I do realise that talk of natural kinds dates back to Aristotle, but I'd better not say too much about ancient philosophers lest I be convicted of practicing history of philosophy without a license.

I am certainly open to the idea that this might be used to explain other philosophical categories besides knowledge. I have some real sympathy with the work of those moral realists who have tried to give naturalistic accounts of human flourishing, and who offer accounts of right action in such terms. (I suppose this is more evidence that I really do have deep affinities with Aristotle!)

Bealer argues that the kind of naturalistic view which Quine holds will rob him of the ability to make the normative claims which (many) naturalists wish to make in epistemology. I don't think this is right about Quine, but I'm certain it's not right about my own view. To the extent that I can show that talk of knowledge is firmly rooted within empirical theories where it plays an important explanatory role, I thereby demonstrate its naturalistic credentials.

Internalist approaches to epistemology, I believe, have a great deal of intuitive appeal. Internalists believe that the features in virtue of which a belief is justified must somehow be internal to the agent. On some views, this amounts to the claim that these features must be accessible to introspection and armchair reflection. On others, it amounts only to the claim that they must be mental features.

Talk of belief in these animals is not some kind of anthropomorphism. We simply cannot explain the kinds of problem solving and behavioral sophistication some species exhibit without supposing that they have genuine beliefs. But once these ethologists finish making the case for animal belief, they quickly move to talk of animal knowledge as well. What I argue is that this is not a mere façon de parler.

So I do, of course, reject much that is central not only to the psychology of Descartes and Kant, but to their epistemology as well. No doubt, the best available theories of today will look primitive in comparison with what we are in a position to understand hundreds of years from now.

I do agree with Stich that a quick move from our evolutionary origins to the reliability of our cognitive mechanisms is not legitimate. As I see it, the case for the reliability or unreliability of various cognitive mechanisms lies elsewhere.