Monday, November 28, 2011

Undoing evidentialism

I'm off to Southampton on Tuesday to give a talk on evidentialism. My focus will be on a pair of claims. First, that your evidence supervenes upon your mental states. Second, that relations between your beliefs and the evidence you have on hand is what determines whether you would believe with justification if you believed on your evidence. I'm skeptical of both claims.

Think about the move away from infallibilism, the view that you have the right to believe p only if your reasons entail that p is true. The move is perfectly sensible. Conee rightly notes that the infallibilist view does seem to lead rather directly to external world skepticism. What the evidentialists do to avoid the skeptical conclusion is reject infallibilism as characterized above but retains another kind of infallibilism that I think is deeply problematic. They think that it's always wrong to believe without sufficient evidence and always right to believe with sufficient evidence. That second claim suggests that normative standing supervenes upon the relations between your beliefs and the evidence you have on hand. To justifiably believe p, your reasons don't have to entail p, but a description of your reasons has to entail that it's right to believe p. It strikes me that there are a number of problems with this approach. Among them, if you think that belief is governed by a truth norm, the second sort of infallibilism entails that the first is correct. The result is that the second sort of normativized infallibilism doesn't lead away from skepticism, but right back to it. Anyway, I'll be discussing further problems for the view and hope to post a paper on the issue soon.

2 comments:

Anonymous said...

A historical background for the problem of evidence

http://waltherpragerandphilosophy.blogspot.com/2011/11/aristotle-on-soul-iii-3-recognizing.html

Dennis Whitcomb said...

Maybe fallibilist evidentialists should say that belief is governed by a "probable truth" norm. Such a norm might say "you should believe p only if p is probably true", where the variety of probability at work is epistemic probability. Would that help with your worries?