Monday, September 7, 2015

Justified belief and conditional evidence

Plausibly, a belief that p is justified only if one has good evidence that p. But what about a case where instead of having evidence for a belief, one has evidence that if one believes it, then it's be true? (I'll call this the Belief Conditional.) For instance, tonight Sam will decide whether to watch Battlestar Galactica or Deep Space Nine. But Sam hates being shown to be wrong. So if she now comes to believe that she will watch, say, DS9, then come evening she will watch, say, DS9 in order to make her earlier belief true. She knows all this. She also hates suspending judgment. So she makes herself believe that she will watch DS9. (She's not deciding what she is to watch. The decision will come tonight.) Once she realizes that she has succeeding in coming to the belief that she will watch DS9, she has evidence that she will watch DS9. But we may suppose that there is a short period of time during which Sam hasn't yet realized that she believes she will watch DS9. During that short period of time, she doesn't have evidence that she will watch it. Instead, she just knows the conditional that if she believes she will watch it, she will watch it. I am inclined to think that Sam's belief that she will watch DS9 is reasonable and justified.

But I am not happy to extend this to a general claim that having justification for a Belief Conditional suffices for justification of unconditional belief. Here's a case that worries me. Suppose that having read a lot of papers defending an error theory about folk psychology, and generally hanging about in unfortunate philosophical company, Fred is in possession of strong evidence that nobody believes anything. But despite the evidence, ingrained habits make Fred continue to believe that someone believes something. (I take it for granted that the error theory is mistaken.) Of course, Fred does know the obvious necessary truth that if he believes that someone believes something, then someone believes something. But nonetheless given the evidence against folk psychology, I am inclined to think that Fred isn't justified in believing that somebody believes something.

I don't know how to distinguish the cases of Sam and Fred. I feel pulled to assimilate one to the other, but I don't know which I should assimilate to which.


Heath White said...

I would think the difference is that what Fred has is strong evidence AGAINST the view that somebody knows something. Whereas Sam does not have any strong evidence for or against the proposition that she will watch DS9. So Sam's belief is sort of self-verifying, and might be justified that way. But Fred's belief is undercut by other things he believes and might be unjustified that way.

Alexander R Pruss said...

Maybe. For the sake of parallelism, I should have said that Fred ends up thinking that the evidence for and against the error theory is balanced. I still have the inclination to think that Fred shouldn't believe that anybody believes anything.

Perhaps I can explain my inclination in Fred's case (both the original and the modified versions). Surely folk psychology is approximately true even if the error theory is true. So even if the error theory is true, probably it still ends up being approximately true that Fred believes there are beliefs. And while it's true that if Fred believes there are beliefs, then there are beliefs, it need not be true that if it's approximately true that Fred believes there are beliefs, then there are beliefs. For it to be approximately true that he is justified in believing in believings, Fred would need to have evidence not just for the Belief Conditional but for the Approximate Belief Conditional (if it's approximately true that I believe p, then p). But just as one should rationally avoid unjustified beliefs, one should perhaps also rationally avoid being in a position where it's approximately true that one has an unjustified belief. Fred fails to do that, so he fails in rationality. (I can say that straight out, because the error theory is in fact false. Otherwise, I'd have to say that he is approximately irrational or something like that.)

(If, on the other hand, folk psychology isn't even approximately true on the error theory, then all bets are off.)

In Sam's case, however, only the Belief Conditional is salient, since the Approximate Belief Conditional is only salient when one takes the error theory seriously.