The offending passage (from "Two Dogmas of Empiricism"):
Kant conceived of an analytic statement as one that attributes to its subject no more than is already conceptually contained in the subject. This formulation has two shortcomings: it limits itself to statements of subject-predicate form, and it appeals to a notion of containment which is left at a metaphorical level. But Kant's intent, evident more from the use he makes of the notion of analyticity than from his definition of it, can be restated thus: a statement is analytic when it is true by virtue of meanings and independently of fact.
Now, I know that this statement is false through and through, and so I wonder why Quine made it. Part of me also wanted to object, "Who does Mr.-Indeterminacy-of-Translation think he is, to say this?" but Quine apparently knew not just two, but a lot of languages, so I don't want to obsess over his indeterminacy-of-translation thesis (over his integrity-or-hypocrisy in advancing such a thesis...). But just so it would be known here clearly:
- Kant had a clear sense of the law of identity, and the law of noncontradiction, as duals (in the modern sense of phrases like "paraconsistent logic is dual to paracomplete logic"). So he not only doesn't leave the containment-talk at the metaphorical level, he never really stated it "metaphorically" in the first place. Now Quine seems to have been dimly aware of Kant's preliminary account of the matter, per the earliest sections of the first Critique. But so did he not know that Kant says:
There exists, however, a formula of this celebrated principle [of analytic truth]—a principle merely formal and entirely without content—which contains a synthesis that has been inadvertently and quite unnecessarily mixed up with it. It is this:—"It is impossible for a thing to be and not to be at the same time." Not to mention the superfluousness of the addition of the word impossible to indicate the apodeictic certainty, which ought to be self-evident from the proposition itself, the proposition is affected by the condition of time, and as it were says: "A thing = A, which is something = B, cannot at the same time be non-B." But both, B as well as non-B, may quite well exist in succession. For example, a man who is young cannot at the same time be old; but the same man can very well be at one time young, and at another not young, that is, old. Now the principle of contradiction as a merely logical proposition must not by any means limit its application merely to relations of time, and consequently a formula like the preceding is quite foreign to its true purpose. The misunderstanding arises in this way. We first of all separate a predicate of a thing from the conception of the thing, and afterwards connect with this predicate its opposite, and hence do not establish any contradiction with the subject, but only with its predicate, which has been conjoined with the subject synthetically,—a contradiction, moreover, which obtains only when the first and second predicate are affirmed in the same time. If I say: "A man who is ignorant is not learned," the condition "at the same time" must be added, for he who is at one time ignorant, may at another be learned. But if I say: "No ignorant man is a learned man," the proposition is analytical, because the characteristic ignorance is now a constituent part of the conception of the subject; and in this case the negative proposition is evident immediately from the proposition of contradiction, without the necessity of adding the condition "the same time." This is the reason why I have altered the formula of this principle—an alteration which shows very clearly the nature of an analytical proposition.
False or not, meaningless or not, it remains that the above is not a metaphor. It is not unclear or underdeveloped, contrary to a preceding remark of Quine's: "... we hear analytic statements defined as statements whose denials are self-contradictory. But this definition has small explanatory value; for the notion of self-contradictoriness, in the quite broad sense needed for this definition of analyticity, stands in exactly the same need of clarification as does the notion of analyticity itself." Well, Quine, you had the clarification right there in the text you were mentioning, so...
A definition is, as the term itself indicates, the representation, upon primary grounds, of the complete conception of a thing within its own limits. Accordingly, an empirical conception cannot be defined, it can only be explained. For, as there are in such a conception only a certain number of marks or signs, which denote a certain class of sensuous objects, we can never be sure that we do not cogitate under the word which indicates the same object, at one time a greater, at another a smaller number of signs. Thus, one person may cogitate in his conception of gold, in addition to its properties of weight, colour, malleability, that of resisting rust, while another person may be ignorant of this quality. We employ certain signs only so long as we require them for the sake of distinction; new observations abstract some and add new ones, so that an empirical conception never remains within permanent limits. It is, in fact, useless to define a conception of this kind. If, for example, we are speaking of water and its properties, we do not stop at what we actually think by the word water, but proceed to observation and experiment; and the word, with the few signs attached to it, is more properly a designation than a conception of the thing. A definition in this case would evidently be nothing more than a determination of the word. In the second place, no a priori conception, such as those of substance, cause, right, fitness, and so on, can be defined. For I can never be sure, that the clear representation of a given conception (which is given in a confused state) has been fully developed, until I know that the representation is adequate with its object. But, inasmuch as the conception, as it is presented to the mind, may contain a number of obscure representations, which we do not observe in our analysis, although we employ them in our application of the conception, I can never be sure that my analysis is complete, while examples may make this probable, although they can never demonstrate the fact. Instead of the word definition, I should rather employ the term exposition—a more modest expression, which the critic may accept without surrendering his doubts as to the completeness of the analysis of any such conception. As, therefore, neither empirical nor a priori conceptions are capable of definition, we have to see whether the only other kind of conceptions—arbitrary conceptions—can be subjected to this mental operation.
Now I also read within the last two or so hours that Quine didn't do much actual reading of previous philosophy. But isn't it hypocritical for such an empiricism-minded writer to not actually empirically check what Kant said about this issue? (It seems as if Quine made an even more drastic empirical mistake, with respect to a rather different topic; but so I wonder whether Quine was actually an empiricist or just pretending to be one?)
