Fumerton on Disagreement

This is the third installment in my series of synopses of essays on the philosophical significance of disagreement.  

In “You Can’t Trust a Philosopher,” Richard Fumerton offers us good news and bad news.  The good news is that epistemic peer disagreement (EPD) concerning philosophical issues does not, in itself, undermine the rational justification of philosophical views.  The bad news is that, in explaining why, we discover a new, distinct defeater of the rational justification of philosophical views.  Fumerton’s argument goes like this:

1.  EPD undermines my rational justification for believing that P if there is good reason to believe that my epistemic peers are reliable when it comes to philosophical truth.

2.  But there is not good reason to believe that my epistemic peers are reliable when it comes to philosophical truth–in fact, there is good reason to believe that they are quite unreliable when it comes to philosophical truth (pervasive, intractable disagreements, patterns of belief, like Brown University grad students more likely to be foundationalists, etc., etc..).

3.  Thus, EPD does not undermine my rational justification for believing that P.

4.  But my reasons for judging my epistemic peers unreliable when it comes to philosophical truth are also reasons for judging myself unreliable when it comes to philosophical truth!

5.  Thus, if I believe that (3) because of (1) and (2), then I should also believe that (5a) I am unreliable when it comes to philosophical truth.

6.  (5a) undermines my rational justification for believing that P!

As Fumerton puts it in the closing section of his essay, we’re “out of the frying pan and into the fire.”  On Fumerton’s view, then, EPD does seem to have led us into a skeptical corner, albeit not as directly as others have suggested.  Fumerton concludes on the somewhat forlorn note that all he can do, so far as he sees it, is continue to try to convince those of his epistemic peers with whom he disagrees that they are somehow epistemically disadvantaged in ways that he is not.

Two Possible Points of Interest:

*Fumerton and the other philosophers I’ve read thus far on the significance of disagreement seem to be assuming that EPD is always epistemically significant at least insofar as it must be reckoned with.  If my EP disagrees with my claim that P, then, Fumerton and others seem to suggest, I have to either suspend judgment that P OR provide (be able to provide?) some plausible account of why I have reason to maintain my claim that P in light of EPD.  This strikes me as interesting because those who argue that EPD does not always undermine rational justification sometimes present themselves as defending common sense and practice, but what they really seem to be defending is an epistemic position somewhere between just forging ahead in the accumulation and defense of one’s philosophical views (the common practice, I would submit) and suspending all judgment on philosophical issues.  They are implying that EPD is something we must reckon with as such–it is not simply my peer’s objections to my position that I must grapple with, but the very fact that s/he disagrees.

*Fumerton’s essay introduced me to the Monty Hall Problem.  Are OPers familiar with this?  Maybe I’ll post it.

Update:  http://www.youtube.com/watch?v=mhlc7peGlGg


This entry was posted in Uncategorized. Bookmark the permalink.

9 Responses to Fumerton on Disagreement

  1. Carolyn Richardson says:

    David, you make the point that Fumerton and others are interested in the epistemic significance of the fact of disagreement PER SE. The DP comes down to the question what I (= any given knower) should do given the fact of my peer’s disagreement. I have a feeling this is an important point, because of the way it pictures the formation of belief (or acceptance). We usually think that whether to believe that P is a matter of whether there is evidence that P, and how good it is. Do I think my neighbour is at home? Well, his car is in the driveway, etc. The disagreement literature is about the significance, not of *evidence* for or against P, but of the fact of my peer’s thinking otherwise. As a consideration in favour of, or against, belief, what other people think is of a different kind from evidence. In this regard, the fact of my peer’s disagreement is like testimony (if you theorize it as being distinct from evidence). I don’t think there are any real, positive parallels between being told something and finding that my epistemic peer disagrees with me, but, as bases for making up my mind, they are both indirect and perhaps second-best to actual evidence…

    • David says:

      Hi Carolyn–

      You write: “The disagreement literature is about the significance, not of *evidence* for or against P, but of the fact of my peer’s thinking otherwise.”

      Could we say that the literature is about whether my ‘peer’s thinking otherwise’ is or is not ‘evidence for or against P’? I guess that’s how I was understanding the debate–whether peer disagreement is ‘epistemically significant’ with respect to justifying one’s belief that P.

      You suggest a possible comparison between the testimony literature and the disagreement literature–that strikes me as very interesting! Both cases involve the epistemic significance of what other people say, albeit in different ways. It might be worth exploring in more detail their similarities and differences. I wonder if this connection has been made in either literature.

      You say that testimony/peer disagreement may be ‘second best to actual evidence.’ I’m in this weird epistemic position where (a) I’m not sure what you mean, but (b) think I agree with you. 🙂

      • Carolyn Richardson says:

        Sorry I took my time responding, David. The point I wanted to make is not all that clear to me. It’s one of those situations where you have the strong sense that some unfortunately inchoate thought is both right and interesting. So, what is the thought? It’s that there’s a difference in kind between the way the following two things serve as considerations in whether to believe that P: what my peer thinks with regard to P and (some class of more direct considerations such as) perceptual evidence that P or not-P. You say that both are evidence, and I see what you mean—they’re both relevant considerations. But I suspect there’s a category of consideration that bears more directly on whether P and is therefore in some sense preferable to testimony and facts about my peer’s thinking. I was wanting to reserve “evidence” for this category of consideration, but that’s not important.

        • Nates says:

          The distinction Carolyn is pointing to reminds me of a distinction in mathematics, which also comes up in Kant — thus my interest in it. It’s the distinction between direct and indirect proofs. A direct proof shows you why the conclusion is true, for it is derived directly from the earlier premises. An indirect proof doesn’t do this. For example, consider a proof by reductio. If I prove that God must exist because denying God’s existence leads to some absurdity, then I now know that God exists, but I still don’t know why God exists (or what makes God’s existence possible), for I have proven the claim only indirectly from the falsity of its negation.

          (Kant was fascinated by these indirect proofs because he saw them being employed constantly in metaphysics in a way that led to various dialectical stalemates — e.g., proving by reductio that the world must both have and not have a beginning in time.)

          Anyway, it seems we can map the same distinction onto our considerations of evidence. Evidence as Carolyn was thinking about it is evidence that directly supports the truth of the claim — or shows why it’s true. So, seeing that the apple is red directly supports my belief that the apple is red. But David is thinking of evidence more broadly, to include both direct and indirect forms. So, the disagreement of my epistemic peers is still evidence that my belief is problematic, even if it doesn’t say anything directly about why the claim is false.

          I don’t think this adds all that much to the discussion, but I spent a chapter of my dissertation on this stuff, and it’s never going to end up in a paper, so this is my chance to use it!

          • juan says:

            I’ve been reading this very interesting thread about EPD, although if there’s anything in philosophy I don’t care about it’s epistemology.

            What I’d like to say, for whatever it’s worth, is that I don’t quite understand what the distinction between direct and indirect proof is (it’s true, I haven’t looked at Kant in a while). I was wondering whether some of you may have the same feeling. Consider the case of mathematics, where you can prove a theorem (like 1+1 = 2, or whatever) from definitions and axioms (directly, as it were) or, instead, indirectly, by reductio, and you show you get some contradiction in the axioms. How are these different? The direct proof is supposed to show me ‘why’ 1+1 = 2. But what does that mean? I can hardly wrap my head around this. It looks to me like ‘believe/know why something is the case’ reduces to believing/knowing that some other facts are the case. In both cases there’s an explanatory relation between a set of beliefs (the premises, as it were) and another belief that follows from them (the conclusion, theorem or whatever you call it). And it seems to me that this explanatory relation is just as strong for direct as for indirect proofs. Direct proofs, on this view, have no advantage left, because the ‘believing/knowing why’ has disintegrated into a set of ‘beliefs/knowings that’. So the status of direct and indirect proofs is the same.
            As for the case of perception, even if seeing the apple is more direct support for my belief that it’s red than hearing people tell me about the apple, it’s still not clear how the ‘why’ question gets in here. Seeing the apple and, on that basis, believing it’s red, does not, I think, tell me why it’s red. No more so than hearing some people praise the redness of the apple tells me why it’s red.
            So, if there’s anything to this direct/indirect distinction in the perception case, perhaps it has to do more with the distance of the stimulus or some such (you may recall Russell’s knowledge by acquaintance vs. k. by description) rather than with any ‘reason why’ the stimulus has property X.

  2. David says:

    Hey Juan!

    This thread has morphed into an interesting question about why we should think direct proofs are superior to indirect proofs. I’m really out of my element here, but what if we say something like this:

    Indirect proofs tell us that a proposition is false, whereas direct proofs tell us that a proposition is true. Knowing what is the case is better than merely knowing what is not the case, and so direct proofs are superior to indirect proofs.

    You may object: Indirect proofs show us what is true BY showing us what is false! If it is false that not-P, then it is true that P!

    In Classical Logic, but not all logics, right? I’ve been reading a book on the philosophy of logic (summer reading), and there are multi-valued logics in which not-P and a contradiction entails not-not-P, but NOT: P. So in the contexts of these multi-valued logics, which reject ‘bivalence,’ direct proof seems clearly superior to indirect proof.

    Why superior? Because more ‘determinate,’ I guess. Anyway, you know logic a hell of a lot better than I do, so enlighten me if I’m terribly wrong here.

    • juan says:

      Hi David

      I see what you mean: it does look like there’s an extra step in going from ~~p to p. And maybe that’s sufficient to make direct proof preferable.

      As for non-bivalent systems, I guess what’s happening there is that what we know as ‘indirect proof’ (in the classical conception) doesn’t work. What you can prove from the negation of a sentence and a contradiction is not the truth of the sentence, but something else, like the negation of the negation of the sentence.
      Consequently, one might reply that we can’t even compare the two methods of proof in this case, because we are not proving the same thing by two methods (which is the case we were interested in, in order to compare which of the two methods is better in arriving at that one thing), but we are instead proving two things (p, and then ~~p, which turn out not to be equivalent in these logics. So we are NOT proving p twice, once directly and once indirectly and then comparing the two).

      I was reading Terence Parsons’ ‘Indeterminate Identity’ a while ago, and in the system he proposes there indirect proof doesn’t work (he has a surrogate rule instead). The reason is that, even if you prove a contradiction from a sentence, it doesn’t mean the sentence is false. All you’ve shown is that it is not true. It might be indeterminate, not necessarily false. So indirect proof in the classical sense fails. But again, here also, since indirect proof doesn’t work, we can’t be comparing two methods of getting at a result (one of the methods doesn’t work at all).

      Of course, none of this invalidates your point that there is indeed some extra work in going from ~~p to p (in systems where you can actually make this move).

      Another further interesting question would be whether the logic needed to formalize knowledge and belief should be one of these multivalued ones or just classical. It looks like classical logic is enough for math, but for epistemology I’m not sure.

  3. Nates says:

    Sorry for hijacking the thread, but this seems interesting too. I don’t think the distinction is meant to depend on rejecting the law of excluded middle. It’s not supposed to be a distinction based on how true the proof is*, but how much explanatory power it has. Kant sometimes calls direct proofs genetic proofs, in that (and here we’re speaking somewhat metaphorically) they show where the truth of the claim originates from.

    I agree with Juan that this idea needs further defense, for the reasons he states. And, in fact, it’s been vigorously contested for centuries along these and other lines. Still, I think there’s something to it. Let me go back to my earlier example. Suppose I have two proofs of the necessity of God’s existence, one directly from the nature of God and the other by reductio from assuming the non-existence of God. If the proofs are legitimate, then they’re equally effective in showing that God exists. But suppose we want to know why God necessarily exists. This must have something to do with the special nature of God compared to other kinds of things. So a proof that appeals directly to the nature of God will bring us more insight into the source of this necessary existence. At any rate, that’s the concept. Whether this distinction holds up to further scrutiny, I’m not sure.

    * However, Kant does say that indirect proofs are less reliable in practice. When we’re engaged in reductio proofs, we always face the danger that we may have neglected an alternative. So, we assume that A and B are the only possibilities, show that B is false, and derive A. But maybe there was an additional possibility C that we failed to think of. This is Kant’s basic diagnosis of why the history of metaphysics is such a mess of contradictory claims. If we stuck to direct proofs, we would avoid this problem.

    • juan says:

      Reading the starred comment, it occurred to me that there is a larger sense of ‘indirect proof’ used in this discussion. Proving A from A v B and ~B is also apparently qualified by Kant as indirect.
      The cases Kant was concerned about in the antinomies, if I remember right, were naturally taken to involve just 2 possibilities. Either the universe has a beginning or it doesn’t, right? It seems like you can’t get more bivalent and intuitive than that. Kant’s solution to these is so striking and original that I’m not even sure it’s in the same order of things as the two contemplated possibilities, so I can’t really fault traditional metaphysics up to Kant for ‘overlooking’ or ‘neglecting’ anything.

      I take the point about the proof of God’s existence, but think also about how natural science proceeds. A lot of scientific research is very indirect, in this larger sense. It involves a lot of trial and error, inductive reasoning and so on. It’s far from the mathematical clarity that Kant was impressed with. And still, today we take science to be a model of ‘explanatory power,’ and we trust science with everything. Also, if you listen to people like Quine, math is not really that different from natural science, and then the distinction between the full-proof proofs of math and metaphysics and those of science seems to dissolve. Of course, that’s only if you’re willing to go as far as Quine.

Leave a Reply

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