Some time ago I argued that a problem with a certain sort of virtue ethics might not be fixable (in “Virtue and Right”). The banner ‘virtue ethics’ covers a variety of views united by not much more than the thought that ethics is in some way or other best approached through the idea of good character and virtue. My arguments set aside many of those views, especially those that, while giving pride of place to virtue, adopt a ‘no theory’ or ‘anti-theory’ stance toward the right. If you are anti-theory, then my train didn’t leave the station. Indeed, perhaps the best version of virtue ethics is one that turns its back on a theory of the right. But the sorts of views I was interested in thinking about were those that aspire to such a theory in order to create an alternative to deontological or consequentialist theories of right action.
Such theories accept some doctrine of the form Rosalind Hursthouse at one time held:
(V) An action A is right for S in circumstances C iff a completely virtuous version of S would characteristically A in C.
V, as I understood it, is a claim about what makes right acts right: they are what a good person would do. It doesn’t necessarily offer an epistemic route to the right or a decision procedure. The problem with V is that it cannot tell us what makes those right actions right that are right precisely because the agent is not completely virtuous or is in some way suboptimal. Those are the actions a good person wouldn’t do. Many of these actions are self-improving actions. The completely virtuous wouldn’t waste their time on self-improvement, even though the rest of us might be doing the right thing by doing so. Self-improvement is a topic in which I happen to have a particular interest, but not all problems with V stem from this. My discussion of V focused on a variety of ways one might fix V, and I concluded that, given the options I could think of, it probably couldn’t be fixed.
I think it’s fair to say that there have since been many imaginative and interesting attempts to fix V. I won’t run through the rationale for V, why ‘completely’ is in there, and so on, nor will I try to cover every reason people may remain un-persuaded. That would take many posts. But there are a few things I might say about some of these attempts.
Before I do, one option might be to treat V in normative theory as analogous to the role idealizations play in, say, economics or decision theory. V might tell us something worth knowing about right and wrong even if it can’t tell us everything about right and wrong, much as theories of idealized utility maximizers with perfect information might tell us something about economics, but not everything about economics. I won’t explore that suggestion here however. I only mention it as a possible approach that one of my colleagues (Paul Weirich) suggested to me.
Some of the tools to fix V that originally seemed to me to be at hand I thought wouldn’t work. An ‘ideal advisor’ move wouldn’t because, for instance, there’s no reason to think being virtuous brings with it skill at advising. Another move I considered was different standards for those who are not fully virtuous. Among the things I disliked about that option was that it makes self-improving actions second-class citizens among the right and the good. Improving oneself can, at least sometimes, be as noble an act as any. So I argued that a ‘different standards’ view would not be satisfactory.
However, one might well think an approach like this can yet be defended. Among the many important ideas in Julia Annas’ 2004 APA Presidential Address she suggested that learning a virtue is gradual and piecemeal, and that this can provide a kind of ‘gradual and piecemeal’ learner’s standard of action.
There are a couple of reasons why I would resist saying this. First, while it might be plausible to apply this to some skills, it is not where precision is important. If the skill is neurosurgery or disarming nuclear weapons, ‘gradual’ and ‘piecemeal’ don’t apply. Learners in such cases act wrongly until they get it right. This seems to be how Aristotle himself thought about ethics:
There are many ways to be in error—for badness is proper to the indeterminate, as the Pythagoreans pictured it, and to the determinate. But there is only one way to be correct. That is why error is easy and correctness is difficult… since it is easy to miss the target and difficult to hit it… ‘for we are noble in only one way, but bad in all sorts of ways’. EN 1106b30—35
Another worry I have with a proposal such as this is that, given the restriction on my topic, it would have to be turned into a theory of right. The way I suppose this would be done is this: two ideals of character, ‘fully virtuous’ and ‘learner-virtuous’. That would yield two theories of right acts, viz., RightF and RightL and the following additional doctrine:
(L) A is RightL for S in circumstances CL iff a learner-virtuous agent would characteristically A in CL.
The problem is that poor students ought to become better (if you are anything like me as a student!). I might lack ‘learner-virtue’, and if so, I ought to acquire it, for example, by applying myself more, paying better attention, and so on. The ‘learner-virtuous’ agent would not characteristically need to do such things. She’s an excellent student and wouldn’t waste time on that. No need to consider the absurd further iteration, ‘learner-learner-virtue’—virtue in learning how to learn how to be virtuous.
It seems that what is called for is a revision of V. There are many possible (various attitudes of the completely virtuous, such as approval, for instance). Consider, however, Valerie Tiberius’ (Philosophical Papers 2006):
(R) A’s φing in c is right iff φing is the action in accordance with the reasons that would guide the action of a completely virtuous person in c.
The idea behind R is that the right conforms to a virtuous person’s reasons. The proposal has its appeal but nevertheless I don’t think it will succeed. If R were taken to mean that the right action is the action that someone guided by virtuous reasons would do, it would not explain cases in which the right thing to do is to make oneself into a person who is guided by virtuous reasons. So R must just say that some actions conform to the reasons guiding the virtuous and others do not, and the ones that do are the right ones.
But a completely virtuous person is not guided by reasons to improve herself, since there are no reasons for her to do so. So self-improving actions can’t be in accordance with such reasons. How, then, can improving oneself be the right thing to do? True, the completely virtuous respond to reasons to be, for instance, compassionate. But would improving myself really be acting in accordance with those other reasons? Naturally we could make the virtuous reasons really general (‘it would be good to do’), but then the theory looks much less appealing compared to the more definite accounts provided by consequentialist or deontological views.
Sean McAleer (JESP, Aug. 2010) offers some imaginative proposals. The first is, rather than defining right in terms of virtue, define wrong in terms of vice:
(VICE) A’s φ-ing in circumstances c is wrong iff a fully vicious agent, acting characteristically, would φ in c.
But suppose A is not vicious. It would be wrong for A to cultivate vice in himself. Yet a fully vicious person would not characteristically try to coarsen his own character. Why waste time on that when there are countless depraved and cruel actions waiting to be done? So it is wrong for A to try to coarsen his own character though the vicious would not characteristically do this.
A second draws on Julia Driver’s idea of “suberogatory” actions, actions that are bad but not wrong. Perhaps not giving a kidney to my child when I can and he needs it is bad, but not wrong. If there can be bad-but-not-wrong actions, why not good-but-wrong or good-but-not-right actions—‘countererogatory’ actions? A selfish billionaire gives $100 to Oxfam. That’s good-but-not-right. She should have given many times that amount.
This is, I believe, just a “different standards” proposal. But even if not, it is implausible. Making yourself a better person is good but not right? C’mon man. More generally, that an agent is less than perfect is no reason by itself to think the actions performed by that agent are also less than perfect. Indeed, there don’t seem to be grounds other than the fact that a fully virtuous person would not characteristically do them for regarding acts of self-improvement as morally lesser than, or not as fully ‘right’ as, other virtuous acts. I suppose sometimes self-improvement doesn’t inspire admiration, for instance, when the person is struggling against or making up for flaws for which he himself is responsible. But actions appropriate for the suboptimal and the like are not limited to such actions. Otherwise harmless people can do better, and it’s not wrong in any sense of that term when they do.
What about Mengzi’s ethics:
(M) A’s φing in c is right if and only if A’s φing in c expresses benevolence (ren), righteousness (yi), ritual propriety (li), or wisdom (zhi).
It’s a stretch to say that what makes cultivating benevolence in oneself the right thing to do is that it expresses benevolence. If you express benevolence, why do you need to cultivate it? I don’t see why it needs to express any of the other attitudes to be right either.
At bottom, the problem is the term ‘express’. Either it means ‘actually produced by’ or ‘characteristic of’. If ‘actually produced by’, then
(M1) A’s φing in c is right if and only if, A is inter alia actually benevolent and A’s benevolence actually causes (motivates, etc.) A to φ in c.
M1 appears to make it impossible to do the right thing for the wrong reason, though some (not me) may be willing to take that idea on board. Still, people who lack these attitudes couldn’t do the right thing. Indeed, very few people are fully benevolent, etc., so very few actions would be right. I suppose that’s possible; maybe doing the right thing is much harder than we think. But it doesn’t seem very plausible to me. ‘Expresses some benevolence, etc.’? OK, if you want a very weak conception of ‘morally right’. I don’t.
Suppose, then “expresses” means “characteristic of”. Then,
(M2) A’s φing in c is right if and only if, an inter alia benevolent person would characteristically φ in c.
Suppose A lacks benevolence. It would be right for him to cultivate it. But that is not what a benevolent person would do—they’re already all full up with benevolence. What about cultivating benevolence out of whichever of the other attitudes one has? Cultivating benevolence will not always be wise, nor righteous, and so on. And, even so, there’s still the case in which one ought to cultivate all of them.
(P) Agent A’s φ-ing in circumstances c is right iff A’s φ-ing respects, promotes, or expresses the virtue most relevant to c.
‘Respect’ and ‘express’ need more explanation. If understood as ‘characteristic of’, the objections to the last proposal apply. But the root difficulty is that this appears to be an ad hoc list of relationships to virtue meant to cover circumstances otherwise bothersome for the proposed standard of rightness. ‘Promote’ is there seemingly to take care of promoting virtue in oneself. If an act doesn’t promote virtue, maybe it respects it, or if not that, then perhaps it expresses it. What unites respect, promotion and expression, other than their being added on to deal with counterexamples?
As I said, there may be other reasons some remain un-persuaded by these worries about V. But to those who want to hold on to some version of V, I don’t find the reasons compelling. Starting with an account good character (or a good life for that matter), and trying to reason one’s way from there to right action seems to me to be going in the wrong direction. Though maybe you disagree.