Sunday, April 27, 2008

Metaethics Diavlog

Wow. This interview between Will Wilkinson and UNC's Geoff Sayre-McCord is incredibly good. Sayre-McCord is a wonderfully clear and careful thinker, and Will asks him excellent, probing questions. (Can you imagine seeing such a philosophically astute discussion on regular TV? It's times like these that I really love the internet!)

One recurring issue concerns the extent to which moral statements are simply redescriptions of natural facts. Does 'Hitler was evil' say anything over and above the fact that he had a callous disregard for others' welfare, etc.? Does the goodness of our social institutions consist in anything more than the descriptive fact that they are conducive to [such-and-such specification of] human flourishing?

The problem with the negative (reductionist) answer is that it risks turning normative disputes into mere semantic disputes. Suppose one were to say: "I grant that Western freedoms are more conducive to personal development, happiness, and all that jazz, but nonetheless they are bad, because it is more important to promote obedience, piety, etc." We don't want to say they've contradicted themselves, as we must if 'good' just means 'conducive to [...]'. Their error is not linguistic. It seems there's a substantive moral question at stake here, viz. how we should organize society, or what is of ultimate value, or some such.

Granted, the tricky thing is to say what this further element of disagreement amounts to. I'm inclined to think it is the question of what moral viewpoint is most reasonable, or what all ideally rational agents would ultimately converge on at the end of inquiry. Depending on our theory of rationality, this might be further reduced to the question of what set of desires/evaluative beliefs is the most internally coherent, unified, and so forth. I think this is some sort of progress. At least it is difficult to re-raise the Open Question Argument at this level: "I grant that X is approved by the maximally coherent evaluative system, and indeed I would endorse it if I were more rational, but nonetheless I think X is wrong!" sounds pretty self-contradictory to me. But in some sense I've just passed the buck from meta-ethics to meta-epistemology, so this picture is still not entirely satisfactory.


  1. The reason I hesistate to agree is, how *is* a desire for piety & obedience incoherent?

    I was very impresed by Michael Smith's book when I read it which you seem to be working alone similar lines to, but from memory he never really did argue why there should be only one coherent set of beliefs for a rational agent.

    And to be honest intutiviely "Hitler had a callous disregard for other people's welfare" sounds at least more what we mean by "Hitler was evil" than "Hitler had views that did not form a maximally coherent evaluative system".

    If Haidt is vaguely right about how the human moral sense works with five foundations, it doesn't seem at all unplausible that we could make many different but coherent sets of moral codes using different combinations of foundations.

    So how do we judge between them?

  2. Well, I think it's a job for first-order normative ethics to establish whether some systematic moral theory is in fact more coherent and reasonable than its competitors. If none is, then my meta-ethical conclusion would be that we cannot "judge between them". From true Rational Pluralism, nihilism (or at least relativism) follows.

    [Though I don't think the Haidt stuff plays any essential role here. It's possible that humans are innately disposed towards certain sorts of incoherent or unjustified views, after all.]

  3. Okay - that seems fair enough, though I guess I'm more skeptical than you that a most coherent theory will be found.

    I agree that its possible that humans may be disposed to incoherent views, but I still feel we can't just cast off our evolved moral intutions altogether. Any system of morality would have to be related to one of those foundations, or I'm not sure what reason we'd have to call it a moral system at all. We may think our intuitions about "good" are flawed, but without starting from at least one, how would you even begin to construct a moral system?

    (This is assuming of course, which I am, there are no objective moral facts, queer or otherwise, floating around outside humanity).

    That's not in itself a problem, as I can't think of a single suggested moral system which didn't relate to either harm, or purity and so on; but if we're at least a little bit tied to our evolved sense of morality, and that has five possible foundations, then it does increase my skepticism of a single coherent solution.

  4. The move to meta-epistemology seems pretty important to me. If we emphasize the apparently normative terms like ‘rational’, then we enter familiar territory: wondering whether these terms e.g. refer to Moorean non-natural properties of rationalness, or what. If we emphasize the apparently non-normative terms like ‘maximally coherent’, then the Open Question Argument seems to retain whatever force it originally had.

    Plus there are plenty of doubts (I think) to be had about convergence. Maybe with normal human beings, but even with rational agents lacking our affective profile (creatures similar to sociopaths)? It looks doubtful.


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.)