Some links

I’ll be presenting a paper called The Bayesian and the Dogmatist (PDF) at a bunch of places over the next few months. The version there is a talk version, which is very rough around the edges. But I hope it’s interesting. It certainly draws together more of the different things I’ve worked on in the past than anything else I’ve done.

Dave Chalmers posted a long list the other day of additions to his list of people with online papers in philosophy. There are a lot of good links there, but I was particularly happy to see that Lloyd Humberstone has an online papers page. There is a ton of interesting material there to work through, and the rewards for so working are very high indeed.

Online Papers in Philosophy, which I used to run, is running as well as ever (probably better) under its new home at Jonathan Ichikawa’s site. Here is its RSS feed. He has forms set up for suggesting corrections and additions to the site, so feel free to go and help out if you’re so inclined.

Finally, here are three new St Andrews related blogs.

Now if only they’d have RSS feeds, I could keep track of some more British philosophy!

4 Replies to “Some links”

  1. Brian,

    In anticipation of your visit here next week, I just read your paper \“The Bayesian and the Dogmatist\”. I like the Keynesian model you offer. But part of the argument for it seems suspect.

    The \“dogmatist\” issue is not my concern here. Rather, your appeal to the fact that P[~EvH | E&K] is less than or equal to P[~EvH | K] in making your case seems problematic. You are, of course, right about the inequality. So even though E may confirm H (i.e. even though P[H | E&K] is greater than P[H | K]), E has to disconfirm this particular consequence of H. But I don’t see this as a problem for the usual Bayesian account.

    I suppose it does show that a Bayesian account of knowledge has to take some care. For instance, it shows that knowledge can’t simply be something like non-Gettierized, significantly high (above some threshold) posterior degree of belief in a true proposition.
    For: It seems intuitively clear that, before learning E, we may not know either ~EvH or H. But after learning E (where E is significantly extensive), we may come to know H due in part to how high P[H | E&K] is; and since H entails ~EvH, we may then come to know it as well. But because P[~EvH | E&K] is less than or equal to P[~EvH | K], we have that P[H | E&K] is less than or equal to P[~EvH | E&K] is less than or equal to P[~EvH | K]. Thus, if P[H | E&K] is above threshold, so is P[~EvH | K]. So, arguably, true belief with belief strength above threshold is not enough for knowledge.

    But even this line may not hold up. The Bayesian thresholder can simply bite the bullet and say that since P[~EvH | K] has to be above threshold (if P[H | E&K] is), ~EvH was already known before E was learned. And although E disconfirms (lowers the degree of confirmation) of ~EvH, it doesn\‘t lower it enough to defeat the knowledge claim.

Leave a Reply