Acting from virtue.
1. Aristotelian and Kantian conceptions of action from virtue
Aristotle distinguishes between acting from virtue and acting merely in accordance with it.(1) This wording, though true to Aristotle, recalls Kant's distinction between acting from duty and merely acting in conformity with it.(2) On the plausible assumption that acting from duty is, often, acting from moral virtue, Kantian actions from duty are often similar in important ways to Aristotelian actions from virtue. Aristotle's concern with virtue and with acting from it went far beyond the moral domain, but there is much to be learned about morality from studying his general conception of acting from virtue.
By way of preliminary explication, although for many virtues, such as fidelity, generosity, and kindness, we speak simply of acting "from" the trait, for others the notion is usually expressed less directly, for instance by locutions like "acting from a sense of justice" and "paying the debt out of an honourable nature". Our task has two main parts: to capture the explanatory force of "acting from", where the source of the action is virtue, and to clarify how such action is from virtue, especially with respect to the normative and character elements this entails. The first task is complicated because, as a disposition, a virtue does not produce action in the "direct" way typical of, say, decisions or volitions, but only through cognitive and motivational processes connected with the virtue. The second task is complicated because many kinds of normative and dynamic elements can figure in connecting the trait with the action. Both tasks are complicated because some virtue names, e.g. "beneficence" and "fidelity", also name passing desires with the appropriate content. Aristotle was aware of this range of problems, but left much to be worked out.
In explaining how virtue differs from craft, Aristotle notes chat while the products of craft determine by themselves whether they are well produced, this does not apply to the products of virtue, since
for actions expressing virtue to be done justly or temperately [and
hence well] it does not suffice that they are in themselves in the
right state. Rather, the agent must also be in the right state when he
does them. First, he must know [that he is doing virtuous actions];
second, he must decide on them, and decide on them for
themselves; and, third, he must do them from a firm and unchanging
Consider this from the point of view of justice, which is adequately representative of a moral virtue (at least in the current sense of "justice", which, though narrower than Aristotle's, will serve adequately). Aristotle's first point is negative. Whereas a statue, for instance, can be beautiful when it is in the aesthetically right state, regardless of how the sculptor produced it, it would be a mistake to say that, regardless of how an action is produced, that action can be performed from justice. It is not sufficient that an action simply be of the right type, e.g. a meting out of equal shares to equally deserving claimants. In short, action from virtue is not a behavioural concept, in the sense of one defined in terms of what is accomplished, as opposed to how. Thus the adverbial forms of virtue terms such as "courageously", "honestly", and "justly"--can apply to actions not performed from the relevant virtues, and even to actions aimed at pretending to manifest those virtues. Given this thin use of virtue terms, the distinction between action merely in conformity with virtue and action from it may be regarded as a special case of a distinction between conduct of a behaviourally specified type, e.g. meting out equal shares, and conduct described mainly in terms of how it is to be explained, e.g. as done from a sense of justice.
Three further points in the passage help to explain its central contrast. Suppose I am the agent. First, I must know that I am, say, meting out equal shares; it will not do if, in signing an order by which I do this, I mistakenly take it to be a check. Call this the recognition requirement. Second, I must decide on this equal distribution and decide on it for its own sake. This implies two conditions: that my action must be (a) decided upon--call this the selection requirement and (b) in a special way motivated by the relevant virtue--call this the intrinsic motivation requirement. Not just any intrinsic motivation will do, however; it would not suffice to make the distribution for its own sake, in the way I do things for their own sake when I do them simply for pleasure, as where I swim simply because I like doing it. Aristotle seems to require that I decide on the action on the basis of a conception of it as, say, just, or as rendering each a deserved share, or as something else that connects my deed with justice as an element in my character. This intrinsic motivation requirement is confirmed (though not entailed) by his third condition: that one must act from a firm and unchanging character. If Jack is usually motivated by greed but, after a moving moral exhortation from a colleague, passes into a temporary just disposition, then even if Jack's resulting just behaviour toward a hated rival meets the other conditions, it does not express the virtue of justice. Call this the character requirement: virtues are elements of character; those elements are "firm and unchanging"; hence, an action from virtue must be from an element with the appropriate entrenchment and stability.
Some of the deontological counterparts of these requirements seem implicit in Kant's conception of acting from duty.(3) Acting from duty, for him, seems necessary for acting from moral virtue. If we may conceive good will as roughly the master virtue, this point is easily explained, at least in so far as acting from duty is a case of acting from good will. A stronger requirement for Kantian action from duty is suggested by Kant's second principle of morality: "An action performed from duty does not have its moral worth in the purpose which is to be achieved through it but in the maxim by which it is determined" (p. 400). Taken in the context of his examples, this requirement suggests that Kant conceives action from duty as motivated solely by a commitment to a suitable principle of duty. Even promoting one's happiness from duty must be grounded in a commitment to the duty of so doing,. and not the happiness to be achieved (p. 399). This position parallels Aristotle's intrinsic motivation requirement. If we suppose that action on a maxim requires a decision favouring that action over one or more alternatives, then Kant, like Aristotle, would have a selection requirement on acting from duty and thereby on acting from moral virtue.
One thesis of Kant's, however, has no exact counterpart in the Nicomachean Ethics: "as an act from duty wholly excludes the influence of inclination . . . nothing remains which can determine the will objectively except the law, and nothing subjectively except pure respect for this practical law" (p. 401). This exclusiveness requirement rules out any motive other than duty as actually motivating an action truly performed from duty. Aristotle does not say that an action cannot express virtue unless nothing else (i.e., nothing besides elements in the virtue) plays a part in motivating it, though he might perhaps have said, what Kant's overall position seemed to allow Kant to say, that an action purely from virtue must meet the exclusiveness requirement. I shall return to this question.(4)
Aristotle's selection requirement may sound psychologically unrealistic, but I do not think he should be read as holding that a virtuous action must arise from a piece of deliberation. For one thing, he stresses that habit is required to develop virtue, and he allows that acting from virtue can be an expression of habit; it can even be a habitual action. Moreover, he describes the grammarian as an analogue of the virtuous agent: the "way" the former produces a grammatical sequence is to produce something "expressing the grammatical knowledge that is in us" (1105a25); presumably, Aristotle conceived such knowledge as capable of yielding action quickly and without our considering alternatives. Similarly, Kant's notion of maxims from which we act does not seem to. require their conscious rehearsal before action. Granted, in reconstructing the genesis of an action with a view to judging that action, Kant imagines agents carefully formulating maxims (as with his famous four examples); but he does not require this kind of reflection for acting from duty in general. It may sometimes be needed to determine one's duty, but not for simply acting from a grasp of what that duty is, as where it is obvious that one is to tell the truth.(5)
2. The motivation and range of action from virtue
My own account of acting from virtue presupposes some of what Aristotle and Kant say, extends other points they make, and sets forth requirements not contained in their treatments of the topic. I begin with the questions of intentionality, deliberateness, and voluntariness.
An action that, under a given description, is performed from virtue, must at least normally also be intentional under that description.(6) To be sure, if I do something knowingly but not intentionally, as where, on a weekend visit, I benefit one child as a foreseen but not intended consequence of giving its sibling a Ping-Pong set, I am not acting merely in accordance with a relevant virtue, say generosity. My benefiting the second child is not merely fortuitous; and more important, this predictable result of my generosity might be both an incentive toward my acting from that virtue again and an indication that I have done something from virtue. A deed not done from virtue may still be at once a result of virtue, a natural sign of it, and identical with an action that, under another description, is performed from virtue.
An action from virtue may, however, be intentional without being premeditated or even deliberate in the strong sense that implies underlying deliberation. Perhaps in some places Kant conceived action from virtue as emerging from pieces of practical reasoning and, on that ground, considered such action deliberate in the strong sense.(7) Depending on the conditions for acting on the basis of practical reasoning, such deliberateness might be entailed by Kantian action from virtue. I prefer to conceive practical reasoning as possible without such deliberation, for instance where one seeks a means, reasons instrumentally the conclusion that it will achieve one's end and concludes in favour or it, but never weighs any alternative.(8)
The matter of voluntariness is more difficult. If "having to do" something, say reprimand a friend, because it is a duty, is a case of not doing it voluntarily, then obviously acting from virtue need not be voluntary. Let us call this discretional involuntariness: you act not at your pleasure but because you "must"; yet you do something that is "up to you". By contrast, volitional involuntariness, the kind reflex "actions" exemplify, bypasses the motivational system, "the will". Discretional voluntariness may be set aside as clearly not necessary for acting from virtue, whereas volitional voluntariness is necessary: actions from an inescapable duty may express virtue, in that special way implied by acting from duty; involuntary "actions" do not go through one's will and cannot express virtue.
The more difficult issue is lack of voluntariness owing to (non-moral) compulsion--coercive involuntariness. If, under threat, I am compelled to keep a promise, can my keeping it be done from virtue, say from fidelity? One would think not; but there is at least one hard case. Suppose that although I would keep the promise owing to the threat, I would also do so from duty, and each reason figures in me as an actual, causally sufficient motive for the deed. Might we say that the action is performed partly from virtue? Arguably, it can express virtue in the right way; it simply expresses fear as well. This is a permissive view, allowing both that actions from virtue be heterogeneously motivated and that motives of virtue not be necessary conditions for the agent's performing the act in question. One might instead hold the strong view that the motive of virtue must be at least a necessary condition for the action. (There the elements of compulsion would not be sufficient: if the compulsive pressure would not succeed without the cooperation of the virtuous motive, then the action is not compelled.) Kant would probably require still more: that a motive of duty operate as both necessary and sufficient in the actual grounding of the action.
Both the permissive and the strong view have some plausibility, though the strong view is less plausible when separated from compulsion, since compulsion may be felt to be a pre-emptive rather than merely cooperating motive, eliminating rather than enhancing the effect of any virtuous motive that is also necessary for the same action. I doubt that compulsion must be pre-emptive;(9) in any case, the wisest course is to allow different degrees to which actions may be performed from virtue, and hence progressively stronger conceptions of acting from virtue. Actions may be performed wholly from virtue when no other kind of motive cooperates; partly from virtue when another kind does; and, in this mixed case, they may be performed from virtue to various degrees depending on the relative weights of the aretaic and non-aretaic motives in producing or sustaining the action. The notion of sustenance is crucial, especially for actions and activities--which I am not here distinguishing from actions--that take a significant amount of time, such as giving a lesson as a fulfilment of the duty to instruct. Self-interest and duty could alternate as sustainers, and an action I am performing from duty at one moment might be performed from self-interest at the next. This possibility is not ruled out by Aristotle's character requirement; for even a firmly entrenched virtue need not preempt a quite different, independent motive that is aligned with it. Entrenchment of a trait in one's character is one thing; exclusivity of its motivational influence on a given action is quite another.
Perhaps even more than Kant, Aristotle expresses what seems a strong requirement on the content of the motivation underlying an action from virtue. When Aristotle says of an action from virtue that one must "decide on it for its own sake", he may appear to be implying that if an action is virtuous under a description, the agent must decide on it under that description (or under the corresponding concept, the one expressed by the relevant description--the requirement would not be linguistic). But surely his examples and overall discussion allow a wider reading: there must be some description of the action relevant to the virtue such that (roughly speaking) the agent decides on the action on account of its fitting that description. This thesis needs both qualification and explication.(10)
If deciding on an action implies selecting it from among options, then the view that an action from virtue must be decided upon--the selection requirement--seems mistaken. There need be no question in the agent's mind of options, e.g. of alternatives to giving each needy person an equal amount of rice. If Aristotle's point is that the action is in some sense "chosen," it is probably true, at least if simply choosing to do something is distinguished from choosing it from among alternatives. But if deciding on an action implies either its prior consideration or, especially, its selection from one or more alternatives, then deciding is not necessary for action from virtue.
The larger question here is not how the elements underlying the action interact in the mind to produce or sustain that action, but what those elements are. What is it that makes a description relevant to the virtue in the right way. That is, what reason-indicating description of a would-be action from virtue is such that if one performs the action for a reason the description indicates, then it is an action from the virtue in question? No name of the virtue need occur in the description; my equal treatment of those in my care can be intentional and done from justice even if I do not, internally or aloud, describe or conceive it specifically as just. Must the description entail that the action is of the kind that "by definition" may be said to instantiate the virtue, e.g. justice (as that virtue applies to action)? I think not. For one thing, it is enough if there is a strong presumption, say a strong probability that justice is served by proportioning the pay one gives to several hourly co-workers to their time on the job; the description of this act would not be, by definition, a case of justice (not an easy condition to satisfy non-trivially, in any case). For another, an action can be done from virtue when there is only good reason for the agent to believe that it meets an appropriate condition. Virtue allows for fallibility, and although there are limits to how far off the mark one can be, action from virtue is consistent with some degree of "unavoidable" error.
Just what is required in the virtuous agent's conception of the action, if that action is to be performed from virtue? The problem is to capture what I shall call aretaic connectedness, the connection between the action and the agent's beliefs and desires, that we must clarify in order to understand action from virtue. Perhaps it is in part the difficulty of explicating this special connection, as distinct from the commonly discussed difficulty of formulating rules for virtuous action, that led Aristotle, and leads other virtue theorists, to try to understand virtuous action, in the sense of the kind of action appropriate to a virtuous agent, in terms of what is consonant with virtuous character, as opposed to understanding such agents in terms of dispositions to perform a kind of action specifiable independently of the virtue, say in terms of the categorical imperative, the principle of utility, or Ross's prima facie duties. Doing the right kind of deed in the wrong way is not virtuous, even if the deed is just the one required by sound principles.
We might begin, in the spirit of Aristotle, with the idea that the relevant range of descriptions is of a kind by which a person of practical wisdom would, in exercising virtue, be motivated, and that these will fall on some dimension from excess to deficiency. This is vague, but if we try to achieve specificity in some of the standard ways, we encounter falsehood; e.g., not just any description implying that the action maximizes human happiness will serve: one could want to do the deed for the wrong kind of reason and thereby fail to qualify, at the time, as a (morally) virtuous agent. For instance, being motivated by considerations of aggregate happiness might lead one to make an optimific but inequitable distribution of rewards to employees. Moreover, virtue theorists, at least, seem committed to denying that aretaic connectedness can be captured by descriptions which, in the light of rules (such as the principle of utility), determine the appropriate action.
One might think that a Kantian approach to acting from virtue could lead us to a solution of the aretaic connectedness problem. But if that approach is to be more than schematic, it must take us from an account of action from a morally acceptable maxim to an account of acting from some appropriately related virtue. It might be argued that all Kantianism needs here is a notion of acting from moral virtue conceived as dutifulness, and it matters little how we distinguish, say, the virtue of justice from that of fidelity, since, independently of these terms, our overall ethical theory will require the same actions in any given circumstance calling for moral decision. But this high-handed approach would leave us with too thin a theory of how to describe, credit, criticize, and even educate, people morally. Even if all we cared about were getting people to do the right thing from some appropriate moral reason, we must surely teach morality in terms of more specific categories--and quite possibly in terms of the "departments" of morality that the virtues can be taken to represent. I believe, then, that for both normative and analytical purposes even a well-developed Kantian ethics needs a better way of clarifying the notions of virtue and of action from it.
3. The cognitive and motivational grounding of action from virtue
I want now to propose an account of acting from virtue built around six notions, corresponding to situational, conceptual, cognitive, motivational, behavioural, and teleological dimensions of such action. These dimensions are, first, the field of a virtue, roughly the kind of situation in which it characteristically operates; second, the characteristic targets it aims at, such as the well-being of others in the case of beneficence and the control of tear in the case of courage; third, the agent's understanding of that field; fourth, the agent's motivation to act in that field in a certain way, where that way is appropriate to the virtue; fifth, the agent's acting on the basis of that understanding and motivation; and sixth, the beneficiaries of the virtue, above all (and perhaps solely) the person(s) who properly benefit from our realizing it: for beneficence, other people in general; for charitableness, the needy; for fidelity, family and friends; etc. These six notions are specially appropriate to explicating action from virtue (and, to a lesser extent, action from emotions and vices), as opposed to actions from very different sorts of dispositions, e.g. boredom, fatigue, and misapprehension. None of those actions has, for instance, a distinct field or target, though emotions, such as love (of one kind) may also have beneficiaries or, like vices, characteristic sufferers, such as the victims of passionate anger; and with fatigue, at least, while the "from" is (as in combination with action-locutions in general) explanatory, it does not imply what it most often does with those locutions: a motivational explanation. (11) Let us first consider the field of a virtue.
The field of justice might be roughly retribution and the distribution of goods and evils; that of fidelity might be conduct required by explicit or implicit promises; and so forth. Such fields may overlap other aretaic fields, but each has some distinctive feature(s). How does a virtuous person understand the field of, e.g., fidelity? It would be natural for the appropriate understanding to manifest itself in believing that promises create a duty to keep them, that working with others generates obligations to them, and so on. But suppose someone did not use the concepts of duty or obligation (at least here) and thought simply that it is good to keep promises and good to criticize people who do not. A virtuous person could be skeptical about moral concepts or think them indistinguishable from aretaic concepts in general. We can imagine someone who, upon making a promise, wants to keep it because that is appropriate to human relationships, and tends to feel disapproval of anyone who does not keep promises, on the ground that the behaviour is inappropriate to such relationships. And of course, a person might want to keep a promise because promise-keeping is commanded by God. A moral field cannot be understood without a sense of its (moral) normativity, but that sense is not restricted either to virtue concepts (as Aristotle may seem to imply) or to hedonic ones (as some utilitarians perhaps tend to think) or to deontological principles (as Kantians may tend to think). It does appear that there are some general requirements for understanding any moral field, e.g. that a kind of impartiality be recognized as necessary, that the well-being of people must be given some weight, and that the relevant norms be, if not "designed" to overrule self-interest, then capable of conflicting with it.(12) It may be that action from virtue requires an exercise of some normative concept, if only that of what is in some appropriate way good or bad; and certainly the possession of a virtue entails a recognitional capacity. A loyal person, e.g., must have a sense of when to act in support of friends: this is part of what it is to understand the field of a virtue, and without it one would not act from virtue. But neither this special requirement for understanding the field of a virtue nor the general requirements for comprehension of a moral field entail that action from a virtue must have any particular motive among those appropriate to its field.
With all this in mind, we can see that aretaic connectedness need not proceed through any direct application of a moral or even a virtue concept (though this apparently does not hold for all normative concepts). This point bears especially on the motivational dimension of acting from virtue. Most important, action from a given virtue need not be internally motivated, i.e. (roughly), performed from a desire to realize that virtue. Let me clarify this by example first, then in general terms. Suppose I see two children dividing apples they have just picked, and I notice that their pickings are about equal. I see one child take far more than half, and I want to intervene. I do so in order to persuade the greedy one to share equally. I may see this persuasion as just; but I may also see it as appropriate to their similar investments of time and energy; as imposing on them the way civilized people should treat each other; as affirming the equality of the two as persons; as promoting harmony between them; or in other normative ways appropriately connected with justice. If I am motivated to intervene on the basis of any of these conceptions, my action seems suitably connected with justice for me to count as acting from it: in the first case the concept of justice applies directly, in the others (on plausible assumptions) indirectly. If I act from any of these motives, I act from justice and, if they are properly grounded in my character, from virtue.(13)
We can discern, then, two ways in which, on the basis of an understanding of the field of a virtue, an agent can act from it: directly and indirectly. Both notions--which we might pair with direct and indirect aretaic grounding need explication. I act directly from, say, justice provided that, first, an adequate concept of justice (whether I would use the word or not) figures centrally in my motivation; second, the content of my motivation is appropriate to justice, as where I want to compensate a victim of wrongdoing; and third, the motivation itself, e.g. a desire to treat people equally in distributive matters, is properly grounded in my character. I act indirectly from justice when an adequate subsidiary concept, such as fairness, is motivationally central in that way, or where (a) my motivation is appropriately subsumable under the relevant virtue concept or a subsidiary one, say where I act in order to divide the children's takings in accordance with their efforts, and (b) the relevant motivation has a specific content appropriate to justice and is sufficiently connected with the relevant aretaic elements to ground the action in them. (The motivation of an action from virtue need not, however, be a standing element in the agent, e.g. a long-term commitment to the moral education of children, as opposed to a desire responding to a unique situation.) The second, indirect case is more complicated. Suppose that need, in addition to effort, is crucial for the justice of the division in question; then my (exclusive) concern with equality of effort will not suffice to subsume, under the concept of justice, my attempt to distribute in accord with effort: I am too far off the mark. I might qualify as trying to act from justice, and even as coming close; but there is a limit to how much one can misunderstand the features of a situation relevant to a virtue and still count as acting from that virtue as opposed to unsuccessfully trying to do the relevant kind of action.
One way to give a theoretical account of acting indirectly from virtue is to assume that an action's being performed from virtue supervenes on natural properties of the action, or at least on some set of properties underlying its virtuousness. The idea would be that an action from virtue is such because of its more basic properties, such as being motivated by a belief that the children should have shares of apples proportionate to their efforts in picking them. The suggestion is meant to be minimally controversial, and for anyone who finds the notion of supervenience unhelpful we could also put the point in terms of a dependence of the virtuousness of an action on other properties of it. Thus, even an intuitionist who thinks that the obligation of beneficence is normatively basic could allow that an action could be performed from beneficence when motivated by properties of the action that are, for the agent, psychologically more basic than its beneficence. The agent might, e.g., conceive the action that in fact is performed from beneficence, not in terms of beneficence but simply as relieving suffering. It is difficult to specify in a general way what properties are basic to an action from virtue, but suppose for the sake of argument that, say, generosity in an action supervenes on its character as a giving of something voluntarily and in the (reasonable) belief that it will benefit the recipient (as opposed to giving it from a sense that it is owed). We can now say that an action is indirectly grounded in a virtue provided it is not directly grounded in it but is based on the agent's believing the action to have a suitable subset of the base properties for that virtue (though not necessarily under this or any other technical description). Roughly, the difference is between aiming at the target of the virtue under the relevant aretaic concept and aiming at it under some appropriate description framed in terms of the base properties of action from that virtue. This need not be all of those properties: if Carol gives more time to students than she thinks they deserve, in order to teach them more, this may be enough to qualify her pedagogical actions as performed from generosity. But one can act from a virtue by acting from its grounds without having that very virtue in mind. This is a pattern characteristic of indirect aretaic grounding.
The case of generosity raises the question whether, for some virtues, indirect aretaic grounding is the only kind possible. Suppose I give a Ping-Pong set to a child, not in order to benefit the child, but simply because, after years of preferring the sibling, I want to behave generously toward this child. Can this act be performed from generosity (and thereby directly grounded in it)? Acting from generosity is surely not entailed by acting from a desire to be generous. Even if one has this virtue, one might be instead acting for the wrong kind of reason, though in accord with the virtue. Surely my action here, unlike a just deed performed from a desire to be just toward those concerned, would not be an action from virtue at all and so would not exhibit direct aretaic grounding.(14)
Suppose, however, that I give the Ping-Pong set both because I want to benefit the child for its own sake and because I want to behave generously. Perhaps I may now be acting from virtue--though not purely so. This would be a case of partial direct aretaic grounding. The action would also be performed in the service of virtue, since I act partly from a second order pro attitude toward generosity. Such higher order attitudes are, however, not necessary for having virtue, nor does acting from them entail acting from virtue. Acting from a virtue requires promoting or otherwise properly dealing with certain elements in its field--its beneficiaries, say children or the needy or the oppressed, where the beneficiaries figure in the target of the virtue. Acting in the service of a virtue requires promoting the virtue. Doing that may affect the beneficiaries little or not at all. There is no need to explicate here all the categories these points bring out. It is enough to have provided a framework for doing that and to have shown that an action's being explicitly directed toward promoting a virtue is neither necessary nor sufficient for its constituting acting from virtue.
The distinction between direct and indirect grounding in virtue is neutral between Kantian and Aristotelian conceptions of virtue and acting therefrom: it applies whether virtues are internalizations of independently knowable moral (or other) principles or whether, on the contrary, moral (or other) principles are knowable only as generalizations from the behaviour of people with virtues of character. The distinction is also neutral with respect to the problems of cooperating motives in ethical theory. My conception of acting from virtue does not require that one be motivated solely by the relevant aretaic ground(s), the ground(s) appropriate to action from that virtue. This exclusivity of motivation is required only for acting purely from virtue. If one acts from both love and a sense of justice, one does not act purely from moral virtue; but love is not a companion that must prevent, as opposed to outshining, one's acting from the virtue of justice.
Imagine, however, that by contrast with a cooperating non-moral motive, I have a further reason; e.g., I am persuading the greedy children to share equally, not for its own sake but in order to promote human happiness, or in order to abide by the will of God. There are at least two relevant possibilities (both also applicable to non-moral virtues): first, that I take the further end to be morally relevant, e.g. to be an appropriate moral ground of the action, and secondly, that I take the further end to be an adequate ground but have no moral conception of how this is so. In the first case, we could say that the ultimate end of my action is moral; and in part in the light of such cases we might adopt what I shall call the moral motivation thesis: that an action from (moral) virtue must be morally motivated, though not always intrinsically so motivated, e.g. performed for the sake of justice. Here my action serves a moral end, but not directly, "in itself"; if it is from, e.g., justice, it need not be internally motivated, although it would be motivated by some consideration consonant with an aretaically internal motive, such as to rectify an unequal distribution.
The second case suggests a moral connection thesis: that given a suitable immediate motive connected in the right way with the moral field of the virtue, the action can be performed from that virtue even if the action's ultimate motivation is not moral.(15) Acting ultimately from love, e.g., seems consonant with acting (partly) from justice, provided the immediate motivation of the action is of the right kind, say a determination to treat people equally. This would be a beneficent, perhaps a natural, kind of justice.(16) If natural justice exists, it shows that action from virtue need not be from a single virtue a point that is in any case implicit in the possibility of acting at once from, say, courage and justice, as where one justly and courageously denies an unfair request made by an intimidating employee.
The moral motivation thesis seems plausible: it appears characteristic of acting from moral virtue that the agent act at least indirectly from a suitable moral motive. The moral connection thesis, which denies that even ultimate moral motivation is necessary for action from virtue, also seems plausible, but is harder to assess (there are of course counterpart aretaic connection and motivation theses, nor should we presuppose a sharp distinction between moral and other virtues). What if moral obligation is rooted in (non-moral) considerations of happiness, or in God's will, conceived non-morally? If this is possible, then it should be possible for considerations of human happiness or of divine will to ground moral actions, and hence to be ultimate, independent motives for actions from virtue. Perhaps we must allow that possibility if we are to have an account of acting from virtue neutral with respect to all of the major moral theories. For suppose that a hedonistic or divine command view of the grounds of morality is correct. Why, then, could I not be acting from moral virtue if I am motivated by hedonic considerations, or by divine command, even if I see no connection between those motives and morality as such (perhaps because I simply do not operate, directly at least, with moral categories)? I am, after all, acting from the grounds of such virtue, and to this extent one might regard my action as (indirectly) aretaically grounded. There would thus be a connection, which I could come to see, between my conception of the action and moral virtue; still, my acting from moral virtue would not require my being, in any direct way if in any way at all, morally motivated.
The right conclusion to draw here may be that it is simply not clear how narrowly we should construe acting from moral virtue (or acting from virtue simpliciter). It is best to distinguish a narrower and broader notion. One might hold that in the case of acting ultimately for a non-moral purpose, say from a desire to treat someone lovingly, the agent, if acting from a moral virtue, is nonetheless not doing so in a moral way. Perhaps so; but must we require of all actions from a moral virtue that they must be performed in a moral way? If acting from virtue were equivalent to acting for the sake of it (in some senses of this phrase), that might be so; but acting for the sake of virtue is not necessary for acting from it. This applies both to acting for the sake of the virtue promotionally, as where one tries to enhance the amount of honesty in the world, and to acting for the sake of it acquisitionally, as where one acts to try to produce the virtue in oneself. The latter case shows that acting for the sake of a virtue is also not sufficient for acting from it.
We should add, then, to the distinction among degrees of acting from virtue which emerge when we consider the cooperation of virtuous with non-virtuous motives, a distinction between (a) acting, to any degree, from a virtue, in the way(s) (conceptually speaking) most closely tied to it as to the grounds of the action, particularly in regard to the action's being conceived in terms of the relevant concept, e.g. justice or fidelity, and (b) simply acting from it in some other way that manifests that virtue. The former cases are most often found among actions that exhibit direct aretaic grounding; one acts both from the virtue and in a certain way under the concept of it. The latter are most often found among actions that exhibit indirect aretaic grounding; one acts from the virtue but not under the concept of it, only under some suitably connected concept. Aristotle sometimes had in mind the former, stronger notion--acting from virtue in the way most closely tied to it as to the grounds of the action; but his overall moral theory, like the most plausible virtue and rule theories in ethics generally, leaves room for the weaker notion. It is clear that in either case the action is rooted in the agent in a way that makes it plausible both to say, with Aristotle, that the action expresses virtue as a feature of character and, with Kant, that it manifests good will. These points, in turn, make clearer an important point suggested above: action from virtue, as opposed to action merely in conformity with virtue, is very important in appraising people: the former, unlike the latter, is commonly a reliable indication of their aretaic character.(17)
4. The moral scope of acting from virtue
In the light of the connections now apparent between virtue concepts and more general ones, both moral and non-moral, we can explore the moral scope of action from virtue and, in particular, whether action from moral virtue is sufficiently comprehensive for a morally adequate life. Suppose (artificially) that one acted only from moral virtue, and always from some moral virtue appropriate in the circumstances. Would this suffice for living a morally adequate life?
An affirmative answer is certainly plausible, at least for those views that take the possession of the moral virtues to be the internalization of some comprehensive set of sound moral standards. If, however, we try to frame a list of moral virtues in terms of which to focus the question, there is greater difficulty. One problem is getting a sufficiently comprehensive list of virtues from the moral domain alone: would justice, fidelity, honesty, and beneficence be sufficient, if broadly construed, or must we add to the agent's repertoire, say, courage and even intellectual virtues, since these seem required for realizing the moral virtues where danger produces fear, or where insufficient information threatens to make reasonable choice impossible? A more difficult problem is how to cover aretaic conflicts: just as obligations of beneficence and fidelity can conflict when beneficently helping someone in distress requires breaking a promise, a virtuous agent can be pulled in two directions by different virtuous tendencies. Here practical wisdom is required in the same way that, for Ross, it seems required to deal with conflicts among the prima facie duties he thought morally fundamental.(18) Practical wisdom is not a specifically moral virtue but a higher order one applicable to reflections and decisions concerning moral and other kinds of virtues. If, as is likely, it is required for a morally adequate life on the part of an otherwise virtuous agent, then the exercise of moral virtue alone is not sufficient for such a life, even if the exercise of virtue, overall, is.
To say, however, that acting from moral virtue is not sufficient by itself for a morally adequate life does not entail that virtue ethics is not sufficient for the action-guiding task of normative ethics. By virtue ethics I mean roughly the kind of ethical position according to which the following two ideas are central: first, the fundamental moral concepts are virtue concepts, as opposed, above all, to rule concepts; and second, the basic normative aims of moral agents are aretaically determined, in the ways we have seen, by the requirements of acting from virtue, as opposed, say, to being dictated by a commitment to following certain deontic rules. Aristotle can be read as holding such a view in some places, for instance in saying that
actions are called just or temperate when they are the sort that a just
or temperate person would do. But the just and temperate person is
not one who [merely] does these actions, but the one who does
them in the way in which just or temperate people do them. (1105b6-9)(19)
Taken as a statement of virtue theory as applied to these traits, this passage implies that what makes an act, e.g., just, is its being the kind a just person as such would perform (in a certain way); we do not explicate what a just person is by first identifying certain types of acts and then characterizing that kind of person in terms of a suitable disposition to perform acts of that kind.(20) This metaphysical conception of virtue ethics is consistent with taking practical wisdom or other higher order virtues as crucial for directing the virtuous agent in everyday life. Thus, a virtue ethics is at least not prevented on that score from providing a basis for the morally adequate life.
There remains, however, a significant problem. Even if the notion of a virtuous person is metaphysically more basic than that of a virtuous action, there is the epistemological difficulty of determining what, or even who, a morally virtuous person is without already knowing what sorts of thing such an agent would do. Can we reasonably take someone as a model of justice or fidelity without relying on some idea of what deeds are appropriate to such a person? If not, how can a virtue theory ever tell us what we should do, even in the matter of building character, if we do not already know? One Aristotelian answer is that if we know our proper function and see how it is properly exercised, i.e., so exercised as to produce a life of flourishing, we can see how the agent in question--the virtuous agent--chooses in matters involving pleasure and pain, which constitute the larger field of moral virtue. In rough terms, the virtuous agent aims at targets appropriate to human flourishing and acts so as to hit a mean between excess and deficiency. Suppose this is correct. There is still a normative notion built into flourishing, and this would seem at best difficult to discern without a sense of what behavioural outcomes are to be sought. Some such outcomes seem essential for hitting the right targets. Are we happy when merely content, or must we perform certain intellectual, aesthetic, and physical tasks with a certain kind of result? Are there not intellectual standards, such as those of logic and mathematics, at least, that must be brought to our activities as guides within which virtue develops? (Aristotle himself must have thought so, for he considered philosophical contemplation the highest happiness and surely saw it as governed by logical and epistemic standards.)(21)
It is true that once we have role models, virtue can be taught by their example and without antecedent (propositional) standards. Historically, then, virtue ethics might operate independently of rule or other non-virtue accounts, such as intuitionism. But conceptually, virtue notions seem dependent on other normative concepts.
This negative conclusion must not be overstressed. It remains quite possible that the moral worth of actions depends on their being actions from virtue: even if virtue concepts cannot by themselves tell us what conduct befits us as moral agents, it may be that the only (or the most) morally creditable way to do the things in question is from virtue. A second major moral thesis is also left open: that even the moral worth--in the sense of goodness--of persons lies in their virtuous character (or lack of it).(22) Together these theses constitute a virtue theory of moral worth, and they may be regarded as partially explicating what it is for character to be morally fundamental. This kind of virtue ethics is consistent both with Kantianism and with other views commonly contrasted with virtue ethics when the latter is construed as embodying a theory of moral obligation.(23)
If there is a conceptual dependence of virtue concepts on other normative concepts, it does not indicate a one-way street. Any moral rules with enough specificity to guide day-to-day behaviour need interpretation and refinement to be useful in making moral decisions. It could turn out that practical wisdom is indispensable in using these rules, and that a basic element in such wisdom is a tendency to seek a reflective equilibrium between plausible rules and virtuous inclinations. Even thoroughgoing virtue theorists can grant rules a place. Such rules as they countenance are generalizations from virtuous conduct, for instance from the choices of the phronimos, rather than, say, formulae for optimizing non-moral good (as for utilitarianism), or specifications of obligatory act-types (as for the Kantian tradition). But these rules still have a degree of authority and can override virtuous inclinations in some cases. At worst, the rules are a generalization from many such inclinations, and these rules may thus imply that inclinations conflicting with them are aretaically deviant. Even making virtues conceptually fundamental need not make them indefeasible sources of moral authority. Similarly, if one could specify the types of actions a virtuous agent should in general perform, practical wisdom and a virtuous disposition would be required for applying the relevant rules in particular cases.
On the broad conception of acting from virtue developed here, it is aretaically grounded intentional action: it is action grounded in virtue either directly, as where the agent acts explicitly in the light of the concept of the virtue in question, or indirectly as where one acts on the basis of a different kind of consideration that is suitably relevant to the virtue in terms of its field and target. Such action is, then,from virtue in being explained by beliefs and desires properly connected with the appropriate aretaic elements in character, and it is virtuous both because of its connection with those elements and because of what kind of action it is. This conception of action from virtue provides a model for understanding moral action in general, conceived as action having moral worth: just as action from moral virtue does not require acting for the sake of moral virtue, and can be grounded only indirectly in it through beliefs and motivation appropriate to the moral virtue in question, moral action, even conceived nonaretaically, need not be performed for the sake of a moral principle or even as an application thereof, and can be grounded only- indirectly in such a principle. Moreover, there are degrees to which one may act from virtue, depending especially on the extent to which non-virtuous motives contribute to the action in question. The moral worth of an action can also depend on the balance among moral and non-moral influences on its performance, and this point applies to non-virtue theories as well as to virtue theories.
One of the largest questions raised by the account of acting from virtue presented here is what values, if any, constrain the development of virtue. If the notion of virtue is not merely historical, not just a notion rooted in the established practices of one or another society, if instead it belongs to a universally valid ethic, then it is apparently not entirely autonomous with respect to moral and other values. Moral virtue seems best construed as a kind of internalization of moral values or perhaps moral principles or other standards of moral conduct. It is not their ground, though it may influence their content through the effort we regularly make to achieve reflective equilibrium between virtuous inclinations and general principles. Moral virtue may ground moral conduct genetically, but not conceptually; and this is confirmed by the way in which we must understand acting from moral virtue: not simply in relation to people with certain traits, but in relation to the reasons for which they act, above all the kinds of reasons pertaining to what is of moral value or to what is morally required by general rules or standards.
In the theory of moral worth, however, virtue is absolutely central. This point is easily obscured by the common attempt to construe virtue ethics as providing by itself an adequate theory of moral obligation. Virtue can be the ground of moral worth even without being the ground of moral rightness or obligatory conduct. Agents cannot truly act morally if, as moral nihilism has it, there are no sound moral standards; but according to both virtue theories and other plausible ethical views, actions gain no moral worth by mere conformity with sound standards of conduct: the right actions performed in the wrong way, and especially from the wrong motives, have no moral worth. The mere existence of objectively true moral standards, even together with our regular conformity to them, would not guarantee moral action--action from virtue or from duty or from any other morally appropriate ground--and might for that very reason bring no moral goodness into the world.
Virtue need not be acquired, moreover, from studying moral values as such; it is normally acquired by imitation and socialization, and it probably cannot be taught without models. These two truths do much to account for the appearance of conceptual autonomy the notion of virtue seems to have. In normal human lives, virtue may be genetically prior to moral principles. It also has a kind of operational autonomy, both in the sense that one can act from virtue without being motivated by any aretaically external standards and even in the sense that one's immediate motivation need not be moral at all. Whether the fundamental moral standards are rules or intuitions or non-moral goods or something else again, virtue is required to realize those standards, and acting from virtue is the main basis of the moral worth of agents.(24)
(1) This is not exactly his wording, but the distinction seems clearly implied in his contrast between merely doing just and temperate deeds and doing them in the "way in which just and temperate people do them". See NE 1105a25-1105b15. Terence Irwin's translation (1985) is used throughout.
(2) See especially Kant (1959, pp. 397-400, and p. 406). Lewis White Beck's translation is used throughout, and the Academy numbers are given in the text.
(3) Kant says, e.g., "virtue signifies a moral strength of will . . . the moral strength of a man's will [as opposed to that of a "superhuman" being] in fulfilling his duty, a moral necessitation by his own legislative reason in so far as reason itself constitutes a power of executing the law." See the "Introduction to the Doctrine of Virtue" (1964, p. 66, A. 404).
(4) For discussion of some of the possible kinds of mixed motivation and some of Kant's theoretical options concerning them, see Audi (1989, ch. 3). I would add, on the bearing of the categorical imperative, that (1) we surely need not be treating people merely as means, or failing to treat them as ends, if we act justly toward them both from a just character and from love, (2) perhaps in such a case one might still regard the relevant maxim of the action as one formulable in terms of the just motive alone. The motive of justice might have to be not only sufficient but primary relative to that of love; but even then the motive of love could play a significant motivational role.
(5) For a defence, see Audi (1989, ch. 3). Note that Aristotle's term prohairesis is usually translated as "decision", when Aristotle himself describes it as "deliberative desire" (113a10). Meyer calls it "the desire most important to virtue and vice of character" (1993, p. 24). Cf. Broadie (1991, pp. 78-9). If it is any kind of desire, "decision" is not quite the right word, since, unlike desires, decisions are made and are events in the ordinary sense entailing change.
(6) Even qualified by "normally" this point may be too strong. If a humble person characteristically and "automatically" does not intervene in an argument between parties who, though competent, know less about the topic, might this be both an action from humility and non-intentional? Supposing the answer is affirmative, it still appears that an action from virtue must have some description under which it is performed from virtue and intentional, e.g. avoiding the appearance of instructing people.
(7) Kant's famous four examples in Sc. 2 of the Foundations, especially in their first occurrence intended to illustrate universalization of maxims, would be a case in point.
(8) A possibility argued for in Audi (1989, ch. 5). Cf. Broadie (1991, pp. 85-9).
(9) What chiefly makes it seem so is that for typical compulsions the agent will be preoccupied with, e.g., avoiding the threatened consequence; but preoccupation with one motivating scenario need not be proportionate to its impact on action: people can do things mainly for prestige or money while stressing to themselves and others that they are acting from charity or friendship.
(10) Cf. the view, commonly attributed to Aristotle, that "Our rational actions [including actions from virtue] are the actions we perform because we think they will contribute to our happiness" (Meyer 1993, p. 25). Rational action is subordinate to our desire for happiness, but I doubt Aristotle implies that all such action is directly aimed at it. For a defence, see Audi (1989, ch. 1).
(11) I thank the Editor for drawing my attention to this contrast (which deserves independent elaboration on another occasion). I must also forgo addressing contrasts between actions from virtue and (certain) actions from vice (e.g. slovenliness) and from emotion (e.g. anxiety). Still, some of what is said below about action from virtue should help in clarifying action from vices and emotions, especially where the latter is intentional and to the extent that vices or emotions are constituted by desires or beliefs or combinations thereof.
(12) One problem is how to characterize the moral point of view. For discussion of this see Kurt Baier's book of that title (1958); Frankena (1973); and Gert (1988). I assume that just as one can take the moral point of view even if one regards it as derivative from that of rationality, people can take the moral point of view even if they consider moral standards theologically grounded, and that there is thus a way to conceive divine commands so that obedience to them can be morally, not just religiously, motivated. A similar problem is how to square the possibility of conflicts between morality and self-interest with the kind of egoism apparently implicit in Aristotle and others who offer a plausible ethical theory from egoistic starting points. The beginning of an answer is that for Aristotle, while moral virtue is essential to our flourishing and hence moral conduct tends to serve self-interest, long-run self-interest has social dimensions and can thereby conflict with moral demands.
(13) Acting from justice would be acting for the relevant reason. Acting for a reason is a complicated notion which I presuppose here, an account of it, and appraisal of other accounts, is given in my (1986).
(14) This sort of problem is insightfully discussed by Williams (1985, pp. 10-1). He has suggested (in conversation) that justice is one virtue such that action from it must be internally motivated. Justice does seem the best candidate for a virtue meeting this condition, but I doubt that it does--unless, perhaps, we restrict it to a specific kind, e.g. the "trait" of being retributively just (Broadie 1991, p. 88). If we construe any trait narrowly enough, actions from it will be correspondingly restricted as to appropriate motivation.
(15) An immediate motive is one that the action is performed (directly) in order to satisfy, as where one drinks simply to slake thirst, but one could drink to rehydrate the body, which one does in turn to avoid suffering. If one does not avoid suffering in order to satisfy a still further motive, this (self-protective) motive is ultimate. Motivational chains can be long, so that the connection between an action and a virtue ultimately grounding it may be extended; and since the (direct) in-order-to relation is non-transitive (in the sense that one can A in order to B and B in order to C, yet not A in order to C), an action's ultimate motives need not underlie it in the way its immediate motives do.
(16) This contrasts with the austere principled kind of justice Kant would have us cultivate and act from. He is not, however, committed to treating acting from virtue as something we can do "at will", nor is there a direct duty to satisfy this description.
(17) Hume goes so far as to say it is the motive, not the action, that (directly?) deserves praise or blame (Treatise, p. 477; cf. p. 464).
(18) See especially Ross (1939, ch. 2). The sorts of problems emerging in the text may indicate one reason why Aristotle might have thought the virtues unified; reflection on the problems certainly suggests that in a virtuous person at least many virtues are interconnected, but that by no means requires a strong kind of unity.
(19) Cf. NE 1129a7-9, which seems at least to reverse the emphasis: "the state everyone means in speaking of justice is the state that makes us doers of just actions".
(20) A contemporary defender of virtue theory especially sensitive to some of the conceptual problems arising here is Jorge Garcia (e.g., Garcia 1989, pp. 277-83).
(21) Some of the large literature on Aristotle's conception of happiness supports my points here. See, e.g., Cooper (1975); Kraut (1979), Maclntyre (1981) Broadie (1991); and Prior (1991).
(22) Moral worth in the sense of dignity is a different, capacity notion; but it is related: it is largely the capacity for good character.
(23) Regarding Kant, we would have to assume that actions from duty, which have moral worth, are also performed from virtue. Some passages in Hume suggest he might be committed to the virtue theory of moral worth, e.g., "'Tis evident, that when we praise any actions, we regard only the motives that produced them ... The external performance has no merit ... all virtuous actions derive their merit only from virtuous motives, and are consider'd merely as signs of those motives" (1888, pp. 477-8). Even a utilitarian can hold the virtue theory of moral worth--though only as a contingent truth--since virtuous character, or its producing actions, need not contribute to intrinsic value. Frankena suggests that "a man and his actions are morally good if it is at least true that, whatever his actual motives in acting are, his sense of duty or desire to do the right is so strong in him that it would keep him trying to do his duty anyway" (1973, p. 70). This differs from the virtue theory of moral worth in at least two ways: the relevant actions need not be performed from virtue (since the actual motives are not crucial), and the content of the relevant motivation is both specifically moral and indeed restricted to the deontic concept of right and duty. (However, I see nothing in Frankena's overall position that requires his holding either the first, permissive, thesis or the second, restrictive one.)
(24) An earlier version of this paper was given at Santa Clara University's Conference on Virtue Ethics in March 1994, and I benefited from discussion with the other speakers, especially Philippa Foot. I also want to thank Norman Dahl, Julia Driver, Philip Kain, Christopher Kulp, Michael Meyer, William Prior, Elizabeth Radcliffe, and the Editor and an anonymous referee for helpful comments
Aristotle: Nicomachean Ethics. Translated by Terence Irwin 1985. Indianapolis: Hackett.
Audi, Robert 1986: "Acting for Reasons". The Philosophical Review 95, 4, pp. 5 1 1-546.
--1989: Practical Reasoning. London and New York: Routledge.
Baier, Kurt 1959: The Moral Point of View. Ithaca: Cornell University Press.
Broadie, Sarah 1991: Ethics with Aristotle. New York and Oxford: Oxford University Press.
Cooper, John: 1975. Reason and Human Good in Aristotle. Cambridge, Mass: Harvard University Press.
Frankena, William K. 1973: Ethics. Englewood Cliffs, NJ: Prentice-Hall.
Garcia, Jorge L. A. 1989: "The Problem of Comparative Value". Mind 98, 390, pp. 277-283.
Gert, Bernard 1988: Morality. Oxford: Oxford University Press.
Hume, David 1888: A Treatise of Human Nature, ed. by L. A. SelbyBigge. Oxford: Oxford University Press.
Kant, Immanuel: Foundations of the Metaphysics of Morals. Translated by Lewis White Beck 1959. New York: Liberal Arts Press.
--: "Introduction to the Doctrine of Virtue". Translated by Mary McGregor 1964. New York: Harper and Row.
Kraut, Richard 1979: "Two Conceptions of Happiness". The Philosophical Review 88, 2, pp. 169 - 197.
MacIntyre, Alasdair 1981: After Virtue. Notre Dame: University of Notre Dame Press.
Meyer, Susan Sauve 1993: Aristotle on Moral Responsibility. Oxford: Basil Blackwell.
Prior, William J. 1991: Virtue and Knowledge. London and New York: Routledge.
Ross, W. D. 1939: The Right and the Good. Oxford: Oxford University Press.
Williams, Bernard 1985: Ethics and the Limits of Philosophy. Cambridge, Mass: Harvard University Press.
|Printer friendly Cite/link Email Feedback|
|Author:||Audi, Robert; Katz, Jerrold J.; Williamson, Timothy; Weiner, Joan|
|Date:||Jul 1, 1995|
|Previous Article:||Language and Time.|
|Next Article:||The Seas of Language.|