Friday, October 26, 2007

Experimental Philosophy

If philosophers are going to appeal to facts about what seems "intuitive", should they first do empirical work to find out whether most folk actually share their intuition? So suggest the experimental philosophers. I'm skeptical, however. The epistemic force of an intuition depends on the coherence of the conceptual scheme that generates it. Philosophers are presumably better than layfolk at thinking clearly about philosophical concepts. So I don't really see that we have much to learn from their untutored intuitions. (Some complain that our intuitions are "corrupted" by theory - but mightn't this be better described as refinement?)

Doris and Stich (2005) 'As a Matter of Fact: Empirical Perspectives on Ethics' puts forward the case for experimental philosophy. Below, I reproduce my comments from a past (off-blog) discussion.

* * *

Insofar as philosophers are concerned with non-contingent matters, it seems that a priori analysis should suffice. Consider the internalist's question: is amoralism possible? We already have the challenge from Hume's imagined "sensible knave" -- what difference do real life psychopaths make? In either case, it seems like a question for conceptual analysis: given such and such a scenario, how are we to describe it? (Does the knave/psychopath really form moral judgments, or merely schmoral ones?)

Now, a central argument of the paper criticizes conceptual analysis on the grounds that empirical work (presumably: the "vignette" method favoured by the new "experimental philosophy" movement) is required to uncover *real* folk concepts. But this doesn't do justice to the normative element of analysis. They write:
Smith can reply that responses like those Nichols reports would not be part of the maximally consistent set of platitudes that people would endorse after due reflection. But this too is an empirical claim... (p.125)

How is this an empirical claim? What people would conclude on ideal reflection depends on what propositions are maximally coherent, etc. There's no experiment we can do to pin down what this is; any amount of actual reflection by third parties can always be rebuffed as insufficient to reach the ideal end-point -- "those participants," one may claim, "have not undergone *due* reflection." Maybe they've reasoned badly. The only way we can judge this is to engage in normative reasoning ourselves, and see whether the participants' answers correspond to what we've already determined to be true from the armchair!

Similar issues arise for the problem of persistent moral disagreement. The authors write:
the argument from disagreement cannot be evaluated by a priori philosophical means alone; what's needed, as Loeb observes, is 'a great deal of further empirical research into the circumstances and beliefs of various cultures'.

But I can't see how that would help, if in the end we can only judge others' rationality according to the substantive conclusions that they reach.

Besides, we should (in principle if not in practice) be able to tell from the armchair whether one position or another is rationally necessitated. Simply imagine all the conceivable cultural disagreements, and the difficulty of adjudicating between them. (Which ones obtain in the actual world seems quite irrelevant.) If we ultimately find a conclusion to be rationally necessitated after all, and others disagree (without providing any new reasons, since - ex hypothesi - we've already considered them all in reaching our previous conclusion), then that simply shows that they haven't engaged in fully ideal reasoning yet.

Matters are different in practice, of course, due to our own fallibility. Arguably, our credence in philosophical claims should be informed by the empirical (meta-)evidence provided by others' judgments of the issue. But can we ever hope to scientifically measure the rationality of their judgments, on purely procedural grounds (i.e. without begging the substantive question at hand)? If not, we may find that the real adjudicating work must still be done from the armchair.


  1. I expect philosophers would be systematicaly biased in some regards so maybe experimental data would not go amiss as long as one debated the application of that data.

  2. Re. intuition corruption: I don't understand your view. Are you saying that philosophical thinking *doesn't always* "corrupt" (i.e. make unusuable for this or that philosophical purpose) intuitions, or are you saying that it *never* does? If it's the former, then you aren't saying much.

    If it's the latter, I think you're committing yourself to a much more controversial idea than you are letting on. The latter claim seems to amount to the assertion that no (interesting) philosophical argument ever makes or relies upon any claim about how things would seem to us if we were unaware of the philosophical implications of such seemings. That's a bold claim! If it were true, then I think a very large number of well-known philosophical arguments (e.g. in ethics, to take what seems to me the most obvious example) would be deemed bogus in one fell swoop.*

    Of course, none of these points show that your view is mistaken. But they do suggest that you shouldn't take yourself to be defending "traditional" philosophy (i.e. the philosophy traditionally done by philosophers who use no experimental evidence), because that sort of philosophy very often relies on the very kind of claim you are saying isn't philosophically relevant.

    *= I figured I would be able to substantiate this claim by opening a philosophy paper at random from my harddrive. I picked Foot's paper, "Utilitarianism and the virtues," and find this sentence:

    I do not mean to go into these matters in detail here, but simply to point out that we find in our ordinary moral code many requirements and prohibitions inconsistent with the idea that benevolence is the whole of morality.

    This claim about "our ordinary moral code" sounds to me like a claim about the beliefs -- intuitions -- of ordinary people. Insofar as it is controversial, the controversy couldn't be resolved by a philosopher simply reflecting on whether this claim describes *her own* beliefs, because she is not an ordinary person.

  3. Hi Richard, Antti Kauppinen defends a view similar to yours in "The Rise and Fall of Experimental Philosophy."

    For a view that defends intuitions, but also some forms of experimental philosophy, you might be interested in my "A Defense of Intuitions."

  4. Many experimental philosophers are not interested in what the intuitions of ordinary people are, but instead interested in what intuitions, themselves, are; how people use intuitions; and what psychological forces are at play when one uses intuitions. It seems (intuitively) obvious to me, that we need to understand intuitions before we haphazardly use them for evidence to philosophical problems. I see most of Knobe's work in this Quinean tradition, as does Chris at Mixing Memory.

  5. Wes - those sound like interesting and important questions. I should clarify, then, that my criticisms are instead limited to the branch of experimental philosophy that would have us do conceptual analysis by means of surveying the folk.

  6. If you want to call it refinement, you owe a theory of what the target of the philosophical study is, and why the kind of training that philosophers get would be likely to produce refinement.

    It's also important to distinguish different subfields of philosophy: for better or worse, intuition mongering dominated epistemology after Gettier (I'm not up on how it is now), so coming up with the account of philosophical refinement seems both difficult and pressing there. This is one reason that it's not helpful to criticize experimental philosophy as a whole--it's a coherent movement insofar as they all want to bring experimental methods into philosophy, but since they work in different subfields, their aims and collateral commitments are going to be very different.

  7. Richard, you say, "my criticisms are instead limited to the branch of experimental philosophy that would have us do conceptual analysis by means of surveying the folk". But your criticism extend to my claims about the partly-empirical work needed to resolve the argument from moral disagreement. I don't understand that to be (in its most common versions, at least) an argument about how best to understand moral concepts, nor do most others who discuss it. Anyway, one central focus of the debate has been whether there is disagreement of a sort that threatens our claim to have epistemic access to moral truth. If some disagreements would disappear when there is agreement on the non-moral facts, for example, then that disagreement shouldn't worry those who claim that we do have such access. Likewise if apparent disagreement is really simply that (because the same principles are being applied to very different situations--as in the now cliche case of Eskimos putting the elderly out to die) it isn't a problem either.

    Unfortunately, in my view, we too often put forward a few stock examples of apparent moral disputes that can be resolved (or shown to not really be disputes at all) in these ways, and that doesn't settle the issue. Boyd, for example, says that he believes MOST moral disputes would vanish if there were agreement on the non-moral facts. How does he know THAT? It seems to me that he's making a claim that's at least partly empirical--one that we'd have to test to evaluate properly.

    But even conceptual analysis seems to me to need SOME empirical backing. As I said in the paper from which Doris and Stich quote:

    "Our language has no bearing on the question of what properties or entities exist. What exists exists whether we talk about it or not, and there are undoubtedly many things for which we have no words. But our language is relevant to the question of whether certain existent properties and entities are the ones we have been talking about when using a certain vocabulary."

    We can (and certainly should) refine the "folk" concepts. But we should be wary of refining too much, lest we change the subject! To find out what the folk concepts are, we need to get out of our comfy armchairs (or better yet, talk some psychologists into doing it!)

  8. Hi Don - you're right, my clarification to Wes wasn't quite accurate. I should've simply said that my criticisms are limited to those elements of experimental philosophy discussed in the main post (which goes beyond just conceptual-analysis-by-survey, but at any rate doesn't claim to be exhaustive).

    On the substantive issues:
    (1) It seems to me that the question whether "moral disputes would vanish if there were agreement on the non-moral facts" (and, crucially, if all parties were sufficiently rational) is simply the question whether the non-moral truths rationally entail determinate moral conclusions. And that's not an empirical question at all, right?

    (2) I agree that it takes empirical work to find out if other folk have the same concepts we do, or if we're just talking past each other/ "changing the subject". Maybe we're speaking different languages. That'd be linguistically interesting. But what does it have to do with philosophy? (I want to analyze free will, say, this very concept I already have a grasp of -- not just whatever other folk refer to by 'free will', since that could be anything at all, and so not necessarily what I'm interested in.)

    DK - Yeah, I'm pretty sympathetic to the "bold claim". It's a tricky issue, though. Internalizing false theories may cause our intuitions to go awry, but this is just as true of the unarticulated "theories" that permeate our background culture. I guess if academic speculations are more likely to be misguided than the inherited "wisdom of the ages" (so to speak) then that'd give us some reason to favour folk intuitions. But I don't see any special advantage to being "unaware of the philosophical implications of such seemings". The epistemic work is instead being done by a kind of Burkean conservatism, and pessimism about the quality of the new theories that seek to supersede commonsense. (Cf. my remarks about "meta-evidence" toward the end of the post.)

  9. 1) Claims about whether M disag. would disappear upon agreement on non-moral facts do NOT boil down to the (non-empirical) question whether the non-moral facts entail determinate moral conclusions. Most philosophers don't believe that if there is moral truth it is entailed by non-moral facts. The question many of us interested in moral disagreement have in mind concerns the moral beliefs people would actually hold if they were better informed.

    The idea, as I tried to suggest, has to do with epistemic access to the supposed moral truth. By analogy, suppose I claim that we have a faculty of vision, which enables to tell us about nearby objects. It would be an embarrassment to such a(n obviously true) "theory" if people looking at the very same things reported wildly different "visual experiences". But if these differences could be explained as owing to, for example, differences of perspective, then the embarrassment would disappear.

    The argument gets pretty complicated, depending on the sort of access hypothesized, and non-moral beliefs are only part of the story. But entailment of the moral by the non-moral is not standardly thought relevant.

    2) On your point about being interested in free will (the concept you already have a grasp on) as opposed to what other people refer to when they use the concept, I'm not fully getting your position, though I've heard people say similar things. Presumably you are not just looking to understand a concept that appears only in your personal idiolect, but a shared one. Suppose I (as a philosopher) want to understand the concept of moral rightness, which (cont'd)

  10. I understand to be the concept of "utility maximizing". And suppose I discover that some actions really are utility maximizing. I will not, thereby, have vitiated realism about moral rightness, but only about rightness in my own idiolect (or concept, etc.) and that of others like me. If I want to understand rightness I want to understand the thing people using the word are talking about. Don't I?

    Sorry to have taken so long to answer (and to take so long in answering!) If I have violated the norms of bloggery forgive me, please, for I am new at it!

    Btw, I have a piece that addresses these (latter) issues in much more depth in Walter Sinnott-Armstrong, ed. Moral Psychology, Vol. 2 (MIT, pp. 355-421). Of particular interest, perhaps, are the comments by Gill and Sayre-McCord and my reply to them. S/M, in particular, takes a position that sounds like what Richard is getting at in his point (2), responding to my earlier post.

  11. Hi Don, thanks for your reply, and for the reference (which I will need to chase up). Let me clarify my take on the two points in question:

    (1) I don't think we have any special "faculty" of moral intuition. All we have is the general faculty of reason. So, on my view, if the moral facts are not a priori entailed by the non-moral facts, then they can't be known at all, and so we (in our ignorance) are condemned to persisting disagreement. Conversely, if the moral facts are entailed by the non-moral facts, then obviously disagreement will not persist upon idealization.

    (2) I really do just care about my personal idiolect.

    Note that even a utilitarian should - reflecting on Moore's open question argument - recognize some irreducibly normative moral concept (it doesn't matter whether they or others call it "rightness") distinct from that of "utility maximizing", even if they turn out to denote the same property.

    This concept I've just described - the irreducibly normative one, which just so happens to be commonly called "rightness" - is certainly well worth understanding. But why would you care about "whatever other people mean by 'rightness'", under this peculiar description? You should care about rightness, the concept itself, and not under the guise of the mere word.

  12. (1) OK, but you're ruling out a lot of views inconsistent with that and not saying why. Some people think we have a priori access to some moral truths, but not through a special faculty. (Even you think we have access to something like that, I bet: truths about valid arguments, eg.--part of that general faculty of reason, perhaps).

    But lets not lose sight of the issue. There is a question WHETHER there are moral facts--of the entailing or non-entailing kind.

    Disagreement is prima facie evidence that there are not, because the best evidence we could have for such facts is that we seem to have epistemic access to some. Failing to secure agreement, even when there's agreement on the NM facts, is evidence that they are not knowable even if entailed. (But if they are not knowable why believe in them?). It's not conclusive evidence, since some people may not be in a position to know them even when ideally informed, and there are other things which could account for the persistence of disagreement.

    Btw, the idealization is very problematic. I hate to keep sending you my papers, but since I suspect at this point the thread involves just you and me, I'd be happy if you looked at "Full-information theories of Individual Good"," which is on a different but related idealization. And if you haven't yet, you might look at my P. Studies paper (on disagreement)--the one Doris and Plakias mention.

    (2) When you say "irreducibly normative," you mean (or are attempting to refer to) something. I don't know what you mean, but I can guess, based on what you say and what I know of others' use of the words. The latter is because I suspect you really do mean to be speaking a (semi) public language and not just employing your own idiosyncratic meanings.

    Ok, I don't hate sending you to my papers.


  13. (1) I'm losing track of where we disagree here. If a moral principle P is knowable a priori then (a fortiori) P is a priori knowable given the non-moral facts. Hence, if some agent fails to believe that P, they must suffer from some cognitive limitation or rational flaw that's impeding their ability to recognize a priori truths.

    So, the question whether fully informed and rational agents would still disagree over moral issues just is the question whether moral truths are a priori knowable (given the non-moral facts). And this is emphatically not an empirical question.

    Having said that, I agree that observing actual people (assuming we're more-or-less rational) provides some (defeasible!) evidence about the ideal case. But it's very indirect -- more like the testimonial evidence you might get about mathematics by asking mathematicians about their beliefs. It doesn't show that empirical work is necessary to make progress on the question. Indeed, any empirical justification we obtain is strictly derivative on the rational work of those surveyed.

    (2) Sure, as a communicating agent I mean to speak a public language. I have certain thoughts I wish to convey, so - for practical reasons - I try to use words the same way as others do, so that you might grasp the thoughts I have in mind. But as a philosopher, it's the thoughts I care about.

  14. I'm sorry to interrupt what is, for me, an illuminating conversation with what may be a perfectly useless question but...

    If we can consider philosophical positions as hypotheses of some sort then couldn't these be tested empirically? What I mean is that if a philosophical position is true, it should entail certain real-world results that are different from other philosophical positions. For example, if I believe that there is no moral faculty then I should be able to develop a test of that hypothesis. One might argue that James Blair presented work that tested just this hypothesis at the SPP conference last year (people with certain brain abnormalities were unable to make certain moral judgments). If we accept that his work has added to the conversation (if you dispute the results or the implications you still have to address or account for this, apparent or real, criticism) then it seems that empirical work has proven to be useful to philosophy.

    Sure, you might be able to get to the same conclusion by purely theoretical, arm-chair methodologies, but the availability of a traditional method isn't an argument against alternatives. At the very least one can, and some might say must, test the truth value of an arguments premises. ex. suppose the claim "humans have intentional states" is part of an argument for assessing moral praise and blame, then it would only make sense to test to see if humans really do have intentional states. Having intentional states won't prove the argument but not having them will clearly disprove it.

  15. Hi Christopher, I guess I see a lot of questions as having a scientific (empirical) component and a philosophical (interpretive) component. Applying this to your example, it is not a philosophical issue whether certain brain abnormalities cause one to make such-and-such judgments. It's a philosophical question whether said judgments are properly described as moral judgments, though. But this is no longer a question that's directly susceptible to empirical evidence.

    Similarly, you might empirically test whether humans have brain states of type X. But it's a matter of pure philosophy whether brain states of type X are properly understood as intentional states!

    (See also my old post: Does philosophy need science?)


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.