(3) Quine 1951, inadvertent re-definition of Kant's term "analytic"; result in ambiguity (or vagueness?)

 

In "Two Dogmas of Empiricism", Quine argues that the distinction between truth 'a priori' (truth by virtue of meaning or logic alone, independently of empirical fact) and truth 'a posteriori' (truth by virtue of empirical fact) does not exist--because it cannot be defined, due to the fact that the notion of 'meaning' cannot be defined.

 

In so doing, he makes use of terms which were originally introduced by Kant. Quine explicitly refers to Kant, and it is quite clear that he means to use the terms in the way Kant used them. However, his terminology reflects a mistake which, among students of Kant--particularly, among beginners in this field--is exceedingly widespread: he confuses the dichotomy "a priori"/"a posteriori" with the dichotomy "analytic"/"synthetic". That means, among other things, that he uses the term "analytic" as referring to that peculiarity of truth for which Kant uses the term "a priori". Here is Kant on "a priori":

Kant: "Erkenntnisse a priori" sind solche, die "von der Erfahrung, und selbst von allen Eindrücken der Sinne unabhängig" sind; Erkenntnisse "a posteriori" hingegen solche, die "empirisch" sind, also "ihre Quellen [...] in der Erfahrung haben". (Kant, KdRV 2(1987), ?[Einleitung]) 

Here is Quine on "analytic":

Quine: "truths which are analytic [are] grounded in meanings independently of matters of fact [while] truths which are synthetic [are] grounded in fact" (Quine 1951, 20).

And here is Kant on "analytic": 

Kant: "Analytische" Urteile sind solche, in denen "das Prädikat B [...] zum Subjekt A als etwas [gehört], was in diesem Begriffe A (in versteckter Weise) enthalten ist"; "synthetische" Urteile sind solche, in denen "B [...] ganz außer dem Begriff A [liegt]". (Kant, KdRV 2(1987), ?[Einleitung])

(According to Kant's exposition, "analytic" truth appears as one subtype of his "a priori" truth.)

 

An example of an "a priori" truth is: "If an object is a giraffe, it is not a crocodile". It is true by virtue of its meaning alone, even though being not a crocodile is not (hiddenly) contained in being a giraffe.

An example of an "analytic" proposition would be "Ophthalmologists are doctors". It is true, because 'x is a doctor' is (hiddenly) contained in 'x is an opthalmologist'. Another popular example is "Bachelors are unmarried"; it is true because 'x is unmarried' is (hiddenly) contained in 'x is a bachelor'.

 

Now, turning to Quine's definition of "analyticity": Clearly, this does not conform to Kant's definition of "analyticity". At the same time, it clearly does conform to Kant's definition of "a priori". Thus apparently, Quine commits the confusion to which so many (other) beginners in the study of Kant's work (including myself, as I must confess to my shame ...) succumb: the confusion of "analytic" with "a priori". 

 

Quine seems to have been confronted with the fact that his interpretation of Kant's "analytic" is faulty, and with how Kant really defines the term. Instead of correcting his terminological fault, he launches a counteroffensive, attacking both Kant and whoever made him aware of his mistake:

[...] 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.

While Quine represent himself as a great expert in Kant interpretation, he obvioiusly is not. Kant's usage of the term "analytic" accords perfectly with the definition he states.

 

(Quine's fault probably goes back (at least) to Frege (intermediated, perhaps, by Carnap), who also seems to confuse "analytic" with "a priori". However, it seems that the large distribution of the mistake is (more immediately) due to Quine's influence.)

(Wikipedia currently suggests that the logical positivists, and particularly Frege and Carnap deliberately "revised" the Kant's notion (rather than just confused Kant's dichotomies). However, as far as I see, the exposition contains no serious evidence for this view.)

 

Due at least in part to the large and strong influence of Quine's "Two Dogmas of Empiricism", the definition of "analytic" as 'true by virtue of meaning alone, independently of matters of fact' (i.e., as 'a priori') became largely established within analytic philosophy--even though this definition is clearly inadequate, given that and how Kant had introduced these terms. Since Kant's work is still available, and experts in Kant's work of course used the correct definition, the term "analytic" became ambiguous between the meaning Kant had defined and the meaning which Quine (along with numerous other modern philosophers) applied to that term. As a result, the notion of "analyticity" is nowadays completely dubious: some define it (adequately) like Kant, some define it like Quine (and Frege, and Carnap).

 

Here is a source which defines the term as Kant did:

analytic–synthetic distinction, the distinction, made famous by Kant, according to which an affirmative subject-predicate statement (proposition, judgment) is called analytic if the predicate concept is contained in the subject concept, and synthetic otherwise. (Cambridge Dictionary of Philosophy, 2nd ed. 1999)

 

Here is a source which prefers Quine's definition:

An “analytic” sentence [...] has historically been characterized as one whose truth depends upon the meanings of its constituent terms (and how they’re combined) alone, as opposed to a more usual “synthetic” sentence [...] whose truth depends also upon the facts about the world that the sentence represents [...]. (Rey, Georges, "The Analytic/Synthetic Distinction", The Stanford Encyclopedia of Philosophy (Winter 2020 Edition), Edward N. Zalta (ed.), forthcoming URL = <https://plato.stanford.edu/archives/win2020/entries/analytic-synthetic/>)

 

Ambiguity is what ordinarily happens when a term is re-defined. But in the case of Kant's "analytic", things got even worse. Confronted with (at least) a duality of definitions, contemporary authors often want to have the cake and eat it, trying to find a phrasing which 'unifies' both meanings, the original of Kant, and the modern one (which undertaking is logically quite odd). [EXAMPLE!]

"Ein Satz ist analytisch, wenn seine  Wahrheit oder Falschheit durch die Bedeutungen der in ihm enthaltenen Ausdrücke festgelegt ist. Um den Wahrheitswert eines analytischen Satzes wie z.B. ‚Alle Junggesellen sind unverheiratet“ einzusehen, genügt es, ihn zu verstehen. Ein Satz ist s., wenn seine Wahrheit oder Falschheit davon abhängt, was in der Welt der Fall ist. S. Sätze wie z.B. ‚Alle Menschen sind sterblich’ müssen wir wohl stets empirisch überprüfen. Wenige zeitgenössische Philosophen folgen Kant in der Annahme, es gäbe s. Wahrheiten a priori ( Transzendentalphilosophie). ([zu finden in; accessed through internet] Christian Nimtz: Analytisch und Synthetisch, in: Stefan Jordan/Christian Nimtz (Hgg.): Philosophie. Hundert Grundbegriffe, Stuttgart: Reclam 2009, 24–26.)

The term is defined according to Quine's notion; but it is then implicitly, but clearly, ascribed to Kant! The author obvious does not at all perceive that the two philosophers define completely different notions.