The outline of the argument against the first dogma

The first dogma Quine examines is the view that there is a ``fundamental cleavage between truths which are analytic ...and truths which are synthetic.'' His argument follows (more or less) this outline.

1. Quine begins by trying to define `analyticity', but this immediately leads him to try identify a group of concepts which includes, `self-contradiction', `meaning' and `synonymy,' all of which seem implicated in understanding analyticity. 20-1/63-4
2. The attempt to define analyticity in terms of self- contradiction fails because, as he puts it they ``are two sides of a single dubious coin.''
3. In attempting to understand `meaning' Quine comes to the conclusion that while it is essential to understanding analyticity, the concept of meaning is itself unclear. One thing he knows is that for Kantian analyticity, `meaning' does not mean `extension/reference,' but must include `intension/sense.' 21-2/64-5
4. However, since neither intension nor extension can account for analyticity, Quine turns to synonymy. He hopes that this will help account for both the classes of analytic statements: logical truths and what I dubbed `definitional
   truths.' Of course, we learn quickly that synonymy itself is ``no less in need of clarification than analyticity itself.'' 22-3/65
5. He considers Carnap's notion of state-descriptions, but rejects them as inadequate to the task. At best they only account for logical truths, if they even do that. 23-4/65
6. Definition is the next stop in trying to understand synonymy. Quine identifies three kinds of definition (lexical, explicative and stipulative), only the last doesn't itself rest upon a prior understanding of synonymy, but it isn't to the point of synonymy as needed for analyticity--which we learn on p. 28 is cognitive synonymy. 24-7/66-8
7. Synonymy has often been defined in terms of substitution salva veritate, but Quine argues that this rests upon an undefined notion of necessity. 27-8/68-9
8. Necessity turns out to rest either upon analyticity or cognitive synonymy. 28-30/69
9. Quine argues that we can only use substitutivity salva veritate with any reasonable success in extensional languages, but here there is no guarantee of cognitive synonymy. 30-1/69-70
10. At this point Quine abandons this line of argument, and turns to analyzing analyticity directly. 31-2/70-1
11. He considers a notion of semantical rules intended to allow us to specify within a language the category of `analytic.' He tries importing these definitions from formal languages, where they seem clear and unambiguous, but finds that in fact at the end of the day we end up relying on the natural language notions of analyticity, synonymy, meaning, self-contradiction, necessity--the intensional terms. 32-7/71-4