A STANDARD MODE OF ARGUMENT in moral philosophy is to argue that an ethical theory should be rejected because it counts as right what is intuitively wrong or as wrong what is intuitively right--that, like a logical system from which one can derive theorems that are false, the theory is unsound.
In "Virtue and Right," Robert Johnson seems to make just this sort of argument against virtue ethics, arguing that "the claim that right actions are those of a virtuous person is ... utterly false." (1) But he can, I think, more profitably be read as making a different sort of argument against virtue ethics, that, like a logic which cannot generate all logical truths as theorems, virtue ethics is incomplete, since there are moral requirements (e.g., duties of moral self-improvement) that cannot be derived from it:
...any theory that relies on [the claim that right actions are
those of a virtuous person] to construct a virtue-oriented
theory of right action "will be unable to explain moral
distinctions we regularly make regarding behavior appropriate
for those who could better themselves. (2)
1. Virtue Ethics and the Charge of Incompleteness
When pressed for an account of right action, virtue ethicists will often respond with something like the following:
Virtuous Agent Agent A's [phi]-ing in circumstances c is right iff
a fully virtuous agent, acting characteristically,
would [phi] in c. (3)
As Rosalind hursthouse is at pains to remind us, standards like Virtuous Agent need to be filled in with a substantive account of what the virtues are and what a virtuous person is like, just as a consequentialist standard needs to be filled in with a substantive account of what the best consequences are, etc. (4) Johnson assumes that, however Virtuous Agent is fleshed out, there will be things that those of us who are not fully virtuous--whom I shall call the subvirtuous (5)--ought to do in order to improve ourselves morally. But since a fully virtuous person would not characteristically perform any of the actions required of the subvirtuous, none of these actions can be right according to Virtuous Agent, so virtue ethics is incomplete. A completely virtuous and thus temperate agent can safely navigate the dangers of the ice cream aisle without falling into gluttony; i, on the other hand, had best stay away, as my desire for Ben & Jerry's Chubby Hubby will likely get the better of me. A completely virtuous and thus honest person would not keep and reflect upon a list of the lies she tells in order to become more honest, but this may be just what a subvirtuous person should do. (6) Since a fully virtuous person would neither avoid morally fraught situations nor take positive steps away from vice and toward virtue, the subvirtuous person's duty to do so cannot be generated from Virtuous Agent--and thus virtue ethics is incomplete.
The charge of incompleteness bedevils not just contemporary virtue ethics but its historical antecedents, as well. Consider Aristotle's attitude toward shame (aidos), for instance. Despite the suggestion that shame is a virtue--the cause of political courage is a virtue, since its cause is shame (7)--Aristotle's official view is that shame is not a virtue, but even so it is conditionally good: "if one were to do disgraceful actions, one would feel disgrace," (8) but of course the antecedent is never fulfilled for a virtuous person. But just as continence falls short of virtue but is morally preferable to incontinence, so shame is not a virtue but is morally preferable to shamelessness. Remembering that our emotions no less than our actions are morally evaluable for Aristotle, when I act wrongly I ought to feel shame. But since the fully virtuous person would not act wrongly and thus will not have done anything to be ashamed of, how can an Aristotelian virtue ethics, construed somewhat anachronistically as accepting Virtuous Agent, ground the rightness of a subvirtuous agent's feeling shame? …