Virtue as Power

misp_217

164..178

MIDWEST STUDIES IN PHILOSOPHY Midwest Studies in Philosophy, XXXV (2011)

Virtue as Power MICHAEL LEBUFFE

M

any discussions of virtue, including prominent recent defenses and criticisms of virtue ethics, take virtue to be what is valuable in a person. In this familiar, roughly Aristotelian sense, virtue is a property of a person, typically presented as a character trait or a disposition, that is opposed to vice and explains why or in what respects a person is good. So understood, the nature and origin of the value of virtue is a central topic. The topic is complex and discussions involve many different terms and issues.1 A simple version of the options that will be useful for my purposes may be put, however, in terms of the relations that one finds between virtue and right action. One might take virtue to be fundamental in the sense that it is intrinsically valuable and its value is the source of the value of other things.2 Thus, we might hold that a courageous act is valuable, if it is, just because it flows from courage and courage is valuable. Alternatively, one might think that something else is valuable and that virtue holds whatever value it has because of its 1. For a sophisticated, historically oriented discussion that has influenced many recent discussions, see Christine Korsgaard, “Aristotle and Kant on the Source of Value,” Ethics 96 (April 1986): 486–505. 2. Rosalind Hursthouse has the most fully developed recent version of a view taking virtue to be the source of our understanding of the value of actions. Rosalind Hursthouse, On Virtue Ethics (Oxford: Oxford University Press, 1999). See especially, 36–39. Other important recent works that take virtue to be fundamental in something like this sense include Michael Slote, Morals from Motives (Oxford: Oxford University Press, 2001); Christine Swanton, Virtue Ethics: A Pluralistic View (Oxford: Oxford University Press, 2003); and Julia Annas, Intelligent Virtue (Oxford: Oxford University Press, 2011), especially 36–40.

© 2011 Copyright the Authors. Midwest Studies in Philosophy © 2011 Wiley Periodicals, Inc.

164

Virtue as Power

165

relation to this something else.3 Such a theorist might argue that courageous acts are intrinsically valuable, and that courage is valuable instrumentally insofar as it is a disposition to act in a valuable way. Finally, one might also hold—against both of the other positions—that virtue holds intrinsic value and that other things do so as well.4 Such a theorist might hold that both courage and courageous action have intrinsic value. My purpose in this paper is to advertise a different sense of “virtue” on which virtue is the power to act.5 Let us refer to the two senses as VirtueV (for value) and VirtueP (for power). (I will also continue to use the unscripted term where I do not mean to invoke either technical sense.) VirtueP does not obviously oppose vice. The power to do a courageous act, depending upon how one understands power, might be at the same time a power to do a cowardly act. As a result, VirtueP does not obviously hold value itself either (although we might eventually and given other assumptions argue that it does). Although I do not wish to evaluate VirtueV or any particular version of it, I will argue here that the notion of VirtueP has or should have a broader appeal than the notion of VirtueV. Unlike that VirtueV, 3. This view is associated with Kant. The opening of Kant’s Groundwork, 4: 393–94, which offers perhaps the best known statement of such a view, is the most apt for the present purpose as well. There, Kant makes the possession of a good will a condition of the value of temperament, character, fortune, happiness, and those qualities considered by the ancients to be unconditionally good. He also makes the value of these instrumental and derivative. These traits, however, should be distinguished from what Kant himself takes virtue to be. Kant’s view about the value of what he takes to be virtue does not clearly make its value derivative. In the Metaphysics of Morals, Kant’s definition of virtue in general in terms of duty (6: 405) suggests that virtue, as he understands it (which should be distinguished from the list of traits that concern the opening of the Groundwork), is understood in terms of a more fundamental notion. I hesitate to conclude from this that virtue in Kant’s formal account has derivative or instrumental value, however. Citations are volume and page numbers of Kant’s Gesammelte Schriften and are drawn from Kant, Practical Philosophy, trans. and ed. Mary Gregor (Cambridge: Cambridge University Press, 1996). John Rawls is a recent figure who is perhaps associated with this view. Rawls’s account, however, resembles that of the Metaphysics of Morals. He writes, for example, “once the principles of right and justice are on hand, they may be used to define the moral virtues just as in any other theory.” So, like Kant’s official account, his account of the value of virtue is difficult to categorize neatly. See John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971), 192. 4. For example, see Thomas Hurka, Virtue, Vice, and Value (Oxford: Oxford University Press, 2001). Hurka introduces his own position in a similar way, pp. 3–4, and defends the coherence of his account as a fully consequentialist view in chap. 1. Another prominent example of such a view is Robert Adams, A Theory of Virtue: Excellence in Being for the Good (Oxford: Oxford University Press, 2006). 5. This notion also has deep historical roots, although the notion of virtue as power is typically combined with the notion of virtue as value in ways that can make them difficult to distinguish. I associate the pure idea, that virtue is power and only accidentally anything morally praiseworthy, most closely with Virgil’s Aeneid. See, for example, Book 5, ll 454–60, where despite shame, disdain, and anger, Entellus is highly virtuous as he strikes on all sides in battle. Similar ideas are present in many places but prominent in Machiavelli, who argues that weak princes are dangerous to a state but that “two virtuous princes in succession are sufficient to acquire the world.” Nicolo Machiavelli, Discourses on Livy, trans. Harvey Mansfield and Nathan Tarcov (Chicago. IL: University of Chicago Press, 1996), I, 20. Spinoza is also quite clear at Definition 8, Part 4 of the Ethics: “By virtue and power I understand the same thing.” This is my translation from Spinoza, Spinoza Opera, ed. Carl Gebhardt (Heidelberg: Carl Winter, 1972).

166

Michael LeBuffe

VirtueP is, or should be, important to a broad range of accounts of morality. For we use VirtueP to describe the facts about a person that we take to be relevant to our evaluation of that person in action—as, for example, when we argue that what a person can do limits the sorts of actions that we might plausibly find the person obliged to do—and a broad range of accounts of morality do or should take facts about a person to be relevant in this way. VirtueP, then, may stake a claim to being a basic moral concept in a way that VirtueV may not. In exploring VirtueP, I will be particularly interested in distinguishing it from VirtueV because the two concepts are sometimes difficult to distinguish in recent theories of virtue. In her important book, Intelligent Virtue, for example, Julia Annas writes: [ . . . ] although it is natural for us to think of a virtue as a disposition, we should be careful not to confuse this with the scientific notion of a disposition, which just is a static lasting tendency. A classic example is that glass has a disposition to break under certain circumstances [ . . . ] [ . . . ] A virtue is not a static condition like this; it is a disposition as a result of which Jane acts and thinks in a certain way, and which is at any time strengthened by her generous responses and weakened by failures to have them. If she is generous, her generous actions and feelings both come from a virtue and fortify it.6 Annas makes the characteristics in virtue of which a person is good also characteristics in virtue of which she acts. It is, of course, not a fault in a theory that it incorporates two different concepts. We may fairly take Annas to be defending the substantial thesis that the real, important understanding of virtue incorporates notions of value and power at the same time: the generous person both is good in that respect and also acts or has a power to act generously. Nevertheless, views like Annas’s do face serious objections. I will want to show that objections raised against full, rich theories of virtue, such as Annas’s, tell against VirtueV but not against VirtueP.7

6. Annas 2011, 8–9. Another important example may be found in Gopal Sreenivasan, “Errors about Errors: Virtue Theory and Trait Attribution,” Mind 111 (2002): 47–68. Sreenivasan discusses the existence of “cross-situationally consistent character traits” where those are defined in terms of behavior. See especially, p. 57. This view is also consistent with the Aristotelian view that there are further necessary conditions (beyond virtue) for virtuous activity. Like Annas, however, it incorporates a kind of activity into the definition of virtue. 7. Note that this close association of good character and activity is not a feature of the view of the most famous proponent of VirtueV. Aristotle distinguishes virtue (arete) from virtuous activity (eudaimonia) and holds that something must be added to virtue to get virtuous activity in an agent: “Hence the happy [that is, virtuous and active] person needs to have the goods of the body and external goods added, and needs fortune also, so that he will not be impeded in these ways.” Aristotle, Nicomachean Ethics, 1153b17 and following, translation by Terence Irwin (Indianapolis: Hackett, 1985). I have added the brackets. Irwin translates “eudaimonia,” “happiness.” Thus understood, a good character (my VirtueV) is necessary but not sufficient for virtuous activity, eudaimonia, which requires further internal and external goods. It is helpful to compare this passage with the opening of Kant’s Groundwork (see note 3 above). See also 1099b15–29.

Virtue as Power

167

Here is a more detailed account of VirtueP. It is that property, the possession of which by an agent is a necessary condition for the appropriateness of judgments that the agent has done well or badly and, more important (since my first, blunt characterization makes VirtueP close to the same thing as agency), the nature and degree of which determines and limits the kinds of judgments that might be appropriate. Suppose, as I think that many do although they might not put it this way, that whether a given person in a given instance does well or badly depends upon a collection of circumstances, both internal and external, in which the person acts. A person’s VirtueP is the sum of relevant internal circumstances. 1. VIRTUE AND AGENCY A set of examples, I hope, will make this conception of virtue more clear and, in particular, will show how VirtueP differs from agency. One might think that what it is to act well, in many contexts, depends upon facts about the agent. Thus, when Susie first enters a pool, she swims well if her face is in the water as she paddles; when she is stronger and more comfortable, she swims well if she makes it a length of the pool; when she is yet stronger and well practiced, she swims well if she breathes rhythmically; and so on. All of the actions, we may assume, are Susie’s. Our assessment of these actions depends, however, not merely upon whether Susie acts, which is a question about agency, but upon what kind of actor Susie is, which is a question about VirtueP. I do not mean to offer any particular criterion for acting well for Susie. Perhaps one might think that if she swims to the best of her ability, then she swims well. Alternatively or in addition, as Annas’s account of generosity suggests, one might think that if Susie swims in such a way that it will help her to improve her ability to swim, then she swims well. Regardless of the particular account one gives, if one takes Susie’s ability to be relevant to the evaluation of her action, then one finds the association between virtue and the value of actions that the notion of VirtueP is meant to convey. It is natural to think that unqualified attributions of value to action—not swimming well but simply doing well—are similarly dependent upon our understanding of the particular agent’s circumstances. Some six-year-olds do well to refrain from hitting. Some eight-year-olds do well to share. Some twelve-year-olds do well to give. These evaluations are similarly dependent upon our understanding of the agents and so similarly depend upon the notion of Virtuep. The difference between VirtueP and agency is already evident from these first examples. Susie is always an agent; her VirtueP varies. One might, however, resist the conclusion that we need to include VirtueP among our list of notions fundamental to the assessment of actions by attributing degrees to agency to different agents. It stretches ordinary, natural, uses of “agent” only very little, for example, to say that a six-year-old is an agent to only a very slight degree, that an eight-year-old is an agent to a greater degree, and so on. This strategy may be useful in particular contexts, as for example in the determination of legal responsibility. The strategy depends, however, on having a view of human character that is somewhat flat and therefore limited. I, although generally as sophisticated as most adults, am not much of a singer. If I hit some note played in the middle of my

168

Michael LeBuffe

voice range on the piano, I have sung well. I frittered away my youth playing ball games, however, so I am far ahead of the curve on throwing things. If I skip a rock on a lake seven times, I have not thrown particularly well. Although a full agent in any ordinary sense, I nevertheless possess a variety of different levels of different abilities, the understanding of which is relevant to the evaluation of how I have done in acting. Once again, it is natural to think that unqualified attributions of value to a person’s actions depend similarly on details about the particular person. Albert, whose honest behavior consistently has been repaid with lies, acts well to remain silent rather than to lie himself. George, who is not similarly scarred, can do better and so does not act well if he remains silent. On the other hand, George finds it very difficult to help others. He does well to give half of what Peter Singer recommends he should give. Albert has no such quirks, and he does poorly if he gives the same amount. To explain these judgments by saying, “George is a full agent in ways that Albert is not and Albert is a full agent in ways George is not,” would be to stretch the concept of agency. The cases are handled much better in terms of VirtueP: George and Albert are virtuous in different ways, and this affects the ways in which we evaluate their actions. Of course, one might reject these judgments and insist that there is a single right thing to do and that a failure to do it is always a failure to do well. On such an analysis—which I associate with the Kantian view that we assume our own freedom in acting as a ground for our action without attending to the question of whether we are actually free—there really is, it turns out, only room for the notion of agency.8 Either George is an agent or he is not. If he is, then the right thing to do is to give the full Singer-recommended amount and that is that. I do not wish to argue that this position is wrong, here, although I think it is. I think that it is clearly better, even if one thinks for independent reasons that there is a right action that is different from the best that the agent can do, to judge the agent’s actions against what that agent is capable of doing. Instead of defending my view, I will argue that the burden of argument falls heavily on proponents of the view that there is a single right thing to do and that is that. The conception VirtueP is useful if people in some cases are simply not capable of acting in the best way, however that is understood, and especially if a given person is, like Albert or George, for some sorts of actions more capable of acting well, and for other sorts of actions, less capable. Let us suppose that the morally best action is the way that the best person would act. (Even if one takes “morally best action” to mean something else, this should be close to a materially equivalent definition.) The claim that I think makes the notion of VirtueP a useful notion is the claim that some people in some circumstances cannot do the morally best action. To support the view that the notion of VirtueP is not useful and that there is only room for agency as the notion relevant to our judgments of agents in acting is either to say that the claim that would make the notion of VirtueP useful is false—that is, that it is false that some people in some circumstances cannot do something as well as the best 8. For the passage from Kant that I have in mind here, see Kant’s note to the Groundwork at 4: 448.

Virtue as Power

169

people in the best circumstances—or it is to say that the claim is true but that nevertheless people ought to do things that they cannot do. Either of these positions is extremely strong. The conception VirtueP allows us to refer to the internal circumstances of an agent relevant to the evaluation of that agent’s action. Even if there is an independent sense in which certain actions, such as honest speech or generous aid, possess value wholly independently of facts about agents, there remains good reason prima facie to think that claims about morality and a particular agent, including claims about obligation, responsibility, and generally whether a person does well or not in a particular action, depend upon virtue so understood. Even if one decides ultimately that the apparently good reason for taking such judgments to depend upon an agent’s internal circumstances does not withstand scrutiny, one will bear the burden of argument in showing that this is so. So, the notion VirtueP remains an important, basic one, in a way, even for those who reject its importance. 2. TWO PROMINENT SORTS OF USES So understood, there are two prominent sort of uses of VirtueP: one that applies in situations in which we evaluate some agent relative to what that same agent might have done and another that applies in situations in which we evaluate one agent relative to other agents. Several of the examples that I have already given are examples of the first, intra-agent sort of use. If Susie is a swimmer of some strength and experience, then I say that she has done well in breathing rhythmically if that is among the best things that she could have done: she might have made some error of the sort that is perfectly understandable in a swimmer of her level, but she did not. Such analyses also apply, especially for moral evaluations, to circumstances in which agents make choices. Given Albert’s history, one might have expected him to lie and that would have been perfectly understandable. However, Albert chose to remain silent. Given his internal circumstances, he acted well. An understanding of what agents are capable of doing, that is, of VirtueP is required here in order to make these sorts of judgments. The inter-agent sort of use can apply simply where comparison of one agent to another is desirable, but it is natural, under further assumptions about choice and power under which agents do not really choose what they do or have the power to do otherwise, and intra-agent uses are therefore limited or ruled out. Supposing that neither Allison nor Mike has a choice about what they do and that it would be best (under some incomplete description of circumstances and supposing no particular agent) for others not to eat any of Susan’s delicious Halloween candy, we can say that Allison, who eats only one piece while Mike eats five, is more virtuous than Mike. She is able to do better than he is able to do.9 Such a use is prominent in Spinoza, who is a necessitarian and a determinist. For example, in giving an account of different cases in which people have ended 9. There are similar uses that are more sophisticated and invoke counterfactuals or generalizations. Thus, we might say that Allison is more powerful than Mike because, in a range of tempting circumstances, she is, or is better equipped to be, more resistant.

170

Michael LeBuffe

their own lives, Spinoza presents Seneca’s action as more virtuous than two similar actions. Seneca, given the choice to take his own life or to be executed, takes his own life. Another person’s own sword hand is twisted against his body. A third person (I gather) is a suicide in the more familiar sense of forming and acting on a desire no longer to live. No one, given Spinoza’s metaphysical views, could have done otherwise than they did. In the other two cases, however, those who die exercise no causal influence in doing what they do, on Spinoza’s account of causal influence. Seneca’s position is certainly not powerful. However, in controlling the one thing that he can control and acting to end his own life rather than passively allowing another to take it, he is one degree of activity over the other two. He is more virtuousp.10 It is in inter-agent uses of the concept of VirtueP that it becomes most closely associated with value. Further controversial but historically common theses, such as the thesis that akrasia is impossible or the thesis that all evil arises from passivity, contribute to the view that a person who is more active has not only the power to do more but also is and does better. Spinoza implies that this is true of Seneca. 3. POWER AND NATURALISM To define virtue as power is to issue a promissory note. The concept depends for its appeal on the fact that we need to point to internal circumstances in order to justify the variability with which we assign value to agents’ actions rather than on any particular account of what those circumstances are, such as an account of human nature or an account of character traits. Susie does well to swim a length of the pool before she tires. Diana Nyad swims poorly if she only makes it that far.The relevant external circumstances of action, we may suppose, are the same. So the difference grounding the difference in our evaluation must be a difference in the agents’ internal circumstances. What those circumstances are, for our purposes, is a further question. Because the notion of VirtueP does not include such detailed answers, it is therefore a great deal thinner than detailed accounts of character or human nature of the sort often associated with virtue ethics and VirtueV, which just is a characterization of those internal circumstances—courage, generosity, honesty, and so on—that make a person a good person.11 I think that the thinness of VirtueP is part of its appeal. Debates about particular accounts of human nature or human 10. See Spinoza’s Ethics, Part Four, Proposition 20, Scholium. This is a controversial reading of a difficult text, which I will not defend here. I do discuss the interpretation in LeBuffe, From Bondage to Freedom: Spinoza on Human Excellence (New York: Oxford University Press, 2010), 191–93. Often, Spinoza discusses virtue in evaluating the state of a person without explicitly comparing that person to others. Given his metaphysical views, I tend to take such uses in an inter-agent sense. Roughly, they mean, in similar circumstances (e.g. confronted by a tyrant’s demand to kill oneself or to be killed), different people act in a variety of ways. To kill oneself is the best (or among the best) of these ways, although, of course, in a complete description of particular circumstances, no one could act in a different way from the way he or she does act. 11. However, some accounts of VirtueV may also be thin. See, for example, Robert Adams, A Theory of Virtue: Excellence in Being for the Good (Oxford: Oxford University Press, 2006), 26. Adams makes his book a study of excellence, but not of particular virtues.

Virtue as Power

171

character can seem to conflate the appeal of the accounts themselves with the appeal of the concept of virtue in such a way that, if the particular accounts are defeated, virtue is defeated as well. VirtueP, though, is independent of any detailed theory of virtue ethics. It is also different from the concept of virtue most closely associated with those views, VirtueV. An important feature of the notion of VirtueP insofar as it does incorporate a substantive thesis—the thesis that our judgments about the value of agents’ actions depend on some facts about the internal circumstances of the agent—is that it is naturalistic. Of course, it might (and should) be a component of comprehensive moral theories that are both naturalistic and nonnaturalistic. By the label “naturalistic,” I mean only that VirtueP may be attributed, if we want to, to all things in nature that might be considered agents, and also that VirtueP can contribute, if we want it to, to the evaluation of all kinds of actions. Let us take agents first. VirtueP may be found in anything that we take to act. Just as we might say that a given human being swims well if she puts her face in the water as she paddles, so we might say that a given tree grows well if it grows three inches in a year, a small stream flows well if it keeps lush the area around it that is usually lush, and a small lion hunts well if it manages to kill an antelope of such and such a size. In any of these cases, saying that the thing in question does well is only to say that it has acted in a way that, as far as it can, approaches some standard of value. The notion of VirtueP is also naturalistic insofar as it treats the moral value of actions in the same way that it treats other kinds of value in actions. It is true, of course, that we need to say a great deal more about many topics before we can vindicate our judgment that Albert does well in remaining silent. Indeed, this point is a point in favor of VirtueP. I have claimed that the conception is or should be one of our basic ethical concepts. That claim is more plausible to the extent that the concept is more compatible with different conceptions of value. So we might say that honesty characterizes the behavior of the best human being and that staying silent is the closest that Albert might have done to approach that ideal which stands for us as a standard of value. Or, just as readily, we might characterize value by saying that, in that situation, the very best thing that we can imagine a speaker doing in terms of the consequences of the action might have been telling the truth, and that, once again, remaining silent is the closest that Albert, given his particular internal circumstances, could have done to approaching that standard of value. At any rate, while it is true that we need to say more about moral value, it is also true that we need to provide an account of other kinds of value, such as swimming well, throwing well, and so on, also in order to vindicate our judgments about particular agents’ performances in these kinds of activities. The situation, with respect to moral judgment, is the same situation that we face in justifying other attributions of value. Moral value, as far as VirtueP is concerned, is therefore not privileged among other kinds of value. I have presented moral value thus far as doing well without qualification. But I need not have. Suppose that we offer a general account of value on which value is promoting happiness. In that case, to judge that Susie does well without qualification will just be to judge that Susie does well at promoting happiness. That is a kind of project that is not any more different

172

Michael LeBuffe

from swimming well as throwing well is from singing well. So, as far as VirtueP is concerned, moral evaluations of agents’ actions are on the same footing as other evaluations. I mention in the introduction to this paper that comprehensive moral theories that emphasize virtue tend to incorporate both VirtueP and VirtueV, and that criticisms of such views might, as a result, seem to undermine both kinds of virtue at once. The naturalism of VirtueP is especially important to the task of showing that such criticisms, while they may give us reason to doubt particular theories of VirtueV and the comprehensive theories of morality that include both conceptions of virtue, do not give us reason to doubt the usefulness of VirtueP. For the most powerful criticisms of comprehensive moral theories that emphasize virtue apply only to nonnaturalistic components of those theories: the special importance of the human being as a kind of thing capable of producing moral value or value of a particular kind; the moral value of particular kinds of actions or of actions arising from particular internal characteristics of agents. Because VirtueP is naturalistic, however, it resists the same objections, and it shows that they apply only in limited ways as criticisms of virtue generally. 4. VIRTUEP AND HUMAN NATURE One powerful objection, which I find convincing, is raised against accounts of virtue that depend upon a concept of human nature. Such accounts are typically perfectionist.12 They take the highest degree of virtue to consist in the perfection of human nature, and they take any person to be virtuous to the extent that that person develops the various capacities that constitute human nature. Thus, one might say that human nature consists in rationality; perfection, or the highest degree of virtue, in the highest degree of rationality; and valuable activity in the development of or manifestation of a high degree of rationality. Such an account allows for a clear distinction between doing something well and doing well in an unqualified sense. Thus, one might say that growing tall, while just fine, is not doing well because height is not a kind of rationality, and so growing tall is not developing a higher degree of perfection. Playing chess well is doing well in an unqualified sense, by contrast, because (one might think) it is the display of a kind of rationality. Perfectionists’ accounts of human nature are, of course, much more thoughtful and detailed than my example. The objection—the best form of which Philip Kitcher raises against the account of human nature developed by Thomas Hurka—is nevertheless telling against a wide range of accounts of human nature that one might put to the service of ethics. On it, there is no ethically meaningful property that is distinctively human. Notably, rationality is not distinctively human. 12. The relationship between virtue and perfection may be construed in different ways. One clear account may be found in Hurka 2001 (53–57). Hurka takes virtues to be good because they make a person more perfect rather than because the person who has the virtue has a particular attitude toward them, such as wanting them or taking pleasure in them. A full, rich virtue ethics can, of course, include an account of human nature without, at the same time, being a perfectionist theory.

Virtue as Power

173

A number of familiar particular examples of less rational human beings and more rational things that are not human beings might be given, which undermine the association of rationality and humanity. The best reason for thinking that this is correct, however, which Kitcher presents, applies to all human beings: it is perfectly plausible to imagine an evolution of the human species that happens in such a way that no individual ever develops a rich mental life.13 One might look elsewhere for properties other than rationality that do turn out to be essentially human. If it does turn out that there are some such properties, Kitcher argues, they will be properties, such as the possession of twenty-three pairs of chromosomes or growing hair, the development of which are not morally important.14 In short, there is no human nature that it would be valuable to develop. Although this objection is convincing as a reason to jettison a wide range of theories of virtue—those on which human nature or a high degree of its development is valuable—it does not touch Virtuep. This is one of the advantages of the naturalism of the conception. Virtuep does not require that agents be human beings or actions be human actions. One might add nonnaturalistic elements to Virtuep, especially if, as in some interperson uses especially one might be tempted to do, one finds moral value to be different in kind from other sorts of value in applying only to human beings or their actions. If one takes unqualified ascriptions of doing well (that is, again, one way of conceiving of doing well morally) ever to be true, certainly one will want to give an account of some sort of what that means. Once again, however, there is no particular claim about the nature or origin of value built into the conception of Virtuep; there is only a claim about the relation between the value of an agent’s action, however that is to be understood, and the power of an agent. So theories of human essence and human perfection are consistent with but wholly complementary to Virtuep. Objections to those theories do not tell against it. What is most important is that, while it is correct I think that there simply is no such thing as a human essence, the principal reason for thinking that virtue is a central ethical concept—that our ascription of value to particular actions depends upon facts about the agent—is not challenged by this objection. 5. VIRTUE AS POWER AND CHARACTER TRAITS Examining a different objection, prominent versions of which have been presented, collected, and polished by Gilbert Harman, does present a challenge for Virtuep, although it is aimed narrowly at full, rich accounts of virtue in which VirtueV figures prominently. Such accounts of virtue ethics typically list character traits, including virtues such as courage and honesty, and vices such as gluttony and anger, that are understood to be stable dispositions in those who have them to act

13. Philip Kitcher, “Essence and Perfection,” Ethics 110 (October 1999): 59–83. Kitcher describes familiar counterexamples in different sections of the essay. For this general argument, see 71–73. 14. Both of these examples are Kitcher’s. See Kitcher 1999, 67 and 70.

174

Michael LeBuffe

in certain ways.15 Such accounts might, for example, explain the fact that one person runs away from a threat while another stands and faces it in terms of the courage of the first person and the timidity of the second. Harman offers a definition of “character trait” similar to that frequently found in accounts of virtue ethics: it is a relatively long-term stable disposition to act in distinctive ways.16 He distinguishes character traits, so understood, from other internal characteristics of agents: Character traits must be distinguished from psychological disorders like schizophrenia, mania, and depression, and from innate aspects of temperament such as shyness or being basically a happy or sad person. Character traits include virtues and vices like courage, cowardice, honesty, dishonesty, benevolence, malevolence, friendliness, unfriendliness, as well as certain other traits like friendliness [sic] or talkativeness.17 Then he argues, in part, that the results of social psychology give no evidence of such traits.18 When external factors are very similar, Harman suggests, the results of social psychological experiments show that people do not differ in the ways in which we would expect them to if they had character traits in different degrees. To take Harman’s favorite example, the famous Milgram experiment showed that, placed in the same, peculiar kind of situation, nearly all people are willing to behave in a manner that a virtue theorist might ordinarily characterize as cruel.19 In the experiment, all participants were willing to inflict what was described to them as a “very strong shock” on a stranger. If the external circumstances of the experiment—a slow progression from acceptable behavior to cruelty, the demand of the experimenter to continue, the difficulty subjects have in understanding what is happening—produce the same behavior (or close to the same behavior, as some subjects went well beyond “very strong shock”) in all subjects, however, there is strong reason to think that external circumstances and not internal ones explain human behavior. 15. See Gilbert Harman, “Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error,” Proceedings of the Aristotelian Society 99 (1999): 315–31. This is a paraphrase of Harman, 317. It applies accurately, I think, to a variety of versions of virtue ethics, including Aristotle’s, as Harman notes. A recent, sophisticated version of such a view may be found in Annas 2011. Annas’s work (8–15) on refining the notion of a disposition is especially interesting to me insofar as it brings the possession of a virtue very close to a power to perform certain sorts of actions. 16. Harman 1999, 317. 17. Harman 1999, 316. 18. See Harman 1999 and Gilbert Harman, “Skepticism about Character Traits,” The Journal of Ethics 13 (2009): 235–42. This is a paraphrase of Harman’s own summary of his view in Harman 2009, 241. 19. See Stanley Milgram, “Behavioral Study of Obedience,” Journal of Abnormal and Social Psychology 67 (1963): 371–78. I will not provide a full account of this experiment here because I take it that it is familiar to most readers. Please see Harman 1999 for a useful summary, emphasizing conclusions relevant to the subject of the character trait of cruelty and the topic of virtue ethics.

Virtue as Power

175

This conclusion is relevant to our evaluation of accounts of virtue ethics that suggest that different virtues and vices cause different behavior in different people. Different people behave in the same way in the same external circumstances, or so the Milgram experiment and others that Harman cites suggest. Social psychology, on Harman’s account, tends then to suggest either that everybody possesses these virtues and vices in a similar degree (for example, the Milgram experiment might suggest that everyone is cruel) or that we would do better not to talk about virtue and vice. In either case, Harman concludes (and, to anticipate, this conclusion will be my principal target here) that our concern that people act well should lead us to focus on external circumstances and their effect on human behavior rather than on character traits: Now, it seems to me that what happened in Nazi Germany and in Bosnia, Somalia, etc., taken together with results in social psychology about the relative explanatory importance of individual character versus the situation in which a human being is placed, indicate that the very natural human tendency to think in terms of character traits leads us in the wrong direction. To the extent that we are interested in improving the lot of mankind it is better to put less emphasis on moral education and on building character and more emphasis on trying to arrange social institutions so that human beings are not placed in situations in which they will act badly.20 Harman’s work is extremely valuable because it raises seriously the question of whether the ancient theory of virtues is compatible with what we know now about human beings. It is not clear that he is correct in arguing that the results of social psychology undermine various claims about the existence and variety of character traits. Even if Harman is wrong, however, the argument presses defenders of virtue to develop clearer accounts of what the central terms mean.21 To the extent that Harman is correct about character traits, any theory of VirtueV ought to be revised. Most discussions of Harman have responded directly to his claim about the relevance of the results of social psychology to virtue ethicists’ claims about character traits. I have a different concern, one that I think shows the importance of an ancient philosophical concept that Harman does not address, VirtueP. The thinness of VirtueP protects it from the main thrust of Harman’s objection, of course. VirtueP involves no substantive claims about stable dispositions to act in certain ways, and therefore also no claims about traditional virtue and vices. On the most basic conception of VirtueP, we are justified in giving an agent’s action a certain kind of evaluation, if we are, in large part by facts about the internal circumstances of the agent. The thought that an agent acts by means of some VirtueV—that she is courageous or cruel, for example—would amount to a positive further thesis about what those circumstances are. If it is false that any agent is courageous, or especially courageous when compared against others, that itself does not touch the core conception of VirtueP. 20. Harman 2009, 241. 21. Sreenivasan (2002) is a detailed argument evaluating Harman’s success that does just this.

176

Michael LeBuffe

In another respect, however, the objection that different people tend to behave in very similar ways in similar external circumstances, if correct, does threaten to undermine the usefulness of the notion of VirtueP. In distinguishing VirtueP from agency, I have argued that the notion of VirtueP helps us to make use of the plausible idea that a given person is, for some sorts of actions, more capable of acting well than others, and for other sorts of actions, less capable than others. Just as I sing worse and throw better than many others, so Albert helps others better than and tells the truth worse than George. If social psychology suggests that there is not this kind of difference among people with respect to the kinds of actions that we find to hold moral importance, then it does tell against the usefulness of the notion of VirtueP in ethics. There would be little point to an idea meant to capture the relevance of differences among agents to the evaluation of actions if there are no relevant differences among agents. The naturalism of VirtueP with respect to actions, on which it does not discriminate between kinds of actions that have a particular moral importance and other kinds of actions, suggests a response. Whether or not we are all equally cruel, generous, and so on in some external circumstances, it is clear that people are different in other ways. Harman does not deny such differences. Indeed, in trying to discriminate among VirtuesV (that is, character traits) and other psychological characteristics, Harman seems clearly to allow that we vary in our temperaments and psychological disorders. Of course, to move beyond the psychological, we vary in a number of talents, such as ball throwing and singing as well. The conception of VirtueP does not recognize the special relevance of character traits or the irrelevance of psychological disorders or innate aspects of temperament, or even differences in physical ability to morality. It is a further question for a philosopher who finds VirtueP important whether only courage, cruelty and so forth are relevant to the moral evaluation of an action. Indeed, it seems quite plausible to say that other internal characteristics regularly are relevant. If, for example, happiness is morally good, depression certainly is a morally relevant characteristic. A depressed man will, insofar as he is depressed, have less power to act well than other people. To the extent that depression affects him but not them, he has and causes less happiness. In inter-agent evaluations, a depressed person is worse, arguably morally worse, than others. That is one of the reasons that depression is such a serious condition. It is plausible, moreover, to offer intra-agent moral evaluations of the actions of people affected by depression that make that condition a critical part of the evaluation: the depressed man does well to get out of bed, get dressed, and get to work. Those same evaluations might not apply to a healthy person who would have a very different VirtueP. A wide variety of human abilities ought, I think, to inform our moral evaluation of agents’ actions, and VirtueP, which puts moral evaluation on the same footing as other kinds of evaluation, leaves us the freedom to capture that intuition. Suppose that three people set out to do well. Each of them knows that a neighboring village has no food and that all of its residents face starvation. Each of them has limited resources: a car, $900, an assistant, a week. So their external circumstances are the same. What each person is able to accomplish will depend upon a wide variety of different abilities, including the abilities to organize, to endure, to

Virtue as Power

177

communicate, to negotiate, to drive. All of these abilities will contribute to what they do, but what they do will be something that has moral value, will it not? Suppose that Ira comes up with ten truckloads of rice and a working farm, Jean comes up with three truckloads of rice, and Sam comes up with 100 takeaway pizzas. On the supposition, which will do for our purposes, that calories indicate value in this instance, it is clear which action is better than which when we measure the results against a standard independent of the particular agents. But how are we to evaluate the actions as the actions of agents? Once again, it seems plausible to say that circumstances internal to the agent, what she is able to do, are relevant to the evaluation of each agent’s action. To take an intra-agent case, if I judge that Jean did poorly and acted badly, it is because I feel that given her capacities as an organizer, a communicator, a negotiator, and so on, she could have delivered much more. One might make an inter-agent evaluation as well: Jean is simply more virtuousp than Ira or Sam if she is able to do more. Because “more” here means morally better actions, the evaluation is once again a moral evaluation that follows from an understanding of a wide range of abilities, not merely or even prominently character traits: Jean does morally better than the others. VirtueV is often understood to be part of a view on which there is only a short list of internal circumstances or a relatively small percentage of internal circumstances relevant to our assessment of the moral value of a person’s action or the moral value of the person. Harman’s focus on character traits in his criticism of virtue theories seems to inherit this view. VirtueP, in its thinness, permits a better view: a wide range of different abilities can contribute to our evaluations of agents and their actions. I think that this is a strength of the doctrine. Many of the real internal circumstances that affect our actions, perhaps most, are both highly changeable and morally important. Harman is right that there would be no point in devoting resources to the development of character traits that do not exist or, if they do, that do not change. Whether or not there are character traits of some kind, I think that Harman’s suggestion that we should not devote a lot of energy to moral education and building character, where that means training people to be honest, generous, and so on, is worth further empirical investigation. If it is correct that such training cannot change people with respect to their tendencies to be cruel, generous, and so on, that is important to know. Perhaps because Harman inherits his targets’ assumption that these traits are all that could be relevant to improving individual human actions, however, he also arrives at an unsatisfying conclusion. Harman lists only two options that we have for improving the lot of mankind: character building education and arranging social institutions. Since we cannot do the former, he argues, we should devote our energies to (“put more emphasis on”) the latter. This discussion of VirtueP gives us reason to bethink Harman’s discussion radically incomplete. If we are interested in improving the lot of mankind, there is a point in devoting resources to the development of a wide range of human abilities that do exist, do contribute the value of agents’ actions, are highly changeable, and indeed, in most cases, are changeable by education and training. Power is the ability to act, and more powerful people have a greater ability to act. Morally good action is not a special kind of action exempt from this general conclusion. So the same kinds of

178

Michael LeBuffe

education and training that make people better at doing different kinds of things also make people better able to act well. 6. CONCLUSION We use the concept of Virtuep to describe facts about an agent that we take to be relevant to our evaluation of that agent’s actions. So understood, Virtuep is a concept related to agency that is compatible with many different comprehensive theories of morality: all of those that regard agents’ internal circumstances as so relevant. Virtuep is compatible, therefore with those theories that include a richer conception of virtue on which it is associated with value or the disposition to act in valuable ways. The two senses of virtue should be distinguished, however. Virtuep does not limit the internal circumstances of an agent that one might find to be relevant to the moral evaluation of the agent’s action by means of an account of human nature or character traits, and it does not imply that there is some particular class of circumstances, such as those captured by the notion of Virtuev, that is relevant. Moreover, in the absence of further theses about value, Virtuep does not make an agent or the agent’s actions morally better as VirtueV does. That work is left for other components of a comprehensive theory of morality.22

22. I am grateful for the opportunity to contribute to this volume. Paul Hoffman was a great, creative, careful philosopher. He was a member of my dissertation committee in the late 1990s, and I was lucky to have his criticism of my work early in my career as well. Most important, Paul was a good friend. In recent years, although I saw Paul frequently at conferences, I most looked forward to surfing with him at San Onofre during the summer.The sessions were easy, comfortable, and pure fun. They were marked by Paul’s intense interests in length of ride and Descartes, both of which tended to get us space in crowded takeoff zones. I cherish their memory.