Simon Keller has an interesting paper called Self-Effacement in Ethical Theory (PDF). It is a common claim that utilitarianisms (or consequentialisms as they are usually called -- somewhat oddly, I always think, because 'utilitarianism' was originally their own label, whereas originally 'consequentialism' was given by opponents as an insult) are self-effacing. Keller defines self-effacement as "if the considerations that it posits in telling that story are not considerations that always should serve as motives for action, according to the theory itself." I don't think this is the best way to do it, because any sufficiently comprehensive ethical theory will discuss lots of issues that no one thinks should be relevant to motivation; but if it is understood narrowly to encompass only considerations that make a moral action good, it's basically right. To say that utilitarianism is self-effacing is to say that what it says makes an action right often can't be a motive for doing that action. So, in other words, being motivated to visit a friend in the hospital simply because it increases the amount of pleasure in the world and reduces the amount of pain is a bizarre and (even most utilitarians would admit) self-defeating motivation. So it's usually held, even by consequentialists, that consequentialism is, or can be, self-effacing; and consequentialists, of course, argue that this is not a bad thing (and sometimes that this is a good thing). It is sometimes argued that deontologism, the view that what makes an action right is its being according to a rule, is also self-effacing; and some virtue theorists argue that virtue theory is not self-effacing, and that, therefore, this is an advantage it has over its rivals. Keller argues that this is not true; not only is virtue theory self-effacing, it is especially problematic in the way it is self-effacing.
His argument is this. Suppose we have three friends, Arthur, Benjamin, and Christine, who have rented the only cabin on a campground. While there, they notice a family of hikers who have been stuck in the rain, who are struggling to put up a tent and who can't light a fire in order to cook any food. So they invite the family in. Each of the three friends has a different motivation, however. Arthur's motivation is "to help out the hikers and relieve their misery." Benjamin's is "to act generously". Christine's is "to do what the fully virtuous person would do." And Arthur, Keller supposes, clearly has the best motivation. It's not that the others have bad motivations -- they are both sincere, and they both clearly have good motivations; but Keller thinks Arthur clearly has the best motivation.
One of the problems with Keller's argument is that it is not at all clear that it is so. If we take the bare motivations, without Keller's commentary (which is arguably tendentious), it isn't clear how their motivations differ in any significant way, at least for this particular case. In sincerely being motivated to act generously, Benjamin would know that this involves helping out the hikers and relieving their misery (otherwise, we'd have no reason to think it would motivate him to this particular action). In sincerely being motivated to act as a fully virtuous person would, Christine knows that this involves helping out the hikers and relieving their misery (for the same reason). So the motivation really appears identical as far as this action is concerned; it's the same motivation under three different descriptions.
Now, it is true that the descriptions are different. But what makes them differ is not what they are motivating someone to do in this case -- which is the same -- but how they place this within a project of moral living. Arthur's motivation, 'to help out the hikers and relieve their misery' has no implications for broader moral life; it is entirely focused on this particular act. Benjamin's motivation, 'to act generously', on the other hand, implies an entire life-project of acting generously, which involves (in cases like the one at hand) helping people and relieving their misery. It is a more sophisticated motivation because, although it doesn't really differ for this particular case, it has implications for Benjamin's wider life. And Christine's motivation, to be a fully virtuous person, is much the same, except that it is one level more sophisticated. It implies an entire life-project of acting virtuously, which includes the entire life-project of acting generously, and includes helping out the hikers and relieving their misery in this particular case. (I have used 'more sophisticated' rather than 'more moral' deliberately. I see no reason to think any of the three friends as having a 'more moral' motivation in this case, because, as I said, the only difference in their motivations is how the immediate motivation, to help out the hikers, relates to their broader, more enduring, motivations. I don't think it is necessary to have the most sophisticated moral motivation in order to be as fully moral as one could demand; what really matters is not sophistication of motivation but consistency in motivation -- but there my virtue-theory inclinations are showing.)
The flaw in Keller's argument is that he assumes that Christine must hold the view that she should act virtuously rather than the view that she should help out the hikers and relieve their misery. But, of course, part of a virtue-theorist's whole point is that not only is the desire to be virtuous a good motivation, it is a meta-motivation that can (and must) subsume other moral motivations. Virtue theory allows for a more powerful and sophisticated theory of moral motivations, and if there's any real problem with self-effacement, it's that it is a sign of weakness in a theory of moral motivation -- the morality of an act and the moral motivation for it become somehow disconnected.
The reason that the desire to be virtuous must subsume other motivations is that the desire to be virtuous is a motivation directed not to this or that particular action, but to one's whole life. It is a motivation that unifies a lot of other motivations -- the motivation to be generous, for instance -- that in turn unify particular-directed moral motivations across time. This is not self-effacement; nothing about this says that acting from the motivation to be virtuous is inconsistent with acting from the motivation to help this particular person in this particular place. In fact, it is explicitly opposite: the motivation to be virtuous is consistent with the motivation to do anything virtuously.
Keller is right that if Christine's motivation to be virtuous does not involve a motivation to help these particular people here and now, that Christine's motivation is not as good as Arthur's. It is possible for such cases to arise. But this is not because the motivation to be virtuous is in general defective in this way, but only because in Christine's case the motivation is only defectively realized -- she has only a defective version of an excellent motivation. Arthur's motivation would be better not because his motivations do not include the motivation to be virtuous (as Keller suggests), but because he has a motivation essential to being a virtuous person in this case, and Christine (by the supposition we are considering here) does not. Christine has a good ultimate end in view; she just doesn't have in view the proximate end that the ultimate end requires in this particular case. Again, we can see that there is no need for self-effacement here; Christine's motivation is incomplete, and self-effacement can only be required where the complete motivation acn't involve the considerations of the ethical theory in question. So even on Keller's assumption we can't get as far as Keller seems to think.
The thing about virtue theory, then, is that it does not require that one act out of the motivation to be virtuous. But this is not self-effacement; self-effacement could only occur if there were some cases in which it would not be virtuous to act out of the motivation to be virtuous. Virtue theorists aren't committed to this, or, at least, Keller does not show it. What he shows is that virtue theory does not require the high-level motivation to be a virtuous person. Precisely because it doesn't, the lack of it in a clearly moral motivation is not a problem for virtue theorists unless it can be shown that being virtuous would require that lack. Keller thinks this is the case, but his reasoning depends on a faulty assumption about what moral motivation in virtue theory would have to involve (the reasoning by Hurka that Keller mentions makes a similar mistake). I don't think it can be shown that the motivation to be virtuous is ever inconsistent with the motivation to do this particular virtue-necessary thing, and this appears, at first glance to be a necessary truth. Self-effacement appears to arise only when an ethical theory has too restrictive an account of moral motivation, one that occasionally puts local motivation in conflict with global motivation; and virtue theory's advantage is precisely that it isn't restrictive in this way at all, because what it is really doing is placing moral action into the project of a complete and ongoing moral life. And the motivation to have a complete moral life is never inconsistent with the motivation to act in this particular (moral) way in this particular case, because it is meta-motivation capable of including the particular motivation. Other views, on the other hand, are self-effacing (when they are) because they flatten out motivational structures and treat motivations as all governed by the same sort of principle. While a virtue theorist might possibly have an account that does this (and I wouldn't be surprised if there are virtue theorist who do try to do this), virtue theory itself does not require it, and in fact sits more comfortably with a more sophisticated account of motivation. And this is really what Keller's argument shows.
So it seems fairly clear that virtue theories are not self-effacing in the way Keller suggests. Now, I'm not convinced that self-effacement is really that much of a problem for anyone. While there are clearly bad motivations, I see no reason why good motivations always have to be the same, and, as I said, if the motivations just differ as to their sophistication, I don't think this makes much difference. Christine's more encompassing motivation isn't a more moral motivation than Benjamin's motivation, and neither are a more moral motivation than Arthur's, despite the fact that Arthur has the least sophisticated motivation. Arthur has the essential motivation, as do Benjamin and Christine, and that's what primarily counts. The advantage Benjamin and Christine have over Arthur is that more sophisticated motivations make it easier to be consistently moral. And I think deontologists, for instance, and perhaps even some consequentialists, can easily avail themselves of an argument similar to the one I have given (although it definitely would depend on the particular view). What self-effacement issues really show is not (contrary to some virtue theorists) that virtue theory is automatically the better account of particular moral motivations, but that virtue theory recognizes a broader range of sophistication in possible moral motivations. This is certainly an advantage, as far as moral psychology goes, but it doesn't automatically show that self-effacing ethical theories are problematic in themselves.
