DARWALL'S EUTHYPHRO PROBLEM
If there is an odd duck among the current accounts of the nature of welfare or well-being, Stephen Darwall's rational care theory has got to be it. For Darwall, to say that something -- e.g., befriending Huang -- is "part of my welfare" is not to say that it fulfills my (fully informed) desires, nor that it gives me more pleasure and less pain than otherwise. Rather, it is to say that befriending Huang is "something that it would be rational to want for [me] for [my] sake" (2002). If this were merely a corollary to the claim that the friendship would be good for me, then it would be relatively uncontroversial: many, I think, would accept that the phrases "what you rationally want for me insofar as you care about me" and "my good" are coextensive. But of course, Darwall presents rational care not as a corollary, but as an analysis of the concept of welfare itself. We explain what it means for X to be good for me by saying that it is what another would rationally want for me for my sake. And this is precisely what makes Darwall's theory so odd, for it seems to have the order of explanation reversed. Normally, we tend to think, another caring person would want me to befriend Huang because it is good for me. But for Darwall, it is only good for me because a caring person would want it (rationally, and for my sake). Is his view defensible? For my own reasons, I am going to jump headlong into the tiny Euthyphro faction in welfare ethics, and argue that it is. We don't care because it's good. Rather, it's good because we care.

Darwall himself has something to say about this problem, but much to the chagrin of his reviewers (e.g., Brad Hooker, Christopher Heathwood), his discussion of it seems cursory in light of the fundamental importance of the problem. For starters, he makes the compelling point that "caring for Jane" cannot possibly be so straight forward a matter as desiring Jane's good, even intrinsically. I might well have an intrinsic desire for her good and yet fail to have the right attitudinal stance toward her. To truly qualify as someone who "cares for Jane," I should not just want her good, but I should also have an intrinsic concern for Jane herself. This helps us to see the difference between wanting X for Jane, and wanting X for Jane for her sake. Accordingly, we don't care for Jane simply because we want her to acquire her good -- we must also have the proper appreciation for her as a person, appreciation which serves as a reason for our care.

Of course, this only shows what's wrong with the claim that we who care about Jane want X for her solely because it is good for her. It could still nevertheless be the case that X's being good for her is a necessary condition -- one among others -- for its being worthy of rational care. And if this is the case, then the concept of rational care still presupposes the concept of welfare, and not the other way around. Darwall's answer (again, easy to miss) is to suggest that care is "something like a psychological natural kind," and therefore that it is not necessary to provide it with a definition: "Just as we can use a term like 'water’ without a prior definition to refer to the natural stuff in the rivers and lakes for purposes of an empirical theory, so likewise might we refer to care for purposes of a metaethical theory of welfare if it is a natural kind."

I suspect this strikes most readers as slippery, especially when we consider Darwall's paradigmatic case of the care as a psychological natural kind: sympathetic concern. When we sympathize with the victim of some kind of tragedy, surely we are responding to reasons for considering the tragedy bad for the person in question. If she has lost her family, we know that her life will always seem somewhat emptier to her, that she won't be able to take as much comfort in the happy illusion of unconditional love as the rest of us do, and that she is likely to undergo years of grief. If she has lost her livelihood, then we know that she will be deprived of a way of life in which she was deeply invested, and will have to reconcile herself to another. These aren't just reasons for sympathizing with her, these are reasons to think that her life has gone poorly for her. What, then, does sympathetic concern add to our concept of welfare?

I would like to suggest that we in the Euthyphro faction could, despite appearances, have it both ways here. We could concede the point that care is responsive to reasons for thinking something good or bad for a person, and yet still maintain that it adds something to our concept of welfare to think of it in his way. And we could do this, I believe, if we are pluralists about the sorts of reasons for thinking something good or bad for someone. That is, we might insist that welfare is a property shared by states of affairs of very different kinds, so that their worthiness for rational sympathy is the only consistent explanation we can offer.

This view is lent some initial plausibility, at the very least, by the intractable sorts of problems that tend to plague more traditional accounts of the nature of welfare. If we are hedonists about welfare, for example, then we are likely to have to bite the bullet on experience machine scenarios, and we will have a difficult time accounting for the sense that a person can be made better or worse when her plans for her children succeed posthumously (my intuitions about these examples are close to Chad's, in his June 1st entry below). On the other hand, if we are proponents of the informed desire view, we are going to have trouble accounting for goods that we do not desire, even rationally and with full information, or informed desires for things (such as world peace in the distant future, etc.) that do not make significant contributions to our well-being. What best captures the common feature of the diverse constituents of well-being -- whether they be posthumous goods, happiness goods, goods of which the benefactor is aware, or goods of which she is not -- is that they are all what a sympathetic observer would rationally want for her. But this is not to deny the view that our caring or sympathizing should be responsive to features that we normally understand as being constituents of well-being. In fact, it is to embrace it.

To defend the rational care theory in this way is to take a page from the unlikely playbook of Thomas Scanlon -- not because he upholds the rational care theory of welfare (he certain doesn't), but because he employs a strategy like this one against a similar "order of explanation" problem in moral contractualism. But to take this line is also to nudge our Euthyphro faction a bit closer to the Confucian philosopher Dai Zhen, who thought that we could never entirely capture what it is that makes something good for us without reference to human deliberation, and a certain kind of sympathetic deliberation in particular. The informed desire theorists already allow that welfare terms contain an ineliminable reference to human deliberation, insofar as they maintain that a person's good is what she would rationally want for herself if she were fully informed about the probability of satisfaction, opportunity costs, etc. The next question for welfare theory, then, is why the deliberation in question must be self-reflective. There are good, intuitive reasons, I believe, to think otherwise.
- posted by Justin Tiwald @ Saturday, June 11, 2005