Putnam on Synonymity and Belief1

By Wilfrid Sellars
University of Minnesota


Published in Analysis 15 (1955): 117-120.

In his recent note on "Synonymity and the Analysis of Belief Sentences"2 -- Hilary Putnam, following Benson Mates,3 develops the following argument:
  1. Suppose we use "Greek" and "Hellene" as synonymous.
  2. Then, on any current theory of synonymity,
    (D)
    All Greeks are Greeks
    (D')
    All Greeks are Hellenes
    are synonymous,
  3. as are
    (S)
    Whoever believes that all Greeks are Greeks believes that all Greeks are Greeks
    (S')
    Whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes.
  4. Now, it is "quite likely" that nobody doubts that whoever believes that all Greeks are Greeks believes that all Greeks are Greeks.
  5. But it is "quite easy" to suppose that somebody does doubt that whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes, though Putnam, for one, does not doubt it.
  6. Accordingly
    (e)
    Nobody doubts that whoever believes that all Greeks are Greeks believes that all Greeks are Greeks
    and
    (f)
    Nobody doubts that whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes
    "may quite conceivably have opposite truth values and so cannot be synonymous."4
  7. But on current theories of synonymity the synonymity of (e) and (f) follows from that of "Greek" and "Hellene".
  8. Consequently, given these current theories and additional premises which seem beyond dispute, the assumption that "Greek" and "Hellene" are synonymous leads to the conclusion that they are not synonymous, -- and similarly in the case of any other pair of different terms.
  9. Thus, in terms of current theories of synonymity "there is but one conclusion to which we can come: 'Greek' and 'Hellenes' are not synonymous, and by the same argument neither are any two different terms". 5

To remedy this situation, Putnam proposes a new theory of synonymity according to which the synonymity of (e) and (f) does not follow from that of "Hellene" and "Greek". For whereas, on his theory, the synonymity of "Hellenes" and "Greeks" entails the synonymity of
    (D)
    All Greeks are Greeks
and
    (D'')
    All Hellenes are Hellenes
it does not entail that of
    (D)
    All Greeks are Greeks
and
    (D')
    All Greeks are Hellenes.

Now the truth of the matter is that the argument summarized above rests on a simple mistake. And if Putnam had asked himself why he does not doubt that whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes , he would undoubtedly have discovered his mistake, and saved himself a considerable expenditure of ingenuity.

The key to the puzzle is the initial stipulation: "Suppose we use 'Hellene' . . . as a synonym for 'Greek'."6 All right, suppose we do -- Putnam, myself and the rest of us. It follows, of course, that
  1. George is a Greek
and
  1. George is a Hellene
asserted by us , necessarily have the same truth value. It follows equally, though it may take a moment to appreciate the fact, that
  1. Jones believes that George is a Greek.
and
  1. Jones believes that George is a Hellene
asserted by us , necessarily have the same truth value. That is to say, they necessarily have the same truth value if we are making a simple use of "Greek" in the one case, and "Hellene" in the other, to formulate what it is that we take Jones to believe. On this assumption, the truth of (3) does not presuppose that Jones uses the word "Greek", nor the truth of (4) that he uses the word "Hellene," nor the joint truth of (3) and (4) that he has two words by which to refer to the inhabitants of Greece. Consider now
  1. Jones believes that all Greeks are Greeks
and
  1. Jones believes that all Greeks are Hellenes.
Here we must be cautious. We must remind ourselves that sentence (6), even as a sentence in our language may well have more than one employment. Thus, when we use (6) to make a true assertion, who is using the words "Greeks" and "Hellenes"? Are we making a straightforward (a pure using use) use of these words to formulate what Jones believes, as we would be making a straightforward use of "disease" and "curable" if we said
  1. Jones believes that all diseases are curable.
Or are we making a disguised mention (a covert mentioning use) of the words "Greek" and "Hellene" as used by Jones , so that (6) is equivalent, in effect, to
    (6')
    The sentence "all Greeks are Hellenes" as used by Jones expresses something that he believes.
Clearly it is only on the former supposition that the question "Does the synonymity of (5) and (6) as sentences in our language follow from the synonymity, in our language, of 'Greek' and 'Hellene'?" is a relevant question to ask.

It can, indeed, be doubted that anyone would ever use (6) without intending to refer to Jones' use of "Greek" and "Hellene". But what can be said is that if (6) is used as on the former supposition, -- if, that is, our sole purpose in using (6) rather than (5) is the stylistic one of not wishing to use the same word twice -- then any grounds we might have for asserting (5) would equally be grounds for asserting (6) and vice versa. Exactly the same considerations recur in the case of
  1. Whoever believes that all Greeks are Greeks believes that all Greeks are Greeks.
and
  1. Whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes.

Putnam tells us that while he does not doubt that whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes, "It is easy to suppose that someone does doubt this." But it should now be clear that it is easy to suppose this only if one is using (9), and in particular the phrase "believes that all Greeks are Hellenes", in such a way that it contains a covert mention of the words "Greek" and "Hellene". 7 For if we make what we have called a 'pure using use' of
  1. X doubts that whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes
it formulates (for us) exactly the same proposition as
  1. X doubts that whoever believes that all Greeks are Greeks believes that all Greeks are Greeks.
and it is not easy to suppose that there is a value of 'x ' for which (10) obtains.

Yet it can be supposed. Thus, even so interpreted as to be relevant to the problem,
    (f)
    Nobody doubts that whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes.
may be false. On the other hand,
    (e)
    Nobody doubts that whoever believes that all Greeks are Greeks believes that all Greeks are Greeks
may be true. But it would be a howler to infer that the combination (e)-true-and-(f)-false could obtain. For given the initial premise of the discussion (the synonymity in our usage of "Greek" and "Hellene") and the relevant interpretation of belief (and doubt) sentences, this would be equivalent to the overtly self-contradictory combination (e)- true-and-(e)-false.

Notes

1 This paper was already submitted to Analysis before I had an opportunity to see Prof. Alonzo Church's "Intensional Isomorphism and Identity of Belief", (Philosophical Studies, 5, 1954) which reaches the same (or similar conclusions by a slightly different route.

2 Analysis, April, 1954, pp. 114-122.

3 "Synonymity", Univ. of California Publications in Philosophy, 25, 1950, pp. 201-226; reprinted in Semantics and the Philosophy of Language, ed. by L. Linsky, Univ. of Illinois Press, 1952.

4 loc. cit. p. 117.

5 p. 117.

6 p. 117 (my italics).

7 Putnam races to meet this confusion. Thus, what he actually writes is:

Now I do not doubt that 'whoever believes that all Greeks are Greeks believes that all Greeks are Hellenes' is true; but it is easy to suppose that someone does doubt this. . . .

By using the semantical form " '. . .' is true" Putnam mentions the relevant expressions and by failing to stipulate that the 'someone' in question uses the words "Greek" and "Hellene" as we do Putnam indeed makes it easy to suppose that this 'someone' is doubtful of what he ('someone') uses (9) to express.


Transcribed into hypertext by Andrew Chrucky, Sept. 25, 1997.