Saturday, October 30, 2010

Fallible but factive

Knowledge ascriptions are factive, but many of us are fallibilists about knowledge. So, I hope that this isn't an argument that would move many people to think that there can be false, justified beliefs:

(P1) It is possible to justifiably believe p even if you have the same justification(s) for believing p that someone does in some ~p-world.
(C) It is possible to justifiably believe p even if ~p.

I think this isn't a good argument [why not?], but I do wonder if it is this sort of argument that helps explain why all of the orthodox accounts of justification allow for the possibility of false, justified belief.

There's a reaction to this sort of argument that is just as bad as the argument. There are passages in McDowell where it seems he urges us to flip this argument on its head. He doesn't use the language of justification, but there's the thought that knowledge is a standing in the space of reasons and the internalist thought that epistemic standing cannot be beyond your ken or blankly external to you. This might add up to a sort of infallibilist view on which (C) is rejected which is taken as reason to reject (P1).

I try to untangle the web a bit in the latest draft of my disjunctivism paper (here).

I've also posted a revised draft of my paper, "Belief's Aim and its Justification". In it, I argue that if reasons for action are facts, justification ascriptions are factive. See here.

Why is it not a good argument? Whether something is justified depends, in part, upon the reasons to/for and the reasons against. Even if two subjects have the same reasons for V-ing, V-ing might be justified only for one of them because they might differ in terms of the cases against their V-ing in the situations they face. So, the argument assumes that there will not be differences in the reasons not to believe when we move from a situation in which a belief is true to a situation in which it is false. Why would anyone think that? Is there no reason not to believe the false?


Dylan said...

As far as why people think there can be false justified beliefs, I think one reason is that people have the intuition that certain false beliefs of people are justified.

Examples: There was a time when people were justified in believing Newtonian physics. There are people who are justified in believing that certain liars are telling the truth. The liars are very good at their art, and the people they trick have every reason to think they're being told the truth.

I would say that this is why I think there can be justified false beliefs. Intuitions about cases, not the argument you suggest.

Clayton said...

Hey Dylan,

I think that that's surely a reason why people believe there can be FJBs and maybe the first reason. I think there are ways of dealing with that intuition and then the fallibility bit comes out. For example, it comes out in Conee's paper on the truth-connection and Cohen's 1984 paper in Phil Studies where he presents the new evil demon objection.

To deal with the intuitions you cite, a two-pronged response is in order. The first is to say that there might be intuitions of a kind that point in one direction and that's consistent with there being intuitions that point in a different direction and we should go with the strongest. The second is to say that the intuitions you cite pertain to personal justification, not doxastic justification. The factivity claim is about ascriptions of doxastic justification, not personal justification.

So, while I think factivity is true, I also think this comes out true:
(1) There was a time when people were justified in believing Newtonian physics.

I don't believe this:
(2) There was a time when people believed Newtonian physics and those beliefs were justified.

That's because I think X is justified only if X can be defended from criticism. To defend a person from criticism is to uphold their rationality and to affirm that they were responsible in how they tried to discharge their duties. To show that a belief is justified is to show that it is permissibly held, held without the believer violating her epistemic obligations. Since you can't do that while believing falsehoods, (2) isn't true but factivity explains why (1) is.

Clayton said...

Quick follow up on the liars example. You wrote, "The liars are very good at their art, and the people they trick have every reason to think they're being told the truth."

That's right, I think. But, suppose in w1, Al is told a lie by a fantastically good liar. In w1, ~p, but Al believes p on the liar's testimony. In w2, Bill is told the truth by a truth-teller that is as similar to the liar as you could possibly make him.

In the example, Al and Bill have every reason to think they're being told the truth. You might even say they have the same reasons. But, they have different reasons not to believe. And that's the gap that I'm worried about. You can't say that if two have the same reasons to believe that they both have sufficient reason to believe, this just neglects differences in reasons not to believe. Again, the intuition that the material quoted is correct is fine. Someone who accepts Factivity can say that it is true, but that it misses something.

Dylan said...


Yes, I think I see the basic idea as to how you're going to respond to the physics example, although I'd have to see it all worked out to really evaluate it.

As far as what you say about the liar example, this is probably a dumb question, but why do Al and Bill have different reasons not to believe?

Clayton said...

Hey Dylan,

The "have reason" locution is probably not the best, but I know some writers (e.g., myself) use it to pick out reasons that apply to people whether cognizant of them or not. So, suppose you accept this view of testimony (not my view, but a view):

KK: You shouldn't accept S's testimony unless you know S knows.

Given KK, someone who is told p by a liar will always have a reason not to believe that those who are told by the knowledgeable truth-teller might not. They might not be cognizant of them, but that's another thing. It's consistent to say both that they have the same reasons to believe but differ in terms of their reasons not to believe.

So, maybe Bill and Al can both say (correctly) that their reason to believe is that someone who seemed honest said p. But, we can say that there were reasons for one to suspend judgment that there weren't for the others (if we hold a view on which you ought not accept S's testimony unless S speaks the truth).

I realize that the crucial claim is (to put it mildly) controversial. But, suppose you think that someone can't have sufficient warrant to assert p unless p is true. It's not wholly implausible given this premise that there's a correlative epistemic obligation not to accept p on the basis of testimony if you yourself shouldn't offer that testimony for purely epistemic reasons. Given the popularity of truth-requiring accounts of warranted assertion, it's sort of surprising to me that there aren't more deontologists about justification who have a truth-requirement on justification-ascriptions.