Feldman on Alston

Rich’s review of Alston’s Beyond Justification is now available here at Notre Dame Philosophical Reviews. It is a very helpful review, raising just the right questions while at the same time recognizing the significance of the book.


Feldman on Alston — 7 Comments

  1. That’s a very interestng review. I’ve got to look closer at Alston’s view on grounds for belief. According to R. Feldman, “Alston holds that something is grounds for a belief only if it makes it probable that the belief is true” (94). If something like that is true, it looks like whether a belief is grounded is (in some cases) entirely up to me. Take my belief in Newcomb cases that the opague box contains $1 mil. It’s up to me whether that belief is grounded. If I freely choose one box, then the evidential probability that the box contains $1 mil. goes up and the belief is grounded. If I freely choose two boxes then belief is not grounded. “Let’s see, do I want the belief to be grounded or not?” That seems mistaken. It doesn’t seem a matter of free choice whether some belief is grounded.

  2. My belief that the opaque box contains $1 million dollars is grounded if I freely choose one box. Similarly, my belief that I have raised my hand is grounded if I freely choose to raise my hand. Is there anything more implausible about the former than the latter? Is there anything implausible about the latter?

  3. My belief that I raised my hand is not grounded if I freely choose to raise my hand. You might have no idea you raised your hand or did it freely. “You raised you hand in class yesterday, I saw you”. “No I didn’t”. But I have no trouble with my awareness or perception that I freely raised my hand grounding my belief that my hand is raised. Similarly, I would have no problem with my perception that the box suddenly contains $1 mil. on my choosing it grounding that it contains $1 mil. I would also have no problem with a back-tracking counterfactual & my choosing one-box grounding my belief that the box contains $1 mil. But the fact is I don’t perceive the $1 mil. in the box and there is no true back-tracking counterfactual that would lead me to believe that my choosing affects what’s in the box. So the fact that choosing one box raises the evidential probablity that it contains $1 mil. doesn’t *seem* to be enough the ground the relevant belief.
    To bring the matter home a little more. Suppose someone informs you that there is a correlation between scratching your chin and having deadly disease D. Do you really believe that deliberately avoiding that behavior would make it more reasonable for you to believe that you don’t have D? Do you think your belief that you don’t have D is grounded as long as you deliberately avoid scratching? Do you think it would be a great idea to keep your mittens on to avoid scratching? I don’t think any of these free choices should affect whether your belief that you have D is grounded.

  4. Mike, to get a problem here of the sort you’re thinking about, you need to read the claim as an “iff” rather than an “only if” claim, right? I’d expect that Alston’s view is that the additional stuff needed will distinguish between inferential and noninferential beliefs, and appeal to the other features that are prominent when Alston addresses the issue of whether a belief meets important epistemic desiderata.

  5. As far as I can tell, there is no special problem for Alston here. One issue is freely controlling whether you have grounds for a belief. Sometimes you can, as Ram said. The example about scratching and disease D is similar to reference class issues. You can also think of this in terms of defeaters: given knowledge of the correlation, knowing that you haven’t scratched gives you a reason to think you don’t have D. But the fact that you have *chosen* not scratch is a defeater for this reason. (There are assumptions buried here, but I won’t try to spell them out.)

    I think that there is a problem with the sort of reliable indicator view Alston defends. If only maple trees look a certain way, then seeing a tree with that look reliably indicates that one is seeing a maple. So, he seems to be committed to the view that anyone who sees a tree with that look has a good reason to believe he’s seeing a maple, even if he has no idea that maple trees do look that way. Note that the correlation here is not accidental (in some sense), since it is a law-like fact that only maples look that way.


  6. “But the fact that you have *chosen* not scratch is a defeater for this reason”.

    It would be good if choosing not to scratch did not ground my belief that I don’t have D. But if there is a correlation between not scratching and not having D, then choosing not to scratch is going to affect the evidential probability that I have D. Just looks ad hoc to treat “choosing not to scratch” as a defeater. It lowers the probability that I have D as much as “learning that I chose not not scratch”.

  7. Jon, if there is some necessary condition on grounds that rules out the disease/scratching sort of case, then I think you’re right. I didn’t happen to see that ruled out in the epistemic desiderata, but I could have missed that.

Leave a Reply

Your email address will not be published. Required fields are marked *