On the Reasons for Indeterminacy of Translation
Author(s): W. V. Quine
Source: The Journal of Philosophy , Mar. 26, 1970, Vol. 67, No. 6 (Mar. 26, 1970), pp.
178-183
Published by: Journal of Philosophy, Inc.
Stable URL: https://www.jstor.org/stable/2023887
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 [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
is collaborating with JSTOR to digitize, preserve and extend access to The Journal of
Philosophy
This content downloaded from
70.81.118.70 on Sun, 29 Jan 2023 21:55:34 UTC
All use subject to https://about.jstor.org/terms
178 THE JOURNAL OF PHILOSOPHY
sions. But more important, this raises a question about the connec-
tion between indeterminacy of translation and of theory; for, what
would such a revision have to do with translation? The stimulus
meaning of "Rabbit" on my theory is not identical with that of
"Rabbit" on a theory that denies them. In terms of the conceptual
apparatus provided by Quine, we must say that I can make this
revision only at the cost of changing the stimulus meaning of "Rab-
bit"; but, in the Quinean framework, it would seem that if that is
permitted, then anything might be revised into anything. If stimu-
lus meaning plays no role as parameter in the equivalence of the-
ories, then what is the correct formulation of the indeterminacy
thesis?
This would continue to hold even if we imagine that some for-
eigner's theory might be translated by either "There are rabbits but
no rabbit-stages" or "There are rabbit-stages but no rabbits." Nei-
ther coincides with my own theory, but it is now proposed that I
adopt one of them. It is not in general true that whenever a for-
eigner's theory may receive either of two translations, I may adopt
either version of his theory. He may be quite mistaken in his views
through lack of experience. If, in the specific case at hand, I ma)'
adopt the other's theory, what makes this possible?
I have tried in this essay to expose some of the puzzles and ambigu-
ities that infect current discussions of indeterminacy. I do not wish
to claim that indeterminacy theses are inevitably or hopelessly ob-
scure; I would welcome a clarified version that would dispel these
confusions.
Georgetown University B. M. HUMPHRIES
ON THE REASONS FOR INDETERMINACY
OF TRANSLATION *
M t or Y gavagai example has figured too centrally in discussions
of the indeterminacy of translation. Readers see the ex-
ample as the ground of the doctrine, and hope by resolv-
ing the example to cast doubt on the doctrine. The real ground of
the doctrine is very different, broader and deeper.
Let us put translation aside for a while and think about physi-
cal theory. Naturally it is underdetermined by past evidence; a
* I am indebted to Burton Dreben for helpful criticism of an earlier draf
this paper.
This content downloaded from
70.81.118.70 on Sun, 29 Jan 2023 21:55:34 UTC
All use subject to https://about.jstor.org/terms
ON THE REASONS FOR INDETERMINACY OF TRANSLATION 179
future observation can conflict with it. Naturally it is underdeter-
mined by past and future evidence combined, since some observable
event that conflicts with it can happen to go unobserved. Moreover
many people will agree, far beyond all this, that physical theory is
underdetermined even by all possible observations. Not to make a
mystery of this mode of possibility, what I mean is the following.
Consider all the observation sentences of the language: all the oc-
casion sentences that are suited for use in reporting observable
events in the external world.' Apply dates and positions to them in
all combinations, without regard to whether observers were at the
place and time. Some of these place-timed sentences will be true
and the others false, by virtue simply of the observable though un-
observed past and future events in the world. Now my point about
physical theory is that physical theory is underdetermined even by
all these truths. Theory can still vary though all possible observa-
tions be fixed. Physical theories can be at odds with each other and
yet compatible with all possible data even in the broadest sense.
In a word, they can be logically incompatible and empirically equiv-
alent. This is a point on which I expect wide agreement, if only
because the observational criteria of theoretical terms are commonly
so flexible and fragmentary. People who agree on this general point
need not agree as to how much of physical theory is empirically
unfixed in this strong sense; some will acknowledge such slack only
in the highest and most speculative reaches of physical theory, while
others see it as extending even to common-sense traits of macroscopic
bodies.
Now let us turn to the radical translation of a radically foreign
physicist's theory. As always in radical translation, the starting point
is the equating of observation sentences of the two languages by an
inductive equating of stimulus meanings. In order afterward to con-
strue the foreigner's theoretical sentences we have to project ana-
lytical hypotheses, whose ultimate justification is substantially just
that the implied observation sentences match up. But now the same
old empirical slack, the old indeterminacy between physical theories,
recurs in second intension. Insofar as the truth of a physical theory
is underdetermined by observables, the translation of the foreigner's
physical theory is underdetermined by translation of his observation
sentences. If our physical theory can vary though all possible obser-
vations be fixed, then our translation of his physical theory can
1 The concept of observation sentence that I developed in ?10 of Word an
Object gains perhaps some further clarification in pp. 85-89 of Ontological Rela-
tivity and Other Essays (New York: Columbia University Press, 1969).
This content downloaded from
70.81.118.70 on Sun, 29 Jan 2023 21:55:34 UTC
All use subject to https://about.jstor.org/terms
I 8o THE JOURNAL OF PHILOSOPHY
vary though our translations of all possible observation reports on
his part be fixed. Our translation of his observation sentences no
more fixes our translation of his physical theory than our own pos-
sible observations fix our own physical theory.
The indeterminacy of translation is not just an instance of the
empirically underdetermined character of physics. The point is not
just that linguistics, being a part of behavioral science and hence
ultimately of physics, shares the empirically underdetermined char-
acter of physics. On the contrary, the indeterminacy of translation
is additional. Where physical theories A and B are both compatible
with all possible data, we might adopt A for ourselves and still re-
main free to translate the foreigner either as believing A or as be-
lieving B.
Such choice between A and B in translation could be guided by
simplicity. By imputing B to the foreigner we might come out with
shorter and more direct translations, and with less in the way of
elaborate contextual paraphrases, than by imputing A to him. That
is one possibility. A second possibility is that both choices, A and B,
require forbiddingly circuitous and cumbersome translation rules.
In this case we might regard the foreigner as holding neither A nor
B; we might attribute to him rather some false physical theory
which we can refute, or some obscure one which we despair of pene-
trating, or we might even regard him as holding no coherent physi-
cal theory at all. But we can imagine also, third, the possibility that
A and B are both reasonably attributable. It might turn out that
with just moderate circuitousness of translation at certain points
-different points-A and B could be imputed about equally well.
In this event no basis for a choice can be gained by exposing the
foreigner to new physical data and noting his verbal response, since
the theories A and B fit all possible observations equally well. No
basis can be gained by interrogation in a theoretical vein, since
the interrogation would take place in the foreigner's language and
so could itself be interpreted according to either plan. In this event
our choice would be determined simply by the accident of hitting
upon one of the two systems of translation first.
The metaphor of the black box, often so useful, can be mislead-
ing here. The problem is not one of hidden facts, such as might be
uncovered by learning more about the brain physiology of thought
processes. To expect a distinctive physical mechanism behind every
genuinely distinct mental state is one thing; to expect a distinctive
mechanism for every purported distinction that can be phrased in
traditional mentalistic language is another. The question whether,
This content downloaded from
70.81.118.70 on Sun, 29 Jan 2023 21:55:34 UTC
All use subject to https://about.jstor.org/terms
ON THE REASONS FOR INDETERMINACY OF TRANSLATION i8i
in the situation last described, the foreigner really believes A or be-
lieves rather B, is a question whose very significance I would put in
doubt. This is what I am getting at in arguing the indeterminacy
of translation.
My argument in these pages has been and will remain directed
to you who already agree that there can be logically incompatible
and empirically equivalent physical theories A and B. What degree
of indeterminacy of translation you must then recognize, granted
the force of my argument, will depend on the amount of empirical
slack that you are willing to acknowledge in physics. If you were
one of those who saw physics as empirically underdetermined only
in its highest theoretical reaches, then by the argument at hand I
can claim your concurrence in the indeterminacy of translation only
of highly theoretical physics. For my own part, I think the empirical
slack in physics extends to ordinary traits of ordinary bodies and
hence that the indeterminacy of translation likewise affects that
level of discourse. But it is important, for those who would not go
so far, to note the graduation of liabilities.
Gavagai, whose troubles I shall now review, lay at an extreme of
the scale. It was an observation sentence. Its stimulus meaning was
inductively well established, we supposed, coinciding with that of
'Rabbit'.2 Where indeterminacy threatened was in trying to settle
upon the divided reference of gavagai as a term: whether rabbits
or rabbit stages or undetached rabbit parts. Readers have responded
with suggestions of how, with help of screens or other devices, we
might hope to give the native informant an inkling of the desired
distinctions and so settle the reference.
Ingenuity in this vein proves unrewarding because of vagueness
of purpose. The purpose cannot be to drive a wedge between stimu-
lus meanings of observation sentences, thereby linking Gavagai
rather to 'Rabbit' than to 'Rabbit stage' or 'Undetached rabbit
part'; for the stimulus meanings of all these sentences are incon-
testably identical. They comprise the stimulations that would make
people think a rabbit was present. The purpose can only be to
settle what gavagai denotes for the native as a term. But the whole
notion of terms and their denotation is bound up with our own
grammatical analysis of the sentences of our own language. It can
2 Strictly speaking, even this induction presupposes something like an ana-
lytical hypothesis in a small way: the decision as to what to take as signs of
assent and dissent. See Word and Object, p. 30; also D. Davidson and J. Hin-
tikka, eds., Words and Objections (Dordrecht: Reidel, 1968), pp. 312, 317, or
Synthe'se, XIX, 1/2 (December 1968): 284, 289.
This content downloaded from
70.81.118.70 on Sun, 29 Jan 2023 21:55:34 UTC
All use subject to https://about.jstor.org/terms
i82 THE JOURNAL OF PHILOSOPHY
be projected on the native language only as we settle what to count
in the native language as analogues of our pronouns, identity, plu-
rals, and related apparatus; and I urged in Word acid Object that
there would be some freedom of choice on this score. Once such
choices are settled, on the other hand, however arbitrarily, the ques-
tion whether the gavagai are rabbits or stages or parts can be set-
tled too, by interrogation.
The most to hope for from the screens and kindred aids, then,
is an indirect hint as to which of various analytical hypotheses re-
garding pronouns, identity, plurals, etc. might in the end work out
most naturally. When this kind of hint is available, should we say
that the supposed multiplicity of choices was not in fact open after
all? Or should we say that the choice is open but that we have
found a practical consideration that will help us in choosing? The
issue is palpably unreal, and the doctrine of the indeterminacy of
translation depends in no way upon it.
The gavagai example was at best an example only of the inscru-
tability of terms, not of the indeterminacy of translation of sen-
tences. As sentence, Gavagai had a translation that was unique to
within stimulus synonymy; for the occasion sentences 'Rabbit',
'Rabbit stage', and 'Undetached rabbit part' are stimulus-synony-
mous and holophrastically interchangeable. The gavagai example
had only this indirect bearing on indeterminacy of translation of
sentences: one could imagine with some plausibility that some
lengthy nonobservational sentences containing gavagai could be
found which would go into English in materially different ways
according as gavagai was equated with one or another of the terms
'rabbit', 'rabbit stage', etc. This whole effort was aimed not at proof
but at helping the reader to reconcile the indeterminacy of transla-
tion imaginatively with the concrete reality of radical translation.
The argument for the indeterminacy is another thing, as seen ear-
lier in this paper.
Over the inscrutability of terms itself there is little room for de-
bate. A clear example from real life was seen in connection with
the Japanese classifiers.3 This example makes it pretty clear, more-
over, that the inscrutability of terms need not always bring indeter-
minacy of sentence translation in its train, however the case may
be in particular with gavagai. Again the questions raised by de-
ferred ostension (op. cit.), e.g., as between expressions and their
Gddel numbers, are strictly a matter of inscrutability of terms.
3 Ontological Relativity and Other Essays, pp. 35ff. Also in this JOURNAL, LXV,
7 (April 4, 1968): 191ff.
This content downloaded from
70.81.118.70 on Sun, 29 Jan 2023 21:55:34 UTC
All use subject to https://about.jstor.org/terms
A NONESSENTIAL PROPERTY I 83
This, not the indeterminacy of translation, is the substance of onto-
logical relativity.
There are two ways of pressing the doctrine of indeterminacy of
translation to maximize its scope. I can press from above and press
from below, playing both ends against the middle. At the upper
end there is the argument, early in the present paper, which is
meant to persuade anyone to recognize the indeterminacy of trans-
lation of such portions of natural science as he is willing to regard
as underdetermined by all possible observations. If I can get peo-
ple to see this empirical slack as affecting not just highly theoretical
physics but fairly common-sense talk of bodies, then I can get them
to concede indeterminacy of translation of fairly common-sense talk
of bodies. This I call pressing from above.
By pressing from below I mean pressing whatever arguments for
indeterminacy of translation can be based on the inscrutability of
terms. I suppose Harman's example 4 regarding natural numbers
comes under this head, theoretical though it is. It is that the sen-
tence '3 e 5' goes into a true sentence of set theory under von Neu-
mann's way of construing natural numbers, but goes into a false one
under Zermelo's way. But a limitation of this example, as Harman
recognizes, is that '3 e 5' rates as nonsense apart from set-theoretic
explications of natural number.
In these pages I prefer not to speculate on how much better one
might do from below, or from above either. My purpose here is to
separate the issues and identify the arguments; and this may be
managed most effectively by leaving the reader to consider what
more might be proved.
Harvard University W. V. QUINE
A NONESSENTIAL PROPERTY *
JS the number of planets necessarily a composite (nonprime)
number? We are sometimes told that the answer depends on
whether de dicto or de re necessity is in question.' If de dicto
necessity is at issue, the answer is supposed to be "no" since it is not
a necessary truth that the number of planets is a composite number.
4 Gilbert Harman, "An Introduction to Translation and Meaning," in Words
and Objections, p. 14. Also in Synthese. ibid.. p. 14.
* Supported in part by the National Science Foundation.
I E.g., Alvin Plantinga, "De Re et De Dicto," Nou's, iII, 3 (September 1969):
235-258.
This content downloaded from
70.81.118.70 on Sun, 29 Jan 2023 21:55:34 UTC
All use subject to https://about.jstor.org/terms
