Sunday, October 19, 2008

Reflecting on Relativism

I previously considered a kind of sophisticated moral relativism according to which 'X is wrong' is true for you iff your idealized self would hold this position. But to truly count as relativism, this must apply even at the level of fully specified token acts. So (for example) we can say that Sally acted wrongly in getting an abortion is true for Anne but not for Sally, if their idealized selves would disagree in this way about the merits of Sally's action. But I wonder: could Anne's idealized self coherently maintain her position, or would the truth of relativism undermine her first-order judgments of impermissibility?

If an act is morally wrong then the actor lacks sufficient moral reason (and arguably sufficient non-moral reason either) to so act. But can it really be a 'relative' matter what (moral) reasons Sally has for acting? Surely the relevant standpoint here is Sally's own. She has (moral) reason to act as her idealized self would (morally) recommend, and - ex hypothesi - her idealized self recommends abortion. So Sally has sufficient reason to get an abortion. So it cannot be morally wrong for her to do this.

At what step can Anne disagree? Anne's idealized self opposes abortion, but that is merely a reason for Anne to not get an abortion; it is not a reason that is relevant to Sally's action. Anne can see that Sally is perfectly justified in her beliefs and actions. So she can't maintain the claim that Sally's act was wrong; all Anne can claim is that it would be wrong for her to perform a similar act of that type. Sally, seeing Anne's standpoint, will agree: it would be wrong for Anne to get an abortion. But now there is no disagreement over token actions. So the moral truth isn't 'relative' after all: it's absolutely true that Sally may get an abortion and that Anne may not.

You might dispute my assumption that what practical reasons S has is a non-relative matter, determined by S's standpoint and no other. But even if you dispute that, so that Anne may (as yet) hold Sally's actions to be unjustified, Anne presumably cannot deny that Sally has rational beliefs (insofar as they would survive idealization). But now we can appeal to Clayton's principle for moving from Theoretical to Practical Reason:
(1) If a subject judges that she should Φ and it’s not the case that she should refrain from judging that she should Φ, it’s not the case that the subject shouldn’t Φ.

Anne must grant Sally the antecedent -- Sally judges that she should have an abortion, and makes no mistake in doing so -- so Anne must also grant the practical consequent, that Sally didn't act wrongly.

We thus find that moral relativism is incoherent. On the assumption that relativism is true and applies to some token act, it turns out that it doesn't apply to that token act after all. Contra the assumption, moral status is perfectly 'absolute': if the agent commits no (theoretical) error in thinking their action permissible, then it's permissible. Since the former is a non-relative matter of fact, so must be the entailed consequent.


  1. That's a really interesting argument! Is the fact that Sally's idealised self believes that Sally should phi meant to entail that Anne('s idealised self) accepts that it's not the case that Sally should refrain from judging that Sally should phi? I guess I'd like to know what you mean by 'idealised self'.

    I'm just wondering whether one could make a relativist move about theoretical reason. Could a relativist about theoretical reason claim it might be false for Anne that it is not the case that Sally should refrain from judging that she should phi (although this is true for Sally)? - and thus restore consistency to the moral relativist position.

  2. Interesting argument!

    I would have thought that the relativist would deny (1), and instead affirm a version with individually-indexed shoulds.

    (1*) If a subject, i, judges that she should_i Φ and it’s not the case that she should refrain from judging that she should_i Φ, it’s not the case that the subject shouldn’t_i Φ.

    This would seem to leave open the possibility that she shouldn't_j Φ (if a subject j judges that i shouldn't_j Φ, and it's not the case that j should refrain from judging that i shouldn't_j Φ).

  3. Conchis - it's important (for the relativist) that Anne and Sally both mean the same thing by 'should', otherwise there is no common claim to which they ascribe different truth values. (See my old post on 'relative truth and disagreement' for more detail.)

    Gabe - interesting suggestion. I guess extending their relativism even to theoretical reason is the relativist's last hope. But it doesn't seem very plausible. My old post 'Why Disagree (Relatively)?' raises some problems here. For example, Anne will acknowledge that Sally's moral beliefs are true for Sally, and so (in this sense) 'appropriate' given Sally's standpoint. Sally's beliefs are fully 'idealized' -- she suffers no ignorance or cognitive defect. So on what grounds could she possibly be open to rational criticism? Can Sally really be required -- even by Anne -- to instead have false (for her) beliefs?

    Re: idealization, we can say that one's idealized self has all relevant factual knowledge, and unlimited logical/cognitive capabilities, so that she is perfectly able to avoid any inconsistency or incoherence in her beliefs, etc.

  4. It seems to me that that consideration against the theoretical relativist is similar to your initial consideration against the moral one - Anne agrees that from Sally's standpoint she is doing what she should, and believing what she should, so on what grounds is she open to criticism? Perhaps the intuition is stronger with regard to theoretical reason, so Clayton's link (assuming the relativist doesn't deny it) gives us a stronger consideration against the moral relativist.

    The argument in the other post seems to appeal to this too - When Bob persuades Anne to reject P, Anne ends up rejecting Pa, because she accepts (P iff Pa). But, from Bob's standpoint, (P iff Pa) is false. So Anne is getting at her false belief because she is reasoning from a false premise. The issue is whether Bob should allow that (P iff Pa) is appropriate for Anne, because it is true from her standpoint.

    Re idealisation - I was wondering whether it imported some normative constraints (e.g. accepting everything she rationally should) which might rule out disagreement (Anne wouldn't accept that Sally's idealised self would accept anything that - from Anne's standpoint - one shouldn't), but would make the definition of relativism circular.

  5. I'm puzzled by this part:

    Anne's idealized self opposes abortion, but that is merely a reason for Anne to not get an abortion; it is not a reason that is relevant to Sally's action.

    Why wouldn't Anne's idealized self be opposed to Sally's abortion as well as her own? Anne appears to have reason to try to prevent Sally's abortion (if not necessarily decisive reason).

    I think the problem here is an implicit assumption of moral absolutism. Of course, if moral absolutism is true, moral relativism is going to look incoherent at some point.

    Clayton's principle is probably valid in the context of an individual with personal goals. However, if Sally's goals are not Anne's goals, then Anne can acknowledge the validity of Sally's reasoning, but reject the conclusion because the facts (i.e., the goals) were wrong. (I think this might be what Gabe just said.) Clayton's principle is not portable from one individual to another unless individual goals are coincident.

    If goals are absolute, then you have your conclusion of incoherence, but if personal goals are arbitrary, then the argument doesn't work (even if Clayton's principle still holds at a personal level).

  6. "Why wouldn't Anne's idealized self be opposed to Sally's abortion as well as her own?"

    She might, but only a non-cognitivist would think that relevant. I'm talking about truth relativism, i.e. a cognitivist view, so the question is instead whether Anne's idealized self would believe that Sally is doing something wrong. My post argues that she could not maintain this claim. (She might oppose the abortion anyway, but that's a non-cognitive stance, not an ethical claim.)

    "Anne appears to have reason to try to prevent Sally's abortion (if not necessarily decisive reason)."

    Perhaps, but again that's simply a different question. Here you are talking about reasons for Anne to act, not reasons for Sally to act.

    It's not entirely clear what you mean by saying that "Clayton's principle is not portable" (for each individual it is true of themselves but not of anyone else?). And it does not say anything about goals (it's instead about the relation between what one should believe and what one should do), so some clarification is needed before I can make sense of the rest of your comment.

  7. Gabe - "Perhaps the intuition is stronger with regard to theoretical reason"

    Yeah, I guess that's what I was relying on. It just seems overwhelmingly plausible that it's appropriate for people to believe the things that are really true for their context of assessment. To do otherwise would involve a kind of mistake on their part.

    My friend Jack sometimes illustrates this with the example of non-actual contexts of assessment. The relativist thinks we each live in our own 'world', in some sense. So consider our ordinary judgments of appropriateness as they relate to people in merely possible worlds. Consider Patty of the purple-grass world. She believes that grass is purple. That seems entirely appropriate, given that grass is purple in her context of assessment. But strictly speaking -- i.e. from my context of assessment -- Patty believes something that is false. Grass isn't purple, it's green. But we can see this claim of 'falsity' is kind of irrelevant to judging Patty's belief. Whether her belief is appropriate or 'correct' in the relevant sense depends not on whether it is true -- i.e. as I assess things -- but whether it is true in her world. If moral relativism is true, and we each live in our own 'moral worlds' in some sense, I don't see why this principle shouldn't carry over.

  8. The relativist can simply question the notion that there is only one idealized self; it seems likely that the idealized self is indeterminate because we all have conflicting beliefs and desires, so any attempt to remove 'any inconsistency or incoherence in her beliefs' must choose which belong in the idealized Sally. This means that the determination of the idealized self is relative, as is whether 'X is wrong' is true of Sally.

  9. Relative to who? I think I'd be more inclined to describe that as a situation in which it is simply indeterminate whether 'X is wrong' is true for Sally.

  10. Yes, this connection between indeterminacy and relativity is not self-evident. What I was really trying to do was express the following intuition:

    Anne would judge the nature of Sally's idealized self differently to Sally - for instance, because she thinks Sally's idealized self would give priority to her concerns about whether the foetus is a person, whereas Sally thinks her idealized self would prioritize her own interests - so the determination of the idealized self is relative to the person carrying out this determination.

  11. But if there really isn't any determinate fact of the matter about the nature of Sally's idealized self, then Anne's and Sally's judgments here are baseless. They should recognize that the other person's judgment is just as likely to be true as their own. So I don't see the basis for saying that their belief is really true ("for them") rather than just being an unjustified belief concerning an inherently indeterminate matter.

    (I hesitate to say that their beliefs are outright false; "not determinately true" may be a more precise designation.)

    After all, what is this "truth" supposed to consist in? I was proposing a view on which truth consisted in what one would believe "at the [idealized] end of inquiry". But if it's indeterminate what you would believe about P at the end of inquiry, then this view implies that it's indeterminate whether P is true. So your relativist is going to have to propose a very different view from the one I was discussing here -- most likely you'd have to revert to the unbearably crude view that collapses truth into actual belief (which of course has its own problems). My argument would then be that there is no way for a sophisticated moral relativism to work.


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.