I’ve gotten so used to pessimism about the quality of APA presentations that I’m perhaps too selective about what I’ll make the effort to go hear, and I was pleasantly surprised to be reminded of this point by some excellent sessions this past week. Three sessions in particular stood out in my mind as particularly worth attending and thinking about after the session was over. Remarks below the fold about sessions by Coffman, Salerno, and Leite.
One very fun session was EJ Coffman’s talk, with Wayne Riggs as commentator, concerning the relationship between luck and control, and the connection between these and freedom. I had read earlier versions of EJ’s paper, but the session and the discussion were first-rate. One of EJ’s endearing traits is solicitious receptiveness to audience perspectives, which sometimes encourages questioners to lose contact with reality concerning when an objection has been decisively answered, but most of the discussion was in terms of really good questions and attempted counterexamples.
Second, Joe Salerno’s new paradox of happiness was really interesting. As occurs usually in the knowability paradox, he endorsed a principle about happiness distributing across conjunction (in knowability, the claim is that if you know a conjunction, you know each of its conjuncts; with happiness, it is that if you are happy that a conjunction is true, you are happy that each conjunct is true). That distributivity claim for happiness is false, I think: it is in conflict with the logic of defeat. If you are happy that you were comforted by your dissertation director while you were anxious about your upcoming defense, it doesn’t follow that you were happy that you were anxious. But the principle is dispensable for purposes of the paradox. If happiness is factive (if you are happy that p, then p), and any truth is part of a conjunction regarding which someone can be happy, there will be problems if there are some truths about which no one is (conjunctively) happy (this point mirrors the Williamson result that the knowability paradox can be generated without distribution, in terms of what is conjunctively known).
Finally, Adam Leite’s session was superb, and the discussion with the audience first-rate. The question that remained for me after the discussion was the following principle that Adam endorsed: if p is false, then it isn’t evidence for anyone. I wondered whether he meant to derive this result by endorsing Williamson’s idea that your evidence is just what you know, but he said “no”. I’m inclined to think that things can be evidence for one even if false if evidentialism is true, since one can reason explicitly to a conclusion from premises derived directly from experience. For example, if my experience is that of seeing a colleague at an APA meeting, I might reason to conclusion that I’ll be able to find them at the book display at a certain time using past experience plus the belief just formed. Presumably, this book display belief can be justified in this way, and will be justified independently of whether the belief that my colleague is at the APA is true. To the extent that we want to explain rationality or justification of belief in terms of evidence, I think we need to reject the claim that false things can’t be evidence, but I can see the beginnings of an argument against evidentialism here as well.