FRAME SEMANTICS AND THE
NATURE OF LANGUAGE*
Charles J. Fittmore
Depurtment of Linguistics
Universi of California} Berk
Berkeles, California 94720
om to ehedule sessions on “formulating the arg” or the es day of
this conference, based undoubtedly onthe simple wisdom that Lnowing what thing
is ie isa prereuiteto aking How got that nay, wat probeby thom ie
hope that the agora Conbtans tight seeoty ed eye ae
omens on the salar of tngoage Schaar ingeing ita soto eh eee
silatrey want aye on ie some he dea the hat sen a
Evolutionary sent they sre tying to conta. Unlonunales he elena
describing this tust scene isa notoriously difficult one: maki clear thee nature of
language mo teva igen 281 tnnkevesbody Hers ease
‘Ast contibuton attests parol he undeatanuing we ted iil eet for
your coniderationa way af taing abou one aspen a he psc commen
tog ia human Language something! wt tle toasting Tae ae
Secite I find tansy move Iprtn! than the formal spuctenee ae oeet
OF meaning oF moe inporam than he many global proper Be who ees
Inte purely compara prapectn, human ngage eae ee an ee
aie communtae noe tchaae fing hela Tk pean
suse supe that might nt ge mean or tatu mph nthe cae
highhghted, in the other papers 10 be read at this eanleiensas een
teen by traning the ppe in pescvng hinge an Comnunicaing a
eas of irpicing expen Ian aerate Tote si that
uncats ar eateguriey ac teed rough he prosersof eaeheg saat peepit
teat mth), words plan thc opet tone the eee
hut of ean monty am Engh fie ina ame Stat
ferent and mtu, how te ame concent coh Sea eke cee
if the communication sn comprehension praccessndte eed ey
tom helgedspeultions on ow the tly et eae igh area ne
‘Noluion toward language and on te eon of lenges
Thinkers who lucked tue evolutionary
ue ssoluionay perspective coud find questions ke
What is the rin of Lngunge? ncn ins way tha ne canoe Ween en
the sinphity and admire the reasonableness of Sanne! Johtnon'sspeculsions os
the ucon Bs Junnson atgucd a flows, Wlngunge wor ieee aan
inven either by chien of Sd Gould not have been mente ines
bess clhlron do not haus enough to invents language: eld nar hone he
imsented by aul, becuse alls sre nor phable enough to aed a tangoeee
divine inspiration, “ Ne edi
orsin
4> spond a past by the National Sees Fauna, grant ny SOLIS
»
Filmore 19764
Fillmore: Frame Semantics 2
‘When the question of the origin of language is considered from an evolutionary
perspective st loses much ofits clarity and simplicity, Should we be looking for the
Fist step in the chain of events that led to what we now see as human language? The
frst siep away (rom what? Or should we be tying to determine the fast step. the step
by which the Binal crtetial property of language was acquired? What might that have
been? Or is t perhaps some kind of kev step in this development that we need, that
step by which sas overcome the last serious barrier to the natural and inevitable
fequence of changes that ended in the kind of language we have today” Or, are these
the right alternatives? Since there is no evidence whatever on the nature of any of the
intermediate stages along. the great distance separating present-day language and
ven the most elaborated of the call systems, and since communication systems
fappear to have had different histories in different branches ofthe evolutionary tee,
fone of these questions may twen out to be useful. Ultimately, I believe, we ean be
satisfied with nothing short of the whole story, and for that we may need 10
Feconsiruct a long and complex chain of events.
‘Linguistic scholars, who have seldom fet called on to make anything more than
slight lip-service acknowledgments of the problem of the origin of language. have
[generally not troubled themselves very much about these uncertainties. For example,
‘one cominion suggestion (ound in the standard treatises and textbooks is that speech
‘ultimately goes back to the involuntary cries of animals; to Sir Alan Gardiner the
prototype was the squeal of a trapped rabbit? A part of this story, as im Sapir's
Yersion, is that the expressive vocalization, initially inseparable rom the experience
‘that caused it, came by some leap of imagination to be used as 2 name for the
‘experience. This story gives the origin of language an account somewhat similar 10
the way ritualized movements among some animals are traced back to more directly
functional movements appearing in acts of combat or surrender.
‘AL least since the time of Gabclents (1891)! there has been, in treatises on
language. a standard ls of factors in the development of man that were hospitable 10
tie bith of language: the upright gait, the enlarged brain, the infant's long
Aaelpkssness, and all the rest. Building pieces of a language-origin scenario around
fone or another of these has been a favorite and inexpensive pastime of grarmarians
and philologists for a long time, One such account, using the enlarged brain as the
explanatory principle, is given by C. D. Buck.* Buck tells us that while many animals,
including primitive man, had eres that expressed emotion, primitive man, because
his brain was lager than thal of any of his competitors, was alone able to gain “an
Awareness of a connection between the sound and something expressed by i” This
Account Seems 10 require two magic wands. one for explaining why the size of the
brain ts relevant (o the ability Lo perceive relationships, and the other for explaining
hhow “an awareness of a connection” makes it possible to go further
‘One of the miost imaginative of the language-origin stories is found in Otto
Jespersen’s 1921 book om the nature, origin, and development of language.* Jes-
Dpersemany have been the est to usc wetual ingwistc data in guiding his spceulations
He compared what be called “savage” langusges with civilized languages, imputing 0
the primordial Lunguage features that predominate in “savage” languages; and he
uber of wellelocumented principles of language change and applied them
forms wend in time to get regularized; the primitive tangwge
backwards. Irregul
‘therefore lacked regular or eecurring patterns of word formation, Long words tend in
to get shortened, and complex phonetic systems tend to get simplified; the
Primitive Lingwage abounded in very long words pronounced with an exuberant
Variety of dificult sounds, “Savage” languages tend (0 have tones, the long and
Phowctically intricate words in the primitive language were sung, not spoken. Many
Words for abstract or general concepts in modern languages originally2 Annals New York Academy of Sciences
or specific meanings: words inthe primitive language were names of highly specie
objet
ithe Jespersen scenario begins with people devising individual courtship and
battle songs. vsing in them as wide a variety of sounds as their vocal equipmem and
their inventiveness would allow them. To the members ofa farilanty proup, cach of
‘hee personal songs came (be asociated with its singst as a kind af Wagocrin
Jeumone, Within the group. one person could ttet io another by imitating he soce
‘The song, thus, became a proper name—and what, Jespetven asks, could te moots
conetete and specific than a. proper name? Onet this naming. Telatonship: got
Established within a group, it became possible (or people to use a proper make
fel to some tat ofthe owner of that mime, oF 0 femind the group of somc reat
that indiidua's history. On this base, thn, the processes of analogy and sinpien
tion did their work
Just one ofthe reasons why all ofthese stoves are unsats(ying is that they end too
z00n The las scene hs got 10 be mote shana state in wbich people name things and
Svagestions about the origin of language ace equently embedded in discussions
of the essential or evteia features of language, And in the way that many langonoe,
iin stories have concentrated on single critcal events, many inguin ino
language essentials are concerned with the discovery of some single alton caster of
Arts hat separates human language from everything ele The aeempt toe sic
feature" defintion of language has had some ofthe same kind of srory thot we oe
in the definition of man, namely one in which the offial defnition undereece