Monday, August 29, 2005

Epistemology without Evidence

There are two ways to challenge evidentialism. One is to argue (as in my previous post) that there can be 'practical' or non-epistemic reasons for belief. The other, which I want to explore presently, is to argue that there can be non-evidential epistemic reasons. There can be theoretical reasons for you to believe that p, which are not at the same time reasons for the truth of p.

How could this be? Well, the normative potency of evidence derives from taking truth as the normative standard against which to assess beliefs. We say "belief aims at truth", and then assess beliefs according to whether they achieve this goal. But there are other normative frameworks we might adopt instead. There is, for example, the standard of coherence, which assesses a set of beliefs for its internal coherence, rather than applying an external standard such as truth. Despite making no reference to truth, the standard of internal coherence is nevertheless clearly an epistemic standard, rather than a practical one.

So, I propose, we have epistemic reason to believe that p if doing so would yield a more coherent belief set. (There are shades of Michael Smith in this suggestion.)

Suppose that you believe that p, and believe that if p then q, and so on this basis you rationally form the belief that q. The two former beliefs provide you with a reason to adopt the latter belief. In doing so, you make your belief-set more coherent. This is epistemically rational.

So there we have a reason for believing. What about reasons for the thing believed, i.e. the proposition that q? If asked, you would likely point to the propositions that p, and that if p then q, as being your reasons for taking q to be true. However, let us suppose that your former beliefs are false: p is false, and if p then q is false. So while you took these to be reasons for the truth of q, in fact they are not. You were mistaken about whether there are reasons for the the truth of q. There are no such reasons in this case.

Thus we find that there can be (even epistemic) reasons for believing, which are not reasons for the truth of the thing believed. Musgrave's distinction is vindicated, even if his particular application of it is not.


  1. The normative standard of coherence is imporant because coherence is often truth preserving. That is why you point to the modus ponens, which is truth preserving. That's the point of logic.

  2. Sure, but what if the original beliefs (premises) are false? I want to say that you still have a reason to make the inference, even though it is no reason for the truth of the conclusion.

  3. I think I want to say that you can have a (perhaps weak) reason to believe the conclusion, due to your beliefs in the premises, even if those prior beliefs are entirely unjustified according to external standards of evidence or truth-indicativeness. The internal standard of coherence provides normative force quite independently of external considerations.

  4. Obviously, I tend to side with Pat here, but I think there is something to arguments that the rules of good reasoning have normative force independently of their relation to an external goal. John Broome's work on reasoning and Niko Kolodny's new article in Mind are a good place to look for interesting discussions of the issues involved.

  5. Richard, did you ever read Williamson's Knowledge and Its Limits? While I'm far less enamoured of it than I was initially, it's still an interesting approach to some of these issues.

  6. Thanks for the references, Nishi, I'll have to look into them. (Aside, Broome will be guest-lecturing our class in the coming weeks, so I'm looking forward to that!)

    Clark, no I still haven't read Williamson. What does he say about this issue?

  7. I'll try to get to that and post something at my blog Richard. It'll probably be a few days or longer though. I'm kind of backlogged.

  8. I hope this thread isn't dead yet. There was something I was hoping to address and as much of this is in fact inspired by the stuff that Nishi did and did with Velleman, it would be terrific if people still had their ears on (does anyone still get that btw, the old trucker lingo when they used CBs?).

    I'm not myself an evidentialist but I am convinced, like Nishi and Velleman, that the aim of belief is truth. I'm also convinced that given this fact, we can evaluate belief regulating mechanisms, processes, etc. in terms of how well they respond to various kinds of inputs and when they are sensitive to inputs that aren't truth-indicative or play the role of truth-indicative considerations, then this is a mark against the mechanism. It seems that there is something right to the following observation: the standards for evaluating that which regulates belief is fixed by the standards appropriate for evaluating belief, which in turn are fixed by observations regarding the aim of belief.

    What worries me is the move to the slightly different point that we can then turn around and evaluate particular attitudes by thinking about how those attitudes relate to actual or counterfactual operations of a well-functioning belief regulating mechanism. Suppose we grant that if your belief regulating mechansims treated evidential and non-evidential considerations on par, the thing would be in bad shape. Fine, but remember that the aim of belief is to hit a target and that is different from (a) stemming from something that is good at hitting targets or (b) hitting a target having stemmed from something that is good at hitting targets. I worry that any argument for evidentialism that follows this route first has an explanation as to how we would arrive at a view whereby we accept either (a) or (b). Moreover, when it comes to actions, it seems a case can be made for thinking that if the evaluation of actions is to be done in terms of an aim we ought to reject any view that has as a condition of justified action or action that hits its aim one that must do so by virtuous means (Moore was acutely aware of this when he wrote Principia Ethica and described examples where the right thing to do given the circumstances would only accidentally be connected up to an external aim).

    A second but related line of argument for evidentialism might take its start with the observation that certain assertions/thoughts of the form 'p but I have no evidence that p' are Moore paradoxical. There is a second step to the effect that MP utterances/thoughts provide us with a guide to what Broome calls 'normative requirements'. The conclusion: we get a wide-scope ought to the effect that one oughtn't both believe p and not have sufficient evidence for p. This gets us something close enough to evidentialism.

    The trouble with this move, however, is similar to the one from the previous paragraph. It seems that while instances of MP might indicate the presence of a wide-scope ought, it might also indicate a cognitive blindspot which might not in turn indicate any wide-scope ought. A cognitive blindspot is some condition one cannot be cognizant of if one is in that condition. For example, if I intend to X by Y-ing, Y-ing and accidentally bringing X about is a blindspot. I can't see that this is my lot but you can. If I assert that by Y-ing I shall accidentally X, you get something like a practical variant on Moore's Paradox. However (arguably) there is no wide-scope ought in cases such as this (I suspect Broome denies this but I've argued at length that this is a mistake).

    At any rate, these seem to me to be the two most promising lines of argument in support of evidentialism but neither seems conclusive. There are also what I take to be straightforward counterexamples to some evidentialist claims but that is for another day. Any thoughts?

  9. I guess I don't really understand what the import of these "normative requirements" is. Even supposing it's part of the concept of belief that it "ought" (in some limited sense) to be governed by evidence, that doesn't seem to say anything about what we ought all things considered to do.

    If a malicious demon threatened to blow up the world unless we broke these "normative requirements", then surely breaking them is precisely what we should do. When people start talking about "reasons" or "requirements" that have no bearing on what we really ought to do, I just don't get what they're talking about, or why they think it's important.

    This is a worry I brought up in my more recent post on Hieronymi's "wrong kind" of reasons. Given that practical (or "extrinsic") reasons quite obviously do exist, what is gained by calling them "the wrong kind"?

    Given how I'm understanding it, evidentialism just seems so obviously false that I feel I must be misunderstanding what evidentialists are trying to claim. They can't really be denying that extrinsic reasons exist in relation to beliefs. This is far too obviously true for anybody to even think about denying. So what are they denying? I just don't get it!

  10. Richard,

    I think that the questions we are interested in might be slightly different. Perhaps the question you are interested in is this one: why think that the epistemic ought fixes the all things considered ought?

    If it doesn't, then the fact that facts about an individual's evidence settles questions about what one ought, epistemically, believe still leaves open an interesting question as to whether non-epistemic reasons might have a bearing on whether all things considered one ought to believe a claim.

    I think I'm interested in a slightly different question: do the facts about an individual's evidence fix the facts about what that individual should believe? Thus, regardless of whether we say that the atc ought = epi ought or not, the evidentialist has told us the wrong story about all the oughts. Given the relation between reasons and oughts, I suspect they've told us the wrong story about epistemic reasons as well. Anyway, I think the stuff about normative requirements is relevant insofar as I think that focusing just on the epistemic ought, evidentialism is often understood as the claim that there are wide-scope oughts relating a fact about belief and a fact about evidence.

  11. Hi Clayton, I take it that by "should believe" you merely mean "epistemically should believe" (otherwise I'm not sure how our questions are different). Which would be fair enough, since that was the original topic of my post here anyway. I've since gone off on a bit of a tangent :)

    But yeah, my real interest here is broader than that -- concerned with ATC oughts, rather than merely epistemic oughts. So far as the "ethics of belief" debate is concerned, I really don't understand why there is any debate at all. Do you have any idea what evidentialists are trying to suggest there? (You did previously mention that you had some sympathy for the "believing" vs. "getting to believe" distinction. But as explained in the linked post, this strikes me as mere semantics. Is there any substantive disagreement going on?)

  12. Okay, I think I've sorted this out now -- and changed my position in the process!

  13. Richard,

    I have a new post over at which may address some of your criticisms. If not, I would be glad to hear from you on where I have misunderstood your arguments.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.