A. Positive Accounts. After MacIntyre, we looked at John McDowell's argument that in a certain sense virtue is knowledge, but a kind of knowledge which resists formalization in any sort of system of principles. We also examined Philippa Foot's positive remarks about the virtues, including the following suggestions: that the virtues are beneficial characteristics which one needs both for one's own good and for that of others; that they are distinguished from other beneficial characteristics such as health or concentration by being excellences of the will (rather than of the body or the mind); that the virtues are "corrective," that is, that we do not regard a characteristic as a virtue unless there is some sort of natural tendency to act otherwise. We also discussed Foot's interesting account of the conditions under which an action has moral worth, namely that it is both in accordance with virtue and done out of virtue.
B. Problems. Robert Louden presented a number of difficulties he believes are faced by virtue ethics. There are many of these. The first, and one of the most important, is the suggestion that virtue ethics is really not of much value for "casuistry" or applied ethics, i.e. detailed moral reasoning about particular cases. The worry is that, because virtue ethics does not supply principles for determining whether actions are right or wrong, it is incapable of giving specific advice about particular cases. Among the other objections raised by Louden are the worry that we don't know how to tell who is virtuous and who is not (leaving the advice "act as a virtuous person would" rather unhelpful), and the worry that for several reasons we need a form of evaluation directed at particular actions rather than at one's character -- for instance, because even virtuous people may act wrongly, and because character may be rather changeable.
C. Responses to the problems. Rosalind Hursthouse attempts to show by example that virtue ethics can be useful in applied contexts even though it does not provide a system of precise rules for determining whether conduct is morally permissible or not. Michael Slote explores some possible directions that might be taken by a rather extreme "agent-based" version of virtue ethics. The possibilities explored by Slote include the "cool" idea of morality as inner strength, and the "warm" ideas of morality as universal benevolence and as caring, or more focused benevolence. Annette Baier suggests that, if "masculine" ethics is characterized by an emphasis on systems of rules which formalize the notion of obligation, and "feminine" ethics is characterized by an emphasis on love and caring, then we might hope for an ethical theory which integrates both of these into a single framework. Her proposal for a guiding idea for such a synthesis is the idea of trust.
2. The argument from queerness. Mackie argues that, if there were any moral properties, they would have to be very strange properties. It is hard to see how we could come to know about them (this is related to Harman's problem); more importantly, it is hard to see how any objective property could have the kind of connection with motivation that moral properties seem to have.
3. The argument from relativity. People in different
cultures and at different times and places have very different moral beliefs.
Mackie argues that this is most plausibly explained by rejecting the idea
that there are objective moral properties, and explaining moral beliefs
instead in terms of social and cultural factors.
2. Naturalistic realism. Boyd and Railton both argue that good is an ordinary empirical property. (Boyd argues that it may not be definable in any straightforward sense, while Railton offers the beginnings of a proposed definition, but both agree that moral properties are the kind of properties that we can learn about by means of the sciences.
3. Emotivism. Emotivism is the most extreme anti-realist
position. Emotivism holds that there are no moral properties; moral
language does not describe any sort of facts about the world, either natural
or nonnatural. Instead, moral language serves only to express our
emotions. It is as though to say "That's good" were tantamount to
making an affirmative grunting noise, while "That's bad" is essentially
making a negative grunting noise. You aren't saying anything, you're
just expressing how you feel.
2. Moderate realism. In the "sensibility theories" of McDowell and Wiggins, we find a more moderate version of realism. On these views, moral properties are like color properties. They are objective properties of real things, but they are properties that we pick out because of their effects on us. Just as we classify red things as red because they produce a certain sort of experience in ideal observers under ideal conditions of observation, so we classify right actions or good states of affairs as right or good because they produce a certain sort of affirmative reaction in ideal observers under ideal conditions of observation. We have spent a substantial amount of time in class attempting to work out some of the details of this sort of view, and see how it makes it possible to respond to the problems with which we began our discussion.