Mind Association

Decipherment, Translation, Interpretation

Author(s): B. C. Hurst
Source: Mind, New Series, Vol. 92, No. 366 (Apr., 1983), pp. 247-252
Published by: Oxford University Press on behalf of the Mind Association
Stable URL: http://www.jstor.org/stable/2253784 .
Accessed: 18/05/2011 07:05

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=oup. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Oxford University Press and Mind Association are collaborating with JSTOR to digitize, preserve and extend
access to Mind.

http://www.jstor.org
Mind (i 983) Vol. XCII, 247-252

Decipherment, Translation, Interpretation

B. C. HURST

There has been a tendency in discussions of translation to isolate

translation into one language from another from the recognition of sounds
or signs as language.' It will be the contention of the present paper that
successful translation cannot be isolated in this way, and that decipher-
ment, translation and interpretation form a continuous process, such that
not merely can translation and interpretation begin only when sounds or
signs are hypothesized as linguistic or communicative, but also our
confidence that sounds or signs are linguistic is a product of successful
translation and interpretation. Furthermore, our ability to recognize signs
as linguistic is related to our knowledge of the historical or social context of
their makers, and it is our knowledge of this context which allows us to
reach determinate translations and interpretations of utterances or inscrip-
tions. The present paper will deal exclusively with examples of written
language, and it will not be possible to discuss issues peculiar to spoken
language. Nevertheless, it may be suggested that issues raised by the
decipherment of texts not only have some bearing on general questions of
translation but also of themselves warrant attention.
In describing an object as a text, the decipherer is claiming that this
object is the product of human action and not of natural forces. That is to
say, he is putting forward the hypothesis that an action, the making of these
marks by a human, took place. He can support this hypothesis by showing
that this action fits more satisfactorily into an historical account describing
changes at the place and period of the making of the marks, than does an
event-the making of these marks by natural (non-human) forces-which
could be the claim of a rival hypothesis.2
In the description of the object as a text, however, the decipherer puts
forward the hypothesis that the producing of the object so described not
only was a human action but an act of communication. That is to say that
the human agent made the marks for the purpose of communicating some

See D. Davidson 'Radical Interpretation', Dialectica27, I973, p. 3I3: 'Having

identified his utterance as intentional and linguistic we are able to go on to
interpret his words: we can say what his words on that occasion meant.' The
purpose of the present argument is to show that the question as to whether an
inscription is intentional and linguistic can only be settled by successful
'interpretation', and cannot be settled prior to that.
2 The nature of this 'fit' and the manner in which historians decide between rival
claims as to what took place or rival claims as to how to describe data (e.g.
archaeological finds) is analysed in my article 'The myth of historical evidence'
in Historyand Theory,Vol. xx, No. 3, I98I.
247
248 B. C. HURST:

information (information being something with propositional content).

This distinguishes a text from a decoration, where the purpose of the
producer of the decoration is not that of communicating information.
It follows that where an object is described as a text, it is not sufficient
that the producer's purpose in inscribing the object was the inscription of
symbols of sounds or words. It is possible (as in some twentieth century
painting) that letters and words can be used for exclusively decorative
purposes, without the purpose of communicating information. Such
inscriptions cannot be considered texts. On the other hand, works of
representational art, although perhaps informative, cannot be held to be
texts. An example of representational art is not a text as it cannot be made to
render any determinate message with a determinate propositional content.
However, representational art, through picture writing, shades into
pictographic script as the decorative (artistic) purpose of the producer
becomes increasingly subordinate to his communicative (informational)
purpose and as the signs become increasingly symbolic rather than iconic.
As the signs become more symbolic, the need for decipherment increases.
As the informational (propositional) content increases, so there is more to
decipher. As the messages become increasingly determinate so the
possibility of decipherment increases. Thus, it is successful decipherment
which makes possible the distinction between script and picture.
This raises the question as to the nature of successful decipherment. The
first requirement for successful decipherment is that the value or function
given to any sign (sign type) cannot be arbitrarily altered for different
tokens of that sign type. This does not mean that any sign has only one value
or function, but it does mean that the rules for changes in that value or
function are formulated (e.g. position in relation to other signs, presence of
demonstratives, etc.). The ranges of values or functions allowed by the
rules and the rules themselves must remain the same throughout the
decipherment of the text. It might seem, on this basis, that the decipherer
could arbitrarily restrict himself to a single text, declaring that other texts
(similar in script but refractory to his decipherment) were written in other
languages or other scripts. However, this position would be difficult to
maintain if it could be shown by historical accounts related to other data
that these other texts were inscribed by members of the same society at the
same period. In such a case, for the decipherment to be considered
successful, the range of values or functions for signs and the rules for their
use would have to apply to all relevant texts, that is all the texts of that
society produced at that period.
A second requirement for successful decipherment is that the sentences
of the deciphered text should not be meaningless. This requirement follows
from the distinction, previously made, between text and decoration, where
texts were distinguished by their communicative purpose. When the
sentences of the 'text' are meaningless, then nothing is communicated.
This requirement as to the sentences of the text not being meaningless is
necessary for the basic processes of decipherment:

(i) the identification of sign types and the individuating of sign

tokens;
 DECIPHERMENT, TRANSLATION, INTERPRETATION 249

(2) the assigning of values or functions to sign types;

(3) the establishment of the identity of each sign in every text (i.e. the
establishing of which type each sign is a token).

A basic problem for the decipherer is that no two signs (even if tokens of
the same type) are identical, as no two scribes write in identical fashion and
no single scribe writes the same sign twice in exactly th'e same way.
Morphology can be only an approximate guide. Thus, from his scrutiny of
the texts, the decipherer will put forward hypotheses, on the basis of
morphological similarity, as to the identity of sign types, and this will allow
him to individuate signs in the texts as tokens of these types. Yet this may
lead to the production of meaningless sentences. It does not follow that he
abandons, immediately, the set of sign types that he has proposed. Instead,
he may decide to change the value of some sign types or change the identity
of certain signs. Both changes are made in order to produce sentences
which are not meaningless, although the latter type of change cannot
wholly ignore morphological considerations. Success in producing
meaningful sentences by making such changes allows the decipherer to
retain his original sign type identifications, failure obliges him to change
them. Thus, the process of identifying sign types and individuating sign
tokens is constantly checked against the production of sentences which are
not meaningless, and so it is in establishing the meaning of the sentences of
the text that the decipherer decides on the number and values of the sign
types, on the individuation of tokens and the identity of signs.
Decipherment of the script of a text cannot proceed prior to and
independent of establishing the meaning of the sentences of the text. Yet
this requirement might not leave the decipherer with a single decipher-
ment, as different identifications of sign types could produce different sets
of meaningful sentences, each set inconsistent with the others. Thus
further requirements are needed for the making of a successful
decipherment.
One further requirement is that the content of the sentences of the text
must be consistent with what is known (from accounts related to other data)
about the society of the inscriber of the text. Thus a decipherment which is
wholly successful in producing meaningful sentences but where one sign or
group of signs is made to stand for 'aeroplane', in a society which is known
(on the basis of statements related to other data) not to have possessed a
technology capable of constructing aeroplanes, cannot be considered a
successful decipherment. This does not mean that knowledge of a society
derived from texts can never be allowed to modify knowledge about that
society related to other (e.g. archaeological) data. The decipherment with
'aeroplane' would be retained were it possible to show that the presence of
aeroplanes would fit more successfully into a history describing that place
and period than would the elimination of aeroplanes from the account of the
period and the changing of the decipherment. So an additional requirement
for successful decipherment is that objects or occurrences referred to in the
text are consistent with the historical account describing changes at the
place and period of its inscription.
Not merely the content of the text's sentences but the script and language
 250 B. C. HURST:

of the sentences themselves, as revealed by the decipherment, must be

consistent with the historical account describing changes at that place and
period, including changes in the script and language in which the text is
inscribed and accounts of changes in other scripts and languages. For
example, an inscription discovered in South America and dated (by radio-
carbon methods) to the pre-Columban period, could not be considered to
be successfully deciphered-evern if the decipherment met all other
requirements-if the language produced by the decipherment were
modern English.
It is not only necessary that none of the sentences of the text should be
meaningless, but that there should be a relationship between the sentences
such that their inscribing can be shown to be the product of a specific
communicative act and that act fits into the historical account describing
changes at the period and place of the inscription. The decipherer needs to
be able to describe the group of sentences forming the text as together being
(a letter', or 'an inventory' or 'a prayer', etc.; that is to say that the
decipherer needs to be able to indicate the purpose of the inscriber in
grouping together these particular sentences and placing them in this
order. For doubt is thrown upon the success of a decipherment, even when
no sentence is meaningless, if the juxtaposition of sentences appears to be
entirely arbitrary and none bears any relation to the next. In such a case,
while no individual sentence is meaningless, the text as a whole can be
considered meaningless, and the description of the act of inscribing as a
communicative act fits into the historical account less satisfactorily than its
description in some other way (doodle, decoration, etc.), and so the
description of the object as a text cannot be maintained.
It is only on the basis of his knowledge of the social conditions in which
the inscription was produced that the decipherer can decide which class of
communicative act the making of the inscription is a member. His
specification of the class of act, however, will affect his identification of
signs, his assignation of values to sign types and their identification. It is not
the case that the decipherer can first identify sign types and signs, then
assign values to the sign types and find out the meaning of the sentences of
the text, and last discover the purpose intended for the text by the inscriber.
On the contrary, the results of all other processes may require modification
in order to conform to what suits the description of the communicative
act.
It might be thought that the present argument leads to the conclusion
that decipherment is indeterminate, for if one purpose is ascribed to the
inscriber one set of values for the sign types results, while by ascribing a
different purpose the result is a different set of values. However, as
mentioned, the description of the act of inscription in terms of the
inscriber's purpose for the inscription must fit into the historical account
describing the changes at the place and period of the act of inscribing. The
indeterminancy of decipherment is dependent upon the description of the
purpose of the inscriber ip producing the inscription being indeterminate,
and the indeterminancy of the latter is dependent on the historical account
of the changes at that place being indeterminate. However, there is reason
for not accepting that an historical account describing the whole past or any
 DECIPHERMENT, TRANSLATION, INTERPRETATION 251

part of the past is indeterminate, so it is not necessary to accept that the

decipherment of a text is indeterminate.1
Moreover, if decipherment were indeterminate, not only could it never
be said that one decipherment rather than another was successful, but it
could never be said that this was a text (that is capable of decipherment)
rather than a non-text (markings without communicative purpose), as the
indetermination would affect the whole interpretation of the;,past. Where
any number of decipherments are equally acceptable, the possibility that
what is being deciphered is not a text remains as one possible
'decipherment'/interpretation of the 'text'/object. To accept the in-
determinancy of decipherment would be to accept that anything or nothing
could be a text, for it is by successful decipherment that the hypothesis that
this object is a text is most fully corroborated.2
Decipherment can be more or less radical. The more radical a decipher-
ment, the less its success is dependent on knowledge of languages and
scripts of already deciphered texts and the more its success is dependent on
knowledge of non-linguistic and non-scriptual elements in the inscriber's
environment. The less radical a decipherment, the more its success is
dependent on the application to the text of knowledge of language and
scripts obtained from other scripts and the less its success depends on
knowledge of non-linguistic and non-scriptual matters. Radicalness,
however, is a matter of degree, the decipherment of no text being entirely
radical or entirely unradical. Even the most radical decipherment (for
example, decipherment of a pictographic or ideographic script developed
in total isolation from any other script) involves the recognition of the script
as a script on the basis of its possessing some features similar to already
deciphered scripts. The decipherer who encounters a script with no
features in any way similar to already deciphered scripts will simply fail to
recognise it as a script and so the issue of its decipherment will never be
raised. On the other hand, even the least radical decipherment (of a text in
the decipherer's own script and language) requires more than the
application of knowledge of script and language, for without a knowledge of
non-linguistic and non-scriptual matters texts might be dismissed as non-
communicative (not texts) as the purpose of the grouping of the sentences
might go uncomprehended: it is knowledge of the social conventions which
makes possible the identification of the purpose of, for example, the writer
of nonsense verse and so allows inscriptions of such verse to be texts, rather
than their merely being dismissed as decorations or doodles. Moreover,
were the decipherer to be dependent exclusively on knowledge from other
already deciphered texts, this would allow the chance order of decipher-
ment to dictate what could be considered correct decipherment. However,
there is no reason why an anomaly in the decipherment of even the most
recently deciphered text cannot be allowed to call in question the method of

I The way in which historians reject indeterminancy and choose between rival
accounts of the past is discussed in my article cited above.
2 Cf. H. Putnam 'The refutation of conventionalism', Nous viii, I974, on the
effect of the indeterminancy of translation on the underdetermining of the
social sciences.
252 B. C. HURST: DECIPHERMENT

decipherment used throughout all previous decipherments. If a change in

the method of decipherment removes the anomaly in the most recently
deciphered text without creating more anomalies in the decipherments of
other texts then such a change is to be accepted.
It might be claimed that where decipherment is least radical it is only
translation or interpretation. Thus it might be said that a text requires
decipherment only when the (would-be) reader cannot
(i) identify the sign types and individuate the sign tokens of the script
in which the text is written;
(2) assign a value or function to the sign types;
(3) identify the signs;
that a text requires translation when the reader can do (i), (2) and (3) but
cannot
(4) understand the meaning of the sentences;
that a text requires interpretation when the reader can do (i), (2), (3) and
(4), but cannot
(5) understand the purpose of the writer in his grouping of the
sentences of the text.
Yet, as the foregoing analysis has shown, it can never be stated that the
decipherment is successful until all the processes are completed, and so
these processes are best seen as merely stages in the overall process of
decipherment.
Thus decipherment, translation and interpretation can never be per-
formed independently of each other or of an account of the history of the
society in which the text was inscribed, as the act of inscribing and the
meaning of the inscription have to be consistent with what is known of the
inscriber and his environment. A change in the account of that history
(resulting, perhaps, from the discovery of non-textual data) may require a
change in decipherment as much as a new decipherment of a text may
require a change in the account of that history. It follows that if we accept
that a determinate account of the past is possible, then we must accept that
determinate decipherment, translation and interpretation are possible.'

DEPARTMENT OF EDUCATION,
BAYERO UNIVERSITY,
PMB 3011,
KANO,
NIGERIA

My thanks to Mr. Gordon Reddiford for comments on an earlier version of this

paper.