How innocent is deflationism? (Q5931299)

The question of the title refers to whether a deflationist theory of truth is conservative over a non-semantical theory of the world, whether the theory of truth has any substantial, non-semantical consequences. If it does, it is not innocent. To address this question one must first be precise about what a deflationist theory of truth is; this occupies Section 2, which follows basic discussion of the problem and explication of the notion of conservativeness. One must also be precise about what the theory is supposed to be conservative over. Must it, for example, be conservative over pure logic, or might it only be conservative over a base theory such as PA of Peano arithmetic? This is the discussion of Section 3. The author concludes that a deflationist theory cannot be conservative over logic itself. Whether it can be conservative over PA depends on the precise form of the theory, but the best candidates cannot. Nevertheless, the author also maintains that this lack of innocence is not a great problem for the deflationist, for the deflationist need not be committed to conservativeness in either of these senses.