation in Egypt during the age of Bacchius:
see G. Bastianini in CPF [1992] 99143.
Medicine and Exegesis 1167
and life, so characteristic of the Hellenistic schools) which would later become
a conventional feature of the encomium of a physician.212
Doxography as a manner of dealing with a scientific problem by starting
from analysis of the earlier theories has Aristotelian roots: the Aristotelian
doxography on the causes of disease, preserved in Anonymus Londiniensis,213
makes extensive use of the diairetic method and draws on well-known
Aristotelian categories.214 Its character is positive, tending to highlight cer-
tain elements as foreshadowing, in an implicit history of scientific author-
oriented progress. However, this kind of doxography has no continuity in
what is known so far about the Hellenistic schools. Rather, it is closer to the
tradition of Theophrastus Physikai doxai, later incorporated into Atius, than
to the knowledge derivable from available evidence on the doxography of the
medical schools.
In this regard, a growing importance of doxography with apologetic or pro-
treptic purposes can be observed for the Hellenistic period, and the following
typologies can be listed:
in the history of the Herophilean school was there such a heavy concentra-
tion of works of this kind as at the end of the 1st century BC: comprehensive
works covering physiological and pathological theories with their dual pur-
pose, apologetic and protreptic. Such works include that of Apollonius Mys
(1st century BC1st century AD), of Heraclides of Erythrae (1st century BC)
and of Aristoxenus (1st century BC1st century AD), all titled On the school
of Herophilus. The common source is Galens treatise Diff. puls.,218 in which
they are presented as a homogeneous group, apparently concentrated around
debate on sphygmology. In von Stadens view, the cause of such a concentra-
tion lay in the growing insecurity of this school within the world of medicine.219
A clear-cut definition of the work Opinions () by Alexander
Philalethes (1st century BC1st century AD), the title of which seems to imply
a doxographical character, cannot easily be given. However, the only source of
the title is De puls. diff. IV 4 (from Aristoxenus?),220 in a passage where Galen
speaks of Alexanders double definition of the pulse, adding that Alexander,
convinced he had some persuasive arguments on the theme, described them
in his work. Nothing suggests that there was a comparison among opinions
but this may be the result of a cut in the information provided by Galen or
in the source he was using. It is nevertheless certain that it was a text con-
taining a number of arguments, and this is not in contrast with what emerges
from the text of Anonymus Londiniensis, where Alexander Philalethes name
almost always appears in conjunction with Asclepiades of Bithynia,221 and
Alexander very likely constitutes the source for the knowledge of Asclepiades.
The material presumed to derive from Alexander is set in a context with a
marked dialectical approach. It may be a coincidence, but when Anonymus
for the first time cites the two authors polemically, taking up a stance against
them, the sentence is introduced by the expression (XXIV 27)
. If the source was the work Opinions, it is conceivable that rather than
being constructed in the form of a doxography of an Aristotelian type, simi-
larly to that on the causes of disease preserved in the first part of the text of
Anonymus,222 it was drawn up according to the thesis method, where a propo-
218 10, VIII 746, 9 ff. K. Galen seems to have used the text of Aristoxenus as his direct source,
although he mentions this author only rarely (von Staden [1999a] 170176).
219 Von Staden [1989] 541, cp. 457.
220 VIII 725, 17726, 12 K. = Heraclides fr. 39 Guardasole. See von Staden [1999a] 165; von
Staden [1989] 532539.
221 An. Lond. XXIV 31, XXXV 22, 54; XXXIX 1 (Manetti [2011a]).
222 Von Staden [1999a] 186, also finds it puzzling that the ancient sources are silent on the
subject of doxographic or historiographic treatises within the school of Erasistratus as
opposed to the evidence of such activity among Herophileans and Empiricists.
Medicine and Exegesis 1169
sition is stated and arguments for and against are put forward.223 But all this
remains undemonstrated.
Finally, the Vita Hippocratis attributed to Soranus records a treatise On the
hairesis of Hippocrates by an Ischomachus (CMG IV, p. 175, 912), who can-
not be precisely dated. As far as we know, only Galen speaks of a Hippocratic
hairesis: this suggests rather a late redefinition of the Hippocratic heritage,
dating perhaps between the 1st century BC and the 1st century AD.224
Within the framework of exegetic activity in the sphere of medicine out-
lined so far, one element is conspicuous by its absence: there is no trace of
any exegetic or doxographic or historiographic activity whatsoever in the tra-
dition headed by Erasistratus,225 the great physician who was a contemporary
of Herophilus, even though he enjoyed continuous reception at least until the
age of Galen.
By the imperial age, the schools of medicine had become distributed among
the centers of ancient tradition such as Cos and Cnidos, Alexandria and
Pergamon and other cities such as Corinth, Smyrna, Laodicea, Ephesus, and
above all Rome, where the innumerable opportunities attracted a growing
number of physicians eager to build up a brilliant career.226 In this context,
the concept of a school should not be taken as meaning a hairesis (i.e. an orga-
nization associated with a specific doctrinal approach) properly speaking:227
223 The text bearing the title Diktyaka by Dionysius of Aege, which we know from Photius
Library (185, 211), has a strongly dialectical structure (each of the chapters is devoted to
demonstration of the truth of a given argument, immediately followed by its confutation),
was previously considered by H. von Arnim to be the work of a physician of the Hellenistic
age, but in actual fact it is difficult to date the text in question (most recently, von Staden
[1999a] 177187, with the previous bibliography).
224 It is not necessary to suppose that Ischomachus lived after Galen (von Staden [1999a]
185 n. 119). Soranus text is not certain ( in the mss.), but if the correction by
C. Keil is accepted, we must take into account that an Ischomachus is also quoted twice
in Erotianus Lexicon (p. 47, 1 and 103, 15 N.), which thus gives us a terminus ante quem.
225 Already von Staden [1999a] 186 ff.
226 Although Greek physicians were present in Rome from as early as the 2nd c. BC, and
Asclepiades of Bithynia had built up a great following between the 2nd and 1st c. BC, the
opportunities opened up by the new Imperial structure changed the situation, cf. Nutton
[2004], chapters 11 and 12.
227 On the concept of hairesis, see von Staden [1983].
1170 Manetti
228 See Johnson [2010] 74 ff. about the reading communities at Rome in Galens times.
229 See Tecusan [2004] fr. 150 (= Dign. Puls. I 1, VIII 770 K), cp. 155, 203.
230 On Quintus and his school, Manetti-Roselli [1994] 15801589; Grmek-Gourevitch [1994]
14911528; on the scanty circulation of the commentaries of Numisianus, for ex. De ordine
libr. suor. 3.6 Boudon-Millot; on the composition of his commentaries, cf. Gal. Libr.
propr. 9.16 Boudon-Millot. See also Sluiter [1999].
231 On the works Galen devoted to the exegesis of authors other than Hippocrates, see infra
4.1.2.
232 Fr. 188 van der Eijk, on variants of Diocles Matters of health.
Medicine and Exegesis 1171
that Diocles On things in the surgery circulated in his own times under various
titles. Yet Galen seems to be fairly untroubled by this circumstance and is not
prompted to devote special attention to the problem.233 He is far more con-
cerned about safeguarding the correct transmission of the recipes, which he
sets out several times in the pharmacological treatises,234 where, however, his
interest is primarily of a practical order.
233 Fr. 160a van der Eijk. For a putative problem of interpretation of the text of Herophilus,
see T39 von Staden = Gal. Dign. puls. i 3, VIII 954 K.
234 See infra in this paragraph.
235 The list was evidently authoritative and widespread, as shown, for instance, by the fact
that Erotianus included the Prorrhetic II in it even though he was convinced of its lack of
authenticity, saying that he would demonstrate it elsewhere (p. 9, 8 N): Roselli [2000] 180
recalls the order in which information on a disease is given in the technical-therapeutic
treatises. It is remarkable that Erotianus, during the course of the work, adopted a
different order of reading.
236 Mewaldt [1909], updated by Boudon-Millot [2008], who quotes Hunain ibn Ishaqs Risala
(82); doubts on the Galenic authorship of the work have been raised by Smith [1979] 169
n. 85, shared by Manetti-Roselli [1994] 1555 n. 95. On the overall discussion of the issue,
Boudon-Millot [2008] 79 ff.
1172 Manetti
his commentary on Epid. II that the question had been discussed previously.237
The list given by Erotianus contains Oath, cited for the first time deferentially
just a few years earlier by Scribonius Largus in his preface,238 and Law, nei-
ther of which were ever mentioned by Galen. But more generally, Erotianus
bears witness to the persistent uncertainty of some titles (as also testified by
Galens Glossary) and also some absences (e.g. De carnibus, De victu, Praecepta,
Epistulae ecc.).239 From as early as the beginning of the Alexandrian period, the
quantity and variety of works attributable to Hippocrates must have resulted
in the need to assess their authenticity, but virtually our only source allowing
a glimpse of the protracted question is represented by Galen. Three categories
of authenticity are widely employed by Galen and had probably been already
employed before him: the most genuine texts by Hippocrates, the genuinely
Hippocratic texts (such as those written by close associates, for instance by his
son Thessalus or his son-in-law Polybus) and the spurious ones.240 The con-
siderable divergences in doctrine and type of writing were addressed firstly by
setting the works in the framework of the biography and genos of Hippocrates.
Thus there had been some proposals to attribute On joints and On fractures
to Galens grandfather, Hippocrates the son of Gnosidicus; similarly the group
of Epidemics II, IV, VI, with its characteristic as a collection of personal notes,
had prompted the suggestion that it was a re-elaboration by Thessalus, the
son of Hippocrates, of material derived from his father;241 Dioscorides (early
2nd c. AD) had proposed identifying the author of On diseases II as Hippocrates,
son of Thessalus, grandson of the great Hippocrates.242 In parallel with this
search for authorship, attempts were made to identify sections regarded as
spurious within authentic works, as in On the nature of man and On regi-
men in acute diseases, or in Coan prenotions which was regarded as a blend of
authentic material derived from other Hippocratic texts (Aph., Prog., Epid.),
with spurious additions.243 Extremely useful information can be gleaned from
the distinction often drawn by Galenbut certainly predating himbetween
works composed for publication (Epid. I, III), unfinished works (Off.) and
notes designed for private use (, Epid. II, VI). Naturally, some works
were judged as entirely spurious (Epid. V, VII, Gland. and Prorrh. I for Galen).244
Some ancient attributions, such as the attribution of Nat. hom. and Octim.
partu to Polybus245 and that of On breaths to Hippocrates, are treated in a dif-
ferentiated manner by Galen, according to whether or not the work in question
hampered the construction of his idealized and updated Hippocrates. Thus
the work On the nature of man formed the mainstay of Galens representation
of Hippocrates: accordingly, Galen decidedly rejects its attribution to Polybus,
and in the case of Octim. partu., a marginal work in his framework, he cites
Polybus as merely one among the possible hypotheses, whereas with regard
to On breaths, which he cites several times, he veils it in a sort of silent censor-
ship, ignoring it throughout his long exegetic activity.246 But overall, the set of
works attributed to Hippocrates, can at the time of Galen still be depicted as
a circle at whose centre there stand the most genuine works, with the others
radiating outwards from the central core, gradually decreasing in authoriality
with increasing distance from the centre. This explains why some works that
are absent from Erotianus list (e.g. De victu) or considered spurious by Galen
(Morb. I) are nevertheless glossed in the lexicons of the two authors.
The lexicography of the imperial age and subsequently of late antiquity
endeavoured to provide a systematic account of the immense erudition accu-
mulated in the previous centuries through work which, in that earlier period,
had focused directly on documents. Now attention was directed to drawing up
lexicons of a monumentally large nature, starting from a huge array of previ-
ous works (for instance the lexicon of Pamphilus, organized by themes and by
alphabetical order within each theme), or collections representing a new genre
Acut., CMGV 9,1, p. 277, 35; Coan prenotions: Gal. In Hipp. Epid. III, CMG V 10,2,1, pp. 13,
5 ff., 62, 7 ff.
244 In Hipp. Art. XVIIIA 379, 614 K; In Hipp. Prorrh. CMG V 9,2, p. 13 27 ff., 67, 29 ff. and passim.
245 Cited both by Aristotle in Hist. an. III 3, 512b12513a7, and also in the doxography of
Anonymus Londiniensis (see supra); De octim. partu is cited, in At. doxographer, Ps.
Plut. and in Clem. Al. as being by Polybus, in what seems to be a peripatetic doxographic
tradition: Anastassiou Irmer [1997] I, 374378. In these works no connection between
Polybus and Hippocrates is mentioned: Smith [1979] 219222 maintains that Polybus was
inserted only later into the Hippocratic authorship as Hippocrates son-in-law.
246 The text of Flat. is reduced essentially to two sentences of generic gnomic content, cited
in an anonymous manner: namely treatment through contraries, which is much cited,
Flat. 6 (VI 92, 1011 L), and the difference between individual constitutions (ibid. 98, 7 L:
see Anastassiou-Irmer [2001] II 1 281287). Another silent omission is that on De morbo
sacro, which is close to Ar. Anastassiou-Irmer [2001] II 1 340341, [2006] II 2, 224230.
1174 Manetti
organized only by theme (Pollux). At the same time, work was also devoted to
producing epitomes and extracts (Vestinus and Diogenianus). Little by little
the function of lexicons underwent substantial change, becoming increas-
ingly independent of the texts from which they drew their origin. The work of
Erotianus, who lived in Rome at the time of Nero, basically represents a trend
towards conservation of the Hellenistic tradition, whereas the Hippocratic
Lexicon of Dioscorides, the editor of Hippocrates during the age of Hadrian,
seems to be influenced by a new perspective. Galens Hippocratic Glossary also
represents a new and independent tool, with its rigorous alphabetical order.
The medical and Hippocratic materials would later be partly incorporated
into the Onomasticon by Pollux (2nd c. AD) and in the Lexicon of Hesychius
(5th6th c.).247
Erotianus work248 has come down to us in a mutilated and drastically revised
version: it has been alphabetized and abridged, but it is nonetheless the main
source for the history of Hippocratic lexicography. In its original structure the
Lexicon presented the glosses in the same order as the flow of the text, thereby
once again bearing witness to the role of glossography as a direct support for
text exegesis. His main source is Bacchius Lexeis, but he also makes use of lexi-
cons of a later date.249 Erotianus Lexicon begins with an extensive introduc-
tion, of an apologetic nature, in which he outlines the history of Hippocratic
glossography (emphasizing the role of grammarians in addition to that of phy-
sicians), and defends Hippocrates against the charge of intentional obscurity.
In this perspective, he places Hippocrates among the ancient authors on a par
with Thucydides and Herodotus.250 His exegetic aim thus focuses on obscure
expressions rather than common words, similarly to the approach that would
later be adopted by Galen in his Glossary, but the linguistic material Erotianus
has gathered together is vast and extremely varied. At the end of the proem
he gives the list of works by Hippocrates, thereby perhaps testifying to the
247 Wellmann [1931] 46; Degani [1995] 505527; most recently Perilli [2006] 174175. In this
volume, see Dubischar 2.2.3 and Tosi 1.1 and 2.1.
248 Ilberg [1893], Nachmanson [1917], Grensemann [1964] and [1968], Jouanna [1989], cf. also
Smith [1979], Manetti [1999b]; Irmer [2007]; Perilli [2008].
249 Wellmann [1931] 29 ff. The fragments of Erotianus edited by Nachmanson, traceable in
scholium form in the Hippocratic mss. M and R., are sometimes of certain Erotianean
ascendancy, such as frr. 8 and 60 (101, 8; 116, 3 N.), but sometimes they are closer to Hesychius
or to one of the sources of the latter: additionally, they sometimes derive from commentaries
by Galen or from other sources (Perilli [2008] 3839): see the case of Metrodorus, cited in
the lemma (fr. 19, 105, 1014 N.), who, if he is the pupil of Sabinus, clearly cannot have
been mentioned by Erotianus (Anastassiou-Irmer [1997] I, 225).
250 Manetti [1999b].
Medicine and Exegesis 1175
Today we know that Galens Glossary is one of his early works.254 It testi-
fies for the first time to the imposition of a rigorous alphabetical order, in the
modern sense, which takes into account not only the initial letter but also the
order of letters in the whole word, and also of subsequent words in the case
of composite lemmata.255 Thus it was not destined to function as an aid to a
reading of the Hippocratic text, but it must instead have been intended for
an independent use, as suggested both by the alphabetical order and by the
absence of references to the Hippocratic places (with the marked percentage
of unlocalizable lemmata).256 Its general character also explains the decision
to include lemmata from works regarded as spurious. Galen draws mechani-
cally on the material he had available, and sometimes derives his exegesis
from broader contexts, e.g. commentaries.257 The redaction is poor from the
linguistic-stylistic point of view and also as regards the content; this has led
to the suggestion that it may have been a rough draft, or a remainder of short
records, like the notes he jotted down upon reading the Empiricists, or pos-
sibly it may have been the product of a collection of short entries drawn up by
another writer to whom Galen had assigned the task, reserving for himself the
composition of the proem and the polishing of the overall text, which he then
never carried out.258 Galen asserts that he will restrict himself to an explana-
tion of the glosses, the latter being taken to signify words that have fallen out
of use, or words reflecting Hippocrates deliberate manipulation of common
usage. The work is explicitly addressed to beginners.259 Here we will mention
254 After the discovery of De indolentia, 35 (Boudon-Millot Jouanna [2010] 12, 1417) in which
it is stated that Teuthras, the intended recipient of the Glossary, died in Rome in the first
great plague (165166 AD), its composition should be dated to the period of Galens first
stay in Rome. On the general characteristics of his manner of doing lexicography, Skoda
[2001].
255 Perilli [1999]; Perilli [2000a]; Ilberg [1888], on the other hand, considered it to be a later
intervention by a scrupulous copyist.
256 Perilli [2006] 176, ascribed the lack of interest in the lemma and in its form to the intention
that it should, precisely, be a tool for general use.
257 The contradictions with his commentaries can be explained because he draws directly
on his sources, such as the encyclopaedic lexicon of Pamphilus (1st c.) or the Lexicon of
Dioscorides: Perilli [2006] 178179.
258 On the readings of the Empiricists, see Deichgrber [1930] 415, 1416. On the mechanical
selection of material derived from Erotianus in the gloss (XIX 107, 14108, 5 K.),
see Perilli [2006] 189192. His mention of this in Libr. propr. is likewise very cursory, and in
fact the work is placed in the appendix to the section on Hippocrates, without awarding
it any special emphasis at all: 9.13, pp. 161, 20162, 3 Boudon-Millot.
259 Gloss. xix 65, 613 K, 67, 1768, 4 K.
Medicine and Exegesis 1177
only en passant the works of a lexicographic nature that Galen devoted to Attic
vocabulary (in alphabetical order), or to the language of Eupolis, or Cratinus
etc., all of which are lost (Libr. propr. 20).
A rather different typology, closer to the thematic collections and the
Onomastica, is found in a work by Rufus of Ephesus (80150 ca.) and in that of
Soranus of Ephesus (second half of 1st c.first half of 2nd c.).260 These are not
lexicons devoted to the exegesis of authors, but technical terminological rep-
ertories destined for use by physicians; however, they do imply a comparative
approach to anatomical terminology, which also takes into account the literary
sources and an ancient branch of knowledge like etymology. In Rufus work On
the names of the bodys parts, which is extant, the layout of the anatomical lexi-
con is organised a capite ad calcem and citations of ancient authors (Homer,
Aristotle, Epicharmus, Empedocles, Sophocles, Zeno the Stoic), are utilized,
certainly by making use of lexicographic sources.261 In contrast, Soranus work
Etymologies of the body of man262 is lost and barely reconstructible, although
numerous fragments are preserved in later Etymologica and other Lexicons.263
After the work of Bacchius of Tanagra, the first mention of an edition of
Hippocrates is a reference to two complete editions, the second of which was
drawn up not long after the first, during the era of Hadrian, namely the edition
of Artemidorus Capito and that of his relative Dioscorides,264 on both of which
we have specific evidence given by Galen. It was noted earlier in this paper
that during the previous centuries innumerable copies of the Hippocratic
texts must have been made, but the general practice, whether it was a ques-
tion of scriptorium or a personal initiative, consisted in producing a copy that
was correct with respect to the model, followed by only occasional checking
of a second manuscript.265 Galen, on the other hand, often speaks of com-
parisons among various different manuscripts, not only in the commentaries
260 For the biography of the two physicians, Hanson Green [1994] 981988; Sideras [1994]
10851088.
261 For ex. the citation of Hom. Od. 9.373374 (141, 612 Daremberg), with the associated
commentary, may have an erudite origin and the citation of Epicharmus (143, 1012)
presupposes recourse to a specific lexicographic tradition. See also the etymology of
the cultual epithet of (229, 13), traced back to the Empedoclean use of
.
262 Hanson Green [1994] 10211023.
263 Also independently in De natura hominis by Meletius, cf. Hanson Green [1994] 10211023.
264 Ilberg [1890], Manetti Roselli [1982] liiilv; Manetti Roselli [1994] 16171633, Roselli
[2012a] and [2012b]. For the complete edition of Hippocrates, see Gal. In Hipp. Off. XVIIIB
631 K; In Epid. VI, CMG V 10.2,2, p. 415, 1721.
265 Cf. supra 3.1.
1178 Manetti
266 Totelin [2009] 8491. On the transmission of Galens pharmacological texts, see von
Staden [1997c] 6671; von Staden [1998b] 8287. In De indolentia 3137 Galen clarifies
the value and the difficulty of the accumulation of materials in this type of text, relating
how he came into possession of ancient parchment codexes of recipes. Scribonius Largus
Comp. 97, speaking of the recipe of Paccius Antiochus, delivered to the emperor in written
form only after Paccius death and thereafter available in the public libraries, confirms
Galens description, underlining the precious nature of such texts.
267 On the composition of works as an exercise, Gal. Libr. propr. 2.4; 3.7; 14, 9 Boudon-Millot:
they include the first commentaries on Hippocrates, On Crises and Difficulties in breathing
(9.1) and the commentaries on Aristotle (14.14).
268 Similarly in De indolentia 21, speaking of the fact that if the fire had broken out two
months later, all his works would have had two copies one in Campania, and the other
sent to his friends in Pergamon for delivery to the public libraries (
), Galen shows that he entrusted the circulation of his works to
personal channels: see, quite recently, Dorandi [2012].
269 This conceptualization was also very useful to him in the distinction between published
and unpublished writings, enabling him to identify and justify the incomplete texts of
Hippocrates, cf. Gurd [2011] 171174.
Medicine and Exegesis 1179
270 Dorandi [2007] 103127, and [2012]; Roselli [2012a]. For Galens use of the difference
between a published work and one that is not for publication within the context of his
specific project of self-characterization, see Gurd [2011], Dorandi [2012].
271 Libr. propr. 1.1 B.-M. However, Gurd [2011] 176180, underlines that it was a question of
correction of mistakes, whereas elsewhere he never admits a veritable work of revision;
the distinction between works composed for publication and those not for publication
was used by Galen to construct the image of his own work as a stable expression of his
doctrine, which he regarded as never having need of revision, thereby revealing a close
affinity with the attitude of the rhetoricians of the Second Sophistic.
272 P. 6, 818 Boudon-Millot Jouanna [2010]
,
,
, ,
.
1180 Manetti
273 Galen uses the term + acc, but in some cases he refers to them with the term
, probably alluding to the exemplars available to him, cf. Roselli [2012a] and
[2012b]. On Artemidorus and Dioscorides, see also Ilberg [1890], Smith [1979] 234140,
Manetti-Roselli [1994] 16171633, Hanson [1998] 4446.
274 Gal. In Hipp. Epid. VI, CMG V 10,2.2, p. 415, 1721; In Hipp. Nat. Hom., CMG V 9,1, p. 13, 1924;
For their textual innovations, cf. for ex. Gal. Gloss. XIX 83, 815 K; In Epid. VI CMG V 10,2,2,
p. 4, 1517; 314, 1824 etc. A more favorable judgment by Galen emerges from the later
commentary on Ar. (Jouanna [1996] 145 n. 278).
275 With regard to what Galen may have meant by ancient readings, cf. infra n. 355. On the
affinity of some reading of Artemidorus and Dioscorides to the parallel edition of Coac.
in comparison to Prorrh., which could demonstrate that their readings originated from
antiquity, in contrast to Galens view, Roselli [2012b] 2526.
276 Cf. the conclusions of Anastassiou Irmer [2001] II 1, 483 and Roselli [2012b] 17.
277 Short versions: Gal. In Hipp. Aph, XVIIIA 5961 K; In Prog. CMG V 9,2, p. 243, 13; 326, 1517;
but omissions are also reported for each of the two. On questions of dialect, see In Epid.
VI, CMG V 10.2.2, pp. 6, 1314; 483, 2830; for grammatical questions In Hipp. Prorrh. CMG V
9.2, p. 122, 2428.
278 Roselli [2012b] 18 nn. 16, 20.
Medicine and Exegesis 1181
the other edition at hand, or possibly both:279 he definitely was able to take a
close look at Dioscorides editionhe describes its physical characteristics
but he only rarely dwells on any physical condition of Artemidorus edition;
consequently, the Artemidorean edition is harder to characterize.280 In only
one case does Galen mention, ambiguously, a particular reading given by
Artemidorus, saying that it was introduced by , but it is not clear if this
was a variant inserted in the margin or included in the text.281 The impres-
sion one derives from Galen is that Artemidorus text was more eccentric as
compared to the standard he would consider to be authoritative. For instance,
several times Galen points out that Artemidorus is the only one whose text
differs from that of all the other witnesses,282 but this naturally is the fruit of
Galens own point of view. In another couple of cases Artemidorus seems to
have endeavored to regularize certain constructions, for instance by recon-
structing narrative nuclei within Epidemics, but elsewhere he appears to be
concerned mainly with producing a reasonably meaningful text.283 In such
cases Artemidorus seems to have drawn up a fair copy which simply contains
the text he had chosen.
Dioscorides edition, on the other hand, offers an example of an erudite
edition, certainly drawing inspiration from the model of the Alexandrian
Homeric editions. The text is equipped with critical signs,284 and is structured
into parts which sometimes bear subtitles; the punctuation is marked in a
279 Roselli [2012b] 23, points out that all the citations concerning the Prorrhetic belong to the
third and last book of the commentary.
280 Probably one should exclude the hypothesis put forward by Smith [1979] 236 n. 82,
who suggested that Galen had mainly used the edition of Artemidorus, which is said to
have incorporated the results of Dioscorides work; rather, what Galen says concerning
the material appearance of Dioscorides edition would appear, if anything, to suggest the
opposite, namely that he mainly used the edition of Dioscorides, who was in the habit
of putting variants in the margin and in one case seems to have annotated precisely the
variant of Artemidorus, see infra and n. 288. But there is no conclusive proof.
281 Gal. In Hipp. Prorrh. CMG V 9, 2, p. 131, 21132, 2. Elsewhere, in the marginal
annotations introduces an explanation, not a variant: see McNamee [2007] 208 (notes
referring to Callim. Coma Berenices), 333 (notes referring to Pind. Pae.).
282 Gal. In Epid. VI, CMG V 10, 2, 2, pp. 176, 19; 309, 13; 395, 39 ff.; 500, 31.
283 In Epid. II, CMG V 10, 1, pp. 158, 4; 233, 20 ff. Alessi [2012] considers it a useful case to
demonstrate that the medieval mss. of Epidemics II depended on the edition of
Artemidorus, but the question remains doubtful.
284 He uses the obelos for expunctions: Gal. In Nat. hom. CMG V 9,1, p. 58, 7; In Epid. VI, CMG
V 10,2,2, p. 283, 19 app. For the features of Alexandrian see Montana 2.4 and
Montanari 1, in this volume.
1182 Manetti
careful and orderly manner, and it has marginal additions, double readings285
and also an indication of accents,286 at least in cases of ambiguity. These char-
acteristics are definitely appropriate for one who cherished the ambition of
being grammatikoteros.287 At times he mentions his sources (a variant found
in two manuscripts) and in one case he seems to have inserted a marginal vari-
ant (which coincides with the text of Artemidorus), introducing it with .288
Taken together, these features prompt the suggestion that Dioscorides edi-
tion was not a text copied ad hoc, like the one mentioned
in On the avoidance of grief;289 rather, one feels that he may have modeled it
on the ancient Homeric editions, building it up through the stratification of
a long-term study, and that it probably constituted a school exemplar, avail-
able for consultation in some library in Rome rather than a copy destined to
the market.
The erudite character of this work by Dioscorides can be perceived even
more clearly from the direct citation of his Lexicon in Galens Glossary or
from its indirect citation in the commentaries. Thus when Galen attributes
to Dioscorides an explanation of his own textual choices, he is very prob-
ably drawing the information from the Lexicon. However, there remain a
few rare cases that suggest the presence of brief annotations by Dioscorides
285 Marginal additions and multiple readings: Gal. In Hipp. Prorrh., CMG V 9, 2, p. 176, 1218;
In Epid. VI, CMG V 10, 2, 2, p. 180, 912; 480, 4043; signs of punctuation: In Epid. VI, 415, 23;
small titles: In Epid. III, CMG V 10,2,1, p. 110, 2.
286 This item of information can be derived from Galens Glossary: it seems evident to me
that where Dioscorides offers a reading, with a specific accent, of an ambiguous lemma
(, ), one should conclude that there was a corresponding sign in his
edition: Gloss. XIX 120, 15121, 2, 148, 89, 154, 9 K.
287 In Hipp. Epid. VI, CMG V 10, 2, 2, p. 83, 1820: an ambition of this kind could be expressed
by Dioscorides only in a preface to his Lexicon, cf. Roselli [2012a] 72 and supra in this
paragraph.
288 In Hipp. Epid. VI, CMG V 10, 2, 2, p. 232, 20 ff., cf. supra n. 280. Another element that
may perhaps be traced back to his diorthosis is an interlinear correction in a passage
of the Prorrhetic. The variant he presents here is as compared to the lemma
, but Galen then specifies that he added supra lineam a lambda between
colons: this was a fairly common sign in the activity of the scriptoria, which restores the
correct text, cf. In Prorrh., CMG V 9, 2, p. 154, 916: see the discussion on the text and the
possible interpretations in Roselli [2012a] 7375, where, however, a scribal correction is
suggested, and Roselli [2012b] 2526.
289 I interpret the expression as a fair copy, not as sur une base saine like Boudon-Millot
Jouanna [2010] 6: see Manetti [2006], Garofalo-Lami [2012] 15, Roselli [2012a] 66.
Medicine and Exegesis 1183
in the margin of the text.290 Among the criteria of his working method, cer-
tain aspects can clearly be noted, such as attention to grammatical problems,
recourse to Homeric exegesis and to literary references (Pindar), and a taste
for rare words.291
290 In Hipp. Epid. VI, CMG V 10, 2, 2, p. 190, 23: in the case of Epid. VI 4.1 he added an before
, saying that he implicitly took it to be or took it as .
This impression seems to be confirmed by the case of the commentary on Epid. II 2,14,
where Galen (In Epid. II, CMG V 10,1, p. 222, 3941) seems to imply that Dioscorides had
a marginal note saying that the passage was in contrast with the previous one and was
not self-sufficient: cf., for a similar case in Ar., Anastassiou Irmer [2001] II 1, pp. 4344.
The hypothesis put forward by Ihm [2002a] 70, who suggests that it may only have been
a case of punctuation, is not convincing. On the problem of whether Artemidorus and
Dioscorides were also commentators, which in my view is to be excluded, cf. also Roselli
[2012b] 17, see Anastassiou Irmer [1997] I xxvi, who refers the reader to Ihm [2002a] 70
and 84. The definition commentator appears only in the Arabic tradition (comm. on Ar.
and Ibn abi Usaibia).
291 See the gloss , XIX 83, 11 K; , XIX 147, 4 K and In Hipp. Prorrh. CMG V
9.2, p. 134, 21.
292 Hanson Green [1994] 10191021.
293 Sideras [1994] 10851088.
294 De ordine libr. suor. 3.11 Boudon-Millot.
1184 Manetti
haps Aiphicianus); later, during his many journeys, he sought out the teachers
of his own teachers, that is to say, the school of Quintus. Quintus was a pupil of
Marinus, a famous anatomist who lived in Alexandria two generations before
Galen and revived anatomical studies after a prolonged period of decline. Galen
read Marinus work on anatomy, of which he composed an epitome, but he
acquired indirect knowledge of at least Marinus commentary on Aphorisms.295
Quintus, Marinus pupil, who was active in Rome, had a fairly large school, but
left no written works.296 Galen explicitly mentions that he sought as teachers
the most famous pupils of Quintus: Numisianus, Satyrus, Pelops, Aiphicianus,
Lycus,297 endeavoring to become personally acquainted with them or at least
with their works.298 The picture that emerges from these passages in Galen
reveals that much of their exegesis was oral, sometimes set down in the form
of notes or short records by pupils for subsequent transmission, or occasion-
ally written down by the teacher and made available to the pupils themselves.299
The difficulty involved in the circulation of these few written texts, and in actu-
ally locating them, was extreme. The basic infrastructure underlying Galens
Hippocratic exegesis resided in this store of materials he had accumulated over
the years, which derived from a school-room exegesis. This approach tended
to adapt to different doctrinal orientations300 and generally corrected the text
at hand wherever the latter diverged from contemporary medical knowledge,
though it did not disregard the techniques of literary exegesis (choice of vari-
ants, comparison with other Hippocratic texts, use of grammatical or rhetori-
cal concepts, etc.).301
Rufus of Ephesus,302 as far as we know, wrote commentaries on Prorrhetic,
Epidemics II and VI, Aphorisms303 and Airs waters places.304 Galen held him in
good esteem, judging that among the recent commentators Rufus was one who
tended to preserve the ancient reading.305 In actual fact, it would seem from
the cases cited by Galen that Rufus was a very learned commentator who made
use of the commentaries drawn up by the Empiricists (Zeuxis for example)
and also of ancient doxographic sources.306 Rufus carefully mentions the vari-
ants, discussing them with a view to establishing their medical significance,
and does not hesitate to make corrections, displaying considerable ability in
making conjectures.307
Sabinus,308 the teacher of Stratonicus of Pergamon, was active in Alexandria;
he appears to have been the most Hippocratic of the commentators, in that
he shares some of the fundamental doctrines of Hippocrates such as the
humoral theory and that of apostasis.309 His work is thus an internal exegesis,
which seeks to give an orthodox interpretation within a teleological vision of
which the Prorrhetic and Coan prenotions), without ordering them into a hierarchy of the
most authentic, in contrast to the practice generally adopted by Galen.
301 Manetti Roselli [1994] see n. 299: in particular 1592
302 Manetti Roselli [1994] 16001606; Sideras [1994], Ullmann [1994].
303 For the discussion of the commentary Rufus wrote on Aph., see Sideras [1994] 10771253,
Ullmann [1994] 1306; most recently, Fischer [2002) 311313 (an observation of Rufus on
Aph. 4.37 contained in Lat. A).
304 The commentary by Galen on Ar. confirms the existence of a commentary by Rufus
(Anastassiou Irmer [2001] II 1, 30), denied by Sideras [1994] 1009 n. 154 (with previous
bibliography).
305 Gal. In Hipp. Prorrh. CMG V 9, 2, p. 73, 10. In Hipp. Epid. VI CMG V 10,2,2, p. 174, 12 etc.
306 Gal. In Hipp. Epid. VI CMG V 10, 2, 2, p. 122, 7 ff., cf. 119, 12 ff.
307 See Gal. In Hipp. Prorrh. CMG V 9, 2, p. 73, 720, where in the commentary on Prorrh. I 59a
he corrects the transmitted reading by changing it into , thus
showing consideration of the similarities in the writing: the text is incongruent since it
defines cooked urine as a negative symptom, whereas this cannot be accepted, given that
coction is considered to be a positive process in physiological processes.
308 Manetti Roselli [1994] 16071614.
309 On the self-styled Hippocrateans, Lloyd [1993] 401 n. 14. The list of Hippocratean
commentators must include the author of a commentary on Oath, known from Arabic
sources, ascribed to Galen (see infra n. 346): if Nutton [2012] is right in his suggestion that
it could also be a work by a contemporary of Galen, like Satyrus, then it should be dated
between 70 AD and 250 AD.
1186 Manetti
310 Galen does not cite Sabinus in the commentary on Aph., but it can be surmised from
his polemical tract Adversus Iulianum, CMG V 10, 3, p. 39, 1240, 4, that Julianus wrote
polemically against Sabinus exegesis of Aphorisms. Also Steph. Ath., In Hipp. Aph. I 1, CMG
XI 1.3.1, p. 30, 11 ff., says that Sabinus considered Aphorisms to be authentic. Apparently a
new fragment of Sabinus comes from Galens commentary on Ar. (Anastassiou-Irmer
[2001] II 1, 48 n. 1) and concerns the explanation of kedmata (Ar. 22), which, nevertheless,
could also derive from Sabinus commentary on Epid. VI (Epid. VI 5.15).
311 NA III 16.
312 See the passages analyzed in Manetti-Roselli [1994] 16081609.
313 In Hipp. Epid. III CMG V 10, 2, 1, p. p. 25, 49; In Epid. VI CMG V 10, 2, 2, p. 137, 21 ff.; In Nat.
hom. CMG V 9, 1, p. 15, 17 ff.
314 In Nat. hom. CMG V 9,1, p. 87, 1888, 11.
Medicine and Exegesis 1187
corrected them, exploiting the vast wealth of materials available in the capi-
tals public or private libraries, and how he devoted attention to conserving
and reproducing his own works.315 Today we have a clearer idea of the day-by-
day working routine that lay behind his literary production, which was impres-
sively wide-ranging.316 In On my own books he mentions many works devoted
to exegesis, not only of Hippocrates but also of Erasistratus, Asclepiades of
Bithynia, Plato, Aristotle, Chrysippus, Epicurus and of schools such as the
Empiricists, but it is not always clear what type of writings were involved. Some
of his production consisted of epitomes and synopsis of works by authors he
was studying, according to a practice that was common as part of the train-
ing of all intellectuals of his era, and which also encompassed the activity of
writing commentaries as a personal exercise.317 The transition from writing
as a study activity ( , ), for personal use, to that destined
to publication ( ), with regard to writings on Hippocrates,
arose later, and sprang from polemical necessities, to counter other interpre-
tations.318 Similar reasons certainly lay at the root of Galens commentary in
three books on the first book of On fevers by Erasistratus, as well as the com-
mentary on the treatise On the pulse by Archigenes and that on the Introduction
and the Kefalaia of the empiricist Theodas.319 Their specific form is not always
clear but with regard to the text on Archigenes it is explicitly said that it con-
tained both exegesis and assessment ( ). Similarly, with
regard to the writings on authors such as Herophilus and Asclepiades we may
surmise that they were polemical treatments, or at best external exegetical
endeavours.320 It should be underlined that no trace remains of exegetic
321 See Gal. Diff. puls. 4.3 (VIII 724, 15 K.) and von Staden [1999a] 176. The following works
can be defined as syngrammata, likewise carrying out an exegesis of the Hippocratic
texts: On the elements according to Hippocrates, On Hippocrates anatomy, Difficulties in
breathing, Against Lycus, Against Iulianus, On coma according to Hippocrates, On crises,
On critical days, On regimen in acute diseases according to Hippocrates, and also That
Hippocrates claims the same opinion in his other books as in On the nature of man (Libr.
propr. 9.12 Boudon-Millot).
322 On Galens commentaries, Smith [1979] 123176; Manuli [1983]; Lloyd [1993]; Lpez Frez
[1992]; Potter [1993]; Debru [1994]; Manetti Roselli [1994]; Mansfeld [1994] 148176;
Sluiter [1995]; von Staden [1995]; Manetti [1998] 1209 ff.; Ferrari [1998]; Hanson [1998];
Sluiter [1999]; Vallance [1999]; von Staden [2002]; Ihm [2002a] and [2002b]; Strohmaier
[2002]; Manetti [2003]; Flemming [2008]; Manetti [2009]; von Staden [2009].
323 Von Staden [1998b] 7273, Flemming [2008] 324332.
324 On the presentation of Hippocrates and Plato as generally agreeing and the problems
involved, see Manuli [1983]; Lloyd [1993] 407 ff.
Medicine and Exegesis 1189
325 Galen asserts that although numerous writers addressed the respiratory function, none
dealt with anomalous breathing, even though Hippocrates wrote important things in
Epidemics: cf. VII 764, 11 ff. K. ,
.
,
... ... ,
.
326 VII 825, 3826, 4 K.
... ,
,
, , ,
, . Cf. on
also Ars, I 305, 58 K., where Galen lists a series of words synonymous with
, alluding to ancient uses.
327 The tradition current at the time was that of creative reinterpretation, which cannot
be defined as a specific feature of Galens exegesis, cf. Lloyd [1993] and Mansfeld [1994]
155161: on this, already Manuli [1983]; Ferrari [1998]. On the adjusting of Hippocratic
thought to the contemporary opinion of his day in order to demonstrate the permanence
of Hippocrates truth, see von Staden [2002] in particular 126 ff.; Flemming [2008]. For
the same reason it is an oversimplification to see Galens Hippocratism merely as a
defensive weapon in his polemics with his rivals (Lloyd [1993] 412) or as rhetorical gloss
or ideological patina (Smith [1979] 175).
328 Erotian. 29, 10 ff.; 33, 14 ff.; 34, 13 N. The affinity was noted by Mansfeld [1994] 149150. On
the use of the verb in Galens commentaries, see Elem. I 484, 24 K.; In Hipp. Aph. XVIIB
677, 1315 K. and von Staden [2002] 118 and n. 31.
1190 Manetti
and seems to share the same idea of exegesis. As Galen explains in the same
passage, exegesis includes a verdict on the truth of the doctrines but not the
demonstration, which he has concentrated into the first book.
It may be helpful to cite a passage from De tremore in which Galen, citing
Ti. 85d where he points out a mistake made by Plato, comments: it is not
opportune now to investigate whether what Plato said is true, because I have
the intention to explain and give a judgment on what is said in the Timaeus
in another work.329 The treatise is not the place for discussion on the truth
of Platos doctrines, and Galen refers the reader to the commentary on the
text in question. The Commentary on the Timaeus of Plato, which is the first
of the lemmatic commentaries written by Galen, is therefore programmati-
cally devoted also to a verification of Platoss medical doctrines.330 These texts
precede the season of Hippocratic exegesis, which begins with Difficulties in
Breathing,331 written in a mixed form, as we have seen, and is then developed
with continuous commentaries. His decision to embark on a systematic com-
mentary of the Hippocratic treatises was also the natural outcome of a career
that was now solidly established on the professional plane. Having reached the
height of his career, Galen could take a broader perspective on his teaching
activity, which had so far probably been limited to oral work332 as a continua-
tion of the activity of his own teachers. Moreover, the society of his time pos-
sessed prestigious models of commentaries of the great authors of the past
in the literary and above all philosophical field; accordingly, the writing of
commentaries represented the crowning achievement of his intellectual stat-
ure as a physician-philosopher.333
334 Gal. In Hipp. Fract. XVIIIB 318, 1322, 2 K. The passage in In Hipp. Aph. XVIIB 561, 4562,
10 K., where Galen seems to attribute to the commentary the additional task of the
demonstration can actually be seen as referring to a specific polemic against Lycus and
does not contradict the position expressed here. On the variety of forms, see the overview
in Sluiter [2000a].
335 Mentioning his work On exegesis, he addresses the definition of obscurity, which is both
a question in its own right and also stands in relation to the user of the ancient text
(Manetti-Roselli [1994] 15571559). The classification of the different types of obscurity
was already a subject treated by philosophical exegetic tradition: on obscurity and its
forms, see Cic. Fin. 2.15 in Ferrari [1998] 1718. On this theme, Manuli [1983] 472; Manetti
Roselli [1994] 1532, 1558, 1607; Mansfeld [1994] 135, 148154, von Staden [2002] 110114.
336 Manetti Roselli [1994] 1559; Manetti [1998] 1211.
337 Manetti Roselli [1994] 1533.
1192 Manetti
338 Gal. Diff. resp. VII 850852 K.; In Hipp. Art. XVIIIA 729, 18 K.; In Epid. VI, CMG V10,2,2,
p. 141; De comate, CMG V 9,2, p. 188. See Manetti Roselli [1994], Manetti [1998]; Sluiter
[1995]; von Staden [2002].
339 For the historical details, see for. ex. In Hip. Epid. VI, CMG V 10, 2, 2, p. 177, 1216. On the use
of etymologies, which Galen programmatically rejects (In Art. XVIIIA 395, 3 ff.; In Fract.
XVIIIB 363, 3 ff. K.), but practices to a certain extent in the commentaries, see Manetti
[2003] 202215.
340 See for ex. In Epid. II, CMG V 10, 1, 1, p. 230, 1219: the text is so corrupt that Galen feels
he should record and explain the most ancient readings locatable in the previous
commentaries of Zeuxis, Herakleides, Bacchius and Glaukias.
341 For ex. In Epid. I, CMG V 10, 1, p. 116; Placit. IX 1.15, CMG V 4,1,2, p. 542, 2527.
342 It is remarkable that in the commentary on Prorrhetic, which he considers spurious, Galen
is polemic against those commentators who maintain that exegesis must be confined to
justifying an authors text, although patently false: In Prorrh. CMG V 9, 2, p. 52.
343 Manetti Rosellli [1994]; Sluiter [1995]; von Staden [2002]; Manetti [2003] and [2009];
Nutton [2009].
344 Cf. supra n. 139: caution should be exercised in using this term, but Galen is the author who
came closest to a reflection on the genres of the secondary literature, in his reflection on
the function and the reading public of his commentaries and his treatises.
Medicine and Exegesis 1193
at his disposal and Galen himself had written many works on the language
of ancient comedy. More and more frequently he displayed a keen interest in
manuscript variants and in interpolations, as well as in questions of authentic-
ity and authorship.353 He was aware of the limits of an interpretation that has a
stochastic character and cannot emend all the corrupted elements;354 this not-
withstanding, he proposed several explanatory models for the process of text
composition and transmission, drawing extensively on the scriptorial practice
of his day. In this framework, Galen put forward interesting reflections on the
typology of the errors he found himself dealing with, and he came close to for-
mulating a criterion like that of the lectio difficilior. In his later commentaries
he became respectful of the tradition he characterized as ancient, to the point
of opting to preserve ancient readings even if they were apithanoi.355
reflected the culture of his time; thus the suggestion that he had little awareness of
diachronicity would seem to be an oversimplification, Manuli [1983] 475.
353 This is influenced to a large extent by the quality of the texts on which the commentary
is being composed: thus from the fourth section onwards of the commentary on Aph.,
emphasis is placed on philological problems, as in the commentaries on the difficult texts
of Epid. II and VI, whereas in In Hipp. Epid. I some passages are not even mentioned
or supplied with a commentary because they are considered completely clear, and
explanations are restricted to the obscure expressions. Finally, it is quite natural that
Galen awards considerable attention to the issue of the authorship of Nat. hom. and to its
spurious parts, given the ideological importance of this text in his vision.
354 In Hipp. Off. XVIIIB 715 K; In Epid. II, CMG V 10, 1, p. 221, 9 ff; 275, 41 ff.: see Lpez Frez
[1992].
355 On the lectio difficilior, see In Epid. VI, CMG V 10, 2, 2, p. 121, 1722; Hanson [1998] 48; Ferreri
[2005]; also, the still useful collection of material in Brcker [1885]. Over time Galen was
able to form an idea in his own mind of what the ancient tradition was like, basing his
interpretation essentially on the Empiricist commentaries and on those of Rufus, see
Manetti-Roselli [1994] 16331635: on nevertheless maintaining the ancient variants, In
Epid. VI, CMG V 10, 2, 2, p. 121, 12 ff.
356 Suid. s.v.: , also known by the alternative title
.
357 Kind [1927] 11161117; Hanson Green [1994].
1196 Manetti
All that can be read is a Vita Hippocratis ascribed to Soranus in some man-
uscripts of the Corpus Hippocraticum, which is probably a very mutilated
excerpt from Soranus fuller work.358 Contrary to the general title of Soranus
work, the Vita Hippocratis makes no reference to Hippocrates writings or doc-
trines, if not in a skeptical tone,359 well suited to the detached manner in which
Soranus generally refers to Hippocrates. Soranus was certainly a well educated
physician with an interest in grammar (cf. supra), but the Methodical school,
of which he was the representative, had rejected en bloc the long tradition of
Dogmatic medicine. Therefore it is in a sense paradoxical that Soranus devoted
attention to the history of medicine, biography and doxography. This may well
have been an instrument of self-assertion, even if there is too little evidence
for it to be judged as purely an exercise in confutation (a hamartography).360
Confirmation of a polemical attitude can be gained from the fragments of
the treatise On the Soul, in 4 books, quoted extensively by Tertullian, which
critically reviewed the opinions of philosophers.361 Soranus Aitiologoumena
is referred to by Caelius Aurelianus362 as a work on the causes of diseases,
perhaps a broad-based doxographical exposition, but it remains dubious, and
nothing precise can be said about its nature.363 The interest in biography, on
the other hand, was consistent with the culture of the 1st2nd century AD,
when antiquarianism was widely practiced.
A later source, but difficult to date, is the so-called Anonymus Parisinus
and his work On acute and chronic diseases.364 He writes about each disease
358 It was included by J. Ilberg in his edition of Soranus, CMG IV 1927, 173178: there is also a life
in Suidas ( 564), a garbled Latin version (Schne [1903]) and a verse biography composed
by Johannes Tzetzes in his Chiliades (VII 986). All of these are likely to rely to some extent
on the Hippocratic Bios by Soranus of Ephesus, according to Deichgrber [1933] 147 and
recently Pinault [1992]; Hanson-Green [1994] 10101018 (see former bibliography): contra
Edelstein [1935] 12941295. Suidas and the Latin life also contain the list of Hippocratic
writings.
359 13, CMG IV, p. 177, 1925.
360 See Mansfeld [1994] 180; van der Eijk [1999c] 397452, in particular 448; hamartography
in Smith [1979] 224225: biography and lexicography acted as a preparation for the study
of the real opinions and writings (in order to refute them).
361 See the new edition by Podolak [2010].
362 Tard. I iii 55 (= CML VI 1.1, p. 460, 1720); see Wellmann [1901] 140155.
363 Hanson-Green [1994] 10341035. The numerous doxographic traces scattered throughout
the Gynaecia treatise and in the work of Caelius Aurelianus point to the existence of a
structured doxography: Cael. Aur. Cel. I xiv 105xvi 165; II xxix 225xl 234, Tard. II i 5562,
see van der Eijk [1999c].
364 Date uncertain: terminus post quem suggested by the quotation of Mnaseas (4060
AD). Edited by Garofalo [1997]; on the history of the criticism, see Garofalo [1992]. The
Medicine and Exegesis 1197
according to the same pattern: first the causes, then its symptoms and finally its
treatment. The section on causes often contains doxographical reports of only
four ancient physicians, Hippocrates, Diocles, Praxagoras and Erasistratus, all
representative of the Dogmatist tradition, sometimes referred to collectively
as the ancients. The author reports their views in a non-evaluative manner:
His attitude seems, on the whole, reconciliatory rather than divisive.365 His
reluctance to commit himself to the causal explanations of the ancients might
be seen as consistent with a Methodist attitude, combined with a didactic pur-
pose. Anonymous sources are unknown.366
authorship has been variously attributed, now to Herodotus the physician belonging to
the Pneumatic school, now to Soranus, now to Themison: see van der Eijk [1999b] 295 ff.,
with the previous bibliography.
365 Van der Eijk [1999b] 314; for the conclusions 325331.
366 It is possible that he relies on a medico-doxographical tradition with regard to the
question of the so-called affected parts, see van der Eijk [1999b] 32223. Further traces
of a doxographic tradition on Hippocrates in Stobaeus (Jouanna [2010]) and in papyri
(Marganne [2010]). See also, for the doxographic elements in Celsus, von Staden [1999b].
367 On the presence of Hippocratic material in Hesychius, via Diogenianus, and the influence
of the lexicons of Rufus and Soranus on Pollux, see Perilli [2006].
1198 Manetti
Another development in the 4th century that reveals the changing perspec-
tive is the work of Oribasius, who had been trained in Alexandria and then
became the personal physician of the emperor Julian. Oribasius constructed
a great medical encyclopedia, only partially preserved, which was based on
the technique of selecting passages from previous authors and then reorganiz-
ing them by theme.368 Thus paraphrases and summaries played an increas-
ingly significant role, where more or less literal citations tended to merge with
reworked and interpretive segments, without any apparent solution of con-
tinuity, and were often preferred because, in particular, Galens works were
too complex and disorganized, too lengthy and physically unwieldy, for the
purpose of the average educated man, let alone the busy, peripatetic, medical
practitioner.369 Indeed, even Galen himself had begun not only to compose
summaries of his own works as well as those of others, but also to write intro-
ductory texts, condensing more complex treatises.370 Finally the scholia, that
is to say, the texts transcribed for exegetic purposes as an accompaniment to a
work by Hippocrates or Galen, often occupying the margins of a codex, were
formulated, at least in the case of Galen, according to the same technique as
adopted by Oribasius in utilizing his sources, i.e. by using passages taken from
the source author, sometimes simply reordered and juxtaposed, as a means of
explaining and completing the annotated text.
From the death of Galen in the early decades of the 3rd century, up to
the end of the 5th century, surviving evidence of medical exegesis is very
scanty, and even the papyrus witnesses from Egypt concerning Hippocrates
and Galen offer few traces of philological work or commentaries. The papy-
rus codex PRyl. 530, dating from the 3rd4th century, has been revealed by a
recent revision to contain a copy of Aphorisms, interspersed with paraphrase-
commentaries.371 This evidence is of particular value because its editorial
characteristics suggest that it could be defined as a commented edition:372 for
instance, the text is clearly separated from the commentary through signs and
indentations and there is, as far as can be perceived from a reading of the text,
368 On the role of Oribasius in the development of Galenic Hippocratism, see Temkin [1932]
38: cf. Orib. Coll. I 1, CMG VI 1.1, p. 4, 1518.
369 Lieber [1981] 170.
370 Garofalo [2000b] 1415, for the synopsis of De methodo medendi; Galen presents
De musculorum dissectione as an abreg of his larger anatomical work (Libr. propr. 4.1
B.-M.). Galen was also in the habit of writing summaries of other authors texts (for ex.
Marinus Anatomy, Libr. propr. 4.9, of Lycus, 4.34; of Heraclides, 12.3, of Platos dialogues,
16.2); for Platos dialogues see Kraus-Walzer [1951], see supra n. 317.
371 D. Manetti-R. Luiselli in CPF [2008] 180197.
372 According to the definition of Montanari [2006a] 1114.
Medicine and Exegesis 1199
an apparent quantitative balance between the authorial text and the interpre-
tive paraphrase. The relationship between the text and the commentarythe
latter being simple and of a paraphrastic nature, independent of Galenis
one of strong integration, with a manner of realization that foreshadows phe-
nomena attested at a later date. Furthermore, this testimony is valuable in that
it fills a gap in our sources, which otherwise provide us with only two names for
this period: Philagrius and Magnus of Nisibis.373
Philagrius,374 who is known mainly from Suidas and from the excerpta
of Oribasius, was active between the 3rd and 4th century in Thessalonica;
and, in addition to many other works, he also composed a commentary on
Hippocrates, though nothing is known of its nature. Magnus of Nisibis,375 a
pupil of Zeno of Cyprus together with Oribasius around 370, was iatrosophistes,
i.e. a professor of medicine; he was active in Alexandria, where he gained con-
siderable popularity and prestige on account of his dialectical powers, but he
faced accusations of lack of professional competence. He composed a com-
mentary on Aphorisms, which is cited a few times by Cassius Felix.376
A page of Hippocrates Aphorisms is also preserved on a rather well made
parchment codex that has come down to us, PAnt. 28, thought to date from
the 5th century, and characterized by the presence of small sub-titles in the
margin, which may, albeit minimally, be suggestive of its possible use as teach-
ing material.377 It is not until the 6th century, with PAnt 183,378 that one finds a
copy of Aphorisms supplied with marginal scholia derived from different mate-
rials, revealing contacts with the exegetic activity of the school of Alexandria
which was flourishing during that same period. In fact Alexandria became the
main centre of philosophical, but also grammatical and medical studies from
roughly the end of the 5th c. onwards when, after the death of Proclus (485),
the Athens school faced a severe crisis.
During this period there arose a gradual unifying tendency among the higher
education systems of philosophy, grammar, rhetoric and medicine.379 In all
disciplines, as a result of the influence of the philosophical teachings of the
school of Ammonius, educational activity became systematized and was orga-
nized according to precise course plans. The outcome was an exegetic method
well adapted to classroom practice, profoundly influenced by Aristotelian
logic, as testified by the use of the syllogistic method in the development of
arguments (to clarify textual difficulties), the differentiation between sub-
stance and accident, recourse to the four philosophical causes forming part
of the Aristotelian framework, the / procedure of commenting, the
(division of a subject into a series of subdivisions and complementary
definitions).380
As far as the field of medicine was concerned, the classics were read and
commented in a certain order, and this was the case both for Hippocrates and
Galen.381 There was a flowering of introductions to the works, which organize
the discussion according to eight traditional points:382 the subject of the book,
its usefulness, its authenticity, its title, its place in the curriculum, subdivision,
the branch to which it belongs (physiology, etiology or diagnostics), the teach-
ing method.
The construction of a curriculum should not be taken as implying a fixed
and exclusive Canon, even though it is commonly referred to by this term.
It arose from the practical requirements of teaching and was a process that
developed from the early 5th century onwards, continuing to evolve further
until the 7th century, and even beyond this period in the Arabic environment.
Among the works of Hippocrates and Galen, those chosen for educational
379 That medicine was studied on the higher levels together with other disciplines such as
grammar, rhetoric, logic, philosophy, etc., is a recurrent phenomenon in the centuries
immediately following Galen as well: Marasco [2010]. Cf. Lamberz [1987] 120.
380 For ex. Westerink [1964]; Duffy [1984]; Wolska-Conus [1992]; Rouech [1999]; Ieraci Bio
[2003]: a general overview in Pormam [2010].
381 This phenomenon is parallel to the formation, in the school of Ammonius, of a
philosophical curriculum based on an ordered selection of works: in order to reach the
level at which Plato was studied the program began with Porphyrys Isagoge and then
continued with the works of Aristotles Organon; see Hadot [1987] 120122.
382 On the development of the prolegomena to the authors, in particular of the octo capitula,
see Richard [1950]; Mansfeld [1994] (the first three chapters).
Medicine and Exegesis 1201
purposes were texts considered (by the Alexandrians, not always by Galen)383
to be appropriate for beginners. The works were then organized into groups
that were divided by degree of background knowledge required, and were to be
read in class, in a specified order under the guidance of the teacher.384 Works
that were not included in the so-called Canon were by no means excluded, but
it was left up to the individual students to peruse them in greater depth sub-
sequently. Our most authoritative source is unains Risala (Report),385 which
describes Alexandrian educational practice.
I list here below the structure of the curriculum of Galenic studies:386
383 While De sectis is the first work also recommended by Galen (De ordine libr. suor. 2.4 p. 92,
7 ff. Boudon-Millot) for those who do not have have a background in philosophy and
dialectics, for the others the order suggested by Galen does not coincide precisely with
that of the Canon although the latter is clearly inspired by Galen.
384 It should be underlined that all the texts that have come down to us reveal a strong link
with oral teaching, and even their title often makes it clear that they derive from notes
prepared for an oral lesson (for the formula cf. Richard [1950]).
385 The text of Hunain ibn Ishaq, who was the head of a school of translators in Baghdad in
the 9th century, is available in a German translation in Bergstrsser [1925].
386 For the variations in the different Syriac and Arabic sources, see Iskandar [1976]; Lieber
[1981] in particular 173; Strohmaier [1994]; Boudon-Millot [2007a] cxivcxxvi. The Canon
of the 16 works is described by Iohannes Grammatikos (see the following note).
387 Iohannes Grammatikos (see infra) in the Prologue of his Overview of the works of Galens
Canon (transmitted only in Arabic) puts the Anatomy in sixth place after Nat. fac. in
agreement with other Arabic sources, perhaps correctly, cf. Garofalo [2000a] 146 n. 36;
Garofalo [2003b] 207208.
388 Lieber [1981] 174: According to Hunayn...the Alexandrians made a great mistake in
limiting their reading to the first book of each section.
389 De sanitate tuenda was subsequently added: Iskandar [1976]; Lieber [1981].
1202 Manetti
390 Sergius introduced, it would appear, some modifications in the Canon. Stephanus of
Athens (infra) cites the 16 works, cf. Westerink [1985] viiviii, Wolska-Conus [1994]
42 n. 34. For Cassius Felix, see supra n. 376.
391 Ali ibn Ridwan in Lieber [1981] 172174; for the variants of the Hippocratic Canon, which
reached a total of 12 works, see nn. 393395.
392 Cf. Ioh. Alex. In Gal. De sectis, 2ra4849 Pritchet [1982], who asserts that Hippocrates
is too difficult for beginners; Temkin [1932] 3234, points out that the rise of the pre-
emiment role of Galen can be dated as early as Oribasius and his medical encyclopedia;
cf. Temkin [1973] 62 ff. In contrast, on the hypothesis of a western Hippocratism free from
the influence exerted by Galen, see Beccaria [1959] and [1961].
393 Iskandar [1976] 235238, 249. Ibn abi Usaibia cites the same works in a different order,
adding De morbis muliebribus and Officina medici; cf. Duffy [1997] 911. On the variants of
the Canon of Hippocrates, see Irmer [1987].
394 Palladius In Hipp. Fract. 18, 520, 5 Irmer and Steph. Ath. In Hipp. Prog. CMG XI 1.2, p. 30,
3132, 37: cf. Anastassiou-Irmer [2012] III, 439440, but see also the following pages (441
457) for all the lists of Hippocratic writings. See first of all the reconstruction in Iskandar
[1976] 235258. Cf. also Brutigam [1908] 4344.
395 On the order that can be reconstructed from the commentary by Stephanus on Aph., see
Westerink [1992] 1112; Brutigam [1908] 4344: Iusiurandum, Aphorismi, Prognostica,
Medicine and Exegesis 1203
402 Wolska-Conus [1989] 5054 and [1996] 4748. In Pal. lat. 1090 commentaries on De sectis
and on Ars medica are attributed to Gesius. Additionally, a commentary by Gesius
on De temperamentis has been reconstructed (Garofalo [2003b] 205, with previous
bibliography).
403 In Hipp. Aph. II 53, CMG XI 1.3.1, p. 256, 38.
404 See most recenty Manetti [1995], with previous bibliograhy.
405 Baffioni [1954] and [1958]; for a comparison among the commentaries on De sectis,
Manetti [1992].
406 Temkin [1935] 405414; Manetti [1992] 216224.
407 Dickson [1998].
408 Pritchet [1982]; for the commentary on De sectis see also Palmieri [1989] 3442; Goel-
Mayer-Staub [2000] 201222. In the Latin commentary on De sectis, the proem, which
seems to be independent of the commentary itself, is attributed to him in only one
manuscript. On the problems of identification of Iohannes Alexandrinus, see Garofalo
[1999] 189193.
409 Today, this still remains available only in Dietz [1834], II, 1204.
410 Edited by Irmer [1977]; on the relations between parallel redactions of Palladius and
Stephanus, Irmer [1975].
Medicine and Exegesis 1205
411 Magdelaine [2003]; see also Duffy [1997] 9, n. 4: the commentary has been reconstructed
partly also through the citations of Aph. by Palladius himself in the commentary on
Epid. VI. For the reconstruction, these citations were examined comparatively with the
commentary on Aph. of Stephanus, who quite probably used Palladius systematically,
under the name of Galen or of the recent commentator: see Wolska-Conus [2000].
A probable testimony of Palladius commentary on Prognostic is identified in the
summary of this text forming part of a work by al-Yaqubi, see Overwien [2011].
412 He has been identified, by Wolska-Conus [1989], with the philosopher Stephanus of
Alexandria; but see also Rouech [1990]; Westerink [19982] 1923.
413 The commentary on Aphorisms is edited by Westerink [19851995]; for the commentary
on Fract., see Irmer [1977], for the commentary on Prognostic, Duffy [1983].
414 Duffy [1997] 13: the Greek excerpta were inserted as marginalia in the text of the Ephodia
in the codex Vat. gr. 300 (12th c.), written in the Reggio Calabria area; Pritchet [1975] for
the Latin version. In the commentary on Epid. VI Iohannes mentions many commentaries
he had composed on other works of the Canon (Brutigam [1908] 51).
415 The commentary on Nat. puer., edited for the first time by Dietz [1834] II, 205235, was
re-edited by Duffy [1997].
416 Fragments of anonymous commentaries on Epid. VI in Duffy [1997] 119125, in Roselli
[1999] and Ieraci Bio [2012]. In the work of Ali ibn Ridwan, segments of two commentaries
on Mul. 111 have been identified, one falsely attributed to Galen, the other attributed to
an Asclepius, Ullmann [1977]; lastly, the fragment of the commentary on Praecepta, in the
scholium of ms. vat. Urb. gr. 68, was analyzed by Brutigam [1908] 54 ff.
417 However, it is not present in the commentary by Stephanus on Gal. Ad Glauconem.
1206 Manetti
cism include discussion, in the first lecture, of the octo capitula (see above),
the frequent repetition of basic information (anatomical, physiological, thera-
peutical), the tendency to using diairesis418 and the use of the problem and
solution approach (-).
For all the authors mentioned in this survey, who essentially survive only
in their works (or in fragments, like Archelaus),419 it is extremely difficult
to reconstruct the outline of a biography and a corresponding chronology.420
Palladius and Stephanus certainly taught in Alexandria,421 just as the name
Iohannes Alexandrinus leads back to the same city. But in all other respects
their figures are evanescent. The relative chronology would appear to suggest
that Palladius was earlier than Iohannes, and that both preceded Stephanus.422
The personality of Stephanus of Athens is somewhat more clearly defined: dat-
ing from between 550 and 650, he is often defined as a Philosopher and identi-
fied, not without controversy, with Stephanus of Alexandria, a commentator of
two of Aristotles works.423 If the identification is accepted, Stephanus,424 born
in Athens, would have arrived in Alexandria between 567 and 572.425 He cites
Asclepius (twice by name, several times as the commentator of this work),
418 On the diairesis, Duffy [1984]; Mansfeld [1992] 326331; Ieraci Bio [2003] 1113; Ieraci Bio
[2007].
419 For the discussion of a series of hypotheses, see, most recently, Manetti [1992] and [1995].
420 Brutigam [1908] 3546, dates them all to the period 550650 AD, but on the basis of
stylistic features he considers Palladius to be earlier than Iohannes. An internal element
helpful in dating Iohannes is his allusion to his teacher as Triseudemon maximus noster
sophista, who is likely to be identifiable with Gesius (Duffy [1997] 12, and Kessel [2012b]):
this would date him to the first half of the 6th c.
421 Brutigam [1908], 36
422 Brutigam [1908] 38; Wolska-Conus [1989] 82 ff.; Irmer [1977].
423 Wolska-Conus [1989], who also identifies Stephanus with Pseudo-Elias, the author of
the Prolegomena philosophiae. The argument put forward by Wolska-Conus is shared by
Rouech [1990] 108128, although Rouech believes that some doubts still remain.
424 I set aside the commentary and the figure of Theophilus, who, according to Westerink
[1985] 1719, is probably one of the revisors of the work of Stephanus: he is identified
with the Theophilus who was the addressee of some letters written by Photius, designated
as Protospatharius, datable to the 9th c. With regard to Theophilus commentary see
Magdelaine [1988] 273284. However, the chronological relation between Stephanus and
Theophilus remains controversial, cf. Wolska-Conus [1994] (T. dated to the 9th10th c.),
Lamagna [2003] 6768. The commentary attributed to Damascius in some manuscripts
has been shown to be a late anonymous abridgement of Galens commentary, see
Magdelaine [1996] 289306.
425 Wolska-Conus [1989] 589: however, this identification is contested by Lautner [1992]
519522.
Medicine and Exegesis 1207
and one may presume that Asclepius was his tutor.426 It may be possible,
through Stephanus citations, to gain anextremely cautiousidea of the
approach probably adopted in his commentary. Asclepius is defined as the
commentator of Hippocrates, who explains Hippocrates from Hippocrates427
and he is often set in opposition to Galen by Stephanus, but this is not suffi-
cient to define him as the representative of an exegetic strand that was a rival
of the more common Galenic Hippocratism, since the same principle (the law
of exegesis is to interpret Hippocrates with Hippocrates) was stated explic-
itly several times by Galen himself.428 Iohannes Alexandrinus and Stephanus429
sketch three or four points that summarize the aims of exegesis: 1) textual
clarity; 2) the ancient authors thought; 3) the advantage or utility that can be
obtained therefrom (this only in Stephanus); 4) discrimination between truth
and falsehood. At the root of this approach stands once again Galen, who, how-
ever, tended to see the last of the four points as extraneous to the commentary
(cf. supra):430 but the change in perspective arose from the need to reconcile
the form of teaching based on the Hippocratic texts with a teaching meth-
odology that would provide genuine professional training. This also explains
why the hypomnema, while maintaining its lemmatic structure, began to show
greater emphasis on a series of theoretical treatments concerning individual
themes (sometimes also accompanied by small subtitles) that were relatively
unconnected to the passage forming the object of the commentary: this was
a format that served to provide students with a systematic doctrinal training,
and it often went beyond the commentary, taking up themes addressed by
Galen in specific treatises. Thus what Galen had deliberately left out of the
commentary, referring readers explicitly to his scientific treatises, is in the end
426 Westerink [1985] 2021, despite still nursing some doubts, is inclined to favor this
interpretation; a much more decisive position is taken by Wolska-Conus [1996], who
identifies Asclepius as the common source of Stephanus and Theophilus (see supra
n. 416) and tries to reconstruct the character of Asclepius exegesis. But it is not correct
to generalize on the basis of certainly selective citations of Stephanus. One should also
keep in mind the commentary on Hippocrates Mul. attribuited to Asclepius, who can be
identified with Asclepius, a professor of medicine who was a fellow student of Asclepius
of Tralles, a commentator of Aristotle, at the school of Ammonius, see Ullmann [1977].
427 In Hipp. Aph. V 27, CMG XI 1.3.3, p. 94.35:
.
428 Wolska-Conus [1996] 4247; [2000] 6566; for Galen, see supra, 4.1.1.
429 Ioh. Alex. In Hipp. Epid. VI, CMG XI 1.4, p. 28, 1418; Steph. Ath. In Hipp. Aph. Prooem. CMG
XI 1.3.1, p. 32, 2025.
430 Duffy [1984] 2223; see Wolska-Conus [1996] 4247, but also the previous articles [1992]
7786 and [1994] 3342.
1208 Manetti
discussion of the octo capitula, which has some points of resemblance to the
introductory parts of Stephanus commentary. It can be described as an erudite
commentary, which at times seeks to make certain themes clear to beginners
in a manner different from Galen. Although the commentary must evidently
be based on an Alexandrian tradition,448 no conclusion as to specific depen-
dencies can be drawn. The second commentary (Lat. B), defined by Beccaria
as being of a much later date,although study is still in progress449is testi-
fied by ms. Bern. 232 of the 10th century and by two later but independent
and more complete manuscripts. It has no prologue and seems to contain an
abridged and simpler commentary.450
Finally, ms. Ambr. G 108 inf., of the 9th century, preserves not only a first
section, with Hippocratic works, composed of the Latin versions of Prognostic,
On weeks and On airs waters and places,451 but also important testimony in the
form of a block of commentaries on the First Collection of Galens Canon
(in the same order), namely De sectis, Ars medica, De pulsibus ad tirones, Ad
Glauconem, attributed to lecture notes of Agnellus iatrosophist (ex vocem
Agnelli) written down by his pupil Simplicius in Ravenna, as declared in the
subscriptio to the first three texts.452 Agnellus Galenic commentaries bear wit-
ness to a unitary course of the overall set of , to be read and explained
according to the prescribed order. The form of Latin used is strongly overlaid
with Greek traits and the texts clearly bear the hallmarks of the Alexandrian
commentaries: the division into lectures (actiones/praxeis), the tendency
towards diairesis, the mention of groups of Galenic works with the titles of the
Canon (Anatomy, On causes, Diagnostics etc.). The commentary on De sectis
offers many close parallels with that attributed to Iohannes Alexandrinus.453
The reception of Greek medicine in the Arabic context after the conquest
of Egypt was extensive and profound. Through the Arabic sources the great
gaps in the Greek tradition can to some extent be bridged. Galen was widely
translated and utilized, to the point that the lemmata of his commentaries
became the almost exclusive source of the Arabic translation of the work of
Hippocrates; however, traces of two commentaries on Hippocrates Mul. 111
are known through the reports given by Ali ibn Ridwan:454 1) attribuited to
Galen, pseudepigraphical, datable between the 3rd and 6th century; 2) attrib-
uted to Asclepius, identified with the physician of this name, a study com-
panion of Asclepius of Tralles, who was a philosopher and commentator of
Aristotles Metaphysics and a pupil of Ammonius.455
The Arabic tradition also provides information concerning works on Galen
attributed to a Iohannes Grammatikos,456 often identified with the Neoplatonic
philosopher John Philoponus. Criticism has long distinguished these two
figures and has identified a physician named Iohannes Grammatikos distinct
from Philoponus.457 Three works by Iohannes Grammatikos have come down
to us: a commentary on De pulsibus ad tirones (Small pulse), one on De elemen-
tis and another on De temperamentis (but it should be noted that Iohannes
Grammatikos is thought to be distinct from the cited Iohannes Alexandrinus
who was a commentator of Hippocrates and of Galens De sectis (see supra).
The Iohannes now kept separate from Philoponus also wrote a Digest of all
the works of Galens Canon:458 Iohannes synopsis of De pulsibus has been
shown to be an epitome of his own commentary, which has come down to
us. It can therefore be surmised that Iohannes Grammatikos is very likely to
have adopted a similar procedure for the other works, and one may find in this
line of reasoning a confirmation of the close complementarity of the two text
typologies.
Gotthard Strohmaier has recently proposed that the author known as
Iohannes Grammatikos who composed a commentary on book XI of
Galens De usu partium, transmitted in an Arabic codex, should be identi-
fied as being John Philoponus, the Neoplatonic philosopher. The text, which
is a remnant of a commentary on the entire work, is written in the form of
454 Ullmann [1977] 245262, but Overwien [2005] 204 ff. shows that other commentaries
were also used, for ex. that of Palladius on Aph. The picture could undergo further change
with the systematic study of the Syriac and Arabic translations dating from earlier than
Hunain, as pointed out recently by Oliver Overwien (Colloque Hippocratique, Paris
810 November 2012).
455 Asclep. In Arist. Metaph., CAG VI 2, p. 143, 31 f.
456 A commentary on De antidotis, under the name of Iohannes Grammatikos, in a Cairo ms.,
is also known (Pormann [2003] 248 and n. 42).
457 Brutigam [1908] 50; Meyerhof-Schacht [1931] 121; Sezgin [1970] 157160.
458 Cf. Garofalo [2000a]; Garofalo [2003b] 207208: it is one of the sources of Galens 16-work
Canon.
Medicine and Exegesis 1213
459 Strohmaier [2003]; the commentary is also cited by Ibn abi Usaibia, ibid. 110 n. 13.
460 Todd [1977] 118120; Todd [1984] 106 ff.
461 Todd [1977] and [1984]: with a better examination of the testimonies, Rouech [1999]. See
also Manetti [1995] 3031, Perilli [2000b] 100116.
462 Perilli [2000b] 101, who cites the passage of Asclepius commentary taken from the lessons
held by Ammonius on Aristotles Metaphysics. Asclepius of Tralles was perhaps also
the author of a commentary on Aph. and on Mul. I (Ihm [2002a] 7376), cf. supra. On
the doubtful information of commentary writing attributed to philosophers who were
exponents of the Alexandrian school (Damascius, David, Elias), see most recently Ihm
[2002a] 7981 and 86.
463 For ex. Westerink [1964] and [1990] xxliii.
464 Cf. for ex. Schiano [2003].
465 Garofalo [1999] 187. It is by now accepted that these were translations from Greek and
not material elaborated by the Arab authors, cf. already Peterson [1974] 113115; Garofalo
[1994] 333; Garofalo [2003b]. Some codices of the Summaria are reproduced anastatically
by Sezgin [2001].
466 Garofalo [1994]; Pormann [2004]. However, there are also summaries of works not
belonging to the Canon, see Savage-Smith [2002].
1214 Manetti
(the diairesis method is the most prominent feature): some materials are sup-
pressed, but one also notes that certain parts have been shifted to different
positions, or there may be additions, or the use of figures (diagrams),467 as
mentioned earlier in connection with the diaireseis of ms. Vindob. gr. 16.468
In conclusion, mention should be made of a phenomenon that is quantita-
tively important at least for Galen, namely the scholia conserved in some man-
uscripts, which represent the final outcome of the reception of exegesis on
the works of Galen. So far two groups of scholia have been identified, which,
however, represent a rather fragmentary situation:
a) Scholia Parisina, preserved in ms. Par. gr. suppl. 634, 12th c., sch. ad De
elementis, De temperamentis, De naturalibus facultatibus, De sectis, Ad
Glauconem.469
b) Scholia Yalensia, preserved in ms. Yalensis 234 (14th c.),470 but also in
Par. gr. 2147 (16th c.) and in Marc. App. V 9:471 sch. ad De naturalibus facul-
tatibus, De locis affectis, De elementis, Ad Glauconem.
467 For the attestation of the diagrams in the Greek manuscript tradition of various authors,
see Ieraci Bio [2003] 17 ff.
468 According to Garofalo [2003a] and [2003b], they are likely to have derived from a Greek
model that was the base-text of the Tabulae of the Vindob. gr. 16 (supra).; on the affinities
between Arabic summaries and the Tabulae, see also Pormann [2004] 1921.
469 Helmreich [1910]; the scholia on De sectis and Ad Glauconem have been published by
Garofalo [2008].
470 Published by Moraux [1977]; to these should be added the scholia to De inaequali
intemperie, which show an affinity in their approach, and which are preserved in ms.
Philips 4614 (= Yale Beinecke Library 1121): Garcia Novo [1999] proposes a datation
between the 3rd and 6th century.
471 This manuscript was added by Perilli [2000b] 92 ff.
472 The function of these collections is not completely clear: Overwien [forthcoming]
expresses doubt as to whether they were genuinely utilized for teaching purposes.
473 On the scholia, see Manetti [1992] and most recently the articles by Lorusso [2005] and
[2010], with careful examination of the codex Par. gr. suppl. 634. But the other collection
of scholia reveals analogous characteristics.
474 Manetti [1992], Lorusso [2005] 45 and n. 10, and in particular [2010].
Medicine and Exegesis 1215
475 In particular Moraux [1977] III 168171 and p. 57. For the mimetic aspect of exegesis,
according to Galen, supra 4.1.2.
chapter 6
1 Pfeiffer [1968] 152170, and Montana in this volume. The bibliography on Eratosthenes is
vast, but to date there is no reliable modern collection of all his fragments.
2 Pfeiffer [1968] 152.
features of a work of systematic exegesis, where words and lines picked out for
comment from the primary text are cited verbatim and are followed by explan-
atory material;10 where such lemmata are used, they are linked by connect-
ing phrases (e.g. , Thereafter he says) to what precedes.11 Instead
the work bears the main hallmarks of a treatise.12 Hipparchus not only takes
on a variety of topics but also cites passages from Aratus Phaenomena inso-
far as they are serviceable for the purposes of his discourse, without strictly
following the order in which they appear in the poem. Moreover he ends his
references to the Phaenomena at line 729 (2.3.37) because he is not concerned
with the calendarial material and the forecasting of weather changes, on
which lines 7581141 of Aratus poem focus. It is significant that Hipparchus
refers to an individual book or section within his work as a (3.1.1a).
In his usage, at least, this word must mean dissertation as he calls Eudoxus
monographs , essays.13 As has been observed, such treatises were
also interpretations, though in a form different from that of the
[i.e. running commentaries].14 Therefore, if the reading in the man-
uscript-transmitted title of Hipparchus work is sound, it must be taken to
have a loose significance. Yet its very presence in that title looks suspiciously
like a sign of late reworking.15 As it happens, two more titles are in evidence:
one of Aratus extant lives, the so-called Vita III, speaks of a work being writ-
ten against Aratus and Eudoxus ( );16 and the
Suda-entry for Hipparchus cites him as having written on the Phaenomena
of Aratus ( ).17 Either title suggests an essay.18
10 On the characteristics of these commentaries, see e.g. Dorandi [2000], and Del Fabbro
[1979]. They are called by modern scholars, and would no doubt have been
given the same designation in antiquity; see P.Amh. II 12, P.Oxy. XXXI 2536. However, as
S. West [1970] 291 has pointed out, the word is used of a wide range of literary
productions, from rough jottings to the history of Polybius.
11 Cf. Hipparch. 1.8.14, 1.8.18, 1.10.110, 1.10.1921.
12 Martin [1956] 27.
13 Hipparch. 1.2.3, 1.3.10, 1.6.1, 1.8.67, 2.3.12, 2.3.2930. For the differences between
and /, see Montana and Dubischar in this volume.
14 Pfeiffer [1968] 213.
15 Manetti [1995] 2324 provides evidence for the use of in the titles of late
(especially sixth-century AD) scholarly writings to mean a set of notes on particular
points of interest. Cf. also Manetti [1992] 212.
16 Arat. Vita III, ed. Jean Martin [1974] 17.910 (right-hand column). Cf. apud Eudoxum et
Aratum in the Latin translation of this passage, ed. Jean Martin [1974] 17.910 (left-hand
column), reprinted from Maass [1898] 149.22150.1.
17 Suda 521, ed. Adler [1931] 2.657.26.
18 Cf. the extant titles of Aristarchus monographs, which include works of polemics ()
as well as treatises on () particular subjects: see Pfeiffer [1968] 213.
hellenistic astronomers and scholarship 1219
19 Martin [1956] 27 n. 1.
20 Cf. also Hipparch. 1.1.4.
21 Cf. the titles collected by Leo [1904] 258 n. 2 (= Leo [1960] 2.391 n. 3).
22 In this connection, it must be observed that Didymus On Demosthenes ( )
is cast in the form of a , but it is debatable whether in effect the work is best
considered a commentary or a monograph; see Harding [2006] 1320. For a typology of
philological writings, see Dubischar in this volume.
23 These fragments are collected by Maass [1898] 324, to which reference is made
throughout in this chapter. Cf. Kidd [1997] 18; Martin [1956] 2227; Maass [1898] xixv.
24 Fr. 1.1213, ed. Maass [1898] 3 (= Hipparch. 1.3.3)
.
25 Sch. A (Did.) Hom. Iliad 2.192b, ed. Erbse [1969] 1.222.
26 Pfeiffer [1968] 216.
27 West [2001a] 5075. He cites the Iliad scholion under discussion on p. 50 n. 14.
1220 Luiselli
28 They are found in Maass collection of Attalus fragments as nos. 11, 12, 14, 15, 17, 18,
2025, 28.
29 Width of intercolumnar space and of upper and lower margins in the Ptolemaic papyri:
Johnson [2004] 119, 133, and Blanchard [1993]. Marginal comments in papyrus rolls from
Egypt: McNamee [2007], and CLGP (in progress); annotations in the papyrus copies of
Aratus Phaenomena: Luiselli [2011].
30 Hipparch. 1.3.3 (= Attal., fr. 1.1013, ed. Maass [1898] 3).
31 Attal., fr. 14, ed. Maass [1898] 10.
32 Attal., fr. 17, ed. Maass [1898] 1314.
33 Attal., fr. 21, ed. Maass [1898] 17.
hellenistic astronomers and scholarship 1221
All these passages suggest that in Attalus work each explanation was preceded
by a verbatim quotation of the relevant verses of Aratus Phaenomena. It fol-
lows that the text consisted of a series of lemmata and explanatory notes.
This format is characteristic of a running commentary (), but there
is evidence to show that it could also be adopted in a treatise (,
). As we have seen, Hipparchus in his monograph makes occasional
use of short phrases to connect quotations from Aratus Phaenomena, which
he adduces for discussion, to that which precedes.35 So far as I can see, there is
nothing that can establish whether Attalus did or did not use connective units
of utterance before each quotation from Aratus poem. Therefore, it remains
unclear whether his exegesis was cast in the form of a commentary or of
a monograph.
It appears from Hipparchus report in 1.8.810 that some of the lemmata in
Attalus work were extensivemuch longer than the lemmata found in papy-
rus commentaries published to date. Perhaps one might compare a special cat-
egory of papyri which is characterized by lengthy lemmata, accompanied by
comparably long interpretations.36
We now move on to Attalus editorial activity. His wording in the preface
(fr. 1), ... , suggests that Attalus did consciously
revise a text of the Phaenomena of Aratus.37 The format and appearance of such
a are a vexed topic. Recent scholars maintain that we should envis-
age a base text of a literary work on which the editor or corrector (diorthotes)
recorded all variant readings and emendations by entering them in the blank
space between the written columns (the intercolumnium), and/or at the top
and bottom (that is, in the upper and lower margins).38 Hellenistic book-pro-
duction supports this view. Therefore, where Attalus says that he has produced
a corrected copy of Aratus poem ( ...
), it is likely that he is thinking of an existing copy of the Phaenomena
which he marked with his own corrections. There is evidence in support of
this conclusion. In 2.3.2123 Hipparchus gives a verbatim report of Attalus
interpretation of lines 712714 of Aratus Phaenomena.39 As we shall see,
Attalus advances a conjecture in line 713. First he writes
, We think the verse should be written as follows,
then he cites lines 712714 in the emended form. It follows that his conjecture
is unlikely to have been put into the text of these lines as cited by Attalus him-
self as a lemma for his comment. This in turn has far-reaching implications for
the topic under discussion. First, the quotations from the Phaenomena which
Attalus prefixed to his exegetical notes did not constitute the edition proper
of the poem as they did not contain a text. Secondly, Attalus
chosen base-text of the Phaenomena, from which he is likely to have taken all
unemended quotations which he cited as lemmata in the course of his exe-
getical work, did not incorporate any corrections; besides being adduced for
discussion in his , all corrections must effectively have been recorded
in the blank space between the written columns (or in the upper and lower
margins) of the copy of the Phaenomena which Attalus used as a base text for
his critical activity. This emended copy of Aratus poem Attalus made available
to the reader.40 This implies an intention to publish (and circulate) his work,
limited though such an initiative may have been.41
Taken together, the two products of Attalus philological activity constitute
a major scholarly enterprise. Inasmuch as (textual criticism) and
(explanation) represent two distinct activities within the domain of
ancient or scholarship,42 Attalus may well be viewed as a scientist
who was (or intended to be) truly a scholar.43 His work takes us straight to
the world of second-century BC Alexandrian scholarship. So far as we know,
38 Montanari [2011b] 23; Montanari [2009d] 403404; Montanari [2002a] 120121, 125; West
[2001a] 39; Montanari [1998d] esp. 49. See also Montana and Montanari in this volume.
39 Attal., fr. 28.2139, ed. Maass [1898] 23.
40 Attal., fr. 1.1213, ed. Maass [1898] 3 (= Hipparch. 1.1.3)
.
41 On publication of literary works in Graeco-Roman antiquity, see Dorandi [2007] 83101.
42 Schenkeveld [1994] 265.
43 For a different view, see Martin [1956] 27.
hellenistic astronomers and scholarship 1223
it is with Aristarchus (c. 216c. 144 BC) that the scholarly genre which we call
was first exploited on a large scale; as Pfeiffer has written, Aristarchus
predecessors had, with very few exceptions, abstained from writing com-
mentaries on the texts they edited.44 As we have seen, the nature of Attalus
exegetical work is beyond secure determination, yet it is significant that his
interpretive efforts focused on a poem which he edited in a revised form.
Besides Attalus and Hipparchus, the text of Aratus Phaenomena was dealt
with by Diodorus of Alexandria in the first century BC. There is evidence to
suggest that he was an astronomer, and a well-known scientist in antiquity
although he is a rather shadowy figure because of the dramatic loss of his
writings.45 It seems as though Aratus was a major target for study and discus-
sion by the leading astronomers of the Hellenistic age. Much attention will,
consequently, be devoted in the next paragraphs to selected, yet remarkable,
aspects of the early history of the textual transmission of the Phaenomena.46
2 Source Criticism
The debt of Aratus hexameter poem to the prose works of Eudoxus was as
favourite a topic for discussion in antiquity as it is nowadays.47 Hipparchus
discusses it at length in his extant monograph, and claims to have conclusively
proved it. As has been observed, he may well have overstated his own merits
in settling the case.48 Yet his demonstrative method is noteworthy. Hipparchus
sets the words () of the texts side by side (1.2.122), and notes their simi-
larities. In doing so he calls attention to the existence of agreements in error
between those texts. This point is illustrated, to take one example, by an obser-
vation which Hipparchus makes on the position of the Kneeler on the celes-
tial sphere. This constellation, which today is named Hercules, was thought
to have a foot close to the Dragons head.49 While identifying the foot in ques-
tion with the left one, Hipparchus notices (1.2.6) that it is the Kneelers right
44 Pfeiffer [1968] 212. On Aristarchus commentaries, see West [2001a] 74, and Pfeiffer [1968]
212ff. For further discussion, see Montanari [2002a] 124125, and Montanari [1998d] 10.
45 Kidd [1997] 4647; Neugebauer [1975] 2.840841; Martin [1956] 3031. Cf. also Toomer in
Hornblower-Spawforth [2003] 473.
46 For a survey of evidence, see Martin [1956] 2231.
47 See e.g. Martin [1998] 1.lxxxvixcvii, and Kidd [1997] 1617 with further bibliography.
48 Hunter [2008] 1.153154.
49 In order to understand this descriptive language, the reader should keep in mind that all
constellations were viewed in antiquity as figures of which individual parts pictorially
represent individual stars or star-groups.
1224 Luiselli
foot that both Eudoxus (fr. 17) and Aratus (lines 6970) place over the Dragons
head. In his opinion, this inaccuracy is a mild one;50 but apparently it is of a
type sufficient for Hipparchus to establish a connection.
Another example is provided by a remark on the position of Cepheus in
relation to the Little Bear. According to Eudoxus (fr. 33) and Aratus (lines 184
185), the tip of the Bears tail, that is to say the star UMi, forms an equilateral
triangle with the feet of Cepheus, which are represented by the stars and
Cep. Hipparchus finds fault with this view since he says (1.2.12) that the dis-
tance between the feet of Cepheus is shorter than the distance between each
foot and the Bears tail. The implication of the observations discussed thus far
is straightforward. Factual errors common to comparanda are meaningful for
establishing a link between them. The aim of Hipparchus is to emphasize the
significance of patterns of agreement in error as markers of close relationship
between the texts compared (cf. 1.2.6).
3 Manuscript Collation
We now focus on the diagnostic skill of Attalus and Hipparchus. We will begin
by examining the criteria by which variant readings were judged. First and
foremost, let us consider what Aratus has to say on the belt of Perseus (i.e. the
star Per) in lines 712714:
,
.54
, .
the circles themselves are without breadth and fastened all to each other,
but in size two are matched with two.68
62 PSI XV 1464, lines 2627 in the improved edition by Lundon [2011b] 7; on o, see
Lundons commentary on p. 16.
63 Matthaios [1999] 566.
64 Matthaios [1999] 579.
65 On this topic, see Matthaios [1999] 382384. Cf. also Lundon [2011b] 1415.
66 See in sch. Aim (Ariston.) Hom. Iliad 2.135a, ed. Erbse [1969] 1.210, and
Matthaios [1999] 382383. The schema Atticum is most strictly followed in Attic. Koine
Greek of the Hellenistic period is less consistent, although it tends to respect the rule
when non-personal neuters are involved; see Mayser [1934] 2/3.2830 ( 151).
67 Martin [1998] 1.28; Kidd [1997] 106.
68 Transl. Kidd [1997] 107.
1228 Luiselli
Hipparchus (1.9.1) reports the existence of two variant readings in line 467,
and . Attalus is said to have preferred (broad)
in view of a scientific theory put forward by some unnamed astronomers,
whom he calls .69 Hipparchus instead pleads for (without
breadth) on the basis of two arguments, of which one is scientific in nature
and the other literary. The scientific justification is intended to challenge the
rival theory: unlike the authorities invoked by Attalus in support of his choice,
students of mathematical astronomy () believe the celestial circles
to have no breadth; and Hipparchus thinks their views are correct inasmuch
as the circles are notional lines.70 The literary aspect of Hipparchus argu-
mentation calls attention to Aratus himself. Examination of a number of pas-
sages of the Phaenomena where the celestial circles are dealt with suggests to
Hipparchus that Aratus in line 467 must be taken to agree with the leading fig-
ures of mathematical astronomy (1.9.913);71 Hipparchus cites in evidence lines
513514 on the equator, lines 497499 on the northern tropic, and lines 541543
and 553558 on the ecliptic. In method this argument is reminiscent of an inter-
pretive principle adopted within the mainstream tradition of Hellenistic schol-
arship. Aristarchus appears to have based his approach to the textual criticism
of the Homeric poems on the assessment of Homeric usage.72 He is even cred-
ited with a famous precept, (elucidate Homer
from Homer).73 It means that Homers own words are regarded as the best
guide to the undertanding of individual readings and passages of the Homeric
poems. Because of its value to both textual criticism and exegesis the criterion
69 Attal., fr. 20.712, ed. Maass [1898] 16 (= Hipparch. 1.9.1). For information on the theory in
question, see Kidd [1997] 349350.
70 Hipparch. 1.9.68. The point is made clear in 1.9.6:
<> ...
(on the whole I think the mathematicians suppose all the said <circles>...to
be without breadth; nor is it possible to see them as having breadth, for, as it happens, each
of them is considered to be characterized by a notional line which is without breadth).
71 In 1.9.9 Hipparchus is very clear about this:
, (Aratus,
too, in accordance with the mathematicians thinks of them as being without breadth; one
could learn this from what he says on the equator).
72 West [2001a] 37.
73 On the Aristarchean origin of this maxim, see Montanari [1997o] 285286, and Montana
in this volume; for doubts, see Wilson [1997a] 90. For discussion of the principle, see
Porter [1992] 70ff. Schublin [1977] 224227 detects parallels in the juridical and rhetorical
traditions.
hellenistic astronomers and scholarship 1229
5 Emendation
Hipparchus (2.3.67) cites both lines in the above form but reports that Attalus
has altered the transmitted version of line 69383 by emending the manuscript
reading to .84 Hipparchus defends the paradosis: the intention
() of the poet, and also the phenomenon, he says, escape the notice
of Attalus and of the others.85 In addition to this statement, Hipparchus seems
to appeal to the authority of manuscripts against the proposed conjecture.86
There is yet another example. Let us return to the belt of Perseus, on which
Aratus focuses in lines 712714. Hipparchus (2.3.19) observes that the trans-
mitted reading in line 713 is likely to be an error because it does not
conform to the way Aratus introduces the Zodiacal rising of constellations.87
He also states that Attalus recognized the error and emended to
80 Attal., fr. 6.46, ed. Maass [1898] 6 (= Hipparch. 1.4.9). Cf. Kidd [1997] 204.
81 Attal., fr. 6.45, ed. Maass [1898] 6 (= Hipparch. 1.4.9).
82 Transl. Kidd [1997] 123.
83 Attal., fr. 27.2224, ed. Maass [1898] 22 (= Hipparch. 2.3.9)
, ,
(Therefore, I think it necessary not to alter the line as Attalus indicates, since
it is written in this way in all manuscripts).
84 Attal., fr. 27.810, ed. Maass [1898] 21 (= Hipparch. 2.3.7). In fact all extant manuscripts of
the poem give , whereas other sources provide evidence for either or ; for
further information, see the apparatus criticus in Martin [1998] 1.42.
85 Hipparch. 2.3.7 ,
. For discussion of the point at issue, see Kidd [1997] 413.
86 See the passage cited above, n. 83. Cf. Martin [1956] 28.
87 Hipparch. 2.3.19 ,
(from the beginning, he places the beginnings of all constellations
on the eastern horizon, not the middle or end of them).
hellenistic astronomers and scholarship 1231
88 Hipparch. 2.3.2021 = Attal., fr. 28.1517, 25, ed. Maass [1898] 2223.
89 Attal., fr. 28.2122, ed. Maass [1898] 23 (= Hipparch. 2.3.21)
(we think we
should write here in a way that is both concordant with the poet and consistent with the
phenomena).
90 Attal., fr. 28.3639, ed. Maass [1898] 23 (= Hipparch. 2.3.23)
.
91 Cf. Porter [1992] 7374; Schublin [1977] 223225, 226. Of course both principles represent
important criteria in modern textual criticism; see e.g. West [1973] 48.
92 Achill. De Inter. 3.1, ed. Di Maria [1996] 63.2123 = Jean Martin [1974] 33.1013.
1232 Luiselli
,
, .
.
. .
, []
, ;
, .
.100
Here the astronomers and the grammarians had extensive and differ-
ing inquiries about the reading. The grammarians said from ignorance:
the sky rotates the axis. But this is a crowning absurdity, for if we have
defined the axis as motionless (Aratus himself openly says: The axis,
however, stays for ever fixed), how can they say that it rotates? Instead
the astronomers aspirate in order that it may become . The
sense is this: the sky moves and revolves round the axis.
103 On the Aristarchean school, see West [2001a] 7983; Pfeiffer [1968] 252279.
104 Nadal-Brunet [1989].
105 Martin [1956] 2224.
106 Toomer [1978].
107 Toomer [1978].
108 The latter view is maintained by Rossetti-Liviabella Furiani [1993] 687.
109 Rossetti-Liviabella Furiani [1993] 692693; Pfeiffer [1968] 266267; Montana in this
volume. Cf. also West [2001a] 80.
110 Pfeiffer [1968] 211212.
chapter 7
Oliver Hellmann
At the beginning of the third book of his work On the Characteristics of Animals,
Aelian2 reports a story about communication between lions and humans. If
lions enter the houses of humans because of hunger, he was told, men and
women react in different ways. Whereas men drive off the beast immediately,
women try to persuade it using words demanding self-control. They argue
like this:
Are not you ashamed, you, a Lion, the king of beasts, to come to my hut
and to ask a woman to feed you, and do you, like some cripple, look to
a womans hand hoping that thanks to her pity and compassion you
may get what you want?You who should be on your way to mountain
haunts in pursuit of deer and antelopes and all other creatures that lions
may eat without discredit. Whereas, like some sorry lap-dog, you are con-
tent to be fed by another.3
1 For valuable comments on this paper I wish to thank most sincerely H. Enders (Heidelberg),
A. Kirichenko (Trier), B. Strobel (Trier) and G. Whrle (Trier).
2 Ael. NA 3.1. For an interpretation of this chapter cf. Hellmann [2008] 190192.
3 Ael. NA 3.1: ,
, ,
;
The lion, it seems, accepts these arguments, since it moves away in shame. For
Aelian, this behavior is no surprise.
Now if horses and hounds through being reared in their company under-
stand and quail before the threats of men, I should not be surprised if
Moors too, who are reared and brought up along with Lions, are under-
stood by these very animals.4
Aelians conclusion makes it quite clear that his opinion is not communis
opinio. He even gives some special reasons for the lions communicative abili-
ties: it has been in contact with humans for a longer period of time. In his eyes,
this seems to be a necessary condition for the ability to communicate, while
difference in species does not seem to be decisiveat least in mammals.5
This episode may be seen as just a part of the ongoing ancient debate about
the communicative abilities of animals. This debate was not limited to the sci-
entific field of zoology, it was part of the philosophical discussion of logosin
its double sense of reason and speechin animals, which was analyzed mas-
terly by Richard Sorabji about 20 years ago.6
The present study will focus on ancient biological texts.7 In the first part,
I will try to provide some insights into the ancient physiological concepts of
speech and animals ability to communicate with humans as well as with indi-
viduals of their own species. From a broader perspective, the question is one of
language and speech as a topic of ancient biology.
But this is just one aspect of the interrelation between biology on the one
side and philology and linguistics on the other. The second aspect is the work
of ancient textual scholarship on biological material. Several Alexandrian
scholars worked on biological texts or created works with biological content.
Andronicus of Rhodes edited Aristotles biological texts for ancient readers,8
and Byzantine authors created voluminous commentaries for a better under-
. . (Transl.
Scholfield).
4 Ael. NA, 3.1:
,
. (Transl. Scholfield).
5 Since all species named are mammals.
6 Cf. Sorabji [1993].
7 This category is not without problems, of course. While there may be no question that
Aristotles History of animals is a biological text, Plutarchs De sollertia animalium may be
classified as a philosophical work too.
8 See Montana in this volume.
On the interface of philology and science 1237
dialektos,18 and the front-teeth are seen as important for articulation especially
in men.19
Subsequently, these definitions are used by Aristotle to distinguish the
major zoological genera. Cephalopods and Crustacea produce absolutely no
sounds, insects and fishes produce sounds only, the oviparous quadrupeds
such as frogs, birds and the viviparous animals with four feet, i.e. mammals,
have voice (phone), dialektos is restricted to men and some kinds of birds.20
In our context, these birds are of special interest. One may note right away
that lips and teeth may not be seen as a sine qua non for articulated speech,
since birds do not have these organs. But the situation is more complex if one
takes into account that Aristotle considers the beak of the birds as a kind of
supplement for the lips and teeth in Parts of Animals 2.16.21 We may leave
this problem aside and return to Aristotles line of argument in the History
of Animals that concentrates on the tongue. The genus of birds emits voice,
and especially those have articulated speech whose tongue is broad and those
which have a fine tongue.22 Tongues of this kind can be found especially in
little birds, particularly the song-birds: The smaller (birds) have a great variety
of tones and are more talkative than the larger ones.23 Furthermore, Aristotle
carefully notes differences in voice according to sex, and declares that birds
speak and sing especially in the mating-season and when they fight.24 These
statements give a good impression of how Aristotle combined theoretical con-
cepts with detailed observation.
The key role of the tongue is emphasized in the Parts of Animals, too, in a
passage where Aristotle deals with the different functions of that organ.
And that is why among the birds those most able to pronounce articulate
sounds have broader tongues than the others. Those of the four-footed
animals that are blooded and live-bearing have little vocal articulation.
This is because they have a tongue that is hard, undetached, and thick.
Some of the birds, however, are quite vocal, and those with crook-talons
have broader tongues. The smaller ones are quite vocal. And though all
also use their tongue to communicate with one another some do so more
than others, so that in some cases they even seem to be learning form
one another.25
The important point here is that some birds not only have the physiological
apparatus to utter articulated speech, but that they also use this ability to
communicate (pros hermeneian), that is, they use it in the same manner as
human language. And that is not all: they also seem to be able to learn this
kind of language.
We may finally take a look at De Anima, a work that is not biologic in the
strict sense, though it may be classified as belonging to natural science. In his
discussion of the five senses, Aristotle deals with hearing (akoe) in De anima
2.8. In this context we find a definition of the term voice (phone) that is of spe-
cial interest for the present discussion. Having pointed out before that phone in
its true sense is solely the sound of an animate being (empsychou),26 Aristotle
gives his physiological definition:
Hence voice consists in the impact of the inspired air upon what is called
the windpipe under the agency of the soul in those parts. For, as we have
said, not every sound made by a living creature is a voice (for one can make
a sound even with the tongue, or as in coughing), but that which even
causes the impact, must have a soul, and use some imagination (phan-
tasia); for the voice is a sound which means something (semantikos)...27
The important point here is the connection between voice and meaning: the
voice has to be semantikos, otherwise it is to be taken as a mere sound, like
coughing. Are animals able to produce such meaningful voices? The answer
must clearly be yes, if one takes into account Aristotles statement in Parts of
Animals 2.17 about communication among birds.28
From the physiological perspective animals are completely capable of artic-
ulate speech in Aristotles eyes. There is no absolute difference to men here,
the difference is one of degree, not of kind.29 But terminology deserves careful
attention at this point! In all the texts cited above, Aristotle never used the
term logos.30 Logos is restricted to men, as is stated clearly in Politics 1.2 and
Generation of Animals 5.7a biological work!31 For Aristotle logos is more than
articulated speech with some meaning:
And why man is a political animal in a greater measure than any bee
or any gregarious animal is clear. For nature, as we declare, does noth-
ing without purpose; and man alone of the animals possesses speech.
The mere voice (phone), it is true, can indicate pain and pleasure, and
therefore is possessed by the other animals as well (for their nature has
been developed so far as to have sensations of what is painful and pleas-
ant and to signify those sensations to one another), but speech (logos) is
designed to indicate the advantageous and the harmful, and therefore
also the right and the wrong: for it is the special property of man in dis-
tinction from the other animals that he alone has perception of good and
bad and right and wrong and the other moral qualities, and it is partner-
ship in these things that makes a household and a city-state.32
Only man uses this kind of logos with its political and ethical function. What
birds are capable of is not with which to communicate
28 See above n. 25. Ax [1978] 256 n. 38 refers as further evidence in the biological writings to
Hist. an. 536a13ff., and 608 a17ff. (which he declares unecht).
29 This was discussed in detail by Labarrire [2004] 4649, cf. also Ax [1978] 257258.
30 Cf. Ax [1978] 259.
31 See Pol. 1.2,1253a718 and Gen. an. 5.7,786b1722.
32 Pol. 1.2,1253a718:
, . , ,
. ,
( ,
), ,
,
, .
(Transl. . Rackham).
On the interface of philology and science 1241
33 Zirin [1980] 344, who notes, that the term dialektos is not used in this passage.
34 Lennox [2001] 110125.
35 See Kullmann [1990], [1998] 55115 and Lennox [2001] 1109. For the influence of the
Aristotelian scientific pattern on Alexandrian scholarship cf. Montana, Nnlist, and
Lapini in this volume.
36 Cf. Plin. NH 711.
37 Plin. NH 11.8: nobis propositum est naturas rerum manifestas indicare, non causas indicare
dubias. (Transl. Rackham). Cf. Lennox [2001] 115.
38 The material has been collected by Fgen [2007] 5357.
39 Plin. NH 8.1: Maximum est elephans proximumque humanis sensibus, quippe intellectus illis
sermonis patrii...(Transl. Rackham).
40 Plin. NH 8.3. alienae quoque religionis intellectu creduntur maria transituri non ante naves
conscendere quam invitati rectoris iureiurando de reditu. (Transl. H. Rackham).
1242 Hellmann
learned the Greek alphabet and even wrote some sentences.41 We hear about
lions that show clementia by refusing to attack human suppliants, though
there was a dispute, as Pliny remarks, whether this kind behavior is caused
by the natural disposition of the animal or simply by chance. As in the case
of Aristotle, animals that imitate human speech or speak like humans can be
found especially among birds.42 Parrots are marked out first for their ability
to speak (sermocinantes). It greets its masters and repeats words given to it.43
A certain kind of magpie is said to talk even more articulately (expressior).44
In this context Pliny inserts a physiological observation, taken, as one may
assume, from Aristotle: All the birds in each kind that imitate human speech
have exceptionally broad tongues, although this occurs in almost all species.45
Yet physiology is abandoned immediately after this remark, and Pliny adds fur-
ther amazing stories of speaking thrushes, starlings and nightingales, ravens
and crows.46 It is rewarding to take a closer look at Plinys diction here. While
he sometimes uses verbs describing human speech as sermocinari or loqui,
other formulations like humanas voces reddere, sermonem imitari or verba
exprimere give the impression that the sounds uttered by these birds cannot
be interpreted as use of language in the strict sense. But a clear answer to this
question is not to be found in the text.
In his work On the Characteristics of Animals, Aelian reports several
instances of animals that possess remarkable communicative abilities too.
One case, the story about lions in Mauretania, has already been mentioned at
the beginning of this study. There are several further species that are capable
of understanding human speech, most prominently elephants who follow the
vocal instructions of their trainers. Not only do they comprehend the language
spoken by the Indians, but they can learn Greek as well!47 The bird called
asterias that lives in Egypt understands human speech as well,48 as do dogs
41 Plin. NH 8.6.
42 Cf. Plin. NH 10,117124. Stories about speaking animals in other genera can be found in
Plinys work, too, but Pliny has his doubts about them; cf. Fgen [2007] 54.
43 Plin. NH 10.117: imperatores salutat et quae accipit verba pronuntiat. (Transl. H. Rackham).
44 Plin. NH 10.118119.
45 Plin. NH 10.119: latiores linguae omnibus in suo cuique genere quae sermonem imitantur
humanum, quamquam id paene in omnibus contingit. Vgl. Arist. Hist. an. 2.12,504b13.
46 Plin. NH 10.119124.
47 Cf. Ael. NA 2.11, 4.24 ( ), 11.14,
11.25 (Greek).
48 Ael. NA 5.36. The species cannot be identified with certainty. It is sometimes identified
with the Bittern (cf. Thompson [1966] 57, who is sceptical about this), or with a starling
(cf. Scholfield [1958] 329).
On the interface of philology and science 1243
and horses.49 Even sows attend the call of the swineherd.50 Several species
seem to be able to speak, namely birds. The raven is described as the bird that
has the largest variety of voices (polyphonotatos); after training, it is able to
emit human voice.51 Jays and parrots are also able to imitate the human voice,52
and the francolin utters sounds clearer and more articulate than any child
and reacts to the maltreatment of deportation from Lydia to Egypt and the
subsequent famine, which killed many inhabitants of Egypt, with the proverb
Three curses on the accursed.53 This story implies that these birds not only
imitate voices but really speak with an intelligent verbal reaction to their situ-
ation! So the reader may be inclined to assume that Aelian is willing to grant
the faculty of speech at least to some kinds of animals. This would be in full
accord with his general thesis, stated clearly in the prologue and the epilogue,
that animals possess a huge number of human virtues and that they even sur-
pass humans in some of these.54 But his position on this point does not seem
to be clearly defined.55 From the beginning of his work onwards, he contrasts
man (anthropos) and animals without reason / language (aloga). The elephant
with its communicative abilities as mentioned above is none the less called an
unarticulated animal (zoon anarthron) in opposition to man, the zoon logikon,56
and even the so called Dog-heads (kynokephaloi), who are able to understand
the Indian language and have a lot of habits in common with menfor exam-
ple they keep goats and sheep and drink their milkare none the less classi-
fied as aloga.57 And Aelian gives his reasons for this classification as follows:
I have mentioned them along with brute beasts, as is logical, for their
speech is inarticulate, unintelligible, and not that of man.58
49 Ael. NA 3.1.
50 Ael. NA 8.19.
51 Ael. NA 2.51:
.
52 Ael. NA 6.19 and 13.18, 16.2; the Corocotta (perhaps Hyaena crocuta) is said to imitate
human speech (NA 7.22), too, but Aelian is sceptical about this and classifies this story as
fabulous ().
53 Ael. NA 15.27.
54 Cf. Ael. NA, Prologue and Epilogue with Hbner [1984] esp. 157163.
55 Cf. Kindstrand [1998] 29662968.
56 Ael. NA 2.11.
57 Ael. NA 4.46 with Gera [2003] 185187.
58 Ael. NA 4.46: ,
. (Transl. Scholfield)
1244 Hellmann
All in all, Aelians position is not without contradiction. On the one hand, he
denies reason (logos) in animals in accordance with the Stoic doctrine; on the
other, he illustrates their intelligence and virtue throughout his work.59
It was Plutarch who vehemently advocated the idea of logos in animals. His
work entitled Whether Land or Sea Animals are Wiser is not limited to the ques-
tion posed in its title but may be interpreted as defending rationality in animals
against the Stoics.60 In the course of Plutarchs defense of animal rationality,
scarcely any argument employed in ancient discussions for or against its exis-
tence fails to appear.61 Right in the middle of the dialogue Plutarch deals with
the ability of birds to speak and learn:
As for starlings and crows and parrots which learn to talk and afford their
teachers so malleable and imitative a vocal current to train and disci-
pline, they seem to me to be champions and advocates of the other ani-
mals in their ability to learn, instructing us in some measure that they too
are endowed both with rational utterance (logos prophorikos) and with
articulate voice.62
The terminology of this passage makes it quite clear that Plutarch is dealing
with the Stoic doctrine here.63 In another story he tries to demonstrate that
birds possess inner reason (logos endiathetos) too. A certain barber in Rome,
he claims to have heard from eyewitnesses, had a jay with a huge range of
tones and expressions, which could reproduce the phrases of human speech
and the cries of beasts and the sound of instruments.64 When a rich man was
buried accompanied by the sound of many trumpets, the jay heard this music
and as a result uttered no sound anymore. So people suspected that he had
been poisoned by rivals or deafened by the musical instrument. But neither of
these suppositions was true, since the bird after an inner retreat (anachoresis
eis heauto) started to use his voice again. Now, however, it imitated only the
melody of the trumpet it had heard before. As Newmyer pointed out, it was
the period of silence and inner meditation on part of the jay that proved, for
Plutarch, that the uttered reason of birds is prompted by inner reason (logos
endiathetos) that inspires and guides the utterance.65 Therefore, in Plutarchs
eyes, birds possess both kinds of logos differentiated by the Stoics.
3. Scholia.81 As one might expect, the majority in this field derives from the scho-
lia to the Birds of Aristophanes.82 Apart from his work On Birds, Callimachus
dealt with fishes, too. In the article on Kallimachos the Suda-lexicon lists
a work On the Change of Name of Fishes (Peri metonomasias ichtyon).83 This
text was probably part of a larger work entitled Local Nomenclature (Ethnikai
onomasiai).84 We have just one fragment from Athenaeus that lists local names
of several kinds of fish.85 As far as we can see, the scope of this work was lim-
ited to onomatological questions.86
Callimachus successor in Alexandria, Aristophanes of Byzantium also
dealt with onomatology in the animal kingdom.87 A section of his great lexi-
cal work entitled Lexeis was dedicated to the study of names of different ages
(Peri onomasias helikion). In this work, humans were dealt with at the begin-
ning (frr. 3790 Slater) followed by domestic animals (frr. 91171) and wild
animals (171219).88 To receive an impression of the work, we may take a look
at the fragments on the names of young children. The first name is brephos,
explained as the child, right after birth,89 the next one paidion, defined as the
child, nursed by the nurse90 and so on. In the section on domestic animals,
the material was organized according to the pattern: herdsmen, herd, old, mid-
dle-aged, and young animals.91 Further information was added at the end, as
the first fragments on the goat (aix) demonstrates:
aipolos: The herdsman of the goats; aipolion: the mass (sc. of goats);
and the full-grown (are called) tragoi and ixaloi. The next age chima-
roi; the youngest (of the goats) eriphoi. The poet in the Odyssey calls the
full-grown progonoi, those after them metassai, the (goats) even younger
ersai.92
81 frr. 424428.
82 frr. 424426.
83 Suid. 227 s.v. , vol. 3, p. 19 Adler.
84 Cf. Pfeiffer [1968] 135.
85 fr. 406.
86 Animals were also treated in Callimachus Collection of Wonders, cf. frr. 7, 8, 16, 25, 26, 27,
43 Giannini.
87 See Montana in this volume.
88 Cf. Nauck [1848b] 339 and the disposition of the fragments in Slater [1986] 2871.
89 fr. 37 Slater: .
90 fr. 38 Slater: (l. ).
91 Cf. Callanan [1987] 85 and Slater [1986] 39 on frr. 91171, who gives a slightly different
pattern excluding the middle-aged animals: pastor, grex, seniores, iuniores, alia.
92 frr. 9199 Slater: . (92) (sc. ). (93.94)
, . (95) . (96)
1248 Hellmann
Then follow further names for goats of different ages.93 One can easily imag-
ine that this treatise would have been of use chiefly for philological research.
Accordingly, Eustatius used it frequently for his commentaries on the Homeric
epics. In addition, the definitions were used in the lexicographical tradition.
Another work of Aristophanes was dedicated to zoology. The Alexandrian
scholar worked out an Epitome of Aristotles zoological writings in four books.94
Parts of this work along with some zoological information from other authors
have been preserved in a Byzantine collection put together under the patronage
of the emperor Konstantinos Porphyrogennetos.95 In this collection, excerpts
from Aristophanes are combined with texts from Aelian, Timotheus and fur-
ther authors. The first book, which, according to the editor Spyridon Lambros,
is excerpted entirely from Aristophanes,96 can be regarded as a general intro-
duction. At the beginning, zoological names of animal groups are explained
to the reader beginning with the cartilaginous fish (selachia) ( 127). After
this, a section dealing with copulation, pregnancy and birth ( 2897) is fol-
lowed by a heterogeneous collection of singular properties of animals and man
( 98155). Books 24 were dedicated to the treatment of separate animal
species. The topic of book 2 is viviparous animals, books 3 and 4know lost
dealt with oviparous animals, beginning with fish, followed by birds.97 This
macrostructure is explained in the important introductory section of book 2
( 23), which comes from Aristophanes. In this context, the reader is also
informed that viviparous animals will be treated in separate groups according
to the form of their feet. In the first section, animals that have toes (polyschide)
are discussed, followed by cloven-hooved animals (dichela) and animals with a
singe hoof (monycha).98 With this structure Aristophanes preserves Aristotles
scientific zoological classification, as Wolfgang Kullmann has pointed out.99
In this composition, the second in number, after giving the name of the
animal, I will try to place under this heading how many parts the pro-
posed animal has, then I will [speak] about its mating and how many
months it is able to be pregnant, and concerning its birth, what kind
of young and how many [of them] it is able to bear. In all cases [I will
explain] the life of the animal named in the heading, what its character is
like, and how many years it is able to live.100
The leopard has saw-like teeth and (feet with) toes. For it has five toes on
the front paws and four on the hind paws. It moves its legs cross-corner-
wise. It has two breasts. When dissected it has all other parts similar to
the dog, but it has a rough tongue like a file, a lung with four lobes and
a stomach like a pigs. Regarding conception and birth, everything is
similar to the dog. The female is more courageous than the male. It is said
that after eating along with other herbs by mistake the so-called leopards
bane (pardaliagches) it becomes healthy, when it eats mans excrements,
whence its hunters put excrements in their traps. In Asia there are leop-
ards, in Europe there are none at all. It is a characteristic property of the
leopard to move the top of the tail while the tail itself does not move.102
110 Strab. 13.1.54; Plut. Vit. Sull. 26. There is a mass of literature on this story, see Dring [1957]
esp. 412425, Moraux [1973] 194, Barnes [1997], Wilker [2002]; see also Montana and
Lapini in this volume.
111 Cf. Barnes [1997] 2 with n. 3. Others have spoken of a cave or cellar.
112 Cf. Barnes [1997] 12.
113 Barnes [1997] 19.
114 Cf. Barnes [1997] 19.
On the interface of philology and science 1253
Andronicus performed his task well. He not only established the form
and canon of Aristotles writings which, with comparatively slight modi-
fications, we still use today, but initiated a way of doing philosophy which
was to predominate among Aristotelians to the end of antiquity and to
spread to the adherents of other schools. (Hans B. Gottschalk)128
Mit der Ausgabe des aristotelischen Corpus durch den Rhodier Andro-
nikos beginnt zweifellos eine neue Epoche in der Geschichte des
Aristotelismus. Ohne das Vorhandensein eines zuverlssigen und ver-
hltnismig leicht zugnglichen Aristotelestextes wre die Ttigkeit
der Kommentatoren, die schlagartig kurz vor der Zeitwende einsetzt,
beinahe undenkbar gewesen. Es ist eben das Verdienst des Andronikos,
diese Wiederbelebung der aristotelischen Studien angeregt, ja berhaupt
ermglicht zu haben. (Paul Moraux)129
list of Aristotles works we find a work On Animals in 9 books (no. 102), which
is commonly believed to be an edition of History of Animals 19, and a separate
treatise On Sterility (no. 107), which has been regarded as identical with History
of Animals 10.133 Since in Ptolemys list we find a zoological work in 10 books
(no. 48),134 most scholars have argued that Andronicus added the last book to
the History of Animals and possibly gave it its title following Aristotles own
references.135 According to Friederike Berger, who has recently analyzed the
textual history of the History of Animals, Andronicus did even more. To the
original edition in 7 books, she believes, he added 3 more (books 810136).137 If
one accepts this, the question arises, what treatises might be hidden behind
Diogenes Laertius title On Animals. Berger assumes that this title subsumes
the Progression of animals and some books of the Parts of Animals and the
Generation of Animals.138 But this assumption raises further problems. As
Peter Beullens rightly remarked in his review: It is difficult to understand how
the nine books under the title in Diogenes Laertius list could refer
to a jigsaw collection (De Incessu Animalium in three books [!], De Partibus
Animalium 24, and De Generatione Animalium 13).139 So it seems more
plausible that On Animals refers to History of Animals 19, as argued above,
and that Andronicus worked on a History that already included nine books.140
We know even less about the other biological treatises. It is obvious that De
Partibus Animalium 1 had a special function in Aristotles biology, as it served
as an introduction to the zoological writings dealing mainly with questions of
methodology.141 At the beginning, it might have been an independent work
and was later added to our Parts of Animals by Andonicus, as Dring remarked:
We possess a treatise which probably received the title, the external form and
the disposition which it now has by Andronicus.142 But we have no further evi-
dence for this, and Aristotle could well have combined Parts of Animals 1 with
133 Cf. e.g. Moraux [1951] 107, Lennox [2001] 115. Balme [1991] 34. For a different view see
Lord [1986] 155, who sees On animals as a combination of Part. an. and Gen. an., and
Berger [2005] 67, who votes for IA, Part. an. 24 and Gen. an. 13.
134 Its Greek title is restored by Dring [1951] 297 as (= ) .
135 Cf. e.g. Balme [1991] 4, Flashar [2004] 253.
136 810 in the traditional order of Theodore Gaza, 7, 8 and 10 according to the manuscripts.
137 Cf. Berger [2005] 57.
138 Cf. above n. 133.
139 Beullens [2006] 307.
140 Cf. Flashar [2004] 253.
141 Cf. Kullmann [1998] 101115, who called it a propdeutische Schrift.
142 Cf. Dring [1943] 537, citation: 8.
On the interface of philology and science 1257
the other books of this treatise himself.143 In the case of Generation of Animals
the last book (5) is of a special character.144 It deals with secondary characteris-
tics such as hair-colour, whereas the first four books discuss the genesis of ani-
mals and their parts, as signified by the title. As in the case of Parts of Animals
1 this book may have been an independent treatise, as Hellmut Flashar has
presumed.145 But again, we do not know who added it to the remaining mate-
rial and created our Generation of Animals. All in all as with Aristotles works as
a whole, we cannot definitely determine what contribution Andronicus made
to the textual constitution of the three main biological works. There are signs
that he edited at least the History of Animals as we know it todaybut there is
no definite proof.
writings from the first century BC to the end of antiquity.157 Not one of the large
group of Neoplatonists dealt with the History of Animals or Parts of Animals.
Why? The answer lies hidden in the Neoplatonic classification of Aristotles
writings and the corpus of Aristotles works read by these Neoplatonists.
We can see this, if we take a look at the introductions in Aristotles philos-
ophy found in a number of commentaries on Aristotles Categories. These
introductions are structured by means of ten questions about how to study
Aristotle. In our context, the second and third questions are of primary con-
cern: 2. How can Aristotles works be classified? 3. Where should one start to
study Aristotles works? With regard to classification, three groups of works
were differentiated by the commentators: the particular writings (merika), the
general (katholou) and the intermediate (metaxy). If we take Simplicius as an
example, the particular writings are defined as those that are addressed to one
person and written about particular things.158 The general writings are subdi-
vided into hypomnematic writings, which the author put together for personal
reminding, and systematic writings, i.e. dialogs and autoprosopa, in which the
author speaks in his own person. The autoprosopa include theoretical writ-
ings, such as the Metaphysics and Physics, practical writings, such as the Grand
Ethics and the Politics, and instrumental writings (organika), such as the First
Analytics, Categories and the Rhetoric.159 There is no general definition for the
intermediate writings (metaxy), instead the History of Animals and Plants are
presented as examples that do not deal with particular things altogether, since
they deal with animal species.160 As Ilsetraut Hadot has shown in her com-
mentary, it is not only the History of Animals that was thought to belong to this
group, but all the other biological treatises, too.161 In accordance with this clas-
sification, Philoponus gives the Generation of Animals as an example for this
group instead of the History of Animals.162 This classification has important
157 For the De anima the situation is different, of course, due to its philosophical impact.
Cf. the useful synopsis on the ancient commentaries in DAncona Costa [2002], 250251,
especially for mentions in Arabic sources. The Commmentary on Generation of Animals
that was attributed to Ioannes Philoponus (ed. Hayduck, CAG 14.3) in reality is a work of
Michael of Ephesus.
158 Simpl. in Cat. p. 4.1012 Kalbfleisch. See Hadot [1990] 6466 and 6364 on the ten questions.
159 Simpl. in. Cat. p. 4.145.2 Kalbfleisch.
160 Simpl. in Cat. p. 4.1213 Kalbfleisch: , ,
.
161 Cf. Hadot [1990] 6970 and 8586.
162 Io. Philop. in Cat. p. 3.2628 Busse. But cf. Moraux [1973], 74 n. 45 who thinks that this may
be a confusion of facts.
1260 Hellmann
If we now move on to the third question about how to study Aristotle, the
Neoplatonists prefer to begin with logic.164 After this one may advance to eth-
ics, followed by physics in the third position.165 The commentators Ammonius,
Philoponus, Simplicius and Olympiodorus give five titles of physical works:
Physics, On the Heavens, On Generation and Corruption, Meteorology and On
the Soul. As Ilsetraut Hadot has rightly pointed out, these five works seem to be
the only five in the field of physics that were incorporated in the Neoplatonist
educational program. The biological works were not part of this program,
since they belongedin contrast to the physical works just mentionedto
the intermediate writings.166
For these works outside the educational canon, one can easily imagine,
there was no great demand for commentaries to aid the readers.
As far as we know, readers of Aristotle had to wait until the Byzantine era
for commentaries on his biological works. It was Michael of Ephesus,167 who
undertook the task of commenting on Aristotles biology. His undertaking was
part of a larger project in the philosophical circle of the Byzantine princess
Anna Comnena (1083after 1148). As R. Browning has shown in his analysis of
her funeral oration by George Tornikes, it was Anna Comnena, who encour-
aged Michaels work on Aristotle, as Tornikes states himself:168
I myself have heard the philosopher from Ephesos blame her as the cause
of his blindness, because he had worked night after night, without sleep,
commanded by her to write commentaries on the works of Aristotle; the
use of candles had caused drying of the eyes.169
The philosopher of Ephesus is, with all probability, Michael. Accordingly, his
work has to be dated to the middle of the 12th century AD.170
Michael of Ephesus commented on a wide range of Aristotelian works. I
limit myself to name those of biological content with their number in the
series of the Commentaria in Aristotelem Graeca: Parts of Animals (CAG 22,2 ed.
Hayduck), Generation of Animals (CAG 14,3 ed. Hayduck),171 Parva Naturalia
(CAG 22,1 ed. Wendland), Movement of Animals (CAG 22,2 ed. Hayduck) and
Progression of Animals (CAG 22,2 ed. Hayduck). The last two treatises are
available in an excellent English translation with introduction and notes by
Anthony Preus, who points out that Michaels commentaries are the only sur-
viving Greek commentaries on these treatises.172 Apart from the works just
mentioned, Praechter has called attention to an ancient list of exegetes of
Aristotle, according to which Michael commented on the History of Animals
too.173 This work has not come down to us, and it has to be noted at this point
that it is not listed in Michaels own overview of his work.174 All we can say is
that Michael seems to have had a History of Animals at hand that included nine
books.175 If we can trust his words at the end of his commentary on the Parva
Naturalia, he was eager to obtain as many of Aristotles works as he could. If
Aristotle was not available, he consulted the works of Theophrastus.176
in the first (book) he spoke on how the student has to be. That he has to
be educated and that one has to speak first about the common proper-
ties of all animals and then about the individual properties of each kind,
in order that one may not be forced to talk many times about them. (He
said) that one had to observe the phenomena about animals first and
then search for their causes, and (he spoke on the question) in how many
ways necessity (was used). After reprehending the divisions of Plato, he
now is about to talk about causes. (For the information) of which and of
how many parts (each of the animals) consists, he says, one has to search
the work On the History of Animals, where he determines in nine books
the things concerning all animals, but the causes of these parts and how
each of these is positioned according to nature, must be looked at (now).182
, , ,
, ,
, ,
. , , ,
,
,
.
183 Examples in Praechter [1906] 880.
184 Mich. Eph. in IA p. 164 Hayduck. The reference to the diagram is p. 164.1314:
. Cf. Preus [1981], 141 and 178179, Stckelberger [1993] 138.
185 Mich. Eph. in Part. an. 4, p. 96.5 Hayduck: .... Cf. Praechter
[1906] 903904.
186 Preus [1981] 23.
187 Cf. the useful overview of his work by Sharples [1987].
188 See the indices in the CAG editions of Wendland and Hayduck.
1264 Hellmann
3 Epilogue
The interrelations between philology and biology are manifold, and the num-
ber of characteristic examples presented in this paper must therefore be in
some way arbitrary. One goal of the selection presented above has been
to demonstrate the key-role of Aristotle in the tradition of ancient biology
and biological writing. By analyzing language from a biological perspective
Aristotle developed a detailed concept of language in humans and animals,
which may be seen as much more than a major contribution to ancient lin-
guistics, since it played an important role in the philosophical debate about
language as a specific difference of humans.
Aristotles great biological texts stimulated philologists to enter the field of
biology. The mass of material he analysed had to be organised, physiological
concepts and aetiological argumentations had to be explained to the readers,
all the more so as Aristotles style does not always make reading his biological
204 For reception of the Parva Naturalia see now Grellard-Morel [2010].
205 I owe this point to Georg Whrle (Trier).
Bibliography
Alessi R., Research Program and Teaching Led by the Masters in Hippocrates Epidem-
ics 2, 4 and 6, in Horstmanshoff (2010) 119135.
, The Arabic Version of Galens Commentary on Hippocrates Epidemics, Book
Two, as a Source for the Hippocratic Text: First Remarks, in Pormann (2012) 7190.
Aletta A., Su Stefano, copista di Areta, RSBN 41 (2004) 7393.
Alexander P. J., Church Councils and Patristic Authority: The Iconoclastic Councils of
Hiereia (754) and St. Sophia (815), Harvard Studies in Classical Philology 63 (1958)
493505, repr. in Religious and Political History and Thought (London 1978) n. ix.
Alexanderson B., Textual Remarks on Ptolemys Harmonica and Porphyrys Commentary
(Gothenburg 1969).
Allan W., Euripides in Megale Hellas: some aspects of the early reception of tragedy,
G&R 48 (1) (2001) 6786.
(ed.), Euripides, Helen (Cambridge 2008).
Allatius L., Procli diadochi paraphrasis in Ptolemaei libros IV. De siderum affectionibus
(Leiden 1635).
Allen J., The Stoics on the Origin of Language and the Foundations of Etymology, in
Language and Learning. Philosophy of Language in the Hellenistic Age. Proceedings
of the Ninth Symposium Hellenisticum, eds. D. Frede, B. Inwood (Cambridge-New
York 2005) 1435.
Allen P., Jeffreys E. (eds.), The Sixth Century: End or Beginning? (Brisbane 1996).
Allen W. S., Accent and Rhythm. Prosodic Features of Latin and Greek (Cambridge 1973).
, Vox Graeca. A Guide to the Pronunciation of Classical Greek (Cambridge 19873).
Alline H., Aristophane de Byzance et son dition critique de Platon, REA 17 (1915)
8597.
Alpers K. (ed.), Theognostos . berlieferung, Quellen und Text der
Kanones 184 (Diss. Hamburg 1964).
, Bericht ber Stand und Methode der Ausgabe des Etymologicum Genuinum (mit
einer Ausgabe des Buchstaben ) (Copenhagen 1969).
, Zonarae Lexicon, RE X/A (Munich 1972) 732763.
, review of L. W. Daly, Contributions to a History of Alphabetization in Antiquity
and the Middle Ages (Bruxelles 1967), Gnomon 47 (1975) 113117.
(ed.), Das attizistische Lexicon des Oros. Untersuchung und kritische Ausgabe
der Fragmente, SGLG 4 (Berlin-New York 1981).
, Klassische Philologie in Byzanz, review of Wilson (1983), CPh 83 (1988) 342360.
, Eine byzantinische Enzyklopdie des 9. Jahrhunderts, in Scritture, libri e testi
nelle aree provinciali di Bisanzio, ed. G. Cavallo, G. De Gregorio, M. Maniaci (Spoleto
1989) 235269.
, Griechische Lexikographie in Antike und Mittelalter, dargestellt an aus-
gewhlten Beispielen in Welt der Information. Wissen und Wissensvermittlung in
Geschichte und Gegenwart, ed. H.-A. Koch (Stuttgart 1990) 1438.
Bibliography 1269
Andorlini I., Lapporto dei papiri alla conoscenza della scienza medica antica, ANRW
II 37, 1 (1993) 458562.
, Codici papiracei di medicina con scoli e commento, in Le commentaire entre
tradition et innovation, ed. M.-O. Goulet-Caz (Paris 2000) 3752.
, Lesegesi del libro tecnico: papiri di medicina con scoli e commento, STCPF 11
(2003) 929.
Andrews H. H., The Letter of Aristeas, in The Apocrypha and Pseudepigrapha of the Old
Testament in English, ed. R. H. Charles (Oxford 1913) 83122.
Angelidi Chr., Le sjour de Lon le Mathmaticien Andros: ralit ou confusion?, in
. Mlanges Ahrweiler I (Paris 1998) 17.
, Observing, Describing and Interpreting: Michael Psellos on Works of Ancient
Art, 2 (2005) 227242.
Angold M., A Byzantine Government in Exile (Oxford 1974).
Annas J., Knowledge and language: The Theaetetus and the Cratylus, in Language and
Logos: Studies in Ancient Greek Philosophy Presented to G. E. L. Owen, eds. M. Scho-
field, M. Craven Nussbaum (Cambridge 1982) 95114.
ANRW = Aufstieg und Niedergang der Rmischen Welt, II 37.1 and 37.2, eds. G. Haase,
H. Temporini (Berlin-New York 19931994).
Antonopoulou Th., The orthographical kanons of Nicetas of Heraclea, JByz 53 (2003)
171185.
, The date of Theognostos Orthography: a Reappraisal, ByzZ 103 (2010) 112.
Apfel H. V., Homeric Criticism in the Fourth Century B.C., TAPhA 69 (1938) 245258.
Apthorp M. J., New Light from Mount Sinai on the Text of the Iliad, ZPE 127 (1999)
141148.
Arabatzis G., (Athens 2006).
Arata L., Bastianini G., Montanari F., Aeschylus 3. P.Oxy. XX 2256. Hypotheseis di varie
tragedie, in Commentaria et Lexica Graeca in Papyris reperta (CLGP) I 1.4, (Berlin-
Boston 2004) 3451.
Arens H., Aristotles theory of language and its tradition: texts from 500 to 1750. Select.,
transl. and comm. (Amsterdam-Philadelphia 1984).
, Sprache und Denken bei Aristoteles, in Auroux (et al.) (2000) 367375.
Argoud G., Guillaumin J.-Y. (eds.), Sciences exactes et sciences appliques Alexand-
rie (IIIe sicle av. J.-C.Ier sicle ap. J.-C.), Memoires / Centre Jean Palerne 16 (Saint-
tienne 1998).
Arnauld A., Lancelot C., Grammaire gnrale et raisonne (Paris 1660); repr. (Paris 1969).
Arnesano D., Il palinsesto Laur. Conv. Soppr. 152, BBGG 53 (1999) 213238.
, Il Copista del Digenis Akritas, Bizantinistica 7 (2005) 135158.
, La minuscola barocca (Galatina 2008).
, Ermogene e la cerchia erudita, in Bianchi (2011) 95111.
Bibliography 1271
Arnesano D., Sciarra E., Libri e testi di scuola in Terra dOtranto, in Del Corso-Pecere
(2010) 425473.
Arnott W. G., Alexis: The fragments. A commentary (Cambridge 1996).
Aronadio F., I fondamenti della riflessione di Platone sul linguaggio: il Cratilo (Roma
2011).
Arrighetti G. (ed.), Satiro, Vita di Euripide (Pisa 1964).
, Il Pap. Oxy. XIII, 1611. Alcuni problemi di erudizione antica, SCO 17 (1968)
7698.
, Fra erudizione e biografia, SCO 26 (1977a) 1367.
, Hypomnemata e Scholia: alcuni problemi, MPhL 2 (1977b) 4967.
, Poeti, eruditi e biografi. Momenti della riflessione dei Greci sulla letteratura (Pisa
1987).
, Riflessione sulla letteratura e biografia presso i Greci, in La philologie grecque
lpoque hellnistique et romaine, Sept exposs suivis de discussions, Vanduvres-
Genve, 1621 aot 1993, Entretiens sur lantiquit classique vol. 40, ed. F. Montanari
(Vanduvres-Genve 1994) 211249; Discussion, ibidem 250262.
Arrighetti G., Mariotti G. C., Montanari, F., Concordantia et Indices in Scholia Pindarica
Vetera (Hildesheim 1991).
Ascheri P. (ed.), I frammenti degli scritti omerici del grammatico Demetrio Issione (Diss.
Genova 2003).
, Demetrio Issione, Aristarco e il duale omerico, in La cultura ellenistica. Lopera
letteraria e lesegesi antica. Atti del Convegno COFIN 2001, Universit di Roma Tor
Vergata, 2224 settembre 2003, eds. R. Pretagostini, E. Dettori (Roma 2004) 335351.
, Un elenco di grammatici nel Palimps. Lipsiensis gr. 2: problemi di identificazi-
one, RFIC 133 (2005) 413442.
, Demetrius [14] Ixion, LGGA (20092).
, Demetrio Issione dialettologo: lAttico e il dialetto degli Alessandrini, in
Montana (2010) 125152.
, The Greek Origins of the Romans and the Roman Origins of Homer in the
Homeric Scholia and in POxy. 3710, in Montanari-Pagani (2011) 6586.
Asmis E., Philodemus poetic theory and On the Good King According to Homer, ClAnt
10 (1991) 145.
Asmuth B., Angemessenheit, in Historisches Wrterbuch der Rhetorik, vol. 1 (Tbingen
1992) 579604.
, Perspicuitas, in Historisches Wrterbuch der Rhetorik, vol. 6 (Tbingen 2003)
814874.
, (et al.), Bild, Bildlichkeit, in Historisches Wrterbuch der Rhetorik, vol. 2
(Tbingen 1994) 1030.
Asper M., Kanon, in Historisches Wrterbuch der Rhetorik, vol. 4 (Tbingen 1998)
869882.
1272 Bibliography
Bailey C., Epicurus. The Extant Remains, translation and commentary (Oxford 1926);
repr. (Hildesheim-New York 1975).
Bakker E. J. (ed.), A Companion to Ancient Greek Language (Chichester 2010).
Bakola E., Prauscello L., Tel M., Greek Comedy and the Discourse of Genres (Cambridge
2013).
Baldi D., Il Codex Florentinus del Digesto, S&T 8 (2010) 99186.
, Ioannikios e il corpus Aristotelicum, RHT n.s. 6 (2011) 1526.
(ed.), Etymologicum Symeonis -E (Turnhout 2013).
Baldwin B., Herodotus and Tacitus: Two Notes on Ancient Book Titles, QUCC n.s. 16
(1984) 3133.
Ballaira G., Tiberii De figuris Demosthenicis (Rome 1968).
Bally C., Manuel daccentuation grecque (Berne 1945).
Balme D. M. (ed.), Aristotle, History of Animals, Books VIIX (Cambridge, Mass.-
London 1991).
Baltes M., Dianoemata: Kleine Schriften zur Platon und zum Platonismus (Stuttgart
1999).
Baltussen H., Philology or Philosophy? Simplicius on the Use of Quotations, in Epea
and Grammata: Oral and Written Communication in Ancient Greece, ed. I. Worthing-
ton, J. M. Foley (Leiden-Boston-Kln 2002) 173189.
, Plato Protagoras 34048: Commentary in the Making?, in Philosophy, Science
and Exegesis in Greek, Arabic and Latin Commentaries, III, ed. P. Adamson, H. Bal-
tussen, M. W. F. Stone (London 2004) I, 2135.
Bandiera A., Per un bilancio della tradizione papiracea delle Storie di Erodoto, in
Akten des 21. Internationalen Papyrologenkongresses, Berlin 1995, eds. B. Kramer
W. Luppe H. Maehler G. Poethke (Berlin 1997) 4956.
Baragwanath E., de Bakker M., Introduction, in Myth, Truth & Narrative in Herodotus,
eds. E. Baragwanath, M. de Bakker (Oxford 2012) 156.
Baratin M., La constitution de la grammaire et de la dialectique, in Histoire des ides
linguistiques. I. La naissance des mtalangages en Orient et en Occident, sous la
direction de S. Auroux (Lige-Bruxelles 1989) 186206 = [1989a].
, La maturation des analyses grammaticales et dialectiques, in Histoire des
ides linguistiques. I. La naissance des mtalangages en Orient et en Occident,
sous la direction de S. Auroux (Lige-Bruxelles 1989) 207227 = [1989b].
, Les difficults de lanalyse syntaxique, in Histoire des ides linguistiques, I. La
naissance des mtalangages en Orient et en Occident, ed. S. Auroux (Lige-Bruxelles
1989) 228242 = [1989c].
, La naissance de la syntaxe Rome (Paris 1989d).
, Remarques sur la place et le rle du concept de latinit dans les grammaires
latines antiques, in Grammaire et rhtorique: notion de Romanit, ed. J. Dangel
(Strasbourg 1994) 5157.
Bibliography 1275
Bastianini G., Gallazzi C. (eds.), Papiri dellUniversit degli Studi di MilanoVIII. Posi-
dippo di Pella. Epigrammi (P. Mil. Vogl. VIII 309), edizione a cura di Guido Bastianini
e Claudio Gallazzi, con la collaborazione di Colin Austin (Milano 2001).
Bastianini G., Sedley D. N., Commentarium in Platonis Theaetetum, in Corpus dei
Papiri Filosofici (CPF) III. Commentari (Firenze 1995) 227562.
Battezzato L. (ed.), Tradizione testuale e ricezione letteraria antica della tragedia greca.
Atti del convegno, Scuola Normale Superiore, Pisa 1415 giugno 2002 (Amsterdam
2003).
, Colometria antica e pratica editoriale moderna, QUCC n.s. 90 (2008) 137158.
Bauer B., Aemulatio, in Historisches Wrterbuch der Rhetorik, vol. 1 (Tbingen 1992)
141187.
Baumbach M., Lukillos, DNP 7 (1999a) 503.
, Luperkos, DNP 7 (1999b) 510511.
, Moiris, DNP 8 (2000) 343344.
, Salustios [1], DNP 10 (2001a) 1269.
, Seleukos [13: S. Homerikos], DNP 11 (2001b) 365366.
, Telephos [3], DNP 12/1 (2002a) 9495.
, Theodoretos [2], DNP 12/1 (2002b) 321.
Baumgarten R., Schule: Elementar- und Grammatikunterricht. Griechenland, in
Handbuch der Erziehung und Bildung in der Antike, eds. J. Christes, R. Klein, C. Lth
(Darmstadt 2006) 89100.
Baxandall M., Giotto and the Orators (Oxford 1971).
Baxter T. M. S., The Cratylus: Platos critique of naming (Leiden 1992).
Bazzani M., Theodore Metochites. A Byzantine Humanist, Byzantion 76 (2006) 3252.
Bearzot C., La storia greca nella Suda, in Il lessico Suda e la memoria del passato a
Bisanzio, ed. G. Zecchini (Bari 1999) 3574.
Bearzot C., Landucci F., Zecchini G. (eds.), LOnomasticon di Giulio Polluce. Tra lessico-
grafia e antiquaria (Milano 2007).
Beaucamp J., Le philosophe et le joueur. La date de la fermeture de lcole dAthnes,
T&MByz 14 (2002) 2135.
Bcares Botas V., Apolonio Discolo, Sintaxis. Introduccin, traduccin y notas (Madrid
1987).
, Dionisio Tracio. Gramtica. Comentarios antiguos (Madrid 2002).
Beccaria A., Sulle tracce di un antico canone di Ippocrate e di Galeno, I Le prime tra-
duzioni latine del corpus ippocratico, IMU 2 (1959) 156.
, Sulle tracce di un antico canone di Ippocrate e di Galeno, II Gli Aforismi di
Ippocrate nella versione e nei commenti del primo Medioevo, IMU 4 (1961) 175.
, Sulle tracce di un antico canone di Ippocrate e di Galeno, III Quattro opere
di Galeno , IMU 14 (1971) 123.
Bibliography 1277
Bergstrsser G., Hunain ibn Ishaq: ber die Syrischen und Arabischen Galen-
bersetzungen, Abh. fr die Kunde des Morgenlands 17, 2 (Leipzig 1925).
Bernab A., Poetarum epicorum Graecorum testimonia et fragmenta, I (Leipzig 1996).
(ed.), Poetae epici Graeci. Testimonia et fragmenta 2, 3, Musaeus, Linus, Epi-
menides, Papyrus Derveni, Indices (Berlin-New York 2007).
Bernard F., Writing and Reading Byzantine Secular Poetry, 10251081 (Oxford 2014).
Bernard W., Sptantike Dichtungstheorien: Untersuchungen zu Proklos, Herakleitos und
Plutarch (Stuttgart 1990).
Bernardinello S., La grammatica di Manuele Caleca, RSBN 8/9 (19711972) 203218.
Bernays J., Aristotle on the effect of tragedy, in Ancient Literary Criticism, ed. A Laird
(Oxford 2006) 158175.
Berndt R., Alexionis grammatici quae supersunt (Knigsberg 1906).
, Die Fragmente des Homererklrers Herakleon (Insterburg 1914).
(ed.), Die Fragmente des Grammatikers Habron, BPhW 35 (1915) Nos. 46, 1451
1455; 47, 14831488; 48, 15141520.
Bernecker R., Intention, in Historisches Wrterbuch der Rhetorik, vol. 4 (Tbingen
1998) 451459.
Bernecker R., Steinfeld Th., Amphibolie, Ambiguitt, in Historisches Wrterbuch der
Rhetorik, vol. 1 (Tbingen 1992) 436444.
Berschin W., Griechisches-Lateinisches Mittelalter (Munich 1980).
Bertagna M. I., Democrito: la lingua e il mondo, Elenchos 28 (2007) 389401.
Bertelli L., Hecataeus: from Genealogy to Historiography, in The Historians Craft in the
Age of Herodotus, ed. N. Luraghi (Oxford 2001) 6794.
Berthold R. M., Rhodes in the Hellenistic Age (Ithaca-London 1984).
Berti E., Alla scuola di Manuele Crisolora, Rinascimento s. II, 27 (1987) 373.
Berti M. (ed.), Istro il Callimacheo, 1: Testimonianze e frammenti su Atene e sullAttica
(Tivoli 2009).
, Istro alla scuola di Callimaco. Produzione e diffusione della ricerca nella Bib-
lioteca di Alessandria, in Le et della trasmissione. Alessandria, Roma, Bisanzio. Atti
delle Giornate di Studio sulla storiografia greca frammentaria, Genova, 2930 mag-
gio 2012, eds. F. Gazzano, G. Ottone (Tivoli 2013) 193210.
Berti M., Costa V., La Biblioteca di Alessandria. Storia di un paradiso perduto (Tivoli 2010).
Bertier J., A propos de quelques rsurgences des pidemies dans les Problemata du
Corpus Aristotelicien, in Baader-Wienau (1989) 261269.
Besharov J., Imagery of the Igors Tale in the Light of Byzantino-Slavic Poetic Theory
(Leiden 1956).
Betegh G., The Derveni Papyrus: Cosmology, Theology and Interpretation (Cambridge
2004).
Bethe E., Quaestiones Diodoreae mythographae (Diss. Gottingae 1887).
Bibliography 1279
Boudon-Millot V., Jouanna J., Galien, Ne pas se chagriner, text tabli et traduit par V.
Boudon-Millot et J. Jouanna, avec la collaboration de A. Pietrobelli (Paris 2010).
Bourdieu P., Reading, readers, the literate, literature, in In other words: essays towards a
reflexive sociology, transl. M. Adamson (Stanford, California 1990) 94105.
Bowersock G. W., Greek Sophists in the Roman Empire (Oxford 1969).
, Hellenism in Late Antiquity (Cambridge 1990).
, Late Antique Alexandria in Alexandria and Alexandrianism: Papers delivered
at a Symposium organized by the John Paul Getty Museum and Getty Center for the
History of Art (Malibu 1996) 263272.
Boys-Stones G. R., Post-Hellenistic Philosophy: a Study of its Development from the Stoics
to Origen (New York 2001).
(ed.), Metaphor, allegory, and the classical tradition : ancient thought and mod-
ern revisions (Oxford-New York 2003a).
, The Stoics two types of allegory, in Boys-Stones (2003a) 189216 = [2003b].
Brancacci A., Les mots et les choses: la philosophie du langage chez Dmocrite, in
Philosophie du langage et grammaire dans lantiquit (Bruxelles-Grenoble 1986)
928.
, Oikeios Logos. La filosofia del linguaggio di Antistene (Napoli 1990).
, Porfirio e Antistene: e in SSR v A 187, in .
Unione e amicizia. Omaggio a Francesco Romano, ed. M. Barbanti, G. R. Giardina,
P. Manganaro (Catania 2002) 409417 = [2002a].
, Protagoras, lorthoepeia et la justesse des noms, in Platon, source des Prsocra-
tiques, textes runis par M. Dixsaut, A. Brancacci (Paris 2002) 169190 = [2002b].
, Dialettica e orthoepeia in Protagora, Mthexis 23 (2010) 5371.
Brandenburg Ph. (ed.), Apollonios Dyskolos. ber das Pronomen, Einfhrung, Text,
bersetzung und Erluterungen (Mnchen-Leipzig 2005).
Braswell B. K., Didymus on Pindar, in Matthaios-Montanari-Rengakos (2011) 181197.
, Didymos of Alexandria. Commentary on Pindar, ed. and transl. with Introduc-
tion, Explanatory Notes, and a Critical Catalogue of Didymos Works (Basel 2013).
Braswell B. K., Billerbeck M., The Grammarian Epaphroditus. Testimonia and Frag-
ments edited and translated with Introduction, Notes, and Commentary (Bern 2008).
Brutigam W., De Hippocratis Epidemiarum libri VI commentatoribus (Diss. Knigsberg
1908).
Bravo Garca A., Prez Martn, I. (ed.), The legacy of Bernard de Montfaucon (Turnhout
2010).
Breglia Pulci Doria L., Storia Universale e Geografia in Eforo di Cuma, in Storiogra-
fia locale e storiografia universale. Forme di acquisizione del sapere storico nella cul-
tura antica, eds. C. Bearzot, R. Vattuone, D. Ambaglio, Bologna 1618 dicembre 1999
(Como 2001) 139164.
Bibliography 1285
Castelli C., Riano e Omero: i Messeniaca tra imitazione e innovazione, Acme 49 (1994)
524.
Cavallo G., Funzione e strutture della maiuscola greca tra i secoli VIIIXI, in La palo
graphie grecque et byzantine (Paris 1977) 95137.
, La circolazione libraria nellet di Giustiniano, in Limperatore Giustiniano.
Storia e mito, ed. G. G. Archi (Milano 1978) 201236.
, La trasmissione scritta della cultura greca antica in Calabria e in Sicila tra i
secoli XXV. Consistenza, tipologia, fruizione, S&C 4 (1980) 157245.
, La cultura italo-greca nella produzione libraria, in I Bizantini in Italia, ed.
G. Pugliese Carratelli (Milano 1982) 495612.
, Culture et circulation des livres lpoque de Justinien, Faventia 9 (1987)
5164.
, Lo specchio omerico, MEFRA 101 (1989) 609627.
, Qualche riflessione sulla continuit della cultura greca in Oriente tra i secoli
VII e VIII, ByzZ 88 (1995) 1322.
, Scritture informali, cambio grafico e pratiche librarie a Bisanzio tra i secoli XI
e XII, in Prato (2000) I 219238.
, Scuole, libri, pratiche intellettuali a Roma tra il V e il IX secolo, in Roma
dallantichit al medioevo, ed. M. S. Arena (et al.) (Roma 2001) 92103.
, Dalla parte del libro (Urbino 2002a).
, Tracce per una storia della lettura a Bisanzio, ByzZ 95 (2002b) 423444.
, Sodalizi eruditi e pratiche di scrittura a Bisanzio, in Bilan et perspectives des
tudes mdivales (19931998), ed. J. Hamesse (Paris 2003) 645665.
(ed.), La cultura bizantina. Lo spazio letterario del Medioevo III/1 (Roma 2004).
, Da Alessandria a Costantinopoli? Qualche riflessione sulla collezione filoso-
fica, S&T 3 (2005) 249263.
, Leggere a Bisanzio (Milan 2007a).
, Qualche riflessione sulla collezione filosofica, in The Libraries of the Neopla-
tonists, ed. C. DAncona (Leiden-Boston 2007) 155165 = [2007b].
, Oralit scrittura libro lettura. Appunti su usi e contesti didattici tra antichit
e Bisanzio, in Del Corso-Pecere (2010) 1136.
Cavallo G., Maehler H., Greek Bookhands of the Early Byzantine Period (London 1987).
, Hellenistic Bookhands (Berlin-New York 2008).
Celentano M. S., Homoeideia, in Historisches Wrterbuch der Rhetorik, vol. 3 (Tbin-
gen 1996) 15251527.
Cellerini A., Introduzione allEtymologicum Gudianum (Rome 1988).
Celoria F., The Metamorphoses of Antoninus Liberalis. A Translation with a Commen-
tary (London-New York 1992).
Cengarle S. A., Attribuzione di un compendio sul dialetto ionico a Manuele Mosco-
pulo, Acme 23 (1970) 7180.
1292 Bibliography
Cingano E., Riflessi dellepos tebano in Omero e in Esiodo, Incontri triestini di filologia
classica 2 (20022003) 5576.
Citti V., Le edizioni omeriche delle citt, Vichiana 3 (1966) 343.
Clark H. H., Brennan S. E., Grounding in Communication, in Perspectives on Socially
Shared Cognition, eds. L. B. Resnick, J. M. Levine, S. D. Teasley, (Washington DC 1991)
127149.
Classen C. J., Sprachliche Deutung als Triebkraft Platonischen und Sokratischen Philoso-
phierens (Mnchen 1959a).
, The study of language amongst Socrates contemporaries, PACA 2 (1959b)
3349; repr. in Sophistik, ed. C. J. Classen (Darmstadt 1976) 215247.
, Rhetorik und Literarkritik, in La philologie grecque lpoque hellnistique et
romaine, ed. F. Montanari (Vanduvres-Genve 1994) 307360.
, Rhetoric and Literary Criticism: Their Nature and Their Functions in Antiq-
uity, Mnemosyne 48 (1995) 513535.
Classen P., Burgundio von Pisa (Heidelberg 1974).
Clauss J. J., Cuypers M. (eds.), A Companion to Hellenistic Literature (Chichester-
Malden, MA 2010).
Clayman D. L., Timon of Phlius. Pyrrhonism into Poetry (Berlin-New York 2009).
Clrigues J. B., Nicphore Grgoras copiste et superviseur du Laurentianus 70, 5, Revue
dhistoire des textes s. II, 2 (2007) 2247.
Clucas L., The Trial of John Italos and the Crisis of Intellectual Values in Byzantium in the
Eleventh Century (Munich 1981).
Cohen-Skalli A. (ed.), Diodore de Sicile. Bibliothque historique, fragments des livres
VIX (Paris 2012).
, Les Excerpta Constantiniana: une conue daprs un modle
juridique?, JB 63 (2013) 3352.
Cohn L. (ed.), De Heraclide Milesio grammatico (Berlin 1884a).
, Untersuchungen ber die Quellen der Plato-Scholien, Jahrbcher fr Clas-
sische Philologie, Suppl. 13 (1884b) 773864.
, Aper [6], RE I/2 (1894a) 2697.
, Apion [3], RE I/2 (1894b) 28032806.
, Aristarchus and Aristophanes, RE II/1 (1895a) 873 and 1002.
, Arkadios [5], RE II/1 (1895b) 11531156.
, Apollonios [79: A. Anteros], RE II/1 (1895c) 135.
, Apollonios [80: A. der Sophist], RE II/1 (1895d) 135136.
, Apollonios [81: A. Dyskolos], RE II/1 (1895e) 136139.
, Apollonios [83], RE II/1 (1895f) 140.
, Archias [21], RE II/1 (1895g) 464.
, Archibios [6], RE II/1 (1895h) 466.
, Archibios [7], RE II/1 (1895i) 466.
1294 Bibliography
Dover K. (ed.), Aristophanes. Frogs, edited with introduction and commentary (Oxford
1994).
Dowden K., The Uses of Greek Mythology (London-New York 1992).
Downey G., Ancient Antioch (Princeton 1963).
Drachmann A. B., Scholia vetera in Pindari carmina (Leipzig 19031927).
(ed.), Isaac Tzetzae De metris Pindaricis commentarius (Copenhagen 1925).
, Die berlieferung des Cyrillglossars (Copenhagen 1936).
Drger P., Untersuchungen zu den Frauenkatalogen Hesiods (Stuttgart 1997).
(ed.), Apollodor, Gtter- und Heldensagen / Bibliotheke (Dsseldorf 2005).
Drossaart Lulofs H. J. (ed.), Aristotelis De somno et vigilia liber adiectis veteribus transla-
tionibus et Theodori Metochitae commentario (Leiden 1943).
, Aristotelis De insomniis et de divinatione per somnum (Leiden 1947).
Drossaart Lulofs H. J., Nicolaus Damascenus, On the Philosophy of Aristotle. Fragments
of the First Five Books Translated From the Syriac With An Introduction and Commen-
tary (Leiden 1969).
Drossaart Lulofs H. J., Poortman E. L. J. (eds.), Nicolaus Damascenus, De plantis, Five
Translations (Amsterdam-Oxford-New York 1989).
Drules P.-A., Dion de Pruse lecteur dHomre, Gaia 3 (1998) 5979.
Dubischar M., Survival of the Most Condensed? Auxiliary Texts, Communications
Theory, and Condensation of Knowledge, in Horster-Reitz (2010) 3967.
Dbner F. (ed.), Scholia in Theocritum (Parisiis 1849).
Dubuisson M., Le latin est-il une langue barbare?, Ktma 9 (1984) 5568.
, Homrologie et politique. Le cas dAristodmos de Nysa in Stemmata:
mlanges de philologie, dhistoire et darchologie grecques offerts Jules Labarbe,
eds. J. Servais, T. Hackens, B. Servais-Soyez (Lige Louvain-la-Neuve 1987) 1524.
Du C. (ed.), Recapturing a Homeric Legacy (Washington 2009).
Duffy J. M., Stephanus the Philosopher, A commentary on the Prognosticon of Hip-
pocrates, CMG XI 1.2 (Berlin 1983).
, Byzantine Medicine in the Sixth and Seventh Centuries: Aspects of Teaching
and Practice, DOP 38 (1984) 2127.
, Ioannes Alexandrinus, In Hippocratis Epidemiarum librum VI commentarii
fragmenta, CMG XI 1.4 (Berlin 1997).
Dufour R., Chrysippe. uvre philosophique (Paris 2004).
Duke E. (ed.), The Grammarian Seleukos of Alexandria. An Edition of the Fragments
(Diss. Oxford 1969).
Duminil M.-P., Hippocrate, Plaies, nature des os, cur, anatomie (Paris 1998).
Dunbar N., Aristophanes Birds. Edited with Introduction and Commentary (Oxford
1995).
Dntzer H. (ed.), De Zenodoti studiis Homericis (Gttingen 1848).
Dupont-Roc R., Lallot J., Aristote: La Potique. Texte, traduction, notes (Paris 1980).
Bibliography 1305
Eisenstein H., Einfhrung in die arabische Zoographie. Das tierkundliche Wissen in der
arabisch-islamischen Literatur (Berlin 1990).
El-Abbadi M., Life and Fate of the Ancient Library of Alexandria (Paris 19922).
, Demise of the Daughter Library, in El-Abbadi-Fathallah (2008) 8993.
El-Abbadi M., Fathallah O. M. (eds.), What Happened to the Ancient Library of Alexan-
dria? (Leiden-Boston 2008).
Eleopoulos N. X.,
(Athens 1967).
Eleuteri P., Francesco Filelfo copista e possessore di codici greci, in Harlfinger-Prato
(1991) 163179.
, I manoscritti greci della Biblioteca di Fozio, QS 51 (2000) 111156.
Ellis J. R., Philipp II and Macedonian Imperialism (London 1976).
Ellis W. M., Ptolemy of Egypt (London-New York 1994).
Else G. F., Burian P. (ed.), Plato and Aristotle on poetry, ed. with introd. and notes
(Chapel Hill 1986).
Empereur J.-Y., Le Phare dAlexandrie. La merveille retrouve (Paris 2004).
, The Destruction of the Library of Alexandria: an Archaeological Viewpoint,
in El-Abbadi-Fathallah (2008) 7588.
Erbse H., Untersuchungen zu den attizistischen Lexika (Berlin 1950).
, Homerscholien und hellenistische Glossare bei Apollonios Rhodios, Hermes
81 (1953) 163196.
, Beitrge zur berlieferung der Iliasscholien (Mnchen 1960).
, berlieferungsgeschichte der griechischen klassischen und hellenistischen
Literatur, in Geschichte der Textberlieferung der antiken und mittelalterlichen Lit-
eratur, I, eds. H. Hunger, O. Stegmller, H. Erbse, M. Imhof, K. Bchner, H. G. Beck,
H. Rdiger (Zrich 1961) 207283.
(ed.), Scholia Graeca in Homeri Iliadem (Scholia vetera), IVII (Berlin-New York
19691988).
, Beitrge zum Verstndnis der Odyssee (Berlin 1972).
, Zur normativen Grammatik der Alexandriner, Glotta 58 (1980) 236258.
, Zur Orthographie Homers, Glotta 72 (1994) 8297.
Ercoles M., Ancient Metrics, in Encyclopedia of Ancient Greek Language and
Linguistics (EAGLL) 3 vols., ed. G. K. Giannakis (Leiden-Boston 2013; online version
under: http://referenceworks.brillonline.com/browse/encyclopedia-of-ancient-greek
-language-and-linguistics; printed version: Leiden-Boston 2014).
Erler M., Philologia medicans. Wie die Epikureer die Texte ihres Meisters lasen, in Ver-
mittlung und Tradierung von Wissen in der griechischen Kultur, eds. W. Kullmann, J.
Althoff (Tbingen 1993) 281303.
Erler M., Schorn S. (eds.), Die griechische Biographie in hellenistischer Zeit. Akten des inter-
nationalen Kongresses vom 26.29. Juli 2006 in Wrzburg (Berlin-New York 2007).
1308 Bibliography
Ernst N., Die D-Scholien zur Odyssee (Diss. Kln 2006); available online at http://kups
.ub.uni-koeln.de/frontdoor.php?source_opus=1831.
Ernst U., Ordo, in Historisches Wrterbuch der Rhetorik, vol. 6 (Tbingen 2003) 416423.
Erren M., Arat und Aratea 19661992, Lustrum 36 (1994) 189284.
Erskine A., Culture and Power in Ptolemaic Egypt: the Museum and Library of Alexan-
dria, Greece & Rome 42 (1995) 3848.
(ed.), A Companion to the Hellenistic World (Malden, MA-Oxford 2003).
Esposito E., Aeschines, in Commentaria et Lexica Graeca in Papyris reperta (CLGP),
I 1.1, eds. G. Bastianini, M. Haslam, H. Maehler, F. Montanari, C. Rmer (Mnchen-
Leipzig 2004) 311.
, Fragments of Greek Lexicography in the Papyri, in Montanari-Perrone (2009)
255297.
(ed.), Aristophanes, in Commentaria et lexica Graeca in papyris reperta (CLGP),
I.1.4, 2nd ed., eds. G. Bastianini, M. Haslam, H. Maehler, F. Montanari, C. Rmer
(Berlin-New York 2012; 1st ed.: Mnchen-Leipzig 2006) (for lexicographical entries).
, Rec. Francesca Schironi, From Alexandria to Babylon, Eikasms 22 (2011)
524529.
Esposto C., Riano studioso di Omero, Eikasms 19 (2008) 560566.
Eusterschulte A., Guthknecht D., Mimesis, in Historisches Wrterbuch der Rhetorik,
vol. 5 (Tbingen 2001) 12321327.
Evans R., A History of Pergamum. Beyond Hellenistic Kingship (London 2012).
Fabricius C., Der sprachliche Klassizismus der griechischen Kirchenvter, JbAC 10
(1967) 187199.
Fait P., Aristotele. Confutazioni sofistiche (Organon VI), traduzione e commento (Roma-
Bari 2007).
Falkenhausen V. von, Tra Occidente e Oriente: Otranto in epoca bizantina, in Otranto
nel Medioevo, ed. H. Houben (Galatina 2007) 1360.
Falkner Th., Scholars versus Actors: Text and Performance in the Greek Tragic Scho-
lia, in Greek and Roman Actors. Aspects of an Ancient Profession, eds. P. Easterling,
E. Hall (Cambridge 2002) 342361.
Fantuzzi M., Il sistema letterario della poesia alessandrina nel III secolo a.C., in
Cambiano-Canfora-Lanza (1993) 3173.
, An Aristarchean Reading of Apollonius Argonautica, SemRom 3 (2000)
313323.
, Euripides (?) Rhesus 5658 and Homer Iliad 8.498501: Another Possible Clue
to Zenodotus Reliability, CPh 100 (2005) 268273.
, Sung Poetry: the Case of Inscribed Paeans, in Clauss-Cuypers (2010) 181196.
, Hunter R., Tradition and Innovation in Hellenistic Poetry (Cambridge 2004),
Engl. ed. of Muse e modelli. La poesia ellenistica da Alessandro Magno ad Augusto
(Roma-Bari 2002).
Bibliography 1309
, Theophrastus of Eresus: sources for his life, writings, thought and influence.
Vol. 8: Sources on rhetoric and poetics: (texts 666713), commentary (Leiden-Boston
2005).
Fortenbaugh W. W., Huby P. M., Sharples R. W., Gutas D. (eds.), Theophrastus of Ere-
sus, Sources for his Life, Writings, Thought and Influence, 1: Life, Writings, Various
Reports, Logic, Physics, Metaphysics, Theology, Mathematics (Leiden-New York-
Kln 1992).
Fortenbaugh W. W., Schtrumpf E. (eds.), Demetrius of Phalerum, Text, Translation and
Discussion (New Brunswick, NJ-London 2000).
, Dicaearchus of Messana, Text, Translation and Discussion (New Brunswick, NJ-
London 2001).
Fortenbaugh W. W., White S. A. (eds.), Lyco of Troas and Hieronymus of Rhodes, Text,
Translation and Discussion (New Brunswick, NJ-London 2004).
Fortuna S., Garofalo I., Lami A., Roselli A. (eds.), Sulla tradizione indiretta dei testi
medici greci, Atti del IV seminario internazionale di Siena, Certosa di Pontignano,
34 giugno 2011 (Pisa 2012).
Fournet J.-L., Hellnisme dans lgypte du VIe sicle. La bibliothque et luvre de Dioscore
dAphrodit. Vol. III (Cairo 1999).
Fowler D., Criticism as Commentary and Commentary as Criticism in the Age of Elec-
tronic Media, in Most (1999) 426442.
Fowler H. N., Plato. The Statesman; Philebus, with an English translation (London-
Cambridge, Mass. 1925).
, Plato. Parmenides; Greater Hippias; Lesser Hippias, with an English translation
(London-Cambridge, Mass. 1939).
, Plato, Theaetetus. Sophist, with an English translation, vol. VII (Cambridge,
Mass.-London 1987; 1st ed. 1921).
Fowler R. L., Herodotos and His Contemporaries, JHS 116 (1996) 6287.
, Genealogical thinking, Hesiods Catalogue, and the creation of the Hellenes,
PCPhS 44 (1999) 119.
(ed.), Early Greek Mythography, I (Oxford 2000).
, Early Histori and Literacy, in The Historians Craft in the Age of Herodotus, ed.
N. Luraghi (Oxford 2001) 95115.
, How to Tell a Myth: Genealogy, Mythology, Mythography, Kernos 19 (2006)
3546.
Franceschini A., Giovanni Aurispa e la sua biblioteca (Padova 1976).
Frnkel H., Noten zu den Argonautika des Apollonios (Munich 1968).
Frantz A., Pagan Philosophers in Christian Athens, PAPhS 119 (1975) 2938.
, The Athenian Agora XXIV. Late Antiquity, AD 267700 (Princeton 1988).
Fraser P. M., Eratosthenes of Cyrene, PBA 56 (1970) 175207.
, Ptolemaic Alexandria, 3 vols. (Oxford 1972).
Bibliography 1313
Frede M., The Origins of Traditional Grammar, in Historical and Philosophical Dimen-
sions of Logic, Methodology and Philosophy of Science: Part Four of the Proceed-
ings of the Fifth International Congress of Logic, Methodology and Philosophy of
Science (London, Ontario, Canada 1975), eds. R. E. Butts, J. Hintikka (Dordrecht-
Boston 1977) 5179.
, Principles of Stoic Grammar, in The Stoics, ed. J. M. Rist (Berkeley-Los Angeles-
London 1978) 2776, repr. in Essays in Ancient Philosophy (Oxford 1987) 301338.
, Essays in Ancient Philosophy (Oxford 1987a).
, The Origin of Traditional Grammar, in Essays in Ancient Philosophy, ed.
M. Frede (Oxford 1987) 338359 = [1987b].
Freese J. H., Aristotle, The Art of Rhetoric (Cambridge, Mass.-London 1975).
Freytag W., Allegorie, Allegorese, in Historisches Wrterbuch der Rhetorik, vol. 1 (Tbin-
gen 1992) 330393.
Fried M. N., Unguru S., Apollonius of Pergas Conica. Text, context, subtext (Leiden-
Boston-Kln 2001).
Friedlnder L. (ed.), Nicanoris reliquiae emendatiores (Knigsberg
1850) (Berlin 18572).
, Aristonici reliquiae emendatiores (Gttingen 1853).
Friedlein G., Procli Diadochi in primum Euclidis Elementorum librum commentarii
(Leipzig 1873).
Fritz K. von, Timaios, RE VI A/1 (1936) 12261228.
Frchtel L., Attizismus, RAC 1 (1950) 899902.
Fryde E. B., The Early Palaeologan Renaissance (1261c. 1360) (Leiden-Boston 2000).
Frye W., De Heraclidae Milesii studiis Homericis, Leipziger Studien zur classischen Phi-
lologie 6 (1883) 93188.
Fuchs F., Die hheren Schulen von Konstantinopel im Mittelalter (Berlin 1926).
Fhrer R., Schmidt M., Homerus redivivus, review of M. L. West (rec.), Homerus, Ilias,
vol. I (Stutgardiae et Lipsiae 1998), GGA 253 (2001) 132.
Fuhrmann M., Das systematische Lehrbuch: ein Beitrag zur Geschichte der Wissen-
schaften in der Antike (Gttingen 1960).
, Xenokritos [4], RE IX/2 (1983) 1533.
Funaioli G. (ed.), Grammaticae Romanae fragmenta (Leipzig 1907; repr. Stuttgart 1969).
Funaioli H., Habron [4], RE VII/2 (1912) 21552156.
, Hyperechios [4], RE IX/1 (1914) 281.
Funghi M. S., Esegesi di testi orfici. PDerveni, in Corpus dei Papiri Filosofici Greci e
Latini, parte III: Commentari (Firenze 1995) 565585.
Funke H., Poesia e storiografia, QS 23 (1986) 7193.
Furlan I., Marcon S., Eleuterii P. (eds.), Oppianus Apamensis, Tratado de caza (Valencia
2002).
1314 Bibliography
Fyfe W. H., Russell D. A., Longinus: On the Sublime, in Aristotle: Poetics, Longinus: On the
Sublime, Demetrius: On Style, eds. S. Halliwell, W. H. Fyfe, D. A. Russell, D. C. Innes,
(Cambridge, Mass.-London 1995).
Fyrigos A., Il fondamento bizantino dellumanesimo italiano, Studi sullOriente cris-
tiano 1 (1997) 4765.
(ed.), Barlaam Calabro: Luomo, lopera, il pensiero (Roma 2001).
GG = Grammatici Graeci (Lipsiae 18671910).
GL = Grammatici Latini ex recensione H. Keilii (Lipsiae 18551880).
Gabba E., Il latino come dialetto greco, in Miscellanea di studi alessandrini in memoria
di A. Rostagni (Torino 1963) 188194.
Gabrielsen V., Bilde P., Engberg-Pedersen T., Hannestad L., Zahle J. (eds.), Hellenistic
Rhodes. Politics, Culture, and Society (Aarhus 1999).
Gaede R., Demetrii Scepsii quae supersunt (Diss. Greifswald 1880).
Gagn R., Dancing letters: the Alphabetic Tragedy of Kallias, in Choral Mediations in
Greek Tragedy, ed. by R. Gagn and M. Govers Hopman (Cambridge 2013) 297316.
Gaiser K., Theophrast in Assos. Zur Entwicklung der Naturwissenschaft zwischen aka
demie und Peripatos (Heidelberg 1985).
Gaisford Th. (ed.), Poetae Minores Graeci II (Leipzig 1823).
, Georgii Choerobosci Dictata in Theodosii canones necnon epimerismi in Psalmos
III (Oxford 1842).
(ed.), Etymologicum Magnum (Oxford 1848).
Galn Vioque G., A missing Vatican manuscript page at Oxford, ByzZ 102 (2009)
627638.
Gallavotti C., Ledizione teocritea di Moscopulo, RFIC n.s. 12 (1934) 349369.
, Tracce della Poetica di Aristotele negli scolii omerici, Maia 21 (1969) 208214.
, La silloge tricliniana di Teocrito e un codice parigino-laurenziano, BollClass
s. III 3 (1982) 322.
, Nota sulla schedografia di Moscopulo e sui suoi precedenti fino a Teodoro
Prodromo, BollClass s. III 4 (1983) 335.
Gallo L., Per un riesame dei frammenti di Damaste di Sigeo, in Eoli ed Eolide tra
madrepatria e colonie, eds. A. Mele, M. L. Napolitano, A. Visconti (Napoli 2005)
145152.
Gamillscheg E., Johannes Chortasmenos als Restaurator des Wiener Dioskurides,
Biblos 55, 2 (2006) 3540.
Gansel C. (ed.), Textsorten und Systemtheorie (Gttingen 2008a).
, Systemtheoretische Perspektiven auf Textsorten: Vorbemerkungen, in Gansel
(2008a) 718 = [2008b].
, Textsortenlinguistik (Gttingen-Oakville, CT 2011).
Garcea A. (ed.),Caesars De analogia, Edition, Translation, and Commentary (Oxford-
New York 2012).
Bibliography 1315
Garcia Novo E., Les scholies marginales au trait de Galien De inaequali intemperie
dans le manuscrit grec Phillips 4614, in Garzya-Jouanna (1999) 175183.
Garin E., Il pensiero pedagogico dellUmanesimo (Firenze 1958).
Garofalo I., Erasistrati fragmenta, collegit et disposuit I. G. (Pisa 1988).
, Prolegomena alledizione dellAnonymus Parisinus Darembergii sive Fuchsii,
in Garzya (1992) 9195.
, La traduzione araba dei compendi Alessandrini delle opere del canone di
Galeno. Il compendio dell Ad Glaudonem, Medicina nei secoli 6 (1994) 329348.
, Anonymi medici De morbis acutis et chroniis (Leiden 1997).
, Le commentaire ravennate au De pulsibus ad tirones de Galien: le commen-
taire et ses sources, in Deroux (1998) 382392.
, La traduzione araba del commento di Ioannes Grammatikos al De pulsibus di
Galeno, in Garzya-Jouanna (1999) 185218.
, Il Sunto di Ioannes Grammatikos delle opere del Canone di Galeno, in Studi
su Galeno. Scienza, filosofia, retorica e filologia Atti del seminario, Firenze 13 novem-
bre 1998, ed. D. Manetti (Firenze 2000) 135151 = [2000a].
, Una nuova opera di Galeno in traduzione araba, la Synopsis del De methodo
medendi di Galeno, SCO 57 (2000b) 819.
, Note sulla tradizione alessandrina del De differentiis febrium di Galeno, in
Garzya-Jouanna (2003) 149164 = [2003a].
, I sommari degli Alessandrini, in Garofalo-Roselli (2003) 203231 = [2003b].
, Gli scoli al De sectis e allAd Glauconem nel Parisinus suppl. gr. 634, Galenos 2
(2008) 61105.
, Gli estratti di Ali ibn Ridwan dal commento di Galeno agli Umori di Ippocrate,
Galenos 5 (2011) 113169.
(ed.), Galeno, Del non affliggersi, in Galeno, Lanima e il dolore. De indolentia. De
propriis placitis, eds. I. Garofalo, A. Lami (Milano 2012) = [2012a].
, Il commento di Galeno al De alimento e gli estratti di ibn Ridwan, Galenos 6
(2012b) 123164.
Garofalo I., Lami A., Galeno, Lanima e il dolore. De indolentia. De propriis placitis, a cura
di Ivan Garofalo e Alessandro Lami (Milano 2012).
Garofalo I., Roselli A. (eds.), Galenismo e medicina tardoantica. Fonti greche, latine e
arabe, Atti del Seminario Internazionale di Siena, Certosa di Pontignano9 e
10 settembre 2002 (Napoli 2003).
Grtner H. A., Zenon [9], RE X A (1972) 140142.
, Thoas [11] RE suppl. XIV (1974) 760.
, Stephanos [7: St. von Byzanz], DNP 11 (2001) 958959.
Garzya A., Il mandarino e il quotidiano (Napoli 1983).
(ed.), Tradizione e ecdotica dei testi medici tarodantichi e bizantini, Atti del Con-
vegno internazionale Anacapri 2931 ottobre 1990 (Napoli 1992).
1316 Bibliography
(ed.), Storia e ecdotica dei testi medici greci, Atti del II Convegno Inernazionale,
Parigi 2426 maggio 1994 (Napoli 1996).
Garzya A., Jouanna J. (eds.), I testi medici greci. Tradizione e ecdotica, Atti del III Con-
vegno Internazionale, Napoli 1518 ottobre 1997 (Napoli 1999).
, Trasmissione e ecdotica dei testi medici greci, Atti del IV Convegno Internazio-
nale, Parigi 1719 maggio 2001 (Napoli 2003).
Gastgeber C., Aus der Bibliothek des Ioannes Chortasmenos, in Alethes philia, II, eds.
M. DAgostino, P. Degni (Spoleto 2010) 409434.
Gtje O., Die Anschlussfhigkeit sprachlicher Kommunikation in Semiosphren, in
Gansel (2008) 203216.
Gaul N., Andronikos Komnenos, Prinz Belthandros und der Zyklop, ByzZ 96 (2003)
623660.
, Frstin Anna, bedenke, in Zwischen Polis, Provinz und
Peripherie, ed. L. M. Hoffmann (Wiesbaden 2005) 663704.
, The twitching shroud. Collective construction of paideia in the circle of
Thomas Magistros, S&T 5 (2007) 263340.
, Moschopulos, Lopadiotes, Phrankopoulos (?), Magistros, Staphidakes, in
Lexicologica Byzantina, eds. E. Trapp, S. Schnauer (Gttingen 2008) 163196.
, Thomas Magistros und die sptbyzantinische Sophistik (Wiesbaden 2011).
Gautier P. (ed.), Michel Italikos, Lettres et discours (Paris 1972).
Geanakoplos D. J., Theodore Gaza, a Byzantine Scholar of the Paleologan Renaissance
in the Italian Renaissance, Medievalia et humanistica 12 (1984) 6181.
Geerlings W., Die lateinisch-patristischen Kommentare, in Geerlings-Schulze (2002)
114.
Geerlings W., Schulze Ch. (eds.), Der Kommentar in Antike und Mittelalter: Beitrge
zu seiner Erforschung, Clavis commentariorum antiquitatis et Medii Aevi vol. 2
(Leiden-Boston-Kln 2002).
, Der Kommentar in Antike und Mittelalter. Band 2: Neue Beitrge zu seiner
Erforschung, Clavis commentariorum antiquitatis et Medii Aevi vol. 3 (Leiden-
Boston-Kln 2004).
Geffcken J., Zur Entstehung und zum Wesen des griechischen wissenschaftlichen
Kommentars, Hermes 67 (1932) 397412.
Gelzer Th., Klassizismus, Attizismus, Asianismus, in Flashar (1979) 141.
Gemelli Marciano M. L., Le metamorfosi della tradizione: mutamenti di significato e
neologismi nel Peri physeos di Empedocle (Bari 1990).
, Die Vorsokratiker, Band I. Auswahl der Fragmente und Zeugnisse, bersetzung
und Erluterungen (Dsseldorf 2007).
Genette G., Palimpsests: Literature in the Second Degree, Stages vol. 8 (Lincoln NE-
London 1997a), engl. transl. (by C. Newman, C. Doubinsky) of Palimpsestes: la
litterature au second degre (Paris 1982).
Bibliography 1317
Gibson S., Aristoxenus of Tarentum and the birth of musicology (New York-London 2005).
Gielen E., Joseph the Philosopher, an Outstanding Outsider: Philosophy and Rhetoric
at the Court of Andronicus II, in Basileia: Essays on Imperium and Culture in Honour
of E.M. and M.J. Jeffreys, eds. G. Nathan, L. Garland (Brisbane 2011) 205215.
, Ad maiorem Dei gloriam. Joseph Rhakendytes synopsis of Byzantine learn-
ing, in Knig-Woolf (2013) 259276.
Giese C., De Theone grammatico (Diss. Mnster 1867).
Gigante M., Teodoro Metochites. Saggio critico su Demostene e Aristide (Milano-Varese
1969).
, Scetticismo e Epicureismo. Per lavviamento di un discorso storiografico (Napoli
1981a).
, Scritti sulla civilt letteraria bizantina (Napoli 1981b).
, Kepos e Peripatos. Contributo alla storia dellAristotelismo antico (Napoli
1999).
Gignac F. T., A Grammar of the Greek Papyri of the Roman and Byzantine Periods, i: Pho-
nology (Milan 1976).
Gill C., Whitmarsh T., Wilkins J. (eds.), Galen and the World of Knowledge (Cambridge
2009).
Glck M., Priscians Partitiones und ihre Stellung in der sptantiken Schule (Hildesheim
1967).
Goddio F., Clauss M. (eds.), gyptens versunkene Schtze (Berlin 2006).
Goel K., Mayer J.-G., Staub H.-H., Das Darmstdter Johannes-Alexandrinus-Fragment,
AGM 84 (2000) 201222.
Goettling C. G. (ed.), Theodosii Alexandrini Grammatica (Leipzig 1822).
Goldhill S., Programme notes, in Performance Culture and Athenian Democracy, eds.
S. Goldhill, R. Osborne (Cambridge 1999) 129.
Goldsmith J. A., Autosegmental and Metrical Phonology (Oxford 1990).
Golitsis P., Un commentaire perptuel de Georges Pachymre la Physique dAristote,
ByzZ 100 (2007) 637676.
, Georges Pachymre comme didascale, JByz 58 (2008) 5368.
, La date de composition de la Philosophia de Georges Pachymre et quelques
prcisions sur la vie de lauteur, REByz 67 (2009) 209215.
, Copistes, lves et rudits: la production de manuscrits philosophiques autour
de Georges Pachymre, in Bravo Garca-Prez Martn (2010) 157170, 757768.
Gpferich S., Textsorten in Naturwissenschaften und Technik, Forum fr Fremd-
sprachen-Forschung vol. 27 (Tbingen 1995).
Goslings F., Observationes ad scholia in Thucydidem (Diss. Leyden 1874).
Gottschalk H. B., Heraclides of Pontus (Oxford 1980).
, Aristotelian Philosophy in the Roman World from the Time of Cicero to the
End of the Second Century AD, ANRW 2.36.2 (Berlin-New York 1987) 10791174.
Bibliography 1319
Griffith M., Public and private in early Greek institutions of education, in Education
in Greek and Roman Antiquity, ed. Yun Lee Too (Leiden-Boston-Kln 2001) 2384.
Griffiths F. T., review of Cameron (1995), AJPh 118 (1997) 339343.
Grintser N. P., Grammar of Poetry (Aristotle and Beyond), in Grammatical Theory and
Philosophy of Language in Antiquity, eds. P. Swiggers, A. Wouters (Leuven 2002)
7199.
Grmek M. D., Gourevitch D., Aux sources de la doctrine mdicale de Galien:
lenseignement de Marinus, Quintus et Numisianus, ANRW II 37, 2 (1994) 14911528.
Gronewald M., Daniel R. W., Ein neuer Sappho-Papyrus, ZPE 147 (2004a) 18.
, Nachtrag zum neuen Sappho-Papyrus, ZPE 149 (2004b) 14.
Groningen B. A. van, , Mnemosyne 16 (1963) 117.
, Euphorion (Amsterdam 1977).
Grossardt P., Stesichoros zwischen kultischer Praxis, mythischer Tradition und eigenem
Kunstanspruch: zur Behandlung des Helenamythos im Werk des Dichters aus Himera
(Tbingen 2012).
Grube G. M. A., The Greek and Roman Critics (Toronto 1965).
Gruen E. S., Heritage and Hellenism. The Reinvention of Jewish Tradition (Berkeley-Los
Angeles 1998).
, Culture as Policy. The Attalids of Pergamon, in From Pergamon to Sperlonga,
eds. N. de Grummond, B. Ridgway (Berkeley 2000) 1731.
, Jews and Greeks, in Erskine (2003) 264279.
, Greeks and Non-Greeks, in Bugh (2006) 295314.
, Jewish Literature, in Clauss-Cuypers (2010) 415428.
Grnbart M., Byzantinisches Gelehrtenelend, in Zwischen Polis, Provinz und Periph-
erie, ed. L. M. Hoffmann (Wiesbaden 2005) 413426.
Grukova J., Untersuchungen zu den griechischen Palimpsesten der sterreichischen
Nationalbibliothek (Vienna 2010).
Grukova J., Bannert H., Demosthenica libris manu scriptis tradita (Wien 2014).
Grynaeus S., Camerarius J., Claudii Ptolemaei Magnae constructionis, id est Perfectae
coelestium motuum pertractationis, lib. XIII. Theonis Alexandrini in eosdem Com-
mentariorum lib. XI (Basle 1538).
Guardasole A. (ed.), Eraclide di Taranto. Frammenti (Napoli 1997).
Gudeman A., Grundriss der Geschichte der Klassischen Philologie (Leipzig-Berlin 1909).
, Grammatik, RE VII/2 (1912a) 17801811.
, Helladios [3], RE VIII/1 (1912b) 102103.
, Herakleon [5], RE VIII/1 (1912c) 512514.
, Herakleon [6: H. von Ephesos], RE VIII/1 (1912d) 514515.
, Herennios [2], RE VIII/1 (1912e) 650661.
, Kallistratos nr. 38, RE X/2 (1919) 17381748.
, Scholien, RE II A/1 (1921) 625705.
, , RE XI/2 (1922a) 19121915.
Bibliography 1321
Haas W. (ed.), Die Fragmente der Grammatiker Tyrannion und Diokles, SGLG 3 (Berlin-
New York 1977) 79184.
Habicht Ch., Athens from Alexander to Antony (Harvard 1997).
, Athnes hellnistique. Histoire de la cit dAlexandre le Grand Marc Antoine
(Paris 2000), French ed. of Athen. Die Geschichte der Stadt in der hellenistischen Zeit
(Mnchen 1995).
Hadas M., A History of Greek Literature (New York 1950).
Hadot I., Les introductions aux commentaires exgtiques chez les auteurs pla-
toniciens et les auteurs chrtiens, in Les rgles de linterprtation, ed. L. Tardieu
(Paris 1987) 99122.
, Simplicius, Commentaire sur les Catgories, Fasc. I (Leiden-New York-Kben-
havn-Kln 1990).
, Simplicius, Commentaire sur le Manuel dpictte (Leiden 1996).
, Der fortlaufende philosophische Kommentar in der Antike, in Geerlings-
Schulze (2002) 183199.
Haeniche E., Quaestiones Apollodoreae II (Diss. Halis Saxonum 1875).
Hagedorn U., Das sogenannte Kyrill-Lexikon in der Fassung der Handschrift E (Codex
Bremensis G 11) (2005); available online at http://kups.ub.uni-koeln.de/frontdoor
.php?source_opus=1813.
Hgg T., Photios als Vermittler antiker Literatur (Uppsala 1975).
, Photius as a Reader of Hagiography, DOP 53 (1999) 4358.
Haikal F. M., Private Collections and Temple Libraries in Ancient Egypt, in El-Abbadi-
Fathallah (2008) 3954.
Hajd K. (ed.), Ps.-Herodian, De figuris. berlieferungsgeschichte und kritische Ausgabe
(Berlin-New York 1998).
Haldon J., Byzantium. A History (Stroud 2000).
Halliwell S., Aristotles Poetics (London 1986), repr. with new preface (Chicago 1998;
London 20003).
(ed.), The Poetics of Aristotle, Translation and commentary (London 1987).
, Aristotles Poetics, in The Cambridge History of Literary Criticism: Vol. 1: Classi-
cal Criticism, ed. G. A. Kennedy (Cambridge 1989) 149183.
, Aristotle. Poetics, ed. and transl. by S. H.; Longinus. On the Sublime, transl.
W. H. Fyfe, rev. D. Russell; Demetrius. On Style, ed. and transl. D. C. Innes, based on
W. Rhys Roberts (Cambridge, Mass.-London 1995).
, The subjection of muthos to logos: Platos citations of the poets, CQ 50 (2000)
94112.
, The Aesthetics of Mimesis: Ancient Texts and Modern Problems (Princeton-
Oxford 2002).
, The Theory and Practice of Narrative in Plato, in Narratology and Interpreta-
tion: The Content of Narrative Form in Ancient Literature, eds. J. Grethlein, A. Renga-
kos (Berlin-New York 2009) 1541.
Bibliography 1323
Harlfinger D., Rashed M., Leonzio Pilato fra aristotelismo bizantino e scolastica latina,
Quaderni petrarcheschi 1213 (20022003) 277293.
Harlfinger D., Reinsch D., Die Aristotelica des Parisinus Gr. 1741, Philologus 114 (1970)
2850.
Harriott R. M., Poetry and Criticism before Plato (London 1969).
Harris E. M., More Chalcenteric Negligence, CPh 84 (1989) 3644.
Harris W. V., Ancient Literacy (Cambridge, Mass. 1989).
Haslam M. W., A New Papyrus Text of Apollonius Sophista, ZPE 49 (1982) 3138.
, A New Papyrus of the Mythographus Homericus, BASP 27 (1990) 3136.
, Homeric Readings lost and found, CPh 89 (1992) 322325.
, The Homer Lexicon of Apollonius Sophista I: Composition and Constituents;
II: Identity and Transmission, CPh 89 (1994) 145; 107119.
, Homeric Papyri and Transmission of the Text, in A New Companion to Homer,
eds. I. Morris, B. B. Powell (Leiden-New York-Kln 1997) 55100.
Hatab L. J., Writing Knowledge in the Soul: Orality, Literacy, and Platos Critique of
Poetry, Epoch: A Journal for the History of Philosophy 11 (2007) 319332.
Hatzimichali M., Greek Scholarship of the Early Imperial Period: the Work of Pamphilus
of Alexandria (Diss. Oxford 2006).
, Ashes to ashes? The library of Alexandria after 48 BC, in Ancient Libraries, eds.
J. Knig, K. Oikonomopoulou, G. Woolf (Cambridge 2013) 167182 = [2013a].
, Encyclopaedism in the Alexandrian library, in Encyclopaedism from the Antiq-
uity to the Renaissance, eds. J. Knig, G. Woolf (Cambridge 2013) 6483 = [2013b].
Haupt M., Excerpta ex Timothei Gazaei libris de animalibus, Hermes 3 (1869) 130.
, Index lectionum aestivarum 1871: De Minucio Pacato qui dicitur et Irenaeus,
in Mauricii Hauptii Opuscula, vol. II (Leipzig 1876) 435440.
Havelock E. A., The Literate Revolution in Greece and its Cultural Consequences (Prince
ton, NJ 1982).
, The linguistic task of the Presocratics, in Language and Thought in Early
Greek Philosophy, ed. K. Robb (La Salle, Ill. 1983) 782.
Haverling G., Un nuovo frammento della traduzione ravennate degli Aforismi di
Ippocrate e del vecchio commento cosiddetto oribasiano, IMU 38 (1995) 307317.
Heath M., Unity in Greek Poetics (Oxford 1989).
, Stasis-theory in Homeric Commentary, Mnemosyne 46 (1993) 356363.
, Was Homer a Roman?, in Papers of the Leeds International Latin Seminar,
10: Greek Poetry, Drama, Prose, Roman Poetry, eds. F. Cairns, M. Heath (Leeds 1998)
2356.
, Longinus, On Sublimity, PCPhS 45 (1999) 4374.
, Theon and the history of progymnasmata, GRBS 43 (20022003) 129160.
, Pseudo-Dionysius Art of Rhetoric 811: figured speech, declamation, and criti-
cism, AJPh 124 (2003) 81105.
Bibliography 1325
Jacob J., Ut pictura poesis, in Historisches Wrterbuch der Rhetorik, vol. 9 (Tbingen
2009) 9971006.
Jacoby F., Hekataios von Milet, RE VII/2 (1912a) 26672750.
, Herodoros [4], RE VIII/1 (1912b) 980987.
, Herodotos, RE Suppl. II (1913) 205520, repr. in Griechische Historiker (Sutt-
gart 1956) 7164.
, Atthis. The Local Chronicles of Ancient Athens (Oxford 1949).
(ed.), Die Fragmente der Griechischen Historiker, IIIb (Supplement), I (Leiden
1954).
, Die Fragmente der griechischen Historiker, 1a, Kommentar (Leiden 1957).
Jacques A. (ed.), Nicandre. Oeuvres (Paris 2002).
Jahn A. (ed.), Joannis Glycae opus de vera syntaxeos ratione (Bern 1839).
James L., Byzantium: a Very, Very Short Introduction, in A Companion to Byzantium,
ed. L. James (West Sussex 2010) 18.
(ed.), Constantine of Rhodes. On Constantinople and the Church of the Holy
Apostles (Ashgate 2012).
Janko R., Homer, Hesiod and the Hymns: Diachronic Development in Epic Diction (Cam-
bridge 1982).
, Aristotle on Comedy: Towards a Reconstruction of Poetics II (London 1984).
, Aristotle: Poetics I, with the Tractatus Coislinianus. A Hypothetical Recon
struction of Poetics II, The Fragments of the On Poets. Translated with Notes
(Indianapolis-Cambridge 1987).
, The Iliad: A Commentary, ed. G. S. Kirk, Vol. IV: books 1316 (Cambridge
1992).
(ed.), Philodemus On Poems, book 1, ed. with intr., transl., and comm. (Oxford
2000).
, Seduta di chiusura, in Omero tremila anni dopo, ed. F. Montanari, coll.
P. Ascheri (Roma 2002) 653671.
, Socrates the Freethinker, in Blackwells Companion to Socrates, eds. S. Ahbel-
Rappe, R. Kamtekar (Malden, MA 2009) 4862.
(ed.), Philodemus On Poems, books 34 with the fragments of Aristotle On poets,
ed. with intr., transl., and comm. (Oxford 2011).
Janz T., The scribes and the date of the Vat. gr. 1291, Miscellanea Bibliothecae Apostoli-
cae Vaticanae 10 (2003) 159180.
Jeffery L. H., The Local Scripts of Archaich Greece. A Study of the Origin of the Greek
Alphabet and its Development from the Eighth to the Fifth Centuries B.C. (Oxford
1961).
Jeffreys E. M., The Attitude of Byzantine Chroniclers towards Ancient History, Byzan-
tion 69 (1979) 199238.
1334 Bibliography
Kaldellis A., The Works and Days of Hesychios the Illoustrios of Miletos, GRBS 45
(2005) 381403.
, Mothers and Sons, Fathers and Daughters: the Byzantine Family of Michael
Psellos (Notre Dame 2006).
, Hellenism in Byzantium (Cambridge 2007).
, Classical Scholarship in Twelfth-Century Byzantium, in Medieval Greek Com-
mentaries on the Nicomachean Ethics, eds. Ch. Barber, D. Jenkins (Leiden-Boston
2009a) 140 = [2009a].
, The Christian Parthenon: Classicism and Pilgrimage in Byzantine Athens
(Cambridge 2009b).
Kalivoda G. (et al.), Rhetorik, in Historisches Wrterbuch der Rhetorik, vol. 7 (Tbingen
2005) 14231740.
Kallendorf C., Brevitas, in Historisches Wrterbuch der Rhetorik, vol. 2 (Tbingen 1994)
5360 = [1994a].
, Enarratio poetarum, in Historisches Wrterbuch der Rhetorik, vol. 2 (Tbingen
1994) 11241134 = [1994b].
(et al.), Erhabene, das, in Historisches Wrterbuch der Rhetorik, vol. 2 (Tbin-
gen 1994) 13571389.
Kambylis A., Eustathios von Thessalonike. Prooimion zum Pindarkommentar (Gttingen
1991a).
, Eustathios ber Pindars Epinikiendichtung: ein Kapitel der klassischen Philolo-
gie in Byzanz (Gttingen 1991b).
Karla G. A., Maximos Planoudes: Dr. Bowdler in Byzanz? Zensur und Innovation im
spten Byzanz, Classica et Mediaevalia 57 (2006) 213238.
Karp A. J., Homeric origins of ancient rhetoric, Arethusa 10 (1977) 237258.
Karpozilos A., (Ioannina
1982).
Kassel R., Aristotelis De arte poetica liber (Oxford 1965).
, Aristotelis Ars rhetorica (Berlin-New York 1976).
Kassel R., Austin C., Poetae Comici Graeci vol. 6.2 (Berlin 1998).
Kaster R. A., Notes on primary and secondary schools in Late Antiquity, TAPhA 113
(1983) 323346.
, Guardians of Language: The Grammarian and Society in Late Antiquity
(Berkeley-Los Angeles-London 1988).
, C. Suetonius Tranquillus, De Grammaticis et Rhetoribus. Edited with a transla-
tion, introduction and commentary (Oxford 1995).
, Geschichte der Philologie in Rom, in Einleitung in die lateinische Philologie,
ed. F. Graf (Stuttgart-Leipzig 1997) 117.
Katsaros B., Literary and Intellectual Life in Byzantine Thessaloniki, in Queen of the
Worthy II, ed. I. K. Hassiotis (Thessaloniki 1997) 178213.
Bibliography 1337
,
, in The Empire in Crisis (?): Byzantium in the
11th Century, ed. V. N. Vlyssidou (Athens 2003) 443471.
Katzung P. G., Die Diapeira in der Iliashandlung (Diss. Frankfurt 1960).
Kavrus-Hoffmann N., From Pre-Boulete to Boulete, in Bravo Garca-Perez Martn
(2010) 5566.
Kazhdan A., Cutler A., Continuity and Discontinuity in Byzantine History, Byzantion
52 (1982) 429478.
Kazhdan A., Epstein A. W., Change in Byzantine Culture in the Eleventh and Twelfth
Centuries (Berkeley 1985).
Keaney J. J., Moschopoulos and Harpocration, TAPhA 100 (1969) 201207.
, Moschopulea, ByzZ 64 (1971) 303321.
, Notes on Moschopoulos and Aristophanes-scholia, Mnemosyne s. IV 25 (1972)
123128.
, Harpocration, Lexeis of the Ten Orators (Amsterdam 1991).
Kemmann A., Evidentia, Evidenz, in Historisches Wrterbuch der Rhetorik, vol. 3
(Tbingen 1996) 3347.
Kemp A., The Tekhn Grammatik of Dionysius Thrax. Translated into English, Histo-
riographia Linguistica XII 23 (1986) 343363; repr. in The History of Linguistics in
the Classical Period, ed. D. J. Taylor (Amsterdam 1987) 169189.
, The Emergence of Autonomous Greek Grammar, in Geschichte der Sprach-
theorie. II, Sprachtheorien der abendlndischen Antike, ed. P. Schmitter (Tbingen
1991) 302333.
Kennedy G. A., Greek Rhetoric under Christian Emperors (Princeton 1983).
(ed.), The Cambridge History of Literary Criticism. Vol. 1: Classical Criticism
(Cambridge 1989a).
, Language and meaning in Archaic and Classical Greece, in The Cambridge
History of Literary Criticism. Vol. 1: Classical Criticism, ed. G. A. Kennedy (Cambridge
1989a) 7891 = [1989b].
, A New History of Classical Rhetoric (Princeton 1994).
Kenyon F. G. (ed.), Classical Texts from Papyri in the British Museum (London 1891).
, Two Greek School-Tablets, JHS 29 (1909) 2940.
Kessel G., The Syriac Epidemics and the Problem of Identification, in Pormann (2012)
93123 = [2012a].
, Triseudemon maximus noster sophista. The evidence of one Syriac text for
the identification of a source used in John of Alexandrias In Epid. VI, in Fortuna-
Garofalo-Lami-Roselli (2012) 123137 = [2012b].
Keydell R., Die Dichter mit Namen Peisandros, Hermes 10 (1935) 301302; repr. in
Kleine Schriften (Leipzig 1982) 361362.
, Peisandros [1113], RE XIX (1937) 144145.
1338 Bibliography
Kraus P., Walzer R. (eds.), Galeni Compendium Timaei Platonis aliorumque dialogorum
synopsis quae extant fragmenta (London 1951).
Krevans N., Sens A., Language and Literature, in Bugh (2006) 186207.
Kroehnert O., Canonesne poetarum, scriptorum, artificum per antiquitatem fuerunt?
(Diss. Knigsberg 1897).
Kroll W, Heron [4], RE VIII/1 (1912) 992.
, Ioannes [21: I. Philoponos], RE IX/2 (1916) 17931795.
, Iulius [530: L. Iulius Vestinus], RE X/1 (1918a) 872.
, Demo (2a), RE Suppl. III (1918b) 331333.
Krumbacher K., Geschichte der byzantinischen Literatur (Berlin 1897).
Kudlien F., Hippokrates-Rezeption im Hellenismus, in Baader-Wienau (1989)
355376.
Kuehn C., Anastasius of Sinai: Biblical Scholar, ByzZ 103 (2010) 5581.
Khn C. G., Claudii Galeni opera omnia (Leipzig 18211833).
(ed.), Claudii Galeni opera omnia, vol. 19. (Leipzig 1830; repr. Hildesheim 1965).
Khnert F., Allgemeinbildung und Fachbildung in der Antike (Berlin 1961).
Kullmann W., Die Probe des Achaierheeres in der Ilias, MH 12 (1955) 253273.
, Bipartite Science in Aristotles Biology, in Biologie, Logique et Mtaphysique
chez Aristote, eds. D. Devereux, P. Pellegrin (Paris 1990) 335347.
, Aristoteles und die moderne Wissenschaft (Stuttgart 1998).
, Zoologische Sammelwerke in der Antike, in Geschichte der Mathematik und
der Naturwissenschaften in der Antike, Vol. 1: Biologie, ed. G. Whrle (Stuttgart 1999)
181198.
, Aristoteles. ber die Teile der Lebewesen, bersetzt und erlutert, Aristoteles.
Werke in deutscher bersetzung, ed. H. Flashar, 17,1 (Berlin 2007).
Kullmer H., Die Historiai des Hellanikos von Lesbos: ein Rekonstruktiosversuch (Diss.
Leipzig 1902).
Krschner W., Die Lehre des Grammatikers Dionysius, in Swiggers-Wouters (1996)
177215.
Kustas G. L., The Literary Criticism of Photius. A Christian Definition of Style, Hel-
lenika 17 (1962) 132169.
, Studies in Byzantine Rhetoric (Thessaloniki 1972).
La Roche J., Die homerische Textkritik im Altertum (Leipzig 1866).
Labarrire J.-L., Langage, vie politique et mouvement des animaux. tudes aristotli
ciennes (Paris 2004).
Labowsky L., Bessarions Library and the Biblioteca Marciana (Roma 1979).
Lachmann K., ber Zenodots Tagberechnung der Ilias, in Betrachtungen ber Homers
Ilias (Berlin 18652) 9096.
Lackner W., Eine Abhandlung des Konstantinos Stilbes zum Problem der Pseudo-
Chrysostomica, JByz 34 (1984) 107121.
1342 Bibliography
Lang M. L., Graffiti and dipinti, in The Athenian Agora: Results of excavations cond. by
the American School of Class. Stud. at Athens. Vol. XXI (Princeton, N.J. 1976).
Lange L., Das System der Syntax des Apollonios Dyskolos (Gttingen 1852).
Langholf V., Die parallelen Texte in Epidemien V und VII, in Corpus Hippocraticum,
Actes du Colloque Hippocratique de Mons (2226 sept. 1975), ed. R. Joly (Mons
1977) 264274.
, Kallimachos, Komdie und die hippokratische Frage, MHJ 21 (1986) 330.
, Beobachtungen zur Struktur einiger Traktate des Corpus hippocraticum,
Sudhoffs Archiv 73 (1989) 6477.
, Medical Theories in Hippocrates. Early Texts and the Epidemics, (Berlin-New
York 1990).
, Structure and Genesis of Some Hippocratic Treatises, in Magic and Rational-
ity in Ancient Near Eastern and Graeco-Roman Medicine, eds. H. F. J. Horstmanshoff,
M. Stol (Leiden-Boston 2004) 219274.
Lanza D., Scientificit della lingua e lingua della scienza in Grecia, Belfagor 27 (1972)
392429.
Laourdas B.,
(Thessaloniki 1960).
Lapini W., Il nome, la maschera e lidiota, Sandalion 15 (1992) 53101.
, Sguardi su Favorino, in C. A. Ciampi, Favorino dArelate. , a cura di
F. Montanari, premessa di S. Settis, saggi di G. Bastianini e W. Lapini (Pisa 2011) 756.
, Testi frammentari e critica del testo. Problemi di filologia filosofica greca (Roma
2013).
Larchet J.-C. (ed.), La vie et luvre thologique de Georges/Grgoire II de Chypre (Paris
2012).
Laspia P., Larticolazione linguistica. Origini biologiche di una metafora (Roma 1997).
Lasserre F., Die Fragmente des Eudoxos von Knidos (Berlin 1966).
, Llgie de lhutre (P.Louvre inv. 7733 v indit), QUCC 19 (1975) 145176.
Lasserre F., Livadaras N. (eds.), Etymologicum Magnum Genuinum, Symeonis Etymo-
logicum una cum Magna Grammatica, Etymologicum Magnum Auctum, III (-)
(Rome 1976Athens 1992).
Lasserre F., Mudry Ph. (eds.), Formes de pense dans la Collection Hippocratique, Actes
du IVe Colloque International Hippocratique, Lausanne, 2126 septembre 1981
(Genve 1983).
Latacz J., Einleitung: zur Homer-Kommentierung. Von den Anfngen bis zu die-
sem Kommentar, in Homers Ilias. Gesamtkommentar. Prolegomena, ed. J. Latacz
(Mnchen-Leipzig 2000) 126.
, Homers Ilias, Gesamtkommentar, Band II.2 (Munich-Leipzig 2003).
Latte K., Zur Zeitbestimmung des Antiatticista, Hermes 50 (1915) 373394; repr. in
Kleine Schriften zu Religion, Recht, Literatur und Sprache der Griechen und Rmer,
eds. O. Gigon, W. Buchwald, W. Kunkel (Mnchen 1968) 612630.
Bibliography 1345
, review of Canfora (1990), in New York Review of Books, June 14 (1990) 2729.
, The Pride of Halicarnassos, ZPE 124 (1999) 114.
, Curses and Divine Anger in Early Greek Epic: The Pisander Scholion, CQ 52
(2002) 114; repr. in The Further Academic Papers of Sir Hugh Lloyd-Jones (Oxford
2005) 1835.
(ed.), Supplementum Supplementi Hellenistici (Berlin-New York 2005).
Lloyd-Jones H., Parsons P. (eds.), Supplementum Hellenisticum (Berlin-New York
1983).
Lodge R. C., Platos theory of education (London 1950).
Loenertz R.-J., Manuel Kalekas, sa vie et ses oeuvres daprs ses lettres et ses apologies
indites, Archivum Fratrum Praedicatorum 17 (1947) 195207.
(ed.), Demetrius Cydones. Correspondance III (Citt del Vaticano 19561960).
Lohmann D., Die Komposition der Reden in der Ilias (Berlin 1970).
Lohr Ch., Theodorus Metochites, Paraphrasis in Aristotelis universam naturalem philoso-
phiam, bersetzt von Gentianus Hervetus, Neudruck der ersten Ausgabe Basel 1559
mit einer Einleitung von Ch. L. (Stuttgart-Bad Cannstatt 1992).
Lomiento L., review of Prauscello (2006), BMCR 2007.04.57.
Long A. A., Language and Thought in Stoicism, in Problems in Stoicism, ed. A. A. Long
(London 1971) 75113.
, Timon of Phlius, Pyrrhonist and Satirist, PCPhS 24 (1978) 6891.
, Stoic Readings of Homer, in Homers Ancient Readers: The Hermeneutics
of Greek Epics Earliest Exegetes, eds. R. Lamberton, J. J. Keaney (Princeton 1992)
4166.
, Allegory in Philo and Etymology in Stoicism: a Plea for Drawing Distinctions,
in Wisdom and Logos. Studies in Jewish Thought in Honor of David Winston, ed.
D. T. Runia, G. E. Sterling (Atlanta 1997) 198210.
Long A. A., Sedley D. N., The Hellenistic Philosophers. Vol. 2: Greek and Latin texts with
notes and bibliography (Cambridge 1987).
Longo O. (ed.), Scholia byzantina in Sophoclis Oedipum Tyrannum (Padova 1971).
Lonie I., The Hippocratic Treatises On Generation, On the Nature of Child, Diseases
IV (Berlin 1981).
, Literacy and the Development of Hippocratic Medicine, in Lasserre-Mudry
(1983) 145161.
Looy H. van, Lrechthe dEuripide, in Hommages M. Delcourt (Bruxelles 1970)
115122.
Lpez Frez J. (ed.), Galeno: Obra, pensamiento y influencia, Coloquio internacional
celebrado en Madrid 2225 de Marzo de 1988 (Madrid 1991).
, Galeno, lector y critico de manuscritos, in Garzya (1992) 197209.
Lord C., On the Early History of the Aristotelian Corpus, AJPh 107 (1986) 137161.
Bibliography 1349
Lorusso V., Nuovi frammenti di Galeno (In Hp. Epid. VI comm. VII; In Plat. Tim. comm.),
ZPE 152 (2005) 4356.
, Tracce di esegesi alessandrina al De methodo medendi di Galeno in codici
medievali, in Boudon Millot-Garzya-Jouanna-Roselli (2010) 109127.
Lowe N. J., Aristophanes books, Annals of scholarship 10 (1993) 6383.
, Comedy and the Pleiad. Alexandrian tragedians and the birth of comic schol-
arship, in Greek Comedy and the Discourse of Genres, eds. E. Bakola, L. Prauscello,
M. Tel (Cambridge-New York 2013) 343356.
LSJ = H. G. Liddell R. Scott, A Greek-English Lexicon. Revised and augmented by Sir
H. S. Jones with the Assistance of R. McKenzie, Oxford, 1940, with a Revised Supple-
ment (Oxford 1996).
Luc S., Attivit scrittoria e culturale a Rossano: da S. Nilo a S. Bartolomeo da Simeri
(secoli XXII), in Atti del congr. intern. su S. Nilo di Rossano (Grottaferrata 1989)
2568.
, Il Diodoro Siculo Neap. gr. 4* italogreco?, BBGG n.s. 44 (1990) 3379.
, I Normanni e la rinascita del sec. XII, Archivio Storico per la Calabria e la
Lucania 60 (1993) 191.
, Il lessico dello Ps.-Cirillo (redazione v1): da Rossano a Messina, RSBN 31 (1994)
4580.
, Il libro greco nella Calabria del sec. XV, in I luoghi dello scrivere da F. Petrarca
allet moderna, eds. C. Tristano (et al.) (Spoleto 2006) 331373.
, Dalle collezioni manoscritte di Spagna: libri originari o provenienti dallItalia
greca medievale, RSBN 44 (2007a) 3896.
, Note per la storia della cultura greca della Calabria medioevale, Archivio
storico per la Calabria e la Lucania 74 (2007b) 43101.
, Testi medici e tecnico-scientifici del Mezzogiorno greco, in De Gregorio-
Galante (2012) 551605 = [2012a].
, Scritture e libri in Terra dOtranto fra XI e XII secolo, in Bizantini, Longobardi
e Arabi in Puglia nellalto medioevo (Spoleto 2012) 487548 = [2012b].
Ludwich A., Aristarchs Homerische Textkritik nach den Fragmenten des Didymos, dar
gestellt und beurteilt, 2 vols. (Leipzig 18841885; repr. New York-Hildesheim 1971).
, Die Homerdeuterin Demo, in Festschrift zum fnfzigjhrigen Doctorjubilum
Ludwig Friedlaender dargebracht von seinen Schlern (Leipzig 1895) 296321.
, Die Homerische Batrachomachia des Karers Pigres, nebst Scholien und Para-
phrase (Leipzig 1896).
, Index lectionum in Regia Academia Albertina I (Knigsberg 19121913).
, Index lectionum in Regia Academia Arbertina II (Knigsberg 1914).
, ber die homerischen Glossen Apions, Philologus 74 (1917) 205247; 75 (1918)
95127; repr. in Lexica Graeca minora, eds. K. Latte, H. Erbse (Olms 1965) 283358.
1350 Bibliography
Meier-Brgger M. (ed.), Homer, gedeutet durch ein groes Lexikon. Akten des Ham-
burger Kolloquiums vom 6.8. Oktober 2010 zum Abschluss des Lexikons des frh-
griechischen Epos (Berlin-New York 2012).
Meijering R., Aristophanes of Byzantium and Scholia on the Composition of the Dra-
matic Chorus, in . Studia ad criticam interpretationemque textuum Graecorum
et ad historiam iuris Graeco-Romani pertinentia viro doctissimo D. Holwerda oblata,
eds. W. J. Aerts, J. H. A. Lokin, S. L. Radt, N. Van der Wal (Groningen 1985) 91102.
, Literary and Rhetorical Theories in Greek Scholia (Groningen 1987).
Meillet A., Aperu dune histoire de la langue grecque (Paris 19657).
Meillier C., Callimaque (P.L. 76 d, 78 a b c, 82, 84 et 111c), CRIPEL 4 (1976) 261286.
Meineke A., Analecta Alexandrina sive commentationes de Euphorione Chalcidensi,
Rhiano Cretensi, Alexandro Aetolo, Parthenio Nicaeno (Berlin 1843).
(ed.), Stephani Byzantii Ethnicorum quae supersunt (Berlin 1849).
Meiner B., Die technologische Fachliteratur der Antike. Struktur, U berlieferung und
Wirkung technischen Wissens in der Antike (ca. 400 v. Chr.ca. 500 n. Chr.) (Berlin
1999).
, Mndliche Vermittlung und schriftliche Unterweisung in der antiken Berufs
ausbildung, in Antike Fachschriftsteller: Literarischer Diskurs und sozialer Kontext
(Palingenesia vol. 80), eds. M. Horster, C. Reitz (Stuttgart 2003) 153175.
Mejer J., Diogenes Laertius and his Hellenistic Background (Wiesbaden 1978).
Meliad C., Posidippo, lepos ellenistico e la propaganda tolemaica, in Posidippo e gli
altri. Il poeta, il genere, il contesto culturale e letterario, Atti dellincontro di studio,
Roma 1415 maggio 2004 (Pisa-Roma 2004) 293216 = [2004a].
, Scoli a Teocrito in P.Oxy. 2064 + 3458, ZPE 147 (2004b) 1526.
, Dionysius [9] Cyclographus, LGGA (2005a).
, Eudaemon, LGGA (2005b).
, Un passo inedito dei Prolegomena Vaticana alla Techne di Dionisio Trace,
SemRom n.s. 2 (2013) 159186.
, Il codice di Antinoe. Contributi al testo e allesegesi di Teocrito, SemRom
(forthcoming).
Menchelli M., Note sulla corsiveggiante del X secolo, BollClass s. II 17 (1996) 133141.
, Appunti su manoscritti di Platone, Aristide e Dione di Prusa della prima et
dei Paleologi, SCO 47 (2000) 141208.
, Un nuovo codice di Gregorio di Cipro, Scriptorium 64 (2010) 227250.
Mercati G., Se la versione dallebraico del codice Veneto greco VII sia di Simone Atumano
(Citt del Vaticano 1916).
, Scritti dIsidoro il cardinale Ruteno (Citt del Vaticano 1926).
, Notizie di Procoro e Demetrio Cidone, Manuele Caleca e Teodoro Meliteniote
(Citt del Vaticano 1931).
Mergiali S., Lenseignement et les lettrs pendant lpoque des Palologues (12611453)
(Athens 1996).
1362 Bibliography
Merkelbach R., Literarische Texte unter Ausschluss der christlichen, APF 16 (1956)
115117.
, Das Knigtum der Ptolemer und die Hellenistischen Dichter, in Alexan-
drien. Kulturbegegnungen dreier Jahrtausende im Schmelztiegel einer mediterranen
Grostadt (Mainz am Rhein 1981) 2735.
Merro G., Apollodoro, Asclepiade di Tragilo ed Eschilo in Scholl. Eur. Rh. 916 e 922,
RFIC 134 (2006) 2651.
Meschini A., Il codice Vallicelliano di Areta (Padova 1972).
Messeri Savorelli G., Pintaudi R., Frammenti di rotoli letterari laurenziani, ZPE 115 (1997)
171177.
, Pintaudi R., I lettori dei papiri: dal commento autonomo agli scolii, in Talk-
ing to the Text: Marginalia from Papyri to Print. Proceedings of a Conference held
at Erice, 26 september-3 october 1998, as the 12th Course of International School for
the Study of Written Records, eds. V. Fera, G. Ferra, S. Rizzo (Messina 2002) 3757.
Mette H. J., Sphairopoiia. Untersuchungen zur Kosmologie des Krates von Pergamon
(Mnchen 1936).
, Parateresis. Untersuchungen zur Sprachtheorie des Krates von Pergamon (Halle
1952).
, Die kleinen griechischen Historiker heute, Lustrum 21 (1978) 543.
, Neoptolemos von Parion, RhM 113 (1980) 124.
Mewaldt J., Galenos ber echte und unechte Hippocratica, Hermes 44 (1909) 111134.
Meyendorff J. (ed.), Grgoire Palamas. Dfense des saints hsychastes (Leuven 1973).
, Spiritual trends in Byzantium in the late thirteenth and early fourteenth cen-
turies, in The Kariye Djami, IV, ed. P. A. Underwood (Princeton 1975) 93106.
Meyerhof M., Von Alexandrien nach Bagdad, SPAW (1930) 389429.
Meyerhof M., Schacht J. (eds.), Galen, ber die medizinischen Namen (Berlin 1931).
Micciarelli Collesi, A. M., Nuovi excerpta dallEtimologico di Orione, Byzantion 40
(1970) 517542.
Michel A., Schisma und Kaiserhof im Jahre 1054: Michael Psellos, in 10541094: LEglise
et les Eglises I (Chevetogne 1954) 351440.
Milanese G., Lucida carmina. Comunicazione e scrittura da Epicuro a Lucrezio (Milano
1989).
Miller G. L., Literacy and the Hippocratic art: reading, writing, and epistemology in
Ancient Greek medicine, JHM 45 (1990) 1140.
Miller T. S., Two Teaching Texts from the Twelfth-Century Orphanotropheion, in Byz-
antine Authors, ed. J. W. Nesbitt (Leiden-Boston 2003) 920.
Mineva E., , 6 (19941995) 314325.
Minio-Paluello L., Iacobus Veneticus Grecus, Traditio 8 (1952) 265304.
Minon S. (et al.), Dion de Pruse, Ilion na pas t prise. Discours troyen 11 (Paris 2012).
Mioni E., Un lessico inedito di Massimo Planude, JByz 32, 4 (1982) 129138.
Bibliography 1363
, Glosse dialettali negli scholia omerici, in Homer, gedeutet durch ein groes
Lexikon. Akten des Hamburger Kolloquiums vom 6.8. Oktober 2010 zum Abschluss
des Lexikons des frhgriechischen Epos, ed. M. Meier-Brgger (Berlin-Boston 2012)
123139 = [2012a].
, La papirologia omerica: temi, problemi, prospettive, in Bastianini-Casanova
(2012) 116 = [2012b].
, Odissea IX 4345, 7592, in Literarische Texte der Berliner Papyrussammlung.
Zur Wiedererffnung des Neues Museums, ed. F. Reiter (Berlin-Boston 2012) 2630 =
[2012c].
, The Peripatos on literature. Interpretation, use and abuse, in Praxiphanes
of Mytilene and Chamaeleon of Heraclea: text, translation, and discussion, eds.
A. Martano, E. Matelli, D. Mirhady, RUSCH 18 (New Brunswick-London 2012) 339
358 = [2012d].
, Gli storici greci e la filologia alessandrina, in Le et della trasmissione. Ales-
sandria, Roma, Bisanzio. Atti delle giornate di studio sulla storiografia greca fram-
mentaria, Genova, 2930 maggio 2012, eds. F. Gazzano, G. Ottone (Tivoli 2013) 132.
, From Book to Edition: Philology in Ancient Greece, in World Philology, eds.
S. Pollock, B. Elman, K. Chang (Cambridge, Mass., forthcoming).
Montanari F. (ed.), Ascheri P. (collab.), Omero tremila anni dopo. Atti del Congresso,
Genova 68 luglio 2000 (Roma 2002), repr. as: Omero tremila anni dopo. Personaggi
e strutture narrative (Roma 2009).
Montanari F., Lehnus L. (eds.), Callimaque: sept exposs suivis de discussions,
Vandoeuvres-Genve, 37 septembre 2001, Entretiens sur lantiquit classique
vol. 48 (Vanduvres-Genve 2002).
Montanari F., Pagani L. (eds.), From Scholars to Scholia. Chapters in the History of
Ancient Greek Scholarship (Berlin-New York 2011).
Montanari F., Perrone S. (eds.), Fragments of the Past. Ancient scholarship and Greek
Papyri, TiC 1 (2009).
Montiglio S., From Villain to Hero. Odysseus in Ancient Thought (Ann Arbor 2011).
Moore J. M., The Manuscript Tradition of Polybius (Cambdridge 1965).
Moraux P., Les listes anciennes des ouvrages dAristote (Louvain 1951).
, DAristote Bessarion. Trois exposs sur lhistoire et la transmission de
laristotlisme grec (Qubec 1970).
, Der Aristotelismus bei den Griechen. Von Andronikos bis Alexander von Aphro-
disias, Erster Band: Die Renaissance des Aristotelismus im I. Jh. v. Chr. (Berlin-New
York 1973).
, La critique dauthenticit chez les commentateurs grecs dAristote, in
Mlanges Mansel (Ankara 1974) I, 265288.
, Unbekannte Galenscholien, ZPE 27(1977) 163.
, Ein unbekannter Lehrer Galens, ZPE 53 (1983) 8588.
Bibliography 1369
Nagy G., Early Greek views of poets and poetry, in Classical Criticism 1, ed. G. Kennedy,
(Cambridge 1989) 177.
, Poetry as Performance. Homer and Beyond (Cambridge 1996).
, The Library of Pergamon as a Classical Model, in Pergamon. Citadel of
the Gods. Archeological Record, Literary Description, and Religious Development, ed.
H. Koester (Harrisburg, Penn. 1998) 185232.
, Reading Greek Poetry Aloud: Evidence from the Bacchylides Papyri, QUCC n.s.
64 (2000a) 728.
, review of M. L. West (ed.), Homeri Ilias, vol. I, (Stuttgart-Leipzig 1998) BMCR
2000.09.12 = [2000b].
, review of M. L. West, Text and Transmission of the Iliad (Mnchen-Leipzig
2001) Gnomon 75, 6 (2003) 481501.
, Homers Text and Language (Urbana-Chicago 2004).
, Homer the Classic (Washington-Cambridge 2009).
Naoumides M., The Fragments of Greek Lexicography in the Papyri, in Classical Stud-
ies presented to Ben Edwin Perry by his Students and Colleagues at the University of
Illinois, 192460 (Urbana-Chicago-London 1969) 181202.
, The Shorter Version of Pseudo-Zonaras Lexicon, in Serta Turyniana (Urbana-
Chicago-London 1974) 436488.
Napolitano F., Esegesi bizantina degli Halieutica di Oppiano, RAAN 48 (1973)
237254.
Nardelli J.-F., review of M. L. West (ed.), Homeri Ilias, vol. II (Munich-Leipzig 2000),
BMCR 2001.06.21.
Natali C., Aristotle: His Life and School (Princeton-Oxford 2013), Engl. ed. of Bios theo-
retikos. La vita di Aristotele e lorganizzazione della sua scuola (Bologna 1991).
Nauck A. (ed.), Aristophanis Byzantii grammatici Alexandrini fragmenta (Halle 1848a;
repr. Hildesheim 1963).
, Ueber die glossographischen Studien des Aristophanes von Byzanz, RhM 6
(1848b) 321351.
, Lexicon Vindobonense, rec. et adn. crit. instr. A. N., accedit appendix dua Photii
homilias et alia opuscula complectens (St. Petersburg 1867).
Negri M., Eustazio di Tessalonica. Introduzione al Commentario a Pindaro (Brescia 2000).
, Pindaro ad Alessandria. Le edizioni e gli editori (Brescia 2004).
Neils J., Theseus, LIMC 7.1 (1994) 922940.
Neitzel S. (ed.), Apions , SGLG 3 (Berlin-New York 1977) 185328.
Nmeth A., Imperial Systematization of the Past: Emperor Constantine VII and his His-
torical Excerpts (Diss. Budapest 2010).
, The Imperial Systematisation of the Past in Constantinople: Constantine VII
and his Historical Excerpts, in Knig-Woolf (2013) 232258.
1372 Bibliography
Nnlist R., The Ancient Critic at Work: Terms and Concepts of Literary Criticism in Greek
Scholia (Cambridge 2009a).
, Narratological Concepts in Greek Scholia, in Narratology and Interpretation:
The Content of Narrative Form in Ancient Literature, eds. J. Grethlein, A. Rengakos
(Berlin-New York 2009) 6383 = [2009b].
, Aristarchus and Allegorical Interpretation, in Ancient Scholarship and Gram-
mar, Trends in Classics Supplementary Volumes 8, eds. S. Matthaios, F. Montanari,
A. Rengakos (Berlin-New York 2011) 105117.
, A Chapter in the History of Linguistics: Aristarchus Interest in Language
Development, RhM 155 (2012a) 152165.
, Aristarch und das Lexikon des frhgriechischen Epos, in Meier-Brgger (2012)
193214 = [2012b].
, Homer as a Blueprint for Speechwriters, GRBS 52 (2012c) 493509.
, Nestor and speaking silence in the Iliad, Philologus 156 (2012d) 150156.
, What does actually mean?, Hermes 143, 2015
(forthcoming).
Nuti E., Reconsidering Renaissance Greek Grammars through the Case of Chrysoloras
Erotemata, GRBS 52 (2012) 240268.
Nutton V., Museums and Medical Schools in Antiquity, History of Education 4 (1975)
315.
, From Galen to Alexander, Aspects of Medicine and Medical Practice in Late
Antiquity, DOP 38 (1984) 114.
, John of Alexandria Again: Greek Medical Philosophy in Latin Translation, CQ
n.s. 41 (1991) 509519.
, Erotianos, DNP 4 (1998a) 9192.
, Gesios, DNP 4 (1998b) 10161017.
, Lysimachos von Kos, DNP 7 (1999a) 608.
, Magnus von Nisibis, DNP 7 (1999b) 698.
, Philagrios, DNP 9 (2000) 778.
, Ancient Medicine (London-New York 2004).
, Galens Library, in Gill-Whitmarsh-Wilkins (2009) 1934.
, The Commentary on the Hippocratic Oath ascribed to Galen, in Fortuna
Garofalo-Lami-Roselli (2012) 91100.
(transl.), Avoiding Distress, in P. N. Singer (ed.), Galen, Psychological Writings
(Cambridge 2014).
OHara J. J., True Names. Vergil and the Alexandrian Tradition of Etymological Wordplay
(Ann Arbor 1996).
ORegan D. E., Rhetoric, Comedy, and the Violence of Language in Aristophanes Clouds
(New York-Oxford 1992).
1376 Bibliography
OSullivan N., Alcidamas, Aristophanes and The Beginnings of Greek Stylistic Theory
(Stuttgart 1992).
Obbink D., What All Men BelieveMust Be True. Common Conceptions and consen-
sio omnium in Aristotle and Hellenistic Philosophy, OSAPh 10 (1992) 189231.
, Allegory and exegesis in the Derveni papyrus, in Metaphor, Allegory, and the
Classical Tradition. Ancient Thought and Modern Revisions, ed. G. R. Boys-Stones
(Oxford 2003) 177188.
, Vanishing Conjecture: Lost Books and their Recovery from Aristotle to Eco, in
Culture in Pieces. Essays on Ancient Texts in Honour of Peter Parsons, eds. D. Obbink,
R. Rutherford (Oxford 2011) 2049.
Oder E., De Antonino Liberali (Diss. Bonnae 1886).
Odorico P., La cultura della , ByzZ 83 (1990) 121.
, Gli gnomologi greci sacro-profani, in Aspetti di letteratura gnomica nel mondo
antico II, ed. M. S. Funghi (Firenze 2004) 6196.
, Cadre dexposition / cadre de pensela culture du recueil, in van Deun-
Mac (2011) 89107.
Ofenloch E. (ed.), Caecilii Calactini Fragmenta (Leipzig 1907).
Oguse A., Le papyrus grec de Strasbourg 364+16, Aegyptus 37 (1957) 7788.
Oikonomides A. N., Marinos of Neapolis: The Extant Works (Chicago 1977).
Olajos Th., Lcho de la posie lyrique grecque antique dans loeuvre historique de
Thophylacte Simocata, in , eds. C. Scholz, G. Makris (Munich
2000) 264271.
Olson S. D., Broken Laughter: Select Fragments of Greek Comedy (Oxford 2007).
Ooms S., De Jonge C. C., The Semantics of in Greek Literary Criticism,
CPh 108 (2013) 95110.
Opelt I., Epitome, RAC 5 (1962) 944973.
Ophuijsen J. M. van, Hephaestion On Metre (Leiden 1987).
, Parts of What Speech? Stoic Notions of Statement and Sentence, or: How the
Dialectician Knew Voice and Begat Syntax, in Syntax in Antiquity, eds. P. Swiggers,
A. Wouters (Leuven-Paris-Dudley, Mass. 2003) 7794.
Oppermann H., Herophilos bei Kallimachos, Hermes 60 (1925) 1432.
Orsini P., Pratiche collettive di scrittura a Bisanzio nei secoli IX e X, S&T 3 (2005)
265342.
Osann F. (ed.), Philemonis grammatici Quae supersunt, vulgatis et emendatiora et auc-
tiora (Berolini 1821).
Oser-Grote C. M., Aristoteles und das Corpus Hippocraticum (Stuttgart 2004).
Otranto R., review of Keaney (1991), Quaderni di storia 38 (1993) 225244.
, Antiche liste di libri su papiro, Sussidi Eruditi vol. 49 (Roma 2000).
Overwien O., Einige Beobachtungen zur berlieferung der Hippokratischen Schriften
in der arabischen und griechischen Tradition, Sudhoffs Archiv 89 (2005) 196210.
Bibliography 1377
, Eine anonyme Vorlesung ber das Prognostikon aus dem sptantiken Alexan-
dria, Galenos 5 (2011) 91102.
, Zur Funktion der Summaria Alexandrinorum und der Tabulae Vindobo-
nenses, in Enzyklopdie der Philologie, ed. U. Schmitzer (Gttingen 2013) 185205.
, Medizinische Lehrwerke aus dem sptantiken Alexandria, LEC (forthcoming).
Paap A. H. R. E., De Herodoti reliquiis in papyris et membranis Aegyptiis servatis (Leiden
1948).
Paassen Chr. van, The Classical Tradition of Geography (Groningen-Djakarta 1957).
Pace G., La struttura della tragedia secondo Giovanni Tzetze, in Volpe Cacciatore
(2003) 111127.
(ed.), Giovanni Tzetzes. La poesia tragica (Napoli 2011).
Pace N., Analisi preliminare e ricerca di possibili citazioni poetiche nel Lexicon
Ambrosianum, in Letteratura e riflessione sulla letteratura nella cultura classica, eds.
G. Arrighetti, M. Tulli (Pisa 2000) 309325.
Pack E., Antiochia, in Cambiano-Canfora-Lanza (1993) 717767.
Pagani F. Damnata verba, ByzZ 102 (2009) 167202.
Pagani L., Asclepiades [1], LGGA (2004).
, Due etimologie di nomi omerici in Asclepiade di Mirlea (Astyanax e Arnaios),
Eikasmos 16 (2005a) 193209.
, Hegesianax, LGGA (2005b).
, Theodorus [1], LGGA (2005c).
, Ptolemaeus [4] Chennos, LGGA (2006a).
, Theodosius, LGGA (2006b).
(ed.), Asclepiade di Mirlea. I frammenti degli scritti omerici, ed. introd. e com-
mento (Roma 2007a).
, Demetrius [12] Pyktes, LGGA (2007b).
, La filologia antica su Teocrito, in La cultura letteraria ellenistica. Persistenza,
innovazione, trasmissione. Atti del Convegno COFIN 2003, Universit di Roma
Tor Vergata, 1921 settembre 2005, eds. R. Pretagostini, E. Dettori (Roma 2007)
285303 = [2007c].
, Posidonius [1], LGGA (2007d).
, Aristeas, LGGA (20112).
, Asclepiades [3], LGGA (2009a).
, review of Nnlist (2009), RFIC 137 (2009b) 201211.
, Tyrannion [1] Maior, LGGA (2009c).
, Il pronome secondo Aristarco e Abrone. Un capitolo della riflessione antica
sulla lingua, in Aner polytropos. Ricerche di filologia greca antica dedicate dagli allievi
a Franco Montanari, ed. F. Montana (Roma 2010) 105123 = [2010a].
, La Techne grammatike attribuita a Dionisio Trace e la nascita della gramma
tica nellantichit greca, RFIC 138 (2010b) 390409.
1378 Bibliography
Prez Martn I., Le conflit de lUnion des glises (1274) et son reflet dans lenseignement
suprieur de Constantinople, ByzSlav 56 (1995) 411422.
, El patriarca Gregorio de Chipre (ca. 12401290) y la transmisin de los textos
clsicos en Bizancio (Madrid 1996).
, El scriptorium de Cora: un modelo de acercamiento a los centros de copia
bizantinos, in / El cielo en la tierra, ed. P. Bdenas, A. Bravo, I. Prez
Martn (Madrid 1997) 203224 = [1997a].
, La escuela de Planudes: notas paleogrficas a una publicacin reciente
sobre los escolios euripideos, ByzZ 90 (1997b) 7396.
, Maxime Planude et le Diophantus Matritensis: un paradigme de la rcupra-
tion des textes anciens dans la Renaissance Palologue, Byzantion 76 (2006) 433462.
, George of Cyprus and Aelius Aristides, in George of Cyprus and his World
(forthcoming).
Perilli L., La tradizione manoscritta del Glossario ippocratico di Galeno (e lordinamento
alfabetico delle glosse), in I testi medici greci. Tradizione e ecdotica, ed. A. Garzya,
J. Jouanna (Napoli 1999) 429456.
, Lordinamento alfabetico di Lessici e Glossari: il caso del Glossario ippocrat-
ico di Galeno, Bollettino della Badia Greca di Grottaferrata 54 (2000a) 2752.
, La fortuna di Galeno filosofo. Un nuovo testimone dei commentari neopla-
tonici (scholia Yalensia) al De elementis, in La filosofia in et imperiale. Le scuole e
le tradizioni filosofiche, Atti del Colloquio Roma 1719 giugno 1999, ed. A. Brancacci
(Roma 2000) 87135 = [2000b].
, Da medico a lessicografo: Galeno e il Glossario ippocratico, in Mller-
Brockmann-Brunschn (2006) 165201.
, Su Esichio e i cosiddetti frammenti del Glossario Ippocratico di Erotiano, Gale-
nos 2 (2008) 3655.
, Scrivere la medicina. La registrazione dei miracoli di Asclepio e le opere
di Ippocrate, in Antike Medizin im Schnittpunkt von Geistes- und Naturwissen-
schaften, Internationale Fachtagung aus Anlass des 100-jhrigen Bestehens des
Akademievorhabens Corpus Medicorum Graecorum/Latinorum im Auftrag der
Berlin-Brandenburgischen Akademie der Wissenschaften, eds. Ch. Brockmann, W.
Brunschn, O. Overwien (Berlin-New York 2009) 75120.
Pernigotti C., Osservazioni sul Tucidide Laurenziano 69, 2, SCO 47 (2001) 227244.
Pernot L., La Rhtorique dans lAntiquit (Paris 2000).
, Mimesis, rhtorique et politique dans lessai de Thodore Mtochite sur
Dmosthne et Aelius Aristide, in Spirito e forme nella letteratura bizantina, ed.
A. Garzya (Napoli 2006) 107120.
Perosa A., Lorenzo Valla, Collatio Novi Testamenti (Firenze 1970).
Perria L., Il Vat. gr. 2200: note codicologiche e paleografiche, RSBN 2021 (19831984)
2568.
1382 Bibliography
Phlmann E., Die Schriftreform in Athen um 403 und ihre Implikationen, in Erzie-
hungs- und Unterrichtsmethoden im historischen Wandel, eds. L. Kriss-Rettenbeck,
M. Liedtke, (Bad Heilbrunn Obb. 1986) 5164.
, Der Schreiber als Lehrer in der klassischen Zeit Griechenlands, in Schreiber,
Magister, Lehrer: zur Geschichte und Funktion eines Berufsstandes, eds. J. G. Prinz
von Hohenzollern, M. Liedtke (Bad Heilbrunn Obb 1989) 7382.
, Zur berlieferung griechischer Literatur vom 8.4. Jh., in Der bergang
von der Mndlichkeit zur Literatur bei den Griechen, eds. W. Kullmann, M. Reichel
(Tbingen 1990) 1130.
, Die berlieferung der Musik in der antiken Welt, Die Musikforschung 44
(1991) 19.
, Einfhrung in die berlieferungsgeschichte und in die Textkritik der antiken
Literatur. Band 1: Altertum (Darmstadt 1994).
, review of Prauscello (2006), Gnomon 79 (2007) 104111.
, Excavation, Dating and Content of Two Tombs in Daphne, Odos Olgas 53,
Athens 1, Greek and Roman Musical Studies 1, 1 (2013) 724.
Phlmann E., West M. L., Documents of Ancient Greek Music (Oxford 2001).
, The Oldest Greek Papyrus and Writing Tablets. Fifth-Century Documents
from the Tomb of the Musician in Attica, ZPE 180 (2012) 116.
Polastron L. X., Livres en feu. Histoire de la destruction sans fin des bibliothques (Paris
2004).
Polemis I., , 45 (1995) 277302.
, The Treatise On those who unjustly accuse wise men, of the past and pres-
ent, ByzZ 102 (2009) 203217.
Polito R., On the life of Asclepiades of Bithynia, JHS 119 (1999) 4865.
Pollmann K., Zwei Konzepte von Fiktionalitt in der Philosophie des Hellenismus und
in der Sptantike, in Zur Rezeption der hellenistischen Philosophie in der Sptantike.
Akten der 1. Tagung der Karl-und-Gertrud-Abel-Stiftung vom 22.25. September
1997 in Trier, eds. Th. Fuhrer, M. Erler (Stuttgart 1999) 261278.
Poltera O., Simonides lyricus. Testimonia und Fragmente, Einleitung, kritische Ausgabe,
bersetzung und Kommentar (Basel 2008).
Pontani A., I Graeca di Ciriaco dAncona, Thesaurismata 24 (1994) 37148.
, La filologia, in Lo spazio letterario della Grecia antica, eds. G. Cambiano,
L. Canfora, D. Lanza, II (Rome 1995) 307351 = [1995a].
, Da Bisanzio allItalia: a proposito di un libro recente, Thesaurismata 25
(1995b) 83123.
, Manuele Crisolora: libri e scrittura, BBGG n.s. 53 (1999) 255283.
(ed.), Niceta Coniata. Grandezza e catastrofe di Bisanzio IIIII (Milan
19992014).
Pontani F., Mos del Brolo e la sua lettera da Costantinopoli, Aevum 72 (1998) 143175.
Bibliography 1385
Rosato C., Considerazioni su alcuni termini della critica testuale greca, Rudiae 11
(1999) 113128.
Roselli A., Citazioni ippocratiche in Demetrio Lacone (P.Herc. 1012), CErc 18 (1988)
5357.
, Epidemics and Aphorisms: Notes on the History of Early Transmission of Epi-
demics, in Baader-Wienau (1989) 182190.
, Appunti per una storia delluso apologetico della filologia: la nuova edizione
di Demetrio Lacone (PHerc. 1012), SCO 40 (1990) 117138.
, Tra pratica medica e filologia ippocratica: il caso della
di Apollonio di Cizio, in Argoud-Guillaumin (1998) 217231.
, Un commento inedito a Hipp. Epid. VI 2.24 (Epid. II 3.7) nel ms. Par. gr. 2230,
in Garzya-Jouanna (1999) 493508.
, Un corpo che prende forma: lordine di successione dei trattati ippocratici
dallet ellenistica fino allet bizantina, in La letteratura pseudepigrafa nella cultura
greca e romana, Atti di un incontro di studi Napoli, 1517 gennaio 1988, ed. G. Cerri
(Napoli 2000) 167195.
, C. Uninterpretazione delle testimonianze su Cal-
lianatte Erofileo, Galenos 3 (2009) 6974.
, Libri e biblioteche a Roma al tempo di Galeno: la testimonianza del de indo-
lentia, Galenos 4 (2010) 127148.
, Galeno e la filologia del II secolo, in Vestigia notitiai, Scritti in memoria di
Michelangelo Giusta, eds. E. Bona, C. Levy, G. Magnaldi (Alessandria 2012) 6380 =
[2012a].
, Galeno e le edizioni ippocratiche di Artemidoro Capitone e Dioscoride, in
FortunaGarofalo-Lami-Roselli (2012) 1527 = [2012b].
Rosn H. B., Eine Laut- und Formenlehre der Herodotischen Sprachform (Heidelberg
1962).
, Herodoti Historiae (Leipzig 19871997).
, Zu Text und Interpretation der grammatischen Abschnitte in Aristoteles
Poetik und zur Umdeutung und Umformung der Redeteileinteilung bis ins orien-
talische Mittelalter, in History and Historiography of Linguistics. Papers from the
Fourth International Conference on the History of the Language Sciences, Trier,
2428 August 1987, eds. H. J. Niederehe, K. Koerner, vol. I (Amsterdam-Philadelphia
1990) 111121.
Rosen R. M., Cratinus Pytine and the Construction of the Comic Self, in The Rivals of
Aristophanes. Studies in Athenian Old Comedy, eds. D. Harvey, J. Wilkins (London
2000) 2339.
, Aristophanes Frogs and the Contest of Homer and Hesiod, TAPhA 134
(2004) 295322.
Rosenthal F., An ancient commentary on the Hippocratic oath, BHM 30 (1956) 5287.
1394 Bibliography
Rosokoki A. (ed.), Die Erigone des Eratosthenes. Eine kommentierte Ausgabe der Frag-
mente (Heidelberg 1995).
Rossetti L., Liviabella Furiani P., Rodi, in Lo spazio letterario della Grecia antica I.2, eds.
G. Cambiano, L. Canfora, D. Lanza (Rome 1993) 657715.
Rossi L. E., La fine alessandrina dellOdissea e lo di Apollonio Rodio,
RFIC 96 (1968) 151163.
, Umanesimo e filologia (A proposito della Storia della filologia classica di
Rudolf Pfeiffer), RFIC 104 (1976) 98117, review of the Italian edition of Pfeiffer
(1968) (Napoli 1973).
Rossum-Steenbeeck M. van, Greek Readers Digests? Studies on a Selection of Subliterary
Papyri (Leiden-New York-Kln 1998).
Rothe S., Kommentar zu ausgewhlten Sophistenviten des Philostratos. Die Lehrstuhlin-
haber in Athen und Rom (Heidelberg 1989).
Rouech M., Byzantine Philosophical Texts of the Seventh Century, JByz 23 (1974)
6176.
, The Definition of Philosophy and a New Fragment of Stephanus the Philoso-
pher, JByz 40 (1990) 108128.
, Did Medical Students Study Philosophy in Alexandria?, BICS 43 (1999)
153169.
Rowe C. J., Plato. Symposium, edited with an introduction, translation and commen-
tary (Warminster 1998).
Ruijgh C. J., review of Teodorsson (1974), Mnemosyne 31 (1978) 7989.
, Le Spectacle des lettres, comdie de Callias (Athne X 453c455b), avec un
excursus sur les rapports entre la mlodie du chant et les contours mlodiques du
langage parl, Mnemosyne 54 (2001) 257335.
Ruiz de Elvira A., Suidas, y no la Suda, Myrtia 12 (1997) 58.
Russell D. A., Longinus, On the Sublime (Oxford 1964).
, Classicizing rhetoric and criticism: the pseudo-Dionysian Exetasis and Mis-
takes in Declamation, in Le classicisme Rome aux Iers sicles avant et aprs J.-C., ed.
H. Flashar (Vanduvres-Genve 1979) 113130.
, Greek Declamation (Cambridge 1983).
, Plutarch. Selected Essays and Dialogues (Oxford 1993).
, Criticism in Antiquity (London 19952; 1st ed. Berkeley-Los Angeles 1981).
, Figured speeches: Dionysius, Art of Rhetoric VIIIIX in The Orator in Action
and Theory in Greece and Rome, ed. C. W. Wooten (Leiden 2001) 156168 = [2001a].
, Quintilian. The Orators Education, books 35, with an English translation
(Cambridge, Mass.-London 2001b).
Russo G., Contestazione e conservazione: Luciano nellesegesi di Areta (Berlin 2012).
Russo L., The Forgotten Revolution. How Science Was Born in 300 BC and Why it Had
to Be Reborn (Berlin-Heidelberg 2004), English ed. of La rivoluzione dimenticata. Il
pensiero scientifico greco e la scienza moderna (Milano 20012).
Bibliography 1395
Rusten J. S., Dicaearchus and the Tales from Euripides, GRBS 23 (1982a) 357367.
, Dionysius Scytobrachion (Opladen 1982b).
Rutherford I. (ed.), Pindars Paeans. A Reading of the Fragments with a Survey of the
Genre (Oxford-New York 2001).
, (et al.), Decorum, in Historisches Wrterbuch der Rhetorik, vol. 2 (Tbingen
1994) 423451.
Rutherford W. G., Scholia Aristophanica (London 18961905).
, A Chapter in the History of Annotation, being Scholia Aristophanica, III (Lon-
don-New York 1905).
Ryan M.-L., Narrative as Virtual Reality: Immersion and Interactivity in Literature and
Electronic Media (Baltimore-London 2000).
Ryder J. R., The Career and Writings of Demetrius Kydones (Leiden-Boston 2010).
Sabbadini R., Guariniana (Torino 1964, 1st ed. Genova 1891 = Catania 1896).
Saffrey H. D., Le chrtien Jean Philopon et la survivance de lEcole dAlexandrie au VIe
sicle, REG 67 (1954) 396410.
, Retour sur le Parisinus Graecus 1807, le manuscrit A de Platon, in The Libraries
of the Neoplatonists, ed. C. DAncona (Leiden 2007) 328.
Sad S., Herodotus and the Myth of the Trojan War, in Myth, Truth & Narrative in
Herodotus, eds. E. Baragwanath, M. de Bakker (Oxford 2012) 87105.
Sakkelion I., . Scholies de Dmosthne et
dschine daprs un manuscrit indit de Patmos, BCH 1 (1877) 1016; 137155, partial
repr. in Latte-Erbse (1965) 140165.
Samama ., Les mdecins dans le monde grec (Genve 2003).
Sandys J. E., A History of Classical Scholarship, 3 vols. (Cambridge 1908; 19213).
Santoni A. (ed.), Eratostene, Epitome dei Catasterismi. Origine delle costellazioni e dis-
posizione delle stelle, Introduzione traduzione e note (Pisa 2009).
Sanz Morales M., El Homero de Aristteles (Amsterdam 1994).
Sartori F., Ermippo di Berito, schiavo e storiografo, Index 10 (1981) 260270.
Sathas K. N. (ed.), V (Paris 1876).
Saussure F. de, Cours de linguistique gnrale [Paris 1916]; ed.crit. Tullio de Mauro (Paris
1974).
Savage-Smith E., Galens Lost Ophthalmology and the Summaria Alexandrinorum, in
The Unknown Galen, BICS Suppl. 77, ed. V. Nutton (London 2002) 121138.
Saxey R., The Homeric Metamorphoses of Andronikos I Komnenos, in Nicetas
Choniates. A Historian and a Writer, eds. A. Simpson, S. Efthymiadis (Geneva 2009)
121143.
Sbardella L. (ed.), Filita, Testimonianze e frammenti poetici (Roma 2000).
Scarpi P. (ed.), Apollodoro, I miti greci (Biblioteca) (Milano 1996).
Schaefer G. H. (ed.), Gregorii Corinthii et aliorum grammaticorum libri de dialectis lin-
guae graecae, quibus additur nunc primum editus Manuelis Moschopuli libellus de
vocum passionibus (Leipzig 1811).
1396 Bibliography
Snell B., Die Nachrichten ber die Lehren des Thales und die Anfnge der griechischen
Philosophie- und Literaturgeschichte, in Sophistik, ed. C. J. Classen (Darmstadt
1976) 478490.
, Tragicorum Graecorum fragmenta, I. Testimonia et fragmenta tragicorum
minorum, editio correctior et addendis aucta cur. R. Kannicht (Gttingen 1986).
Snodgrass A., Homer and the artists: text and picture in early Greek art (Cambridge
1998).
Sodano A. R., Porphyrii in Platonis Timaeum Commentariorum Fragmenta (Naples 1964).
, Porphyrii Quaestionum Homericarum liber I (Naples 1970).
Sonny A., Ad Dionem Chrysostomum analecta (Kiev 1896).
Sorabji R., John Philoponus and the Rejection of Aristotelian Science (London 1987).
(ed.), Aristotle Transformed. The Ancient Commentators and Their Influence
(Ithaca, N.Y.-London 1990).
, Animal Minds and Human Morals. The Origins of the Western Debate (London
1993).
Sosower M. L., Palatinus Graecus 88 and the Manuscript Tradition of Lysias (Amsterdam
1987).
Sowinski B., Stil, in Historisches Wrterbuch der Rhetorik, vol. 8 (Tbingen 2007)
13931419.
Spadaro M. D., Sulle composizioni di Costantino il filosofo, SicGymn 24 (1971) 175205.
Spang K., Dreistillehre, in Historisches Wrterbuch der Rhetorik, vol. 2 (Tbingen 1994)
921972.
Spanoudakis K. (ed.), Philitas of Cos (Leiden-Boston-Kln 2002).
Speck P., Die kaiserliche Universitt von Konstantinopel (Munich 1974a).
, review of Lemerle (1971), ByzZ 67 (1974b) 385393.
, A more charitable verdict, Klio 68 (1986) 615625.
, Die byzantinische Renaissance, in Selecta colligere I, eds. R. M. Piccione,
M. Perkams (Alessandria 2000) 1732.
, Understanding Byzantium (Aldershot 2003).
Spengel L. von, Rhetores Graeci, 3 vols. (Leipzig 18531856).
Speyer W., Die literarische Flschung im heidnischen und christlichen Altertum
(Mnchen 1971).
, Das Buch als magisch-religiser Krafttrger im griechischen und rmischen
Altertum, in Das Buch als magisches und als Reprsentationsobjekt, ed. P. Ganz
(Wiesbaden 1992) 5986.
, Italienische Humanisten als Kritiker der Echtheit antiker und christlicher Litera-
tur (Stuttgart 1993).
Spreca S., Aspetti di archaiologhia nelle Storie di Eforo: Atene e Delfi, in Costruzione
e uso del passato storico nella cultura antica. Atti del Convegno internazionale di
studi, Firenze 1820 settembre 2003, eds. P. Desideri, S. Roda, A. M. Biraschi, (coll.)
A. Pellizzari (Alessandria 2007) 191202.
1406 Bibliography
Staden H. von, Hairesis and Heresy: the Case of the Haireseis Iatrikai, in Jewish and
Christian Self-Definition, III, eds. B. Meyer, E. Sanders (Philadelphia 1983) 76100.
, Herophilus. The Art of Medicine in Early Alexandria (Cambridge 1989).
, Galen as Historian: His use of sources on the Herophileans, in Lpez Frez
(1991) 205222.
, Lexicography in the Third Century B.C.: Bacchius of Tanagra, Erotianus, and
Hippocrates, in Tratados Hipocrticos. Actas del VIIe Colloque International Hip-
pocratique (Madrid 2429 de septiembre de 1990), ed. J. A. Lpez Frez (Madrid
1992) 549569.
, Science as text, science as history: Galen on Metaphor, in van der Eijk-
Horstmanshoff-Schrijvers (1995) 499518.
, Character and competence. Personal and professional conduct in Greek
medecine, in Mdecine et morale dans lAntiquit. Dix exposs suivis de discussions,
eds. H. Flashar, J. Jouanna (Genve-Vanduvres 1997) 157210 = [1997a].
, Galen and the Second Sophistic, in Aristotle and After (BICS Suppl. vol. 68),
ed. R. Sorabji (London 1997) 3354 = [1997b].
, Inefficay, Error and Failure: Galen on , in Galen on
Pharmacology: Philosophy, History and Medicine, ed. A. Debru (Leiden-Boston-Kln
1997) 5983 = [1997c].
, Andras de Caryste et Philon de Byzance: Mdecine et Mcanique Alexand-
rie, in Argoud-Guillaumin (1998) 142172 = [1998a].
, Gattung und Gedchtnis. Galen ber Wahrheit und Lehrdichtung, in Gattun-
gen wissenschaftlicher Literatur in der Antike, eds. W. Kullmann, J. Althoff, M. Asper
(Tbingen 1998) 6594 = [1998b].
, Rupture and Continuity: Hellenistic Reflections on the History of Medicine,
in van der Eijk (1999a) 143187 = [1999a].
, Celsus as Historian?, in van der Eijk (1999a) 251294 = [1999b].
, A Woman does not become ambidextrous: Galen and the Culture of the Sci-
entific Commentary, in The Classical Commentary: Histories, Practices, Theory, eds.
R. Gibson, C. Kraus (Leiden 2002) 109139.
, Galens Alexandria, in Ancient Alexandria between Egypt and Greece, Colum-
bia Studies in the Classical Tradition 26, eds. W. V. Harris, G. Ruffini (Leiden 2004)
179215.
, Interpreting Hippocrates in the 3rd and 2nd centuries BC, in Mller-Brock-
mann-Brunschn (2006) 1547.
, Staging the Past, Staging Oneself: Galen on Hellenistic Exegetical Traditions,
in Galen and the World of Knowledge, eds. C. Gill, T. Whitmarsh, J. Wilkins (Cam-
bridge 2009) 132156.
Stadter Ph. A., Planudes, Plutarch and Pace of Ferrara, IMU 16 (1973) 137162.
, Herodotus and the Athenian Arche, ASNP s. III 22 (1992) 781809.
Bibliography 1407
Stork P., Ophuijsen J. M. van, Dorandi T. (eds.), Demetrius of Phalerum: The Sources,
Text and Translation, in Fortenbaugh-Schtrumpf (2000) 1310.
Stramaglia A., Libri perduti per sempre: Galeno, De indolentia 13; 16; 1719, RFIC 139
(2011) 118147.
Strano G. (ed.), Leone Choirosphaktes. Corrispondenza (Catania 2008).
Strecker K. (ed.), De Lycophrone, Euphronio, Eratosthene comicorum interpretibus (Diss.
Greifswald 1884).
, Zu Erotianus, Hermes 26 (1891) 262307.
Strodel S., Zur berlieferung und zum Verstndnis der hellenistischen Technopaignien
(Frankfurt 2002).
Strohmaier G., Der Syrische und Arabische Galen, ANRW (1994) 19872017; repr. in Hel-
las im Islam, ed. G. Strohmaier (Wiesbaden 2003) 85106.
, Galen als Hippokrateskommentator. Der griechische und der arabische
Befund, in Geerlings-Schultze (2002) 253274.
, Der Kommentar des Johannes Grammatikos zu Galen, De usu partium (Buch
11), in einer unikalen Gothaer Handschrift, in Hellas im Islam, ed. G. Strohmaier
(Wiesbaden 2003) 109112.
, Galens Non Uncritical Commentary on Hippocrates Airs, Waters, Places, in
Philosophy, Science and Exegesis in Greek, Arabic and Latin Commentaries, II, eds.
P. Adamson, H. Baltussen, M. W. F. Stone (London 2004) 19.
Strootman R., Literature and the Kings, in Clauss-Cuypers (2010) 3045.
, Kings and Cities in the Hellenistic Age, in Nijf-Alston-Williamson (2011)
141153.
, Dynastic Courts of the Hellenistic Empires, in Beck (2013) 3853.
Strout D., French R., Phaeinos [2], RE XIX/2 (1938) 15051506.
, French R., Phrynichos [8], RE XX/1 (1941) 920925.
, French R., Pius [2], RE XX/2 (1950) 18911892.
Stroux J., De Theophrasti virtutibus dicendi (Leipzig 1912).
Struck P. T., Allegory, aenigma, and anti-mimesis: a struggle against Aristotelian rhe-
torical literary theory, in Greek literary theory after Aristotle: a collection of papers in
honour of D. M. Schenkeveld, eds. J. G. J. Abbenes, S. R. Slings, I. Sluiter (Amsterdam
1995) 215234.
, Birth of the Symbol: Ancient Readers at the Limits of Their Texts (Princeton 2004).
Strzelecki L., Quaestionum orthographicarum specimen, Eos 44, 1 (1950) 89110.
Stckelberger A., Aristoteles illustratus. Anschauungshilfsmittel in der Schule des
Peripatos, MH 50 (1993) 131143.
Sturz F. W. (ed.), Etymologicum Graecae linguae Gudianum (Leipzig 1818).
, Orionis Thebani Etymologicon (Leipzig 1820; 2nd ed. Hildesheim-New York
1973).
Bibliography 1409
Taylor D. J., Declinatio: A Study of the Linguistic Theory of Marcus Terentius Varro
(Amsterdam 1975).
, Rethinking the History of Language Science in Classical Antiquity, Historio-
graphia Linguistica 13 (1986) 175190.
, Rethinking the History of Language Science in Classical Antiquity, in The His-
tory of Linguistics in the Classical Period, ed. D. J. Taylor (Amsterdam-Philadelphia
1987) 116.
Tchernetska N., The Tischendorf Greek palimpsests, ARF 2 (2000) 107126.
Tecusan M., The Fragments of the Methodists, I (Leiden-Boston 2004).
Temkin O., Geschichte des Hippokratismus im ausgehenden Altertum, Kyklos 4 (1932)
180; Engl. ed.: History of Hippocratism in late antiquity, in The double Face of
Janus (Baltimore 1977) 167177.
, Studies in Late Alexandrian Medicine, I: Alexandrian Commentaries on
Galens De sectis ad introducendos, BHM 3 (1935) 405435.
, Galenism: Rise and Decline of a Medical Philosophy (Ithaca 1973).
Tenbruck F. H., Zur Soziologie der Sophistik, in Moderne Sophistik, eds. R. Bubner,
K. Cramer, R. Wiehl (Gottingen 1976) 5177.
Teodorsson S.-T., The Phonemic System of the Attic Dialect 400340 B.C. (Gothenburg 1974).
Tepedino Guerra A., Il contributo di Metrodoro di Lampsaco alla formazione della
teoria epicurea del linguaggio, CErc 20 (1990) 1725.
Tessier A., Scholia metrica vetera in Pindari carmina (Leipzig 1989).
, Tradizione metrica di Pindaro (Padova 1995).
, Demetrio Triclinio (ri)scopre la responsione, in La colometria antica dei testi
poetici greci, eds. B. Gentili, F. Perusino (Pisa-Roma 1999) 3149.
, Filologi bizantini di epoca comnena, Incontri triestini di filologia classica 3
(20032004) 114.
(ed.), Demetrio Triclinio. Scolii metrici alla tetrade sofoclea (Alessandria 2005).
, review of B. Gentili, L. Lomiento, Metrics and Rhytmics. History of Poetic Forms
in Ancient Greece (Pisa-Roma 2008), English transl. of Gentili-Lomiento (2003),
BMCR 2009.11.31.
, Musica antica e sistemazione colometrica?, QUCC n.s. 92 (2010a) 1116.
, Schicksale der antiken Literatur in Byzanz: Maas e Pasquali giudicano la filo-
logia dei Bizantini, MEG 10 (2010b) 269276.
Teteriatnikov N., The Dedication of the Chora Monastery, Byzantion 66 (1996) 188207.
Theodoridis Ch., Drei bersehene Bruchstcke des Apollodoros von Athen, Glotta 50
(1972) 2934.
(ed.), Die Fragmente des Grammatikers Philoxenos, SGLG 2 (Berlin-New York
1976).
, Vier neue Bruchstcke des Apollodoros von Athen, RhM 122 (1979) 917.
, Der Hymnograph Klemens terminus post quem fr Choiroboskos, ByzZ 73
(1980) 341345.
1412 Bibliography
(ed.), Photii Patriarchae Lexicon, vols. IIII (Berlin-New York 1982, 1998, 2013).
, Drei neue Fragmente des grammatikers Apion, RhM, 132 (1989a) 345350.
, Zur Schreibernotiz im Etymologicum Genuinum (s.v. ), RhM 132 (1989b)
409410.
, Ein bersehenes Fragment des Aristophanes von Byzanz, RhM n.s. 152 (2009)
396398.
Thiel H. van (ed.), Homeri Odyssea (Hildesheim-Zrich-New York 1991).
, Zenodot, Aristarch und andere, ZPE 90 (1992) 132.
, Der Homertext in Alexandria, ZPE 115 (1997) 1336.
, Scholia D in Iliadem (20142); available on Internet only, at http://kups.ub.uni
-koeln.de/1810/).
(ed.), Homeri Ilias (Hildesheim-Zrich-New York 20102; 1st ed. 1996).
, Aristarch, Aristophanes Byzantios, Demetrios Ixion, Zenodot. Fragmente zur
Ilias gesammelt, neu herausgegeben und kommentiert, IIV (Berlin-Boston 2014).
Thiel R., Simplikios und das Ende der neuplatonischen Schule in Athen (Stuttgart 1999).
Thiemann C., Heliodori colometriae Aristophaneae quae supersunt una cum reliquis
scholiis in Aristophanem metricis (Halle 1869).
Thierfelder A., Beitrge zur Kritik und Erklrung des Apollonius Dyscolus, ASAW 43,
2 (Leipzig 1935).
Thiermann P., Die Orationes Homeri des Leonardo Bruni Aretino (Leiden-New York 1993).
Thomas R., Oral Tradition and Written Record in Classical Athens (Cambridge 1989).
, Literacy and orality in ancient Greece (Cambridge 1992).
, Herodotus in context: ethnography, science, and the art of persuasion (Cam-
bridge 2000).
, Literacy in Ancient Greece: functional literacy, oral education, and the
development of a literate environment, in The Making of Literate Societies, eds.
D. R. Olson, N. Torrance (Malden, Mass.-Oxford 2001) 6881.
, Writing, Reading, Public and Private Literacies: functional literacy and dem-
ocratic literacy in Greece, in Ancient Literacies: the Culture of Reading in Greece and
Rome, eds. W. A. Johnson, H. N. Parker (Oxford 2009) 1345.
, Education, Greek, in Oxford Classical Dictionary, eds. S. Hornblower, A. Spaw-
forth (Oxford 20124) 487489.
Thompson DArcy W., A Glossary of Greek Birds (Hildesheim 1966).
Thompson D. J., The Ptolemies and Egypt, in Erskine (2003) 105120.
Thorn-Wickert L., Manuel Chrysoloras (Frankfurt am Main 2006).
Threatte L., The Grammar of Attic Inscriptions. 1. Phonology (Berlin-New York 1980).
, The Inscribed Schist Fragments from the Athens Academy Excavations (Athens
2007).
Thumb A., Die griechische Sprache im Zeitalter des Hellenismus. Beitrge zur Geschichte
und Beurteilung der (Strassburg 1901).
Bibliography 1413
Tihon A., Le Petit commentaire de Thon dAlexandrie aux Tables faciles de Ptolme,
Studi e testi 282 (Vatican 1978).
, Lastronomie byzantine (du Ve au XVe sicle), Byzantion 51 (1981) 603624.
, Le Grand commentaire de Thon dAlexandrie aux Tables faciles de Ptolme,
Livres II et III, Studi e testi 340 (Vatican 1991).
, Le Grand commentaire de Thon dAlexandrie aux Tables faciles de Ptolme,
Livre IV, Studi e testi 390 (Vatican 1999).
Till D., (et al.), Poetik, in Historisches Wrterbuch der Rhetorik, vol. 6 (Tbingen 2003)
13041393.
Tinnefeld F., Zur intellektuellen Polemik des Nikephoros Gregoras, in van Deun-Mac
(2012) 345359.
Tischendorf L. F. C., Die Manuscripta Tischendorfiana, Serapeum. Zeitschrift fr Biblio-
thekwissenschaft, Handschriftenkunde und ltere Litteratur 8 (1847) 4961.
Tissoni F., Le Olimpiche di Pindaro nella scuola di Gaza a Ferrara (Messina 2009).
Tittmann J. A. H. (ed.), Johannis Zonarae Lexicon (Amsterdam 1967; 1st ed. Leipzig 1808).
Titze F. N. (ed.), Manuelis Moschopuli opuscula grammatica (Leipzig 1822).
Todd R. B., Galenic Medical Ideas in the Greek Aristotelian Commentators, SO 52
(1977) 117134.
, Philosophy and Medicine in John Philoponus Commentary on Aristotles De
anima, DOP 38 (1984) 103110.
Tolkiehn J., Lexikographie, RE XII/2 (1925) 24322482.
Tomberg K.-H., Die Kaine Historia des Ptolemaios Kennos. Eine literarhistorische und
quellenkritische Untersuchung (Bonn 1968).
Tom P., Nuovi contributi per lOrthographia di Giovanni Tortelli (forthcoming).
Tomin J., Socratic gymnasium in the Clouds, SO 62 (1987) 2532.
Too Yun Lee, The Idea of Ancient Literary Criticism (Oxford 1998).
, Legal instructions in classical Athens, in Education in Greek and Roman Antiq-
uity, ed. Yun Lee Too (Leiden-Boston-Kln 2001) 111132.
Toomer G. J., Apollonius of Perga, in Dictionary of Scientific Biography I (1970), 179193.
, Hipparchus, in Dictionary of Scientific Biography 15, Suppl. 1, ed. C. C. Gillespie
(New York 1978) 207224, also available on-line in the Complete Dictionary of Scien-
tific Biography, at http://www.encyclopedia.com/topic/Hipparchus_(astronomer)
.aspx#1.
Torre C., Tra Oriente e Occidente: i Giambi di Eugenio da Palermo, MStudStor 14
(2007) 177213.
, Un intellettuale greco di epoca normanna: Filagato da Cerami e il De mundo
di Aristotele, MStudStor 15 (2008) 63141.
Tosi R., Studi sulla tradizione indiretta dei classici greci (Bologna 1988).
, La tradizione proverbiale greca e Aristofane di Bisanzio (osservazioni sulle
), in Tradizione e innovazione nella cultura greca da Omero allet
1414 Bibliography
ellenistica. Scritti in onore di Bruno Gentili, ed. R. Pretagostini (Roma 1993), III,
10251030.
, Aristoph. Byz. fr. 6 Sl., Eikasms 5 (1994a) 233235.
, La lessicografia e la paremiografia in et alessandrina ed il loro sviluppo suc-
cessivo, in La philologie grecque lpoque hellenistique et romaine, Entretiens Hardt
40, ed. F. Montanari (Vanduvres-Genve 1994) 143209 = [1994b].
, Callimaco e i Glossografi omerici, Eikasms 8 (1997a) 223240.
, Osservazioni sul rapporto fra Aristofane di Bisanzio e lAntiatticista, in .
Scritti in onore di Giuseppe Morelli (Bologna 1997) 171177 = [1997b].
, Appunti sulla filologia di Eratostene di Cirene, Eikasmos 9 (1998a) 327346.
, Harpokration [2], DNP 5 (1998b) 164165.
, Hesychios [1], DNP 5 (1998c) 514515.
, Iulius Pollux [IV 17], DNP 6 (1999) 5153.
, Onomastikon (I.II.), DNP 8 (2000a) 12141217.
, Orion [3], DNP 9 (2000b) 32.
, Pamphilos [6], DNP 9 (2000c) 214215.
, Lessicografia foziana e riscoperta dei classici, in Bisanzio nellet dei Mace-
doni, eds. F. Conca, G. Fiaccadori (Milano 2007) 257266.
Totelin L. M. V., Galens use of Multiple Copies in his Pharmacological Treatises, in
Authorial Voices in Greco-Roman Technical Writings, AKAN-Einzelschriften vol. 7,
eds. L. Taub, A. Doody (Trier 2009) 8192.
Tovar A., Nicetas of Heraclea and Byzantine Grammatical Doctrine, in Classical Stud-
ies Presented to B. E. Perry (Urbana 1969) 223235.
Toye D. L., Dionysius of Halicarnassus on the First Greek Historians, AJPh 116 (1995)
279302.
Trabattoni F., La filosofia antica (Roma 2002).
Trapp M. B., Allegory, Greek, in Oxford Classical Dictionary, eds. S. Hornblower,
A. Spawforth (Oxford 20124) 6263.
Treadgold W., The Macedonian Renaissance, in Renaissances before the Renaissance,
ed. W. Treadgold (Stanford 1984) 7598.
, The Byzantine Revival (Stanford 1988).
, The Break in Byzantium and the Gap in Byzantine Studies, ByzF 15 (1990)
289316.
Trpanier S., Empedocles. An Interpretation (New York-London 2004).
Triantafyllidis M., . (Thessaloniki 19381981).
Trojahn S., Die auf Papyri erhaltenen Kommentare zur Alten Komdie. Ein Beitrag zur
Geschichte der antiken Philologie (Mnchen-Leipzig 2002).
Trombley F. R., Hellenic Religion and Christianization c. 370529, I (Boston-Leiden 2001).
Tsames D. G. (ed.), (Thessaloniki 1985).
Tsantsanoglou K., Punctuation, in A History of Ancient Greek. From the Beginnings to
Late Antiquity, ed. A.-F. Christidis (Cambridge 20102) 13261333.
Bibliography 1415
Tsirpanlis Z., Byzantine Humanism and Hesychasm in the Thirteenth and Fourteenth
Centuries, Gregorios o Palamas 80 (1997) 309322.
Tsitsiridis S., :
(Athens 2001).
Tulli M., Il giudizio di Platone sull esegesi allegorica, in Interpretazioni antiche e
moderne di testi greci. Ricerche di filologia classica III (Pisa 1987) 4552.
, Omero nel III libro delle Leggi, in Platos Laws: From Theory into Practice, eds.
S. Scolnicov, L. Brisson (Sankt Augustin 2003) 227231.
Turner E. G. (ed.), The Hibeh Papyri, II (Oxford 1955).
, P.Oxy. 2404. Aeschines, In Ctesiphontem 5153, P.Oxy.XXIV (London 1957)
130132.
, Lrudition alexandrine et les papyrus, CE 37 (1962) 135152.
, Greek Papyri. An Introduction (Oxford 1968; 19802), Italian transl. (by M. Man-
fredi): Papiri greci (Roma 1984).
, Greek Manuscripts of the Ancient World (Oxford 1971), Second Edition Revised
and Enlarged by P. J. Parsons, BICSSuppl. 46 (London 1987).
, Athenian Books in the Fifth and Fourth Centuries B.C., 2. Aufl. (London 1977).
Turyn A., The Manuscript Tradition of the Tragedies of Aeschylus (New York 1943).
, Studies in the Manuscript Tradition of the Tragedies of Sophocles (Urbana Ill.
1952).
, The Byzantine Manuscript Tradition of the Tragedies of Euripides (Urbana Ill.
1957).
, Codices Graeci Vaticani saeculis XIII et XIV scripti (Vatican City 1964).
, Dated Greek Manuscripts of the Thirteenth and Fourteenth Centuries in the
Libraries of Italy (Urbana Ill. 1972).
Ucciardello G., Astyages, LGGA (2006a).
, Esegesi linguistica, glosse ed interpretamenta tra hypomnemata e lessici. Mate-
riali e spunti di riflessione, in I classici greci e i loro commentatori, ed. G. Avezz,
P. Scattolin (Rovereto 2006) 3583 = [2006b].
, Seleucus [1], LGGA (2006c).
, Glaucus [1], LGGA (2007a).
, Nuovi frammenti di oratori attici nel Vat. gr. 7, Miscellanea Bibliothecae
Apostolicae Vaticanae 14 (2007b) 431483.
, Lupercus, LGGA (2008).
, Hypomnemata papiracei e lessicografia (Messina 2012).
Uhl A., Pherekydes von Athen. Grundriss und Einheit des Werkes (Diss. Mnchen 1963).
Uhl S., Servius als Sprachlehrer. Zur Sprachrichtigkeit in der exegetischen Praxis des
sptantiken Grammatikerunterrichts (Gttingen 1998).
Uhlig G., Noch einmal und zum ersten Male ,
Neue Jbb. fr Phil. und Pdagogik 121 (1880) 789798.
(ed.), Dionysii Thracis Ars Grammatica, Grammatici Graeci I/1 (Leipzig 1883).
1416 Bibliography
, Between poetry and rhetoric: Libanios use of Homeric subjects in his pro-
gymnasmata, QUCC 95 (2010) 131152.
Weber G., Dichtung und hfische Gesellschaft. Die Rezeption von Zeitgeschichte am Hof
der ersten drei Ptolemer (Stuttgart 1993).
(ed.), Alexandreia und das ptolemische gypten. Kulturbegegnungen in helle-
nistischer Zeit (Berlin 2010a).
(ed.), Kulturbegegnungen in Alexandreia und im ptolemischen gypten.
BegriffeProblemePerspektiven, in Weber (2010a) 929 = [2010b].
Weber H., Phon smantik: Zur Geschichte und Problematik semantischer Wortdefi-
nitionen, in Sprechen und Hren. Akten des 23. Linguistischen Kolloquiums, Berlin,
1988, ed. N. Reiter (Tbingen 1989) 389398.
Webster T. B. L., Studies in later Greek comedy (Manchester 19702).
, Monuments illustrating Old and Middle comedy (London 19783).
Wehrli F., Zur Geschichte der allegorischen Deutung Homers im Altertum (Borna-Leipzig
1928).
(ed.), Die Schule des Aristoteles: Texte und Kommentar, 7: Herakleides Pontikos
(Basel 1953; Basel-Stuttgart 19692e).
(ed.), Die Schule des Aristoteles. Texte und Kommentar, 9: Phainias von Eresos,
Chamaileon, Praxiphanes (Basel-Stuttgart 1957; 19692c).
(ed.), Die Schule des Aristoteles, Texte und Kommentar, 1: Dikaiarchos (Basel-
Stuttgart 19672).
(ed.), Die Schule des Aristoteles. Texte und Kommentar, 4: Demetrios von
Phaleron (Basel-Stuttgart 19682).
(ed.), Die Schule des Aristoteles. Texte und Kommentar, 5: Straton von Lampsa-
kos (Basel-Stuttgart 19692a).
(ed.), Die Schule des Aristoteles. Texte und Kommentar, 8: Eudemos von Rhodos
(Basel-Stuttgart 19692b).
(ed.), Die Schule des Aristoteles. Texte und Kommentar, 10: Hieronymos von Rho-
dos, Kritolaos und seine Schler, Rckblick: der Peripatos in vorchristlichen Zeit, Reg-
ister (Basel-Stuttgart 19692d).
, Der Peripatos bis zum Beginn der rmischen Kaiserzeit, in Die Philosophie
der Antike, B. 3: ltere Akademie. AristotelesPeripatos, ed. H. Flashar (Basel 1983)
459599.
Weidemann H., Grundzge der Aristotelischen Sprachtheorie, in Sprachtheorien der
abendlndischen Antike, ed. P. Schmitter (Tbingen 1991) 170192.
Weiss R., The Translators from the Greek at the Angevine Court of Naples, Rinasci-
mento 1 (1950) 195226, repr. in Weiss (1977) 108133.
, Medieval and Humanist Greek (Padova 1977).
Weissenberger K., Prosa, in Historisches Wrterbuch der Rhetorik, vol. 7 (Tbingen
2005) 321348.
1420 Bibliography
, The Athenian Institution of the Khoregia: the Chorus, the City and the Stage
(Cambridge 2000).
Winter F., Schulunterricht auf griechischen Vasenbildern, BJ 123, 2 (1916) 275285.
Wipprecht F., Quaestiones Palaephateae (Diss. Bonnae 1892).
Wisse J., (et al.), Affektenlehre, in Historisches Wrterbuch der Rhetorik, vol. 1 (Tbin-
gen 1992) 218253.
Wissmann J., Athenas unreasonable advice: the education of Telemachus in ancient
interpretations of Homer, GRBS 49 (2009) 413452.
Witty F. J., The Other Pinakes and Reference Works of Callimachus, Library Quarterly
43,3 (1973) 237244.
Whrle G., Delectare, in Historisches Wrterbuch der Rhetorik, vol. 2 (Tbingen 1994)
521523.
, Movere, in Historisches Wrterbuch der Rhetorik, vol. 5 (Tbingen 2001)
14981501.
Wolf F. A., Prolegomena ad Homerum, (Halle a.d. Saale 1795). Engl. ed.: Prolegomena to
Homer, transl. A. Grafton, G. W. Most, J. E. G. Zetzel (Princeton 1985).
Wolf H., : In Claudii Ptolemaei quadri-
partitum enarrator ignoti nominis, quem tamen Proclum fuisse quidem existimant
(Basle 1559).
Wolf P., Vom Schulwesen der Sptantike. Studien zu Libanius (Baden-Baden 1952).
Wolfdorf D., Prodicus on the correctness of names: the case of , and
E, JHS 131 (2011) 131145.
Wolska-Conus W. (ed.), Cosmas Indicopleusts. Topographie Chrtienne, I (Paris 1968).
, Lenseignement du droit Byzance au XIe sicle, T&MByz 7 (1979) 1107.
, Stphanos dAthnes et Stphanos dAlexandrie. Essai didentification et de
biographie, REByz 47 (1989) 589.
, Les commentaires de Stphanos dAthnes au Prognostikon et aux Aphorismes
dHippocrate: de Galien la pratique scolaire alexandrine, REByz 50 (1992) 586.
, Stphanos dAthnes (dAlexandrie) et Thophile le Prtospathaire, commen-
tateurs des Aphorismes dHippocrate, sont-ils indpendents lun de lautre?, REByz
52 (1994) 568.
, Sources des commentaires de Stphanos dAthnes et de Thophile Proto-
spathaire aux Aphorismes dHippocrate, REByz 54 (1996) 566.
, Un Pseudo-Galien dans le commentaire de Stphanos dAthnes aux Apho-
rismes dHippocrate: , REByz 56 (1998) 578.
, PalladiosLe Pseudo-Galien ( )dans le commen-
taire de Stphanos dAthnes aux Aphorismes dHippocrate, REByz 58 (2000) 568.
Woodbury L., Parmenides on names, HSPh 63 (1958) 145160.
, Aristophanes Frogs and Athenian Literacy: Ran. 5253, 1114, TAPhA 106 (1976)
349357.
Bibliography 1425
, The judgment of Dionysus: books, taste, and teaching in the Frogs, in Greek
tragedy and its legacy: essays presented to D. J. Conacher, eds. M. Cropp, E. Fantham,
S. E. Scully (Calgary 1986) 241257.
Woodhead W. D., Etymologizing in Greek literature from Homer to Philo Judaeus
(Toronto 1928).
Woodhouse C. M., Georgios Gemistos Plethon. The Last of the Hellenes (Oxford 1986).
Woodruff P., Socrates and the new learning, in The Cambridge Companion to Socrates,
ed. D. R. Morrison (Cambridge 2011) 91110.
Wouters A. (ed.), The Grammatical Papyri from Graeco-Roman Egypt. Contributions to
the Study of the Ars Grammatica in Antiquity (Brussels 1979).
, The Chester Beatty Codex AC 1499. A Graeco-Latin Lexicon on the Pauline Epis-
tles and a Greek Grammar (Leuven-Paris 1988).
, The grammatical papyrus P. Berol. inv. 9917, in Akten des 21. Internationalen
Papyrologenkongresses (Berlin, 13.19.8.1995), eds. B. Kramer, W. Luppe, H. Maehler,
G. Poethke (Stuttgart-Leipzig 1997) 10211033.
, A Manual on the Parts of Speech, in Literarische Texte der Berliner Papyrus-
sammlung. Zur Wiedererffnung des Neuen Museums (BKT X), ed. F. Reiter (Berlin
2012) 135154 and pll. XIXXXIV.
Wouters A., Swiggers P., Ladverbe chez les grammairiens latins de lAntiquit tardive,
Beitrge zur Geschichte der Sprachwissenschaft 17 (2007) 75118.
, New Papyri and the History of Ancient Grammar: The Chapter in
P.Berol. 9917, in Matthaios-Montanari-Rengakos (2011) 313330.
Wright M. R., Empedocles. The Extant Fragments, edited with an introduction (New
Haven-London 1981).
Wright M., The tragedian as critic: Euripides and early Greek poetics, JHS 130 (2010)
165184.
, The Comedian as Critic: Greek Old Comedy and Poetics (London 2012).
Wst E., Munatios [3a], RE Suppl. VIII (1956) 359361.
Xenis G. A. (ed.), Scholia vetera in Sophoclis Electram, SGLG 12 (Berlin-New York 2010a).
, Scholia vetera in Sophoclis Trachinias, SGLG 13 (Berlin-New York 2010b).
Young D. C., Pindar, Aristotle, and Homer: A Study in Ancient Criticism, ClAnt 2 (1983)
156170.
Yunis H., Introduction: why written texts?, in Written Texts and the Rise of Literate Cul-
ture in Ancient Greece, ed. H. Yunis (Cambridge 2003) 114.
Zagklas N., A Byzantine Grammar Treatise attributed to Theodoros Prodromos, Grae-
coLatina Brunensia 16 (2011) 7686.
Zanker G., Realism in Alexandrian Poetry: a Literature and its Audience (London 1987).
, Callimachus and his Critics: a Response, Prudentia 29 (1997) 3446.
Zanker P., Die Maske des Sokrates: das Bild des Intellektuellen in der antiken Kunst
(Mnchen 1995).
1426 Bibliography
Abas1065 n. 30 Danaids649
Ab al-Faraj ibn al-Tayyib1265 Diktyoulkoi648
Academy of Plato13, 2223, 24, 25, 27, Lycurgus138 n. 360
4748, 75, 79, 144 n. 383, 193, 1040 n. 109, Pers.157
1050 Aesop358, 406
accent784 aesthetics709, 713, 729, 743, 751752, 753
basic word accent945 Aether1014
recessive939941 Aetius335, 394
accentuation204, 252, 261, 262, 924927, Agamemnon135, 136, 152, 689, 692705,
934948 1024
accidence see accidents Agatharchides335
accidentia see accidents Agathias314, 320
accidents784788, 790, 990, 991, 1003 Agathocles of Cyzicum106107, 137 n. 355,
accompaniment vs replacement861 146 n. 391, 154
accusations among philosophers and agonistic text types560, 568570, 586, 591,
schools10381056 593
Achaeus174 agreements in error12231224
Achean wall1109 Ajax1024
Achilles134136, 152, 673705, 717, 733, 750, akamatos933
1024 Akestorides, cup by18
Achilles Tatius330, 333, 358, 363 akolouthia see order
Achilles [Tatius], Aratus commentator Al-Mamun328
1217 n. 5, 1231 Albertus Magnus1264
Acropolites, Constantine405 Albinus340
Acropolites, George405406 Alcaeus119122, 134 n. 344, 136 n. 352,
Actium193 137 n. 358, 222, 249, 374, 1116
Acusilaus of Argos10591060, 1098 Alcibiades113
Ademollo F.1012 n. *, 1018 n. 27, 1021 n. 34, Alcidamas4142, 54, 721, 748, 1034
1023 n. 41, 1027 n. 56 Alcman120, 137, 152, 160, 171, 178, 182
Adrastus of Aphrodisias1054 Alesse F.1048 n. 139
adverb779780, 782, 785, 787, 881 Alexander Aetolus81, 101, 102 n. 164, 641, 749
Aelian418 Alexander Numenius800
Aelius Dionysius325, 332, 584 n. 218, 632, Alexander of Aegae225
829 Alexander of Aphrodisias340, 375,
Aelius Stilo Praeconinus, L.165, 623 401 n. 524, 580, 1053, 1053 n. 158, 1055,
Aeneas1121 12631264
Aeschines174, 325 Alexander of Cotiaeion238239, 296
In Ctes.649 Alexander of Nicaea346, 348, 351
Aeschylus10, 14, 15, 19, 28, 29, 34, 36, 57, 85, Alexander the Great63, 68, 70, 72, 76,
119, 121, 138, 157, 338, 347, 393, 423, 425, 77 n. 58, 80, 81 n. 82, 85 n. 105, 114, 115,
427428, 648, 1034 163, 1115
Alexandria61 n. 4, 70, 7482, 83, 86, 8890, analogy125 n. 306, 140142, 150 n. 411,
100, 109111, 114, 118, 126, 130, 148, 154, 160 n. 457, 166, 174 n. 541, 176, 181, 182,
157, 164, 167, 170 n. 507, 180, 196, 198, 205, 254, 262264, 527, 806, 807,
197 n. 40, 204 n. 64, 206, 207, 209, 211, 809, 810, 811, 814, 815, 817, 818, 820, 822,
212, 213, 214, 215, 216, 219, 221, 223, 225, 823, 827, 832839, 841, 891, 950, 952,
226, 227, 229, 232, 233, 249, 251, 254, 956, 959962, 964, 966, 968, 970971,
257, 261, 267, 268, 269, 270, 272, 274, 277, 975
284, 287, 303305, 311, 333 n. 192, 341, analogy-anomaly controversy263, 801,
383, 642, 643, 715, 720, 729730, 731, 734, 812, 838839
735, 737, 738, 741, 749, 753, 819, 1037, conditions/criteria of808, 809, 813, 815,
1234 834838
Alexandrian ideology, Alexandrianism definitions of832833
85, 209 perfect analogies808
Alexandrian War89, 170 Ananias of Shirak303
Alexion255 Anastasius I270, 271
Alexis360 Anastasius of Sinai306308, 400, 633
Alexius I Comnenus355, 366, 369, 377 Anaxagoras of Clazomenae802, 1035 n. 89
allegoresis86, 106, 107, 116, 136 n. 352, Anaxarchus1115
151 n. 415, 152 n. 420, 154, 162, 1095 Anaximander of Miletus89, 18, 37, 38,
allegory31, 3539, 46, 48, 52, 361364, 376, 1058 n. 3
378379, 382, 388, 390392, 397, 402, Anaximenes of Lampsacus794, 11141115
420, 427, 675680 Anaximenes of Miletus89, 18, 736 n. 103
rhetorical allegory (metaphora) anchoring practices900922
136 n. 352 ancora648
alloiosis10011011 Ancyra207
alphabet5, 6, 8, 1112, 21 Andania1124 n. 68
Greek232 Andromenides5758
alphabetical order623, 626627, 635 Andron of Ephesus1061
alphabetization547, 555, 572, 582 Andron of Halicarnassus1062
Amarantus of Alexandria244 Andronicus Comnenus392
ambiguity in language10291056 Andronicus II Palaeologus403, 409, 411
Amerias624 Andronicus of Rhodes169 n. 503, 1040, 1051,
Ammonius, father of Tryphon180 1051 n. 152, 1052, 1053, 1053 n. 160, 1055,
Ammonius, grammarian (Aristarchus pupil) 1236, 12521255
96, 100, 132133, 156, 159, 568, anecdotes, chreiai205
584 n. 216, 658, 659, 660 Anecdoton Parisinum93 n. 133, 165 n. 486
ad Il. (P.Oxy. 2.221)173 n. 530 Angeli da Scarperia, Iacopo444
Ammonius, philosopher402, 1032 Anna Comnena366, 370, 372, 375, 377
Ammonius Saccas265 annotations549551, 553557, 563, 581, 586,
[Ammonius]469, 484485, 495, 635 590, 598
amphibolia10431056 anomaly124, 125, 136, 141, 150 n. 411, 805, 821,
Amphimachus1043, 1044, 1045 838
Amsler M.1021 n. 34 anonymous professor349350
anachronism711, 721, 729, 732, 740, 745, 747, Anonymus Iamblichi1016
749 Anonymus Parisinus383
Anacreon103, 119, 120, 137 n. 358 Anselm of Canterbury437
anagnosis, word division95, 137 n. 355 Anselm of Havelberg395
analogia see analogy anthologies247
general index 1429
Boccaccio, Giovanni439 215 n. 100, 231, 233, 241, 242, 256, 285,
Boethius409, 415 306, 312, 320, 382, 392, 546 n. 5, 550,
Boethus, lexicographer280281, 332, 629 567, 598, 614, 953, 974, 624, 628, 710,
Boilas, Eustathius358 714 n. 38, 718, 729, 730, 731, 734, 735, 741,
Boissonade J. F.842 744, 746, 749, 986, 1037
Bonazzi M.1053 n. 158 Aitia1037
book(s)60, 64, 65, 74, 75 n. 52, 77 n. 58, Diegeseis580, 587
78, 80, 8299, 110, 116, 118, 121, 122,
130 n. 332, 133, 167170, 172 565
book titles107 n. 200 565
collecting books64, 75 n. 52, 77, 78, 565
82, 84, 87, 94, 144, 168, 172 565
copying books 84, 85, 9092, 99, 565
104 n. 180, 167169 Pinakes570, 572573, 597
philosophical collection (of books)305,
340341 572
retrieval of books64, 84, 90, 99 569
Boudon Millot V.1182 n. 289 Callisthenes of Olynthus1115
boulema12291231 Callistratus of Alexandria92 n. 128,
Bourdieu P.5 105 n. 187, 126128, 136 n. 351, 155 n. 436,
Brancacci A.1016 n. 17, 1024 n. 43 156, 159, 666, 781
breathings924927, 928930 Callistus, Andronicus449, 452
bridging technique, in etymology903 Camariotes, Matthew448
Browning R.12601262 Camaterus, John393
Brucheion78, 82, 89 canon(s) of authors and works62, 85,
Bruni, Leonardo441442 120 n. 273, 137, 1093
Bryennius, Joseph436 Carbones, George428
Bryennius, Manuel407 n. 570, 410, 430 Carians1044
Bryson of Heracleia1038 Carson A.1035 n. 86
Burgundio of Pisa394, 395 Carystius157
Byzantine era194 case, ptosis770, 878, 986, 991, 995996, 999,
Byzantine, Byzantium100, 118, 553, 10021005
558 n. 67, 559, 565, 578579, 584 n. 218, Cassander77, 78, 83
584 n. 220, 586587 Cassio A. C.1053 n. 158
Byzantinists194 Cassiodorus313
Cassius Dio314 n. 80, 325, 341, 346, 357, 382,
Cabaces, Demetrius Raul448 411
Caecilius of Cale Acte195, 282, 986, Cassius Longinus264, 265266, 278, 628,
10011011 629, 632
Caesar, G. Iulius89, 163, 820, 836, 841 n. 239 Castor of Rhodes1077
Caesarea207, 287 Catalogue of the Ships1043, 10971098,
Caesaropapism194 1108, 1117, 11221123
Calchedon1016 catalogues571572, 580, 586587, 598
Calecas, Manuel437 Catrares, John428
Caligula, Emperor221, 226 censorship in Byzantium308, 414
Callimachus61 n. 3, 69 n. 22, 81, 84 n. 97, Cephalas, Constantine348349, 354, 413
85 n. 107, 9698, 100102, 105, 107111, Cephisodorus48
113, 118, 123, 124, 126, 138, 147, 179, 214, Chaeremon206, 225226, 225 n. 165, 252
1434 general index
gnomologies309 grammatistike203
Gomperz T.1019, 1019 n. 28 Grapti, Theophanes and Theodorus309
Gorgias of Leontini20, 4142, 54, 365, 623, Gramanns Law929930
724, 1015 Grassus, John401
Gottschalk H. B.1054 n. 163, 1254 Greece1014, 1039
Grfenhan A.188189 Greek Anthology314, 330, 348349, 386,
Geschichte der klassischen Philologie im 413414, 418, 426
Alterthum188189 Greeks1014
Graham D. W.1020 n. 32 Gregoras, Nicephorus405, 415, 431435,
grammar, grammatike techne/grammatike, 438
ars grammatica10, 16, 1922, 29, Gregory kouboukleisios356
3644, 48, 51, 5557, 67, 75, 93, 95, Gregory of Campsa348
112 n. 234, 125, 128, 140, 141 n. 377, Gregory of Corinth181, 261, 272, 367,
148150, 153, 160165, 171, 172, 176, 373375, 479480, 495, 831, 832
180183, 185, 185 n. 1, 186, 187, 190203, Gregory of Nazianzus302, 312, 320, 345,
203 n. 61, 205, 210, 212, 216, 236, 250, 359360, 363364, 369, 373, 387, 413,
252, 601606, 611, 613614, 618, 759797 422, 440
grammar and other disciplines793795 Gregory of Nyssa308, 406
grammar and syntax852 Grensemann H.1127
grammar as an art760761, 783 Grintser N. P.1038 n. 103
grammar as hexis516522, 534535, Groningen B. A. van1042 n. 121, 1145 n. 87
540 Grosseteste, Robert403
chair for grammar206, 207, 208, 209, Guarino da Verona441443
227 Gudeman A.189
grammar, definition of191, 198, 199, 200, Grundriss der Geschichte der Klassischen
203 n. 62, 515544 Philologie189
grammar, institutionalization of515 Guillelmus de Moerbeke1264
grammar manual/book759, 782783, Gurd S. A.1042 n. 120
788, 790792, 794 Guthrie W. K. C.1015 n. 12
grammar, object and status of515544 Gutzwiller K. J.1246
grammar, parts of199, 205, 273, 523524,
526528, 532, 537538, 542543, 760, Haag E.1020, 1020 n. 33
773774, 783, 788, 796 Habron180 n. 579, 251252, 609, 611, 843
grammar, systematization of198 Hadot I.12591260
grammarian, grammatikos, grammatistes, Hadrian, Emperor207, 228, 234, 235, 246,
grammaticus66, 73 n. 42, 105 n. 188, 256, 282, 294, 665
112, 149 nn. 407408, 157, 162, 168, 185, hagiographies324, 354, 363, 409, 419
185 n. 1, 186, 190, 192, 193, 195, 196, Halliwell S.1017 n. 22
201203, 203 n. 61, 204, 207, 209, 210, Hanson A. E.1049
213, 216, 234236, 249, 266, 267, 517518, hapax legomena134
537 harmony985, 989, 991, 992
grammata see letters Harpocration, lexicographer174, 176 n. 548,
grammatical figures10011011 229, 284, 325, 332, 415, 419, 630
grammatical lexica631633 Harpocration of Argos280, 629
Grammatici Graeci186 Haslam M.647, 662
grammaticus see grammarian Hecataeus of Miletus10581059, 10951097,
grammatike see grammar 1101, 11031104, 1112
grammatike techne/grammatike see grammar Hegesianax of Alexandria1065, 1082
grammatikos see grammarian Helen1015, 10991100
1442 general index
396, 402, 407, 409, 413, 417, 426, 442, 376, 377, 378380, 383, 386, 387,
503504, 614, 715, 731, 733, 1014, 1015, 388391, 394, 397, 400, 401402, 406,
1034, 10951096, 10981099, 1101, 1118 409, 413, 416, 417, 419, 420421, 427428,
[Hesiod]73 n. 40, 111, 118 436, 438439, 442, 443, 444, 447, 448,
Hesychius of Alexandria628, 630 450, 453, 461, 464466, 468, 474,
Hesychius of Miletus100, 222, 227, 229, 234, 488489, 494, 496, 499502, 512,
278, 289290, 314, 315, 342, 354, 355, 546547, 551554, 559, 564, 568570,
465, 471, 473, 494, 510, 513, 636 574, 577 n. 179, 587, 593594, 614, 642,
Epist. ad Eulogium175, 178 n. 566 643, 647, 652, 653, 656, 657, 658, 659,
Hexapterygos, Theodore401 660, 661, 665, 666, 667, 668, 669, 670,
Hicks R. D.874 n. 30, 1013 n. 4, 1031 n. 68, 671 n. 91, 706755, 816, 824, 825826,
1039 n. 107, 1049, 1049 n. 140 830, 841, 841 n. 244, 842, 844, 1017,
Hieronymus of Rhodes72 n. 38, 73 n. 40, 119, 1024, 1034, 1038, 1095, 1098, 11001102,
163 11051107, 1110, 11151118, 11201121,
Himerius334 1123
Hipparchus of Nicaea98 n. 156, 12171232, Homeric language811, 815, 829830,
1234 848
Comm. in Arati et Eudoxi Phaenom. Homeric scholia325, 329330, 338339,
560561 347, 354, 356, 368, 377, 389, 397399,
Hippias of Elis20, 3940, 115 n. 249, 747, 402, 407, 428, 436, 439440
1024, 1098, 1106 Homeric text67, 11, 20, 27, 2932, 33,
Hippias of Thasus38, 1029 n. 64, 1034 3641, 4548, 5152, 53, 5759
Hippocrates, Hippocratic corpus9, 34, Il.230, 240, 247, 253, 277, 647, 652
9597, 117 n. 260, 165, 176, 281282, 306, Od.214, 230, 240, 247, 252, 253, 647, 652,
347, 374, 558559, 562563, 582584, 1024
665, 1050, 1053, 11361140 rhetorical criticism of673705
Arabic reception of12001215 Homerus Sellius579
commentaries on11531164, 11831195, homonyms3839, 50
12021208 homonymy, avoidance of915
Hippocratic biography1133, 11641169, Horace712, 715, 724, 731, 736 n. 103, 743, 744,
11951197 753
Hippocratic ekdoseis11481149, 11771183 Horapollo249, 271
Hippocratic glossaries11481153, humanism327, 364365, 429
11711177 humanism, Christian332, 386, 414, 419
Hippocratic language and the Hunger H.189
grammarians11431147 Hussey E.1035 n. 89
Hippocratic translations12091211 Hymn to Apollo, Homeric1110
Hippodamus of Miletus1038 Hypatia360
Hipponax137, 147 n. 398, 382 Hypatius of Ephesus316317
historia, historiography279, 950, 959960, hyperbaton9971000, 1003
962963, 968, 972977 Hyperechius201, 269270
historical present10071008 Hyperides174, 334
historice200 hyphen927
Holobolus, Manuel407, 408409, 415 hypodiastole927
Homer217, 229, 230, 231, 236, 237, 238, hypomnema see commentary
239, 247, 249, 255, 256, 277278, 287, hypomnema on Thucydides book 296
290, 306, 308, 309, 310, 312, 320, 329, hyponoia36, 37, 1035
330 n. 173, 338, 339, 341 n. 231, 344, 347, hypotheseis of Attic plays123 n. 290,
354356, 360, 361, 365, 368, 369, 374, 170 n. 506
1444 general index
206, 207, 282, 654, 670, 712, 714, 730, 738, Nicephorus of Constantinople313
741, 742, 1037 Nicetas of Heraclea367
music, musical rules and theory8, 10, 16, 18, Nicetas, Patriarch395
19, 21, 24, 40, 46, 47, 53, 55, 57, 58, 60, 63, Nicetas, Psellus companion362
64, 66, 75 n. 50, 114 n. 247, 129 n. 329, Nicholas-Nectarius of Otranto397398,
121, 143 n. 382, 182, 246247, 982, 401402
986988, 999 Nicholas of Damascus353
musical notation and scores63, 120 n. 276, Nicholas of Myra371
121 nn. 278 and 281, 129 n. 329 Nicholas of Rhegium403
Muzalon, Theodore405 Nicholas I Mysticus, Patriarch349
Mynosses1112 Nicholas III Grammaticus, Patriarch367
Mysia103, 143, 155 Nicholas V, Pope452453
Mythographus Homericus559, 580, 587, 739, Nickau K.643
10861088 Nicocrates of Cyprus167
mythology709, 732, 739, 744, 752, 754 Nicoll W. S. M.1042 n. 122
mythology and etymology see etymology Nicomachus of Gerasa303, 405
and mythology Nicomedia207
Night1014
Nagy G.662 Nomion1043, 1044
name, onoma29, 3234, 3741, 4244, 47, Nonnus of Panopolis306, 348, 413
5051 notes, marginal or interlinear see annotation
names between nature and convention noun762763, 766, 769772, 775, 777779,
10181056 784786, 791; see name, onoma
Nardelli J.-F.662 Novokhatko A.1015 n. 13, 1016 n. 17,
narrative voice711, 714, 718, 728, 730, 736, 1017 n. 20, 1017 n. 23, 1018 n. 26,
739740, 745 1024 n. 45, 1028 n. 60, 1033 n. 76,
Nastes1043, 1044, 1045 1034 n. 83, 1034 n. 84, 1037 n. 99
natural word accent945 number986, 991, 1002, 10051006, 1011
natural word order9981000 numerus versuum645 n. 16, 653, 656, 657
Nauck A.842 Nnlist R.1037 n. 97
Neleus of Scepsis78, 167169 Nysa167 n. 495
Nemesius395
Neoplatonic school191, 247 Obbink D.1013 n. 3
Neoplatonism in Byzantium300, 302, 340, obelos104 n. 180, 119, 551553, 652, 653, 654,
361, 364, 377, 397, 405, 427, 447, 449 656, 665, 666, 668
Neoptolemus719, 753 oblique cases, accent of943945
Neoptolemus of Parion628, 631 Ocean1102
Nero, Emperor224, 225, 226, 230, 279 Octavian Augustus, Emperor78, 145, 170,
Nerva, Emperor230 172, 180, 193, 213, 214, 220, 229, 666
Nestor680, 693 Odysseus73, 110, 115, 674, 680683, 687688,
Newmyer S. T.12441245 1024, 11171118
Nicaea207, 398401, 403, 405, 408 Old Testament86
Nicander of Colophon180 n. 575, 215, 348, Olympian Feasts114
368, 413, 443, 624, 625 Olympiodorus304, 340, 560, 1020
Nicander of Thyateira632 92 n. 130, 134,
Nicanor of Alexandria131 n. 335, 154 n. 431, 12281229
173, 236, 256257, 614, 650, 718 n. 48, onoma see name
740 Onomacritus31
1452 general index
Phryne157 Lach.1020
Phrynichus Arabius, lexicographer Phd.1049
124 n. 301, 186 n. 6, 293, 295, 296, 325, Resp.1038, 1049, 1056
332, 346, 422, 584 n. 218, 628, 632, 829, Tht.648, 1050
830 Ti.1038, 1050, 1056
physis/thesis (nomos), anthitesis799, Plato, playwright113
801806 Platonius
Pilatus, Leontius439440 568
Pinakes, pinacography98, 99, 102, 107109, 568
118, 119 n. 268, 123, 149 n. 406, 169, 570, Pleiad, tragic poets102, 126, 127
572573, 590, 593, 597 Pleias, costellation162
Pincipate194 Pletho, Georgius Gemistus447449
Pindar96, 103, 119122, 127, 129 n. 326, Pliny the Elder
129130 n. 330, 137, 138 n. 359, 152, 162, De dubio sermone820
171, 173, 174 n. 533, 178 n. 563, 182, 241, Pliny the Younger234
301, 374, 383, 386387, 393, 394, 413, 417, plot716, 717, 719, 722, 728, 739, 742, 743, 744,
423, 425, 447, 449, 490, 504505, 673, 752
730, 733, 753, 829, 841 n. 244, 1021 Plotinus245, 405 n. 552, 450
Pisander, mythographer10791081 Plutarch335, 343, 356, 358, 364, 411412, 430,
Pisistratean recension of the Homeric poems 438, 442444, 562, 564, 716, 721, 723,
140 n. 371 725, 746, 752, 1029, 1031, 1044, 1045
Pisistratus of Athens167, 383 Gryllus681683
Pius239240 [Plutarch]
plagiarism10381056 Vit. Hom.698
Planudes, Maximus261, 264, 382, 404405, poetic language vs scientific language
409415, 416 n. 621, 419, 422424, 426, 10341056
436, 438 poetics706755
[Plato]739 poetry195, 196, 209
Plato, philosopher35, 3940, 4247, 48, Hellenistic215, 216
58, 59, 63 n. 45, 78, 96 n. 145, 115 n. 251, Homeric218
119 n. 268, 140 n. 370, 149 n. 408, 156, Pohlenz M.1020 n. 31
164 n. 482, 165, 237, 245, 263, 280281, poietes ama kai kritikos185 n. 3
290, 293, 304, 308, 325, 330, 337, 340, polemics see antigraphe
344, 346, 347, 358, 364366, 368, 376, Polemo of Ilium147, 177, 1120 n. 63
396397, 401, 406, 407, 411, 417, 420,
431, 432433, 436, 438, 441443, 447, 570
450, 520, 534, 538, 558, 560, 583, 664, 570
669 n. 88, 673674, 675, 679, 709, 715, Poliziano, Angelo427 n. 693, 455
718719, 723, 724, 726, 727, 728, 729, Pollux, Julius Polydeuces284, 286, 294296,
731, 733, 736, 737, 739, 752, 753, 755, 343, 468, 493, 623625, 829
760763, 783, 793794, 801, 803, 829, 584
843, 844, 1013, 1014, 1016, 1020, 1021, 1021 Poltera O.1035 n. 85
n. 34, 1022, 1023, 1027, 1027 n. 57, 1030, Polus of Acragas1064
1035, 1038, 1040 n. 109, 1040 n. 122, 1049, Polybius347, 354, 728, 11181120
1050, 1054, 1056, 1098, 1100 Polybus34
Chrm.1020 Polycrates of Samos167
Cra.1018, 1018 n. 27, 1020, 1021 n. 34, 1027 polysyndeton993, 994, 996, 997
Euthphr.1020 Pompey the Great89, 167 n. 495
general index 1455
Pontani F.1024 n. 43, 1024 n. 44 Protagoras of Abdera10, 20, 24, 34, 35,
Pontus146 3637, 38, 4142, 716, 1015, 1016,
Porphyry of Tyre265, 301, 305, 322, 361, 364, 1016 n. 17, 1017, 1017 n. 19, 1018, 1038,
377, 380, 402, 427, 450, 629, 630, 675, 10981099
678, 680, 693, 1024, 1056 n. 170 Antilogies1038
Progymnasmata691 proverbs, apophthegmata205, 216, 217, 232;
Quaest. Hom.564 see paroemiography
Port-Royal Grammar851 Psellus, Michael356357, 360365, 372, 373,
Posidippus of Pella81 n. 87, 647 375, 379, 390, 393, 394, 400, 427
Posidonius of Apameia164 n. 483, 778, 1050 Ptolemies7577, 81 n. 87, 8385, 89 n. 117,
Praechter K.1261 90, 114, 126, 130, 144, 145, 154, 716, 738,
Praxiphanes of Mytilene17, 5657, 7273, 1037
108 nn. 208209, 137, 141 n. 374, 152, 163, dynastic crisis in 145/14476, 99,
569, 985, 994, 1032 143 n. 382, 157
predicate, predication761, 773, 775776 Ptolemy, Aristonicus father or son
predication see predicate 170 n. 508, 219220
prepon92 n. 129 Ptolemy Chennus334, 382, 389, 10881089
preposition781782, 784, 787, 882 Ptolemy, Claudius300 n. 13, 305, 326, 330,
prestige, cultural63, 75, 80, 146 n. 392 340, 382, 396, 410, 411, 426, 430, 432,
Preus A.1261 435, 441, 444, 568
Primavesi O.1051 n. 149 Ptolemy Epithetes105 n. 186, 106 n. 192, 117
Priscian258, 261, 300, 310, 313, 415, 607 n. 49, n. 261, 137 n. 355, 654 n. 48
608 n. 58, 609, 780 Ptolemy of Ascalon154, 159 n. 456, 218, 252,
problem-solving93 253254, 255, 273, 285286, 287,
Proclus270, 300, 303, 330, 340, 364365, 377, 612613, 615, 818
401, 406, 407, 560, 1027 Ptolemy Pindarion218, 811, 815, 830, 842,
Chrestomathy 335 848, 955
Procopius of Cesarea207, 299, 320 Ptolemy the Peripatetic529, 532534
Prodicus of Ceus20, 24, 35, 3839, 802, 1015, definition of grammar529534
1016 n. 15, 1018, 1023, 1023 n. 42 Ptolemy, unknown author of a lexicon635
Prodromus, Theodore369, 371, 372, 393 Ptolemy I Soter71, 72, 7678, 80, 81 n. 82,
progymnasmata322, 406407, 422, 445 83 n. 95, 642
pronoun781782, 784, 787, 868 Ptolemy II Philadelphus61 n. 4, 72, 78,
proparoxytonos938939 8284, 87, 88 n. 116, 94, 101, 126, 167,
proper name777779, 791, 795 169, 642
properispomenos938939 Ptolemy III Euergetes8385, 109111, 121,
proposition776779 516
prose writers, scholarship on9497, 107, 119 Ptolemy IV Philopator100, 111
n. 268, 130, 138140, 160 n. 463, 162, 174, Ptolemy V Epiphanes118, 145
183 Ptolemy VI Philometor130, 148
prosodia see prosody Ptolemy VII Neos Philopator130
psile prosodia926 Ptolemy VIII Euergetes II76, 130, 143
prosody131, 140, 149, 165, 166, 181, 182, 204, n. 382
205, 229, 252, 253, 254, 260, 261, 262, 271, Ptolemy IX Soter II100
272, 809, 810, 816, 817, 818, 821, 824, 831, ptosis see case
923948 Puglia E.1041 n. 118, 1042 n. 123, 1043 n. 125,
interaction with other branches of 1044, 1045, 1046, 1046 n. 133, 1048,
grammar947948 1048 n. 136
1456 general index