Final papers should be handed in on or near the last class today. Graduate students can take longer than that, but should at least show me a draft or talk to me about their ideas before the end of the semester. In general, I'm happy to either look at drafts of work or meet to talk about papers any time.

A good paper will be a detailed examination of some specific topic we covered in the class. ("The nature of self-knowledge", e.g., does not count as a specific topic.) Here are some ideas of questions you could try to answer in your paper:

1. Does the phenomenon of self-verifying thoughts have anything much to do with self-knowledge? How far can Burgean basic self-knowledge get us?

2. Is Shoemaker's functionalist explanation of self-knowledge satisfying? Can it explain the (supposed) necessary truth that self-blindness is incompatible with possession of appropriate conceptual capacities?

3. Evaluate in detail one of Shoemaker's arguments that self-blindness with respect to a particular mental state is impossible. Does the argument show that it is in the nature of the mental state in question to be known?

4. Do we need a notion of narrow content to explain self-knowledge of our own intentional states? Discuss one issue with defining such a notion.

5. Do cases of `slow switching' show that externalism is incompatible with self-knowledge?

6. Is the externalist committed to the claim that we can know that water exists on the basis of introspection alone?

7. Evaluate in detail either Putnam's or Burge's argument for externalism. What exactly does the argument establish?

8. How does Peacocke think that we can come to know what we believe? Does he give a satisfying explanation of cases in which we seem to be able to obtain knowledge of a belief without first making a conscious judgement with the content of the belief in question?

9. Explain and respond to Wright's argument that Wittgenstein's rule-following considerations make any substantial epistemology of self-knowledge impossible.

10. How should a proponent of response-dependence analyses of concepts respond to Johnston's `missing explanation' argument?