This responds to Quine's "Notes on Existence and Necessity" and to Quine's #97, including extended discussion of terminology. Contains response to #97 on analyticity.

Response to #100.

continuation of #105.

Quine has just joined the Navy and is working in Washington D.C. This is Quine's first letter to Carnap after receiving from Carnap his "Introduction to Semantics", it has been 18 months since they have spoken together.

"much of the theory is decidedly to my liking, despite dissention on certain points. However, I do feel that the points where I dissent are peculiarly crucial to semantics"

Quine mentions discussions in 1941 on "the program of finitistic constitution system" (which is an epistemological desiderata) but considers that the semantic questions can be considered independently of this (and of epistemology generally).

Quine mentions two ways in which epistemological issues entered into semantics (in that previous discussion):

• (a) Quine and Tarski questioned the precise nature of "your" (i.e. Carnap's) distinction between analytic and synthetic and in the course of such discussion it began to appear increasingly that the distinguishing feature of analyticity, for you, was its epistemological immediacy in some sense. Then we urged ... some sort of finitistic logic. This seemed odd to me, and the first time I have seen it suggested that analyticity has anything to do with "epistemological immediacy". Carnap rejects this in his response (#100)
• (b) I argued, supported by Tarski, that there remains a kernal of technical meaning in the old controversy about reality or unreality of universals.

(b) is taken to suggest consideration of a finitistic constitution system.

He then considers (in reverse order) how the problems of semantics can be progressed independently of these two epistemological issues.

• (b) just accept, probisionally, the "rudimentary Platonism" embedded in rudimentary logic and classical mathematics I don't myself, find any Platonism whatever necessary in the interpretation of classical mathematics, and would be surprised if Carnap did
• (a) Quine finds a provisional acceptance of the term analytic less acceptable, "in accepting the term `analytic' we take on an unexplained notion to which we were not committed hitherto" If "analytic" were defined in terms of semantics, and semantics defines independently of the term "analytic" then there would be no problem. As it happens, Carnap uses the term analytic in describing how to formulate a semantics, and this may seem superficially less satisfactory, however it is obvious how these can be disentangled. Quine seems intent on making an issue out of analyticity.

There follows a long passage explaining at greater length why Quine has a problem with analyticity here (pp. 296-299). At the close of this he mentions his paper "Notes on Existence and Necessity" (not by name).

On analyticity

"To your letter of January 5th, p.1,(a). I do by no means regard epistemological immediacy as the characteristic of `analytic' (in pragmatics)(I emphasised this against Schlick, I do not remember where), but rather that here the truth is independent of the contingency of facts (this of course, should be made more precise)."

... of your article and letters. ...

"(4) In re penult. paragraph of p. 5 of your letter of Jan. 21: My error. As to your phrase `independent of contingency of facts', though, this is a phrase I cannot better clarify to myself intuitively than by explaining it as meaning `analytic', so it doesn't help."