Our beliefs are states of a faculty that aims to represent the truth by means of them. This is a part of what makes these mental states be beliefs (as opposed to, say, desires). Now on evolutionary theories of teleology, this will have to be explained in something like this way: There is a faculty that we have that (a) in fact tends to hold true beliefs, and (b) that it tended to hold true beliefs conferred a survival advantage on our ancestors that led to us having it. This is what makes the faculty count as aimed at truth.
But if one has a deflationary view of truth, one doesn't get to say this. On a deflationary theory of truth, truth is wholly characterized by something like instances of the Tarski schema, such as:
- "Snow is white" is true if and only if snow is white.
- (b') a survival advantage was conferred on our ancestors by the fact that the faculty tended to hold a belief provided that: (i) it was the belief that snow is white, and snow is white, or (ii) it was the belief that snow is not white, and snow is not white, or (iii) it was the belief that tigers are dangerous, and tigers are dangerous, or (iv) it was the belief that tigers are not dangerous, and tigers are not dangerous, or (v) it was the belief that there is life on other planets, and there is life on other planets, or (vi) it was the belief that there is no life on other planets, and there is no life on other planets, or ....
The anti-deflationist does better. For she sees truth as a genuine property and can say (b) as is, if she takes truth to be a fundamental property, or can replace (b) by some natural analysis. But the deflationist does not see truth as a genuine property and hence cannot use (b), and thus cannot give an evolutionary account of the truth-directedness of our faculties.
So what? Well, I think it is undeniable that our doxastic faculties are truth-directed. It is, I think, particularly crucial for naturalists to be able to say this, since it will be needed for an account of intentionality. But a naturalist (in the contemporary, non-Aristotelian sense) can only give an evolutionary account of teleology. Thus, a naturalist cannot be a deflationist about truth.
But neither can a naturalist be an anti-deflationist about truth. For if a naturalist were an anti-deflationist about truth, she would have to hold that truth is to be analyzed in terms of the sorts of properties science speaks of. But science speaks only of first-order properties, and truth is not a first order property (here's one quick argument based on an idea by Leon Porter: if it were, then the liar paradox could be formulated in entirely first-order scientific terms, yielding a contradiction in the language of science).
So naturalism is in trouble.