... xxx ...

Quine's (1951) inductive argument against 'a priori truth' is strongly inexhaustive, hence perfectly unvconvincing

In his celebrated article "Two Dogmas of Empiricism", W.V.O. Quine pursues two aims (which, he believes, are interrelated).

(1) He wants to demonstrate that the distinction between 'a priori' truth (truth by virtue of meaning alone, as opposed to truth by contingent fact) and 'a posteriori' truth (truth by contingent fact) does not exist.

(2) He denies that theories are reducible to sense data reports.   

We are here concerned with the former task.

To avoid confusion, we must first of all make ourselves aware of a terminological peculiarity in Quine's text. Although Quine uses terms which go back to Immanual Kant, and although Quine explicitly refers to Kant in introducing the terms, his usage of these terms does no conform with Kant's. In particular, Quine uses the term "analytic" to refer to what in Kant's writings is referred to by another term, namely, "a priori truth", which is truth by virtue of meaning alone, as opposed to truth by virtue of matters of contingent fact. So although Quine consistently speaks about "analytic" truth, and is consistently interpreted as dealing with "analyticity", his real subject matter is "a priori truth". (Click here for more details.) In order to minimise confusion as far as possible, we shall now turn over to Quine's usage and speak of "'analytic' truth", thereby referring to truth by virtue of meaning alone.

Quine's argument against 'analytic' truth is based on an inductive argument. The problem with this argument is that when an argument works by means of induction, then in order for this argument to be convincing, the induction must be exhaustive. In the present case, Quine argues that the distinction between "analytic" and "synthetic" truth is illusory, because the notion of "analyticity" (truth by virtue of meaning alone) cannot be defined; and that the notion of "meaning" cannot be defined because of all the particular proposals for a definition of "meaning", each is defective; which, again, Quine starts showing taking single proposals one by one. This argument will be convincing only if and insofar as really all proposals for a definition of "meaning" are taken into consideration, and shown to be defective.

inductive arguments must be executed that this inductive  whose execution is strongly inexhaustive.

The description of analyticity as truth by virtue of meanings startes us off in pursuit of a concept of meaning. But now we have abandoned the thought of any special realm of entities called meanings. (Quine 1951, 23)

Clearly, the argument Quine adumbrates here (a bit vaguely and ambiguously, but perhaps clearly enough when the context is taken into consideration) will be convincing only if he means that the concept of meaning can not (not at all) be defined. This, it is insinuated ('We have abandoned'), he had demonstrated before. What he did before was: reject single attempts to define 'meaning'. The structure of the argument employed is obviously inductive: The inability of definition of 'meaning' is supposed to be demonstrated by means of the rejection of the single proposals, respectively. To be convincing, the induction must of course be exhaustive: each proposal of a meaning theory must be enter the examination. Now the amazing fact is: Although at Quine's time there was a high two-digit number of proposals on the market (cf. Ogden & Richards 1923), Quine considers only two--Frege's faulty equation of 'meaning' with extension, and (by far too swiftly) the equation of 'meaning' with 'ideas'. Now, these two proposals obviously miss the mark of exhaustivity dramatically. Quine's insinuation that he had 'abandoned' the idea that meanings exist reflects a perfect lack of understanding what it would mean to show, as Quine full-bodied declares he has done, that the concept of meaning is inexplicable. Apparently, Quine either overlooks the fact that his argument is inductive, or else he fails to understand what it takes to render an inductive argument convincing.