Do you want to embarrass a subjectivist about practical reason or well-being? C’mon, you know you do. Well then, all you have to do is ask them if the heroes they so eagerly cite as motivating and explicating and defending subjectivist accounts are themselves subjectivists. The answer is that often they are not. Hume seems, as is being gleefully pointed out frequently these days, to have been a complete skeptic about normative reasons. Recall he held that even acts and passions mislead by false factual beliefs still lack representative content as so still cannot be comfortable to or contrary with reason, except in an improper way of speaking.
Bernard Williams defended internalism, the view that a necessary condition on having a pro tanto normative reason to O is that one would be motivated to O after sound deliberation. But subjectivism, at least as I understand the view, is a claim about what makes it the case that we have reasons, not merely an account of what tracks true reason claims. Thus Williams’ internalism does not state a distinctively subjectivist position (albeit, when pressed he defends the view in ways that I think are clearly subjectivist friendly). A non-subjectivist could accept Williams’ internalism and yet hold that it is not an agent’s contingent desires that make it the case that she has a reason to O (even if they are a necessary condition on having a reason to O). Indeed, Korsgaard explicitly does so.
Peter Railton’s “full information” account of an agent’s good is, I would submit, despite significant criticism, still the state of the art subjectivist-friendly account. (David Lewis defends a somewhat similar account in a subjectivist-friendly way and Michael Smith adopts features of such a view in his objectivist account of practical reasons). But Railton disowns the subjectivist interpretation of his own view. He writes:
I propose to say that what makes some or other end or activity be part of an individual’s good is not the fact that he would, were he ideally informed (and so on), desire that his actual self pursue it, but rather the existence of the reduction basis for that counterfactual, namely, the particular constellation of law-governed features of the actual individual and his circumstances in virtue of which these claims about idealized hypothetical desires hold. Thus, the truth-condition of the claim that such-and-such is good for a given individual is directly given by the existence of this constellation of features, without detour through idealized desires. We may, then take an individual’s desires, as they approach idealization in the limit, to be indicators of his good, of the presence of the sort of fit discussed previously between an individual and an end or activity. (“Facts and Values,” p. 62-3. See also “Moral Realism,” p. 12 and note 17. Page numbers are to the articles as they appear in Railton’s Facts, Values, and Norms, Cambridge, 2003).
Thus, Railton’s view is that it is not the informed desires that make it the case that something is part of one’s good, but rather than one’s informed desires for 0 are explained by a set of complex, relational, dispositional facts which themselves are the truth-maker for facts about what is part of one’s good. And because of this, I think it right to say that Railton’s view is not subjectivist.
I want to attempt to begin to put pressure on such a view—pressure that would tend to show that Railton should have embraced the subjectivist interpretation of his view. This is usually the best that subjectivists can do with their would-be heroes; namely argue that the would-be hero was doing great until the unjustified move which kept them from being a subjectivist proper.
I will offer two lines of thought aimed at suggesting that Railton should have embraced the subjectivist interpretation of his view, but it is really the second line that I am interested in thinking about here.
The first thought is that Railton has no account of how to characterize the reduction basis of the agent’s idealized desires except in that way. That is, there is no independent account of what the truth-maker of a person’s good claims is, on Railton’s story, except whatever natural facts serve as the reduction basis of the idealized agent’s wants for the non-idealized agent to want. But then it is hard to see how the dependent facts could explain the independent facts.
The second thought is that either the desire is too distinct from its physical causes or it is identical to it. My claim will be that in either case it more plausible to hold that it is desires that ground claims about what is good for a person than to hold that the desires merely track, without grounding, such claims. Consider first the first option. Suppose that, metaphysically, one could have the latter physical facts without the former desire facts (think of Chalmers’ and Levine’s arguments that there seems to be an explanatory gap between the physical and the mental). In this case it would seem that the physical, absent the desire, would not ground claims about what was in a person’s good. Consider a simple case of a matter of mere taste. Suppose A+ desires that A desire diet Coke rather than diet Pepsi. Suppose too that the physical structure of A and A+ could be in place and yet such a desire not exist, indeed perhaps such a physical structure could be in place in a zombie with no mental life at all. Would we still, in such a case, say that the presence of the physical facts, absent the desire facts, grounds claims about what is good for this individual? Presumably not. Presumably we think a thing can only have a good if it has some kind of mental life (or, at a minimum, that some aspects of our good are due to our having a mental life). In any case we think that what grounds the reason to drink diet Coke over diet Pepsi is that one likes the one more than the other and no physical facts could ground this reason in the absence of the desire facts. The moral: if the desire is too distinct from the physical facts such that one can pull them apart modally, then it is the desire, rather than the physical facts that seem to have the better claim to ground the reason.
Alternatively, suppose one could not have the relevant physical facts without also having the desire facts. That is, suppose the former metaphysically determine the latter. Why might this be so? Well the most likely answer is that the desire-based facts just are identical to some complicated physical facts. And if this were so then there could be no competition between the desire-based facts and the physical facts to explain and ground claims about a person’s good. Just as there can be no competition between “its being water” explaining certain things and “its being H2O” explaining certain things, in this case too there is no competition between the desire and the physical structure in grounding claims about a person’s good.
I do not see this problem for Railton’s non-subjectivist interpretation of his own view as decisive. I am sure one could find fancy moves borrowed from the philosophy of mind that could save the objectivist interpretation Railton prefers. So my hopes for my “argument” are limited. I hope to have made a case that Railton’s objectivist interpretation of his view was optional and undermotivated and is, at least at first sight, the less intuitively plausible interpretation than the subjectivist interpretation. If we end up thinking the objectivist interpretation of his view puts the view in its most attractive light we will need to be shown how the desire can plausibly be seen as neither too close to the reduction base nor too far away from it as well. The subjectivist interpretation seems to have an easier time of it here.