From Duty and for the Sake of the Noble: Kant and Aristotle on Morally Good Action
From Duty and for the Sake of the Noble: Kant and Aristotle on Morally Good Action
Abstract and Keywords
Aristotle believes that an agent lacks virtue unless she enjoys the performance of virtuous actions, while Kant claims that the person who does her duty despite contrary inclinations exhibits a moral worth that the person who acts from inclination lacks. Despite these differences, this chapter argues that Aristotle and Kant share a distinctive view of the object of human choice and locus of moral value: that what we choose, and what has moral value, are not mere acts, but actions: acts done for the sake of ends. Morally good actions embody a kind of intrinsic value that inspires us to do them from duty (in Kant) or for the sake of the noble (in Aristotle). The chapter traces the difference in their attitudes about doing one's duty with pleasure to a difference in their attitudes towards pleasure itself: Aristotle sees it as a perception of the good, while Kant thinks of it as mere feeling.
Philosophers have long supposed that Aristotle and Kant disagree about many fundamental issues in moral philosophy. Aristotle tells us that an agent lacks virtue unless he enjoys the performance of virtuous actions, while in the Groundwork Kant seems to claim that the person who does her duty in the teeth of contrary inclination displays an especially high degree of moral worth. Aristotle argues for the virtuous life by attempting to prove that, given the human telos, some form of the virtuous life is the happiest that we can live. Kant scorns appeals to happiness as irrelevant to morality and bids us remember the special vocation of an autonomous being. Aristotle emphasizes the difficulty of formulating general principles of action, and the important role of judgment and perception in practical deliberation. Kant, on the other hand, provides us with a method for testing proposed maxims to see whether their actions are permissible, forbidden, or required. And finally, Aristotle has lately been categorized as a “virtue theorist” who holds that an action's value consists in its being the expression of a virtue; while Kant is supposedly a deontologist who thinks that the value of an action rests in its conformity to a rule.
Yet behind these contrasts, apparent and real, is one undeniable similarity. Aristotle and Kant both believe that in human beings, reason can be practical. This is a view about what specifically human action is, or about how human action is different from that of the other animals. It is the view that human beings exercise choice, in a specific sense that I will explain in this essay, in the determination of our actions. Since moral or ethical value pertains only to human action, it seems natural to think that it is somehow related to, or supervenes on, the specific character of human action. And I think that we do find this idea in both Aristotle and Kant. Both of them believe that the moral value of an action is a function of the way in which it is chosen.
(p.175) I believe that these claims about the practical employment of reason are deeper, both in fact and in Aristotle and Kant's theories, than philosophers have generally recognized. To say that human beings are rational is not just to say that we are rule‐following or logical, but rather to say that we are capable of authentic mental activity, of an engagement with the world that goes beyond mere reaction. In Aristotle's account of theoretical reason, the ultimate expression of our rational nature is our participation in the active intellect that imparts form and intelligibility to the natural world. In Kant's more skeptical account, it is the mind's attempt to construct a systematic, unified, and intelligible world out of the confused mass of phenomena that are presented to it. To say that reason is also practical is to say that our actions, the expressions of our wills, can in a similar way be fully active, self‐generated, or in Kant's special sense, spontaneous.1 And if morality is the full expression of practical reason, then this is the distinguishing feature of the moral agent: that her actions are more truly active, more authentically her own, than those of agents who fall short of moral goodness. To have the distinctively moral attitude, then, is to have an active as opposed to a merely reactive relationship to the world around us.2
At the same time, both of these philosophers were aware that the human mind (unlike the divine one) is also passive or receptive with respect to the world. A central concern of both Aristotle's book On the Soul and Kant's Critique of Pure Reason is to explain the respective contributions of activity and passivity to our mental lives. And this is a central concern in the ethical writings of both philosophers as well. Aristotle and Kant of course acknowledge that passions, inclinations, and impulses, as well as reflective deliberation and choice, play an important role in the determination of action. And I believe that for both, the concern of a theory of virtue, in particular, is to explain how that role may be accommodated in a theory of rationally governed and so authentically self‐generated conduct.
These are large claims, and obviously I cannot undertake to defend them in any adequate way here. I offer them as background to the argument I am going to make, which concerns the very first contrast that I mentioned. In the first (p.176) section of the Groundwork, Kant claims that a person who helps others with pleasure from motives of natural sympathy displays no moral worth, while a person who lacks any natural inclination to help others but nevertheless does so, from the motive of duty, does display moral worth (G 4:398). This appears to be in stark contrast with Aristotle's claim that it is the mark of a good person to take pleasure in moral action (NE 1.8 1099a16–21). In this essay I will argue that this apparent contrast does not reflect any ethical disagreement between the two philosophers at all. There is a disagreement at work here, but it is psychological rather than ethical. My argument will take the following course. In section 1, I will look at Kant's view of what gives an action moral worth, as presented in the first section of the Groundwork, and in the course of that explain why Kant says what he does about the naturally sympathetic person. In section 2, I will argue that Aristotle holds an essentially similar view about what gives an action moral worth. Both philosophers, I will argue, think that what gives an action moral value is the fact that it is chosen for its intrinsic rightness. Finally, in section 3, I will return to the question of the value of acting from natural inclination, and try to explain the real source of Kant and Aristotle's apparent difference on this point.
1. Acting from Duty
Section 1 of the Groundwork opens with a claim that Kant believes his readers will accept, namely, that the only thing in the world that has unconditional value is a good will. The good will is good “only because of its volition” (G 4:394), which means that it is in actions expressive of a good will, morally good actions, that we will see this unconditional value realized. Now the project of section 1 of the Groundwork is to discover the principle of the good will, for this will be the moral law. Kant's idea is this. Good‐willed actions are good because of the way that they are willed, or, as I will put it, chosen. So once we know how they are chosen, we will know what makes them good. Since the moral law tells us to perform good actions, it will tell us to perform actions that have that feature—whatever it is—that makes actions good. Since you and I already know how the investigation turns out, I can perhaps try to say this more clearly. Kant thinks that what makes an action good is that its maxim qualifies to be a universal law. So what he is going to try to show is that the principle of a good will is that of choosing actions whose maxims qualify to be universal laws. That's what good people think about when they choose their actions—whether their maxims qualify to be universal laws.
Now in order to bring this out, Kant says, he is going to look at a particular class of good actions, namely, those that are done from duty. Duty is the good (p.177) will operating under “certain subjective limitations and hindrances, which, however, far from concealing it and making it unrecognizable, rather bring it out by contrast and make it shine forth all the more brightly” (G 4:397). In order to discover what is distinctive about good‐willed actions, what their principle is, Kant proposes to compare actions done from duty with actions done from other kinds of motives, to see what makes them essentially different. He mentions three other kinds of actions. Actions which are recognized as contrary to duty are set aside. It is worth attending to Kant's own words here; he says: “I here pass over all actions that are already recognized as contrary to duty, even though they may be useful for this or that purpose, for in their case the question whether they might have been done from duty never arises” (G 4:397; first emphasis mine). I take Kant to be saying that any value these actions may have must come from their utility. Kant also sets aside, and for the same reason, actions that are in accordance with duty but that are not chosen for their own sakes. The prudent merchant who always charges honestly because a good reputation helps his business exemplifies this category. Kant clearly takes it to be obvious, just as Aristotle does, that a morally good action must be chosen for its own sake. But being chosen for its own sake is not sufficient to make an action morally good. This point is brought out by the next three examples, in which people who act from duty are contrasted with people who do the same actions from direct or immediate inclination. It is possible to do an action for its own sake just because it is what you like to do. The naturally sympathetic person's action falls into this category. Kant says:
Some readers have supposed that what Kant is saying here is that the sympathetic person is really acting for the sake of his own pleasure; that is, that his real purpose is to please himself. According to this view, Kant believes in some sort of psychological hedonism about non‐moral motives, and so supposes that our inclinations are all selfish.3 Such a reading would be (p.178) inconsistent with several of the things Kant says, some of them in the passage I just quoted. Kant characterizes the sympathetic person as “amiable” and without any motive of self‐interest, for instance. Most importantly, however, it flies in the teeth of the conclusion Kant draws from these examples, which is this:
there are many souls so sympathetically attuned that, without any other motive of vanity or self‐interest they find an inner satisfaction in spreading joy around them and can take delight in the satisfaction of others so far as it is their own work. But I assert that in such a case an action of this kind, however it may conform with duty and however amiable it may be, has nevertheless no true moral worth but is on the same footing with other inclinations. (G 4:398)
What makes this clear is precisely the fact that a person who does a beneficent action from immediate inclination and a person who does one from duty have the same purpose—namely, to help someone. Both of these people help others for its own sake. Kant goes on to assert that what gives an action moral value, then, is not the agent's purpose, but rather the “maxim” or “principle of volition” on which it is done.4
That the purposes we may have for our actions, and their effects as ends and incentives of the will, can give actions no unconditional and moral worth is clear from what has gone before. (G 4:400)
(p.179) To understand these claims it is necessary to understand the psychology behind them. According to Kant, our nature presents us with what he calls “incentives” (Triebfedern) which prompt or tempt us to act in certain ways. We might say that the incentives present certain actions along with their ends to us as eligible. We are, at least in part, passive with respect to these, although that is a remark I will qualify later. Among these incentives are our ordinary desires and inclinations. Now the incentives do not operate on us directly as causes of action. Instead, they are considerations that we take into account in deciding what to do. If you decide to act on an incentive, you “make it your maxim” to act in the way suggested by the incentive. How do you decide that? In the Groundwork and even more specifically in Religion Within the Limits of Reason Alone, Kant suggests that there are two principles of volition or choice that might govern this decision: morality or self‐love (REL 6:36).5 If you are operating under the principle of self‐love, your choice is to do what will gratify you, what will satisfy your desires. Kant's point about the naturally sympathetic person is that he is acting under this principle of volition or choice. The trouble with him is not that he wants to help others only because it pleases him to do so. The trouble is that he chooses to help others only because he wants to. His action is chosen as a desirable one, one which he would enjoy doing.6
The person who acts from duty, by contrast, chooses the action because she conceives it as one that is required of her. And here we must be careful to draw the lesson from what has gone before. The point is not that her purpose is “to do her duty.” As I said before, she chooses the action for its own sake: her purpose is to help. The point is that she chooses helping as her purpose because that is what she is required to do. Kant takes this to be equivalent to being moved by the thought of the maxim of the action, the principle of doing it, as a kind of law. The dutiful person takes the maxim of helping others to express a requirement. Rather subtly, the contrast between doing the right thing from duty and doing the right thing from immediate inclination is also supposed to show that seeing a maxim as a law is attending to something about its form (p.180) rather than about its matter. The person who acts from immediate inclination and the person who acts from duty in a sense act in accordance with the same material principle, which Kant specifies as “to be beneficent where one can” (G 4:398). But the person who acts from self‐love sees such action as desirable, while the person who acts from duty takes that principle to be a law. So the dutiful person's principle of volition is to act on those maxims that have the form of a law. This also makes a difference in the kind of value that these two agents accord to the action. For the person who acts from inclination, the action has an extrinsic value, a value that it inherits from his own desires. But the person who acts from duty sees the value of the action as intrinsic: lawlike form is a property that is internal to the maxim and so is in the action itself.7
Later in the Groundwork (G 4:421–423), and also in the second Critique (C2 5:67–70), we learn more about what this thought involves. To think about whether your maxim has the form of a law is to think about whether you could will to be part of an order of things in which everyone acted in the way specified in the maxim. So it is to think about what sort of world this would be if everyone acted as you propose to act. The dutiful person helps, then, because the vision of a world in which people do not help one another is in a certain way unacceptable to her, and she is moved by that fact. I do not mean, of course, that she is moved by some thought about the consequences of what she actually does—she does not see doing a beneficent action as a way of producing a world in which everyone helps. It is rather that, if we could not choose to live in a world in which no one helps, if we find that we must will that people should help, then the principle of beneficence must be a law by its very nature, a law in itself. Seen this way, it looks as if the difference between the two characters is that the dutiful person has a further thought about helping, or takes a more reflective stance towards it, than the naturally sympathetic person. Helping is not just something that it is nice to do, but something that one must do, because of the sort of action that it is. Since the good person chooses her actions by attending to their lawlike form, that is what the moral law instructs us to do—to choose those maxims which have lawlike form.
(p.181) There is a hitch in this argument, which I will come back to in a moment. At this point I want to focus your attention on two important features of the account. First, Kant gives us what we might call a double‐aspect theory of motivation. An agent's motivation to act involves two things—the incentive that presents the action along with its end as eligible, and the principle of volition that governs the agent's choice to act on that incentive. Second, moral value rests specifically in the principle of volition that is exercised in the choice of the action. Moral value supervenes on choice.
This has several important implications. One is that on this account the presence or absence of a natural inclination makes no difference to the moral value of the action. It is obviously possible to choose an action because you see it as intrinsically required while also thinking that it will be a pleasant thing to do. Kant chooses to discuss cases of good‐willed motivation in which no inclination is present—that is, cases of action from duty—for exactly the reason he says he does, because in such cases the operation of the moral principle is especially perspicuous. Relatedly, as I said before, the problem with the naturally sympathetic person is not that he has an inclination and it is not that his inclination is covertly selfish, that his own pleasure is his real purpose. His inclination is disinterested, which is why Kant says he is amiable. The problem is that he chooses to help others only because he has this inclination. His principle of volition is the problem—it is the principle of doing what he likes to do.
Now in one way this makes it look as if Kant is, after all, saying that the naturally sympathetic person is covertly selfish. If a person chooses to satisfy his inclinations because it gratifies him, doesn't that after all show that he looks to himself? In sorting this issue out I think it will be useful to make a distinction that Kant doesn't make. Kant thinks that when we choose an action we employ some principle of volition. But obviously he does not mean that we always consciously recite this principle in our minds or even that we are always aware of it as we make the choice.8 Sometimes, the principle is just implicit in the way we make the choice. Now this suggests that we can distinguish between more and less reflective versions of both of the characters we are considering here. The unreflective sympathetic person may simply be thinking “I want to help” or perhaps just “this person needs help” and he is moved by that thought. As Kant imagines him, it is his natural inclination to make others happy that interests him in helping, which is why Kant thinks he is implicitly or tacitly acting under the principle of self‐love. (p.182) The more reflective sympathetic person who consciously employs the principle of self‐love entertains a further thought, but it is a thought about himself, not a thought about helping: “Doing this sort of thing makes me happy, makes me feel good, so I will.” Or we might even imagine that he does a calculation of prudence, and works out that, of all the activities that he finds attractive, helping others will make him happiest: he enjoys it, it makes people like him, and it lacks some of the untoward side effects that other pleasant pursuits may have.9 In this case, the pleasure he takes in helping may be disinterested, but the decision to pursue that pleasure is not. Similarly, a person may act from duty in a completely unreflective way, simply thinking of an action as required, without thinking much about why it is so or even without really thinking that there is a reason why it is so. We might think here of some ordinary conscientious person who has simply accepted the conventional or religious moral system according to which he was brought up. But of course there are more sinister entries into this category: the Nazi soldier who thinks of “duty” as carrying out the orders of his superiors comes immediately to mind. This, as I have tried to bring out, is not how Kant is thinking of the person who acts from duty. For Kant, to act from duty is not just to be moved by a blank conviction that an action is required, but rather to be moved by a more substantial thought which inherently involves an intelligent view of why the action is required.10
With this distinction in hand it is possible to make certain points. First of all, if we imagine the reflective versions of these two characters as I have just described them, it is not hard to accept the claim that the person who acts from duty exhibits a moral worth which the person who acts from inclination (p.183) lacks. The agent who consciously employs the principle of self‐love in his choice does seem to look to himself; in fact he seems to choose beneficence as one might choose a hobby. Second, I think it is pretty clear that many of the readers who find what Kant says here wildly counterintuitive are in fact comparing the unreflective versions of both of these characters. If we compare the person who helps impulsively, thinking nothing but “this person needs help” and being moved by that thought, with a person for whom duty is just blind obedience to an abstract rule, then the first of these two characters seems much more attractive than the second. The right thing to say to such readers, of course, is that this is simply not the comparison that Kant has placed before us.
But it does raise a question about the comparison that Kant has placed before us. I think that the comparison that Kant has placed before us is between the more attractive members of each of these two pairs. Kant means to compare the unreflective sympathetic person, who thinks simply “this person needs help,” with the more reflective person who acts from duty with some comprehension of why helping is required. But it is also clear that Kant thinks that the unreflective sympathetic person is tacitly or implicitly acting on the principle of self‐love. At least, this is what I am supposing Kant means when he says the action is on a level with other actions from inclination (G 4:398). So Kant's view seems to be that if you act unreflectively, the principle of self‐love is your principle of volition by default. Why does Kant think this?
One answer that I think we should reject is that your tacit principle of volition is what you would say about your choice if you were invited to reflect on why you made it. According to this view, if we asked the impulsively sympathetic person why he helps, and he started to think about it, all he could say is: “I just like to; it gives me pleasure.” Perhaps this is why he helps, but once he starts to reflect on his reasons, it will be natural for him to switch from a merely theoretical self‐scrutiny to a more practical form of reflection. The question “why do you help people?” is naturally understood as a request for justification and so transmutes into the question why one should help people. So the claim is not that “it gives me pleasure, I like it” is what the unreflective sympathetic person would say if he started to think about why he helps. If he starts to think about why he helps, something altogether different will happen.11 In fact Kant's argument relies upon this point, since he thinks that the pursuit of reflection—that is, enlightenment—will lead us (p.184) to a recognition of the categorical imperative as the law of our own autonomy and so to the good.12
So I think that all Kant means is this: so long as you haven't reflected on why you help, you are just following your inclinations where they lead. And so long as you are just following your inclinations where they lead, your choice is implicitly governed by the principle of doing what you are inclined to do, what you like. To say that the naturally sympathetic person acts from self‐love is not to assign him an unconscious ulterior motive, or a secret selfish thought. It is, precisely, to record the fact that he hasn't thought, that he is allowing his choices to be governed by his natural inclinations, and so is simply following where nature leads.
Now I want to come back to the hitch I mentioned in the argument. I said earlier that the conclusion is that the good‐willed person attends to the lawlike form of her maxim in making her choices, and therefore that that is what the moral law tells us to do: to respect, or attend to, lawlike form. But there are, in fact, two senses in which a maxim may have lawlike form. A maxim may be one that can be willed as a universal law—it qualifies to be a law. Action on such a maxim is permissible. Or the maxim may be a law in the sense that it must be willed, that is, it expresses a duty. The conclusion Kant is looking for in the first section of the Groundwork is that a good action is one whose maxim qualifies to be a law. The principle of a good will is to act on a maxim only if it can serve as a universal law. But Kant has chosen to focus on the more specific category of duties, actions whose maxims must be willed as laws. So the hitch is this: it is unclear how Kant wants us to make the step from the fact that people who act from duty choose their maxims because they see them as principles which must be willed as laws to the conclusion that the principle of a good will is that of acting only on maxims which can be willed as laws.
I do not know of a smooth way to rescue the presentation of the argument in the first section of the Groundwork, but it seems clear enough that these two ideas are related. One connection between them is revealed in the negative way in which the categorical imperative is here formulated—“I ought never to act except in such a way that I could also will that my maxim should become a universal law” (G 4:402). Using this formulation we discover that a maxim must be regarded as a law by discovering that the opposite maxim—the maxim of not doing the action in question—cannot be regarded as a law. That Kant has this connection in mind is clear from the fact that, in two of (p.185) the examples, it is the same person who first acts from immediate inclination and, later, when he has lost his inclination through sorrow or adversity, acts from duty.
Consider, for instance, the naturally sympathetic person. At the beginning of the story, he is a happy person, and full of spontaneous sympathy for others. He sees people living on the street and he feels sorry for them. He gives them food or money, and he likes to see the relief and gratitude in their eyes. He enjoys spreading joy around him as he goes about his own business. He's sympathetic and so their delight gives him direct pleasure. Then bad things happen to him. Maybe his wife gets cancer, or his child runs away, or his work, which once seemed promising, comes to nothing. And there is no pleasure in anything for him any more. He is absorbed in his own sorrows, and he has no sympathy for anyone else. But since he has always given to those in need before it occurs to him that he might, and now for the first time he thinks about it. Or maybe one day he passes by someone whom he has often helped before and the person says: “Aren't you going to help me today? You helped me last week, and I am just as hungry as I was last week.” And now our hero says to himself, “After all, it is not just because it gives me pleasure that I should help. These poor people are living on the street, and they don't have enough to eat. Someone must help them! What sort of world would we live in if no one helped people who are in need?” Moved by this thought, he helps. And now for the first time his helping has moral worth.13
This is the kind of story Kant has in mind in the first section of the Groundwork. But what I have just said might lead to a misunderstanding. I do not mean to suggest that the only reason a good‐willed person asks whether her maxim can be a law is as a way of ferreting out those maxims which must be laws. The categorical imperative test is not a kind of Geiger counter for discovering whether there are any duties in the neighborhood. I do not know (p.186) how to fit this point into the argument of the first section of the Groundwork, but I believe that Kant's thought is that a reflective person asks herself whether the consideration on which she proposes to act may really be treated as a reason to act. To ask whether a consideration is a reason is to ask whether it may be taken as normative. And that, in turn, is to ask whether the maxim of acting on that consideration can be regarded as a kind of law. When we experience some incentive to act—say a desire or inclination—you might say that our nature makes a proposal to reason. The proposal is a maxim and it includes a purpose: do this for the sake of that, or do this for its own sake. Reason steps back and considers the proposal, that is, it considers the action as a whole, including the purpose, and determines whether it is a good thing, a thing to be done, or not. Its decision is an act of volition, performed in accordance with a principle of volition. So to choose an action is to be moved by the conception of the impulse to do it as a reason. And its being a reason is an intrinsic property, a property of the maxim's form. Reason says yes to the proposal if it can recognize its own form, the form of normativity or law, in the maxim. In that case, the reason for action and so the action itself, having reason's endorsement, are good.
But actions done from duty are reason's own actions in a special way. To see this, recall once again the person whose natural sympathy is blunted by sorrow, but who still helps from the motive of duty. He tests the maxim of not helping, and he finds he must reject it. He is thereby moved to help. What is the incentive in this case? Kant's answer is that it is the feeling of respect for law. The very thought that shows him his duty—the thought that one must help those in need—in this case operates as the incentive.
Kant thinks that we cannot say how it is possible for reason to provide an incentive, since that is identical to the question how reason can be practical. But in the second Critique he undertakes to describe what happens in us when we are so moved. Let me just quickly sketch this account. Human beings, according to Kant, have a natural tendency to treat our desires and inclinations as authoritative—that is, to think that the fact that we want to do something is in and of itself a reason for doing it. Kant calls this tendency “self‐regard” or “self‐love” and it is more or less identical with the tendency to operate under the principle of self‐love as I have described it. There are two strands to this tendency—the selfishness that makes us long for the satisfaction of our inclinations, and the self‐conceit that inclines us to take the bare fact that we want to do something as a justification for doing it. When the moral law commands us not to do an action to which we are inclined, it thwarts the inclination, and it humiliates our self‐conceit. These feelings are painful. At the same time, however, we experience an awareness of our freedom, which is (p.187) revealed by our capacity to set inclination aside. We experience freedom as a sense of independence from the neediness of inclination, a sense that is akin to pleasure in that it resembles the divine bliss. The complex mix of affect that results is the feeling of respect for law. Respect for law is not a desire to obey the moral law, or more generally a feeling that exists independently of the law and interests us in it. It is the law itself, the very thought of a requirement, operating as an incentive.14
When we are motivated by respect for law, the rational will provides not only the ground of choice but also the incentive to act in accordance with that ground. Since the incentive as well as the volition are reason's own productions, a person who is motivated by duty is to an especial degree active and truly spontaneous. She is not reacting to nature's proposals at all, but actively imposing on her own actions, and through them on the world, a kind of shape or form that is the dictate of her own mind. This is the fullest expression of autonomy, and it is this that gives her actions their special moral worth.
2. Acting for the Sake of the Noble
Aristotle, I will now argue, holds a similar conception of what gives actions moral value. That is, he also holds a double‐aspect account of motivation; and he holds that what gives actions value is the way that they are chosen. Three aspects of his theory may be mentioned in support of these claims: first, the possibility of continence; second, his account of the role of choice (prohairesis) in human action; and third, his claim that a good action is done for the sake of to kalon or the noble.
The bare possibility of continence, of course, shows that Aristotle thinks that human agents have the power to step back from our inclinations and decide whether to act on them or not. We sometimes decide not to act on our inclinations; they do not simply drive us into action, or we could not do that. But one might be tempted to think that according to Aristotle such rational control only needs to be exercised by the continent, since the virtuous person's passions are in order, and can be trusted to direct her automatically to the good. But that cannot be right. At least, the difference between continence (p.188) and virtue cannot lie in whether the exercise of reason is involved in the action. For Aristotle makes it clear that what makes continence and virtue both good states is the fact that both of them involve the right kind of choice.
What Aristotle says about choice is initially one of the more puzzling parts of the Nicomachean Ethics. Choice, he tells us, is voluntary, but it is not the same as the voluntary since the latter is a wider category. Children and animals do things that are voluntary, but they do not act from choice, and some adult actions—those done “on the spur of the moment”—are not chosen although they are voluntary (NE 3.2 1111a4–10). For an action to be voluntary it is enough that the moving principle is in the agent (NE 3.1 1101a14–17); for choice, something more is needed. After exploring various possibilities, Aristotle decides that since the object of choice is something in our own power that is desired after deliberation, choice must be the deliberate desire of something in our own power (NE 3.3 1113a10–12). An action is chosen when we have exercised rational deliberation in determining what we are to do, and we are moved by that deliberation to act.
But of course Aristotle also says, notoriously by now, that “we deliberate not about ends but about what contributes to ends” (ta pros to telos: NE 3.3 1112b12). If we take this to mean that rational deliberation is always instrumental, we will be led to conclude that choice pertains only to actions undertaken for instrumental reasons, or perhaps some natural extension of that category. Chosen actions, that is, would be those that we have determined are necessary and desirable because they will help us to realize ends other than those actions themselves. It would then be the intellectual ability to engage in such rational calculation—instrumental reasoning and other things that are like it—that distinguishes adult human beings from children and animals, and, as a result, chosen actions from the merely voluntary.
This view, however, sits very uneasily with certain other important claims Aristotle makes about choice. For instance, Aristotle tells us that virtue is a state of character concerned with choice (NE 2.6 1106b36); and that choice is more closely bound up with virtue and discriminates character better than actions do (NE 3.2 1111b5–6). He even says that the virtues are choices or involve choice (NE 2.5 1106a3). He also tells us that the continent person acts in accordance with his choice, while the incontinent person does not (NE 3.2 1111b14–15; see also 7.3 1146b22–24). We can hardly suppose that Aristotle is suggesting that instrumental reasoning or some natural extension of it is the surest sign of a person's character, or that he thinks that incontinent people do not engage in any calculation about how to achieve their goals. What is more, Aristotle says that an action is not virtuous unless it is chosen (p.189) for its own sake (NE 2.4 1105a30–32; 6.5 1140b5–10). So whatever he means when he says that deliberation concerns what is towards the end rather than the end itself, he cannot mean that an action is never chosen for its own sake.
The interpretative crux here is of course the much‐debated question just what sort of a limitation Aristotle means to be imposing on deliberation and choice when he says that they do not concern ends but only what contributes to ends. Some commentators have focused their attention, usefully, on the idea of what is “towards the end” or contributory to the end, emphasizing that this should not be taken to refer only to instrumental reasoning in the narrow sense. Constitutive reasoning should certainly be included, and perhaps we may also include those more distinctively moral forms of reasoning that tell us, say, that an action falls under a principle, conforms to the orthos logos or right reason, is in the mean, exemplifies a virtue, or whatever.
This is part of the answer, but it is also important to look closely at Aristotle's conception of an end. Aristotle tells us that wish (boulesis) relates to the end, and that wish is for the good or the apparent good (NE 3.4 1113a15). He says that choice is “near to” wish, and that we choose to get or avoid something good or bad (NE 3.2 1111b20; 1112a4). Wish, however, belongs to the rational part of the soul (OS 3.9 432b5–6). An end, therefore, is not merely a goal, something with a view to which some agent acts. To be an end, something must be conceived as good, where that conception in turn is an act of the rational part of the soul. And to be chosen, to be an object of deliberate desire, an action must be one that contributes to an end in this sense, one that contributes to what is conceived as good. If we then also take “what contributes to the end” in the widest possible sense, the puzzle about virtuous actions being chosen for their own sake dissolves. The deliberation that shows us an action contributes to the end may be instrumental, constitutive, or moral (that is, reasoning about what is in the mean, or in accordance with the orthos logos). That doesn't matter. What matters is that the deliberation shows us that the action is in some respect good. It is the fact we have engaged in rational deliberation to arrive at the idea that the action is good, and been motivated by that deliberation, not the form of the rational deliberation, that is definitive of choice.15 So what Aristotle means is that distinctively human (p.190) actions, chosen actions, are ones which on deliberation we conceive to be good, and desire to do under that conception. That is why chosen actions are the best indicators of character—because they embody, express, or reflect the agent's conception of the good. The incontinent person, incidentally, does not act from choice even if he does engage in some sort of calculation about how to satisfy his vicious desires, because his calculations are not about what contributes to an end at all. Since the goal he is pursuing is not even an apparent good—he knows it is bad—it is not in Aristotle's sense an end.16
I take it that Aristotle and Kant, therefore, share a view about the distinctive character of human action, or at least—to add Aristotle's characteristic qualification—adult human action. Human action, to put it simply, is action that is governed by reason: that is, it is chosen. To say that an action is chosen is to say that it has the endorsement of the agent's reason, that it is conceived as good, and that it is by that conception that the agent is moved.
Kant, as we saw, moves from this picture of human action to a picture of moral value. A morally good action is one chosen because it is intrinsically good, because it has the intrinsic form of a law. Is there anything similar to this in Aristotle? If we do not assume in advance that what these two philosophers are saying must be different, one thing looks immediately similar. Aristotle insists that virtuous action must be in accordance with the orthos logos, the right reason or right rule. In fact he says it must not merely be in accordance with it but from it: “for it is not merely the state (p.191) in accordance with right reason, but the state that implies the presence of right reason, that is virtue” (NE 6.13 1144b26–27). This suggests that Aristotle thinks a good action is one whose agent sees it as the embodiment of right reason, just as Kant thinks that a morally worthy action is one whose agent sees it as an embodiment of the very form of law. I will come back to this point.
First, however, I want to consider the important argument that can be drawn from Aristotle's view that morally good actions are done “for the sake of the noble” (e.g. NE 3.7 1115b12; 3.8 1116b3; 3.9 1117b9; 3.9 1117b17; 3.11 1119b15; 4.1 1120a23; 4.2 1122b6). Aristotle tells us three different kinds of things about why good actions are done by virtuous agents. First of all, in at least some cases the actions are done for some specific purposes. For instance, Aristotle tells us that the courageous person who dies in battle lays down his life for the sake of his country or for his friends (NE 9.8 1169a17–30); in the same way, it seems natural to say that the liberal person who makes a donation wants to help somebody out; the magnificent person who puts on a play wants to give the city a treat, and so on. At the same time, Aristotle says that virtuous actions are done for their own sake; indeed, action is distinguished from mere production or “making” (poiein) by the fact that “good action itself is its end” (NE 6.5 1140b5–10). And finally, virtuous actions are done for the sake of the noble.
If we oversimplify Aristotle's moral psychology these will look like three competing accounts of the purpose or aim of virtuous action. If we take Aristotle to hold a double‐aspect theory of motivation, however, there is no problem at all. When we say that the courageous person sacrifices himself in battle for its own sake, we need not be denying that he sacrifices himself for the sake of his country. It is the whole package—the action along with its purpose, sacrificing your life for the sake of your country—that is chosen for its own sake. As for nobility, Aristotle seems to think of it very much as Kant thinks of good will—it is the specific kind of intrinsic value that moral actions and those who perform them possess. This thought is supported by the account of nobility in the Rhetoric, where Aristotle says that the noble is “that which is both desirable for its own sake and also worthy of praise” (RHE 1.9 1366a33). The Rhetoric account also confirms the claim that nobility is a property that attaches to an action along with its purpose, for in it Aristotle assigns nobility particularly to actions done for certain purposes, such as to benefit others. In fact, Aristotle suggests here that he shares Kant's view that moral value is exhibited in a special way in actions from which we are sure the agent gets nothing. He says that nobility is exhibited in actions which benefit others rather than the (p.192) agent, and actions whose advantages will only appear after the agent's death, since in these cases we can be sure the agent himself gets nothing out of it (RHE 1.9 1366b338–1367a5).17
Now I can be more specific. The view which I take Kant and Aristotle to share is this: when human beings act, we are not driven or directly caused to act by desire, passion, inclination, or instinct. Some incentive, to use Kant's language, presents a certain course of action to us as eligible—it suggests to us that we might undertake a certain action in order to realize a certain end. But reason gives us the capacity to stand back, form a view of this course of action as a whole, and make a judgment about its goodness. This isn't a judgment about whether doing this action will serve some further purpose, about whether it is useful. It is a judgment about its goodness considered as an action, not as a mere production. Both Aristotle and Kant would say that to value an action merely as a form of production, as consequentialists later did, is not yet to value it in its specifically ethical character as an action at all. As Aristotle says, “Making and acting are different . . . so that the reasoned state of capacity to act is different from the reasoned state of capacity to make (NE 6.4 1140a4–5) . . . while making has an end other than itself, action cannot; for good action itself is its end” (NE 6.5 1140b5–10). This is why techne and praxis, art and action, are different things (NE 6.4). It is with that same thought that Kant sets aside cases like that of the prudent merchant who is honest because it is useful as being wholly irrelevant to his attempt to analyze the moral value of actions. People who view actions merely as useful are not thinking of them, or valuing them, as actions at all. (On this view, we might say that consequentialism is not an ethical theory because it fails to address the subject, which is the goodness of action as such, not as a form of production.) So the capacity to choose is a capacity to make a reflective judgment about the value of an action as such and to (p.193) be moved by that judgment to perform or avoid the action. Importantly, this is at the same time a form of self‐command, a capacity to give shape to our own characters and identities. When the agent asks whether the action is a good one she is also asking: do I wish to be a person who is so moved, a person who does that sort of act for that sort of end? To relinquish this prerogative of self‐command for the sake of some mere experience or gratification is in Kant's language heteronomous and in Aristotle's base. To exercise it, especially under circumstances that make it difficult, is to act from duty and so to display that special form of moral worth that Aristotle calls nobility.
Now I want to raise some questions about how far this comparison can be pushed. To act from duty, as we have seen, is to do an action because you think its maxim has the form of a law, that it is intrinsically right or good. Aristotle, by contrast, does not tell us much about what property of an action “nobility” names. He certainly does not attempt to analyze the motive of nobility to arrive at a formulation of the moral principle, in the way Kant analyzes the motive of duty to show us what the principle of a good will is. Aristotle is famously skeptical about the possibility of articulating general principles that will guide our moral reasonings in any very exact way (NE 2.9 1109b13–26).18 Still it does seem natural to identify an action's nobility with the fact that it is in accordance with the orthos logos, the right reason. Its being in accordance with the orthos logos is what makes it intrinsically right, and it is to this intrinsic rightness that the virtuous person responds. If this is right, a noble action, like a good‐willed action, is one that embodies a principle of reason.
It is even possible to argue that nobility is a formal property. Elsewhere I have argued that we can appeal to Aristotle's concept of form to explain what (p.194) Kant means by the form of a maxim.19 In Aristotle's metaphysics, a thing is composed of a form and a matter. The matter is the material, the parts, from which it is made. The form of a thing is its functional arrangement. That is, it is the arrangement of the matter or of the parts that enables the thing to serve its purpose, or to do whatever it characteristically does. Now a maxim also may be seen as having parts. Since every human action is done for an end, we may say that a maxim of an action characteristically has two parts, the act and the end.20 The form of the maxim is the arrangement of its parts. In particular, it is the functional arrangement, the arrangement that enables the maxim to do its job, which is to be a law. A maxim passes the categorical imperative test only if everyone with that purpose could do that action—that is, if the parts are combined so that the maxim can be universalized and so can serve as a law. Now when Aristotle specifies the orthos logos, he always gives us a list of what we might also think of as the parts of the action. The action that is in accordance with the orthos logos is done in the right way and at the right time, directed to the right objects, and so on. So we might think that its overall rightness consists in the way its parts are combined, that is, in its form. The parts are combined in a way that enables them to function, taken together, as a reason for action.21
(p.195) Now I want to push the comparison one step further. Kant's analysis of the motive of duty turns on a comparison between two different ways in which we might choose a morally good action for its own sake—from duty or from immediate natural inclination. Does Aristotle similarly think that there is another way to value an action for its own sake, apart from valuing it for its nobility? Is there a character in Aristotle who, like Kant's naturally sympathetic person, simply enjoys doing the actions that are morally good, without quite grasping the reasons why they are morally good?
Of course there is. Aristotle says:
For all men think that each type of character belongs to its possessors in some sense by nature; for from the very moment of birth we are just or fitted for self‐control or brave or have the other moral qualities; but yet we seek something else as that which is good in the strict sense—we seek for the presence of such qualities in another way. For both children and brutes have the natural dispositions to these qualities, but without thought these are evidently hurtful. Only we seem to see this much, that, while one may be led astray by them, as a strong body which moves without sight may stumble badly because of its lack of sight, still, if a man once acquires thought, that makes a difference in action, and his state, while still like what it was, will then be virtue in the strict sense. (NE 6.13 1144b3–14)
And here we have an alternative description of the naturally sympathetic person of Kant's example. Humanity, of course, is not an Aristotelian virtue, (p.196) but that is not what concerns us here, and for the rest of the essay I will ignore that complication. If it were, Aristotle would say that Kant's naturally sympathetic person has a natural virtue.
3. Acting from Natural Inclination
This brings me back to the more specific question with which I began, the question of Kant and Aristotle's attitudes towards somebody like the naturally sympathetic person, and the more general question of the role of natural inclination in the moral life. Now at this point I hope you will see that as far as the case Kant actually discusses in section 1 of the Groundwork is concerned—the case of the unreflective or unreasoning sympathetic person—there is going to be little disagreement between Aristotle and Kant. Both think that his motivational state is both incomplete and unreliable until he reflects on the reasons why he should be beneficent, until his actions imply the presence of right reason. What he needs in order to become a good person is to think, and to act as a result of his thinking.22
In connection with this point, it is worth noticing that the inclination to which Kant compares natural sympathy is the inclination to honor. Kant says:
The choice of honor as the comparison is important because elsewhere Kant calls the love of honor a “semblance” of morality; in the same place he describes those moved by the love of honor as “morally immature” (IUH 8:26). In the discussion of punishment in The Metaphysical Principles of Justice, Kant suggests that people who commit murder from motives of honor, such as young officers who become involved in duels, should perhaps not be subject to capital punishment. Legislation itself, Kant urges, is responsible for the fact that these people are still morally backwards, so that the incentives of honor are not yet attached to the proper principles (MPJ 6:337). In the Anthropology Kant calls the love of honor “the constant companion of virtue” (ANTH 7:257). Honor, as Kant conceives of it, seems to be a natural tendency to live up to certain standards of conduct, not for the sake of any gain from following them but for their own sake, and out of a kind of pride. It is not yet mature virtue, for the laws of honor do not spring from the honorable person's own will, and he is concerned with what others think of him; yet it does makes him receptive to the more mature state of autonomy. In a similar way, we might take sympathy to be a natural tendency to respond to the plight of others in ways that are prescribed by the Formula of Humanity. The sympathetic person is a Menschenfreund, a friend to humanity. In the Anthropology, Kant says that it was wise of nature to give us the predisposition to sympathy, as a “temporary substitute for reason” (ANTH 7:253). All of this suggests that sympathy and honor are Kantian natural virtues, corresponding to the real virtues of humanity and autonomy respectively and making us receptive to the development of those real virtues. If this is right, Kant and Aristotle need have no disagreement about this kind of case at all.
an action of this kind [that is, like the naturally sympathetic person's], however it may conform with duty and however amiable it might be, has nevertheless no true moral (p.197) worth but is on the same footing with other inclinations, for example, the inclination to honor, which, if it fortunately lights upon what is in fact in the common interest and in conformity with duty, deserves praise and encouragement but not esteem; for the maxim lacks moral content, namely doing such actions not from inclination but from duty. (G 4:398)
I don't think that this is quite right—I think there is still some disagreement—but its nature is best brought out by asking the more interesting question whether they would disagree about the case of actions which do have moral worth, about whether those must be done with pleasure or some other appropriate feelings. So I want to turn to that question.
Now a preliminary point is that we must not exaggerate the views of either philosopher if we are to get this right. Kant thinks that in order to be receptive to moral reasons we must cultivate the virtues, and cultivating the virtues is a matter of adopting certain obligatory ends, such as one's own (p.198) perfection and the happiness of others. At this point I come to an issue I mentioned at the beginning of the essay—the fact that our mental lives have a passive or receptive as well as an active dimension. There is an important difference between giving an account of what sorts of reasons for action morality prescribes and giving an account of how we become receptive to those reasons. There are two problems of receptivity. One is how we are motivated by the dictates of reason when those dictates are presented to us, whether by the arguments of others or simply by the workings of our own minds. To some extent this is just the problem of how pure reason can be practical, which Kant takes to be insoluble; to the extent we can say anything about it, it is the problem Kant is addressing in his account of how the thought of the law gives rise to the incentive of respect. The second problem is how we come to think about our duties at all, how we come to notice which reasons we have. The negative character of the Formula of Universal Law reveals this problem in an especially acute way. Under the Formula of Universal Law you arrive at the duty of helping when you consider the maxim of not helping, but it is only under extremely unusual circumstances that you would consider the maxim of not helping. The naturally sympathetic person, whose mind becomes clouded by sorrow, is in such circumstances. As I portrayed him, he considers the issue of helping for the simple reason that he used to help, or perhaps because someone reminds him of that fact. But what if the idea of helping simply doesn't occur to you one way or another? As Kant himself says in The Metaphysics of Morals, “Maxims are here regarded as subjective principles which merely qualify for a giving of universal law, and the requirement that they so qualify is only a negative principle . . . How then can there be, beyond this principle, a law for the maxims of actions?” (MM 6:389) What he is asking is how there can be a law that says we must have certain maxims. This is the problem that Kant addresses in The Metaphysical Principles of Virtue. Kant argues that we have a duty to cultivate moral ends and the feelings that are naturally attendant upon having those ends so that we will notice the occasions of virtue.23
Sympathy is naturally associated with having the happiness of others as your end, which is required, and in The Metaphysical Principles of Virtue, Kant does not scruple to say that sympathetic feeling is a duty. Being sympathetic helps us to be aware of those cases when our assistance or support will be called for. And if we cultivate moral ends and the feelings that are naturally (p.199) attendant upon having such ends then in the normal course of events we will also take pleasure in successful virtuous action. If I have made your happiness my end and I do something that successfully promotes it I will of course take pleasure in that fact. It doesn't matter whether my original impetus for making your happiness my end was natural inclination or the rational acknowledgment of the value of your humanity; if it really is my end it will normally give me pleasure to promote it. This is an ordinary fact about human motivation: once you have backed a certain horse, for whatever reason, you are going to be thrilled if it wins. Since virtue requires the adoption of ends, it requires, indirectly, the development of a range of feelings, the feelings associated with having those ends. The method, not surprisingly, is habituation. Kant says:
Kant both requires and expects that the virtuous person will in this way at once become receptive to the occasions of virtue and, at the same time, able to take pleasure in virtuous action. He even says:
Beneficence is a duty. If someone practices it often and succeeds in realizing his beneficent intention, he eventually comes actually to love the person he has helped. So the saying “you ought to love your neighbor as yourself” does not mean that you ought immediately (first) to love him and (afterwards) by means of this love do good to him. It means, rather, do good to your fellow human beings, and your beneficence will produce love of them in you (as an aptitude of the inclination to beneficence in general). (MM 6:402)
But what is done not with pleasure but merely as compulsory service has no inner worth for one who attends to his duty in this way and such service is not loved by him; instead, he shirks, as much as possible occasions for practicing virtue. (MM 6:484)
On the other hand, even Aristotle must admit that at least in very hard cases it is only successful virtuous action that will necessarily bring us pleasure, and that in only a limited way. I have a specific hard case in mind. In the Book 9 account of the relation between virtue and self‐love, Aristotle makes the outrageous suggestion that the person who dies in battle gets the greater good because he prefers a short and noble life to years of humdrum existence (NE 9.8 1169a22–24). In the Book 3 account Aristotle is more honest. He says:
After which Aristotle concludes, rather lamely:
Hence also courage involves pain, and is justly praised, for it is harder to face what is painful than to avoid what is pleasant . . . death and wounds will be painful to the brave man and against his will, but he will face them because it is noble to do so or because it is base not to do so. And the more he is possessed of virtue in its entirety and the happier he is, the more he will be pained at the thought of death; for life is best worth living for such a man, and he is knowingly losing the greatest of goods, and (p.200) this is painful. But he is none the less brave, and perhaps all the more so, because he chooses noble deeds of war at that cost. It is not the case, then, with all the virtues that the exercise of them is pleasant, except insofar as it reaches its end.
And it is worth remembering that two of the cases of action from duty that Kant discusses in section 1 of the Groundwork, that of a person who wants to commit suicide because of the acuteness of his misery, and that of a person in the grip of some great sorrow, are tragic cases. Aristotle firmly repudiates the Stoic view that virtue is sufficient for happiness even at moments like this, although I suppose he might still want to say that there is some pleasure to be taken in the virtuous action at hand (NE 1.8 1099b1–8). But then Kant would say that, too—acting from respect for law does always have a pleasant dimension, although the pleasure is of a rather rarefied kind.
But it is quite possible that the best soldiers may not be men of this sort but those who are less brave but have no other good. (NE 3.9 1171a33–1171b20)
Aristotle and Kant might still disagree about one case. There are two characters who are beneficent from duty in the Groundwork examples: the one whose mind is clouded by sorrow and another, whom I haven't discussed yet, who is temperamentally cold. This person seems to be incapable of enjoying beneficent action. I suppose that Aristotle would characterize him as continent rather than virtuous, and would think that this is a less good state, and maybe, although I am not sure of this, judge that he is a less good person. Kant doesn't make the distinction between continence and virtue. But by now I hope it is clear that if he did, he would not say that continence is a better state, or that the cold person is a better person, than the virtuous person who also enjoys beneficence. What Kant says about the cold person in the Groundwork is only that he has a moral worth which the unreflective sympathetic person lacks; he does not compare him either positively or negatively to someone who helps from the motive of duty and also enjoys it. Aristotle does not, as far as I know, ever make the parallel comparison, which would be between merely natural virtue and continence. I assume he would agree with Kant, though, that continence is better than merely natural virtue, since the continent person has the first principle, and this is the important thing.24 There remains only this difference: Kant would certainly not say that the cold person, provided he (p.201) somehow managed to do his duty, was any less good, or was in a less morally good state, than the person who does his duty and also enjoys it.
But the reason why this one difference still remains throws light, I think, on the question why Kant doesn't characterize sympathy and honor as natural virtues, even though he comes very close. What is at work here is a difference between Kant's and Aristotle's views of what inclination is, which in turn depends on a difference in their views of what pleasure and pain are. The difference is that Aristotle thinks of pleasure and pain as something like perceptions of the reasons for actions, while Kant apparently does not believe that pleasure and pain in general play this role. Respect for law comes closest to doing this, since it is a feeling produced by the activity of reason itself, but the pleasures and pains that are associated with ordinary inclinations do not.
Let me first mention the textual evidence for these claims and then say why I think they make a difference. Aristotle tells us that passions are “feelings that are accompanied by pleasure or pain” (NE 2.5 1105b19). The accounts of pleasure in the Nicomachean Ethics itself are mostly concerned with the pleasures we take in activities as we do them; the question of what it means for some state of affairs to seem pleasant or painful to us in the way that is involved in passion is a little different. In On the Soul, Aristotle explains the relationship between pleasure and pain and passion this way:
To take pleasure in something is to perceive it as good or bad, that is, as a reason for pursuit or avoidance. This is why Aristotle insists, throughout the ethics, that it is so essential to get our pleasures right. Aristotle says that when we go wrong in our “wishes”—that is, our conceptions of the good—the error is due to pleasure, “for it appears a good when it is not. We therefore choose the pleasant as a good, and avoid pain as an evil” (NE 3.4 1113a35–1113b1). I don't think Aristotle means merely that we are inclined to count pleasant things among the good things. I think he means that when something is pleasant it literally looks good to us. Aristotle's own favorite comparison of virtue to health can be used to illustrate the point. The healthy person's appetites, which are in a mean, are a reliable guide to what is good for her, that is, to what (p.202) will preserve her in health. The amount she enjoys eating and exercising are actually the amounts she needs, so that her perception of the good—of what she has reason to do—is correct. Hunger tells her something—that she is in need of nourishment, that she has a reason to eat—and if she is in good condition, hunger is right. Since the appetites and passions all involve pleasure and pain, this means that what it is to be in the grip of a passion is to see a situation as being a reason for pursuit or avoidance of a certain kind.25 To be angry is to perceive a reason to fight, or, as Aristotle puts it: “anger, reasoning as it were that anything like this must be fought against, boils up straightaway” (NE 7.6 1149a32–33); to be scared is to perceive a certain situation as a reason to flee, and so on. And since the soul never thinks without an image, as Aristotle says in the passage above, our conceptions of good and evil must be accompanied by images of our circumstances as pleasant or painful in certain ways. These images provide the material with which the intellect works in conceiving the good. It is because of the way the mind works that the virtuous person must experience pleasures and pains in the right way in order to think correctly about practical matters: thinking of something as good is inseparable from imagining it, so to speak, as pleasant.
To perceive, then, is like bare asserting or thinking; but when the object is pleasant or painful, the soul makes a sort of affirmation or negation, and pursues or avoids the object. To feel pleasure or pain is to act with the sensitive mean towards what is good or bad as such. Both avoidance and appetite when actual are identical with this: the faculty of appetite and avoidance are not different, either from one another or from the faculty of sense perception; but their being is different.
To the thinking soul images serve as if they were contents of perception (and when it asserts or denies them to be good or bad it avoids or pursues them). That is why the soul never thinks without an image. (OS 3.7 431a7–16)
Now the merely continent person's contrary passions make it difficult for her to maintain the required images, which is why, as Aristotle says, it is the same person who is both continent and incontinent (NE 7.1 1145b10–11). Mere continence is an unstable state, for the tendency to incontinence, its inevitable partner, can bring about a battle between intellect and passion for control of the agent's perceptual imagination. This is why Aristotle says that it is not knowledge proper but perceptual knowledge that is dragged about by passion in incontinent action (NE 7.3 1147b15–17). The virtuous person's reason, by contrast, is in unchallenged control of her perceptual imagination. And this is Aristotle's solution to the problem of receptivity. In the fully virtuous person, the entire appetitive part of the soul serves as a kind of sensorium for reason.26
(p.203) Kant, by contrast, denies that pleasure and pain tell us anything about anything. He says:
Now Kant shares Aristotle's view that inclination involves pleasure: he defines desire in the narrow sense as a determination of the faculty of desire that is caused by pleasure, and an inclination as a habitual desire (MM 6:212). But since Kant thinks that pleasure and pain are mere feeling, that they are, to put the point a little bluntly, stupid, he also thinks that inclination is stupid. The fact that you have an inclination for something does not tell you anything about that thing or even anything about your own condition. It only signals the thing's relationship to you.27
The capacity for having pleasure or displeasure in a representation is called feeling because both of them involve what is merely subjective in the relation of our representation and contain no relation at all to an object for possible cognition of it (or even cognition of our condition). While even sensations, apart from the quality (of, e.g., red, sweet, and so forth) they have because of the nature of the subject, are still referred to an object as elements of our cognition of it, pleasure or displeasure (in what is red or sweet) expresses nothing at all in the object but simply a relation to the subject. (MM 6:211–212)
And this makes for an important difference between what Kant says about the naturally sympathetic person and what Aristotle would say about him if humanity were an Aristotelian virtue. We have a reason to help human beings who are in need, and Aristotle's account of inclination allows him to see our natural inclination to help as an inchoate grasp of that reason. It is the kind of perceptual starting point from which, in his methodology, we (p.204) can work up to a more conceptual grasp of the first principles or reasons involved (NE 1.2 1095a31ff., and many other places).28 When Aristotle says that the state of the authentically virtuous person is “not the same as that of the naturally virtuous person but like it,” I take him to mean that the authentically virtuous person perceives the reason for action too but perceives it in that special way in which, according to Aristotle, you perceive matters that you also understand.29
But Kant cannot see natural sympathy as an inchoate grasp of the fact that there is a reason to help. He thinks that an inclination signals only a certain subjective suitability between the sympathetic person and the promotion of the happiness of others, a fitness of sympathetic action to gratify this particular person. This is the real reason why Kant describes this person as acting implicitly or tacitly under the principle of self‐love, rather than as having a natural virtue. For Kant, sympathy is not a proto‐virtue but merely a kind of substitute for virtue which nature has given us in the meantime. And this makes it look as if the inclinations and feelings which we are required to develop in order to solve the problem of receptivity will also have to be regarded as mere tools and helps.
The question which of these conceptions of inclination is correct is an extremely difficult one. The intuitive appeal of Aristotle's conception, at least about certain cases, is obvious. Sympathy for the troubled or the needy, in particular, presents itself to us as a response to the fact that there is a reason to help them. Such sympathy is painful, not pleasant, and if we regarded it merely as a source of feeling we would take an aspirin to make it go away. We don't do that, because of what we think sympathy reveals to us—that we have a reason to relieve someone's distress.30 Of course, as the accusation (p.205) of “sentimentality” shows, we also do sometimes dismiss inclinations, pains, and pleasures as mere feeling. There are people—I am one—who take our natural sympathy with the other animals, our acute sense of their pain and vulnerability, to be perceptions of the reasons we have to be merciful and protective towards them. And there are other people who dismiss this as mere sentimentality, as just so much personal feeling that doesn't mean a thing. But the very fact that this is offered as a criticism, or as debunking, shows that we do not in general take our pains and pleasures to be meaningless. We take them, as Aristotle thought, to be indications of what is good or bad, and what we have reason to do.31
But intuition by itself cannot settle the question in Aristotle's favor. Much more work in the philosophy of mind would be needed to show how Aristotle's view could possibly be true.32 That we are attracted to a view like Aristotle's, however, does seem to me to explain why we are uncomfortable with what Kant says about the naturally sympathetic person in the Groundwork. Aristotle seems to give us a superior account of what is going on in this pre‐moral case, and, if he does, he may also be able to give us a superior account of how receptivity works in the case of fully realized virtue as well.
But I do not think that this marks a difference in the basic ethical outlooks of Aristotle and Kant. Although there is a difference in the way these two philosophers propose to solve the problem of receptivity, the problem of receptivity arises for both of them because of the deep similarity in their general conception of what ethics is all about. Human action is not like anything else: as human beings we choose our actions, and, because of that, it is possible for us to transcend mere reactivity in our relationship to the world. The most general and substantive question of ethics is what we should do with this power, which actions we should choose. The more specific question (p.206) of virtue, the question to which Aristotle gave most of his attention, is the question how the receptive part of our nature needs to be configured if this kind of transcendent choice and action is to be possible. It is the question, that is, of what we have to be like, in order to choose autonomously, and for the sake of the noble.33
(1) Spontaneity, in Kant's sense, means having an original source in the agent's own mind or will, rather than in some external cause.
(2) It's worth noticing the comparison with Nietzsche, who also places a high value on this attitude, although of course with more ambivalence about whether it may be identified with the moral attitude. In Essay One of The Genealogy of Morals, Nietzsche proposes that the values “good” and “bad” were born from the spontaneous evaluative acts of the noble or master types, as an expression of the value they set on themselves, while the opposed values “evil” and “good” were the result of reaction against the masters on the part of the oppressed and enslaved. (See Walter Kaufmann and R. J. Hollingdale, trans., On the Genealogy of Morals, in On the Genealogy of Morals and Ecce Homo, especially sections 10–11, pp. 36–43.)
(3) Terence Irwin, in his “Kant's Criticisms of Eudaemonism” (in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty), suggests that Kant has a hedonistic conception of desire and therefore of happiness, and that this is one basis for his criticism of his eudaemonist predecessors. It will be evident that I cannot agree with this. Irwin himself acknowledges that Kant's criticism of eudaemonism need not depend on this thesis, however, since its essence is that action governed by considerations of one's own good is essentially heteronomous. Irwin thinks that this criticism is not decisive, since one may argue that eudaemonistic principles do not derive their authority from our inclination to achieve happiness. One may instead suppose, as Butler and Reid did, that the principle of pursuing our own good has an authority of its own, just as the categorical imperative does. I believe that this argument misses the main thrust of Kant's objection to eudaemonism, although I think that Kant himself is partly responsible for the misconception. There are two elements to Kant's notion of heteronomy: (a) the law is not the will's own law, but rather is given to it from outside, and (b) the will therefore can be bound by that law only through an inclination or an interest, which renders the imperative to follow the law hypothetical. As this way of putting the point makes clear, Kant himself argues as if, and may have thought, that these two elements are inseparable, and he therefore sometimes emphasizes the second element, which Irwin takes to be the essence of heteronomy. But I think that the real essence of heteronomy lies in the first element: the problem with the eudaemonistic principle is that it is not the will's own law. The possibility of the two elements of heteronomy coming apart is illustrated by a case I discuss later in this essay, the case of someone who is motivated by considerations of honor. One is not bound to considerations of honor by inclination or interest: one is honorable for its own sake, moved by a conception of how one ought to act. Yet this kind of action is still not fully autonomous, because the laws of honor are not the will's own laws. Only the categorical imperative, which describes the activity of a free will as such (a free will as such chooses a maxim it regards as a law), is the will's own law. (For a fuller account of this point see my “Morality as Freedom,” CKE essay 6.)
(4) Kant evidently thinks that there are three ways to value, and therefore to choose, an action: as useful (as the prudent merchant values honesty); as good for its own sake, in the sense of being immediately desirable (as the sympathetic person values beneficence); and as morally required (as the dutiful person values beneficence). If the argument of this essay is correct, this coincides with Aristotle's view that there are three objects of choice—namely, the advantageous, the pleasant, and the noble (NE 2.3 1104b30–31).
Addition 2008: When I wrote this footnote I had failed to realize that the conception of action that I attribute to Kant and Aristotle in this essay makes the idea of choosing an action as useful unclear. If an action is an act done for the sake of an end—that is, the description of an action includes its end—then actions are done for their own sake. On this point, and why Kant himself may have failed to see it, see my “Acting for a Reason,” Essay 7 in this volume, especially pp. 222–3.
(5) This is a little oversimplified: In Religion within the Limits of Reason Alone, Kant argues that we are in general influenced by both moral incentives and incentives of self‐love. Whether one has a good will depends on which of these is made the condition of the other (REL 6:36). This suggests that the moral principle will be something like: “Do your duty, and what you like if that is consistent with your duty,” while the principle of self‐love will be: “Do what you like, and your duty if that is consistent with your happiness.” This complicates the picture in ways that I want to leave aside here, however, since these formulations presuppose a certain view of the role of natural inclination in the moral life, the basis of which I will call into question in section 3.
(6) For a rich and subtle account of Kant's views on the operation of the principle of self‐love, see Allen W. Wood, “Self‐Love, Self‐Benevolence, and Self‐Conceit,” in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty.
(7) I do not agree with J. B. Schneewind's view that “for Kant nothing possesses the kind of intrinsic value that G. E. Moore thought would belong to a beautiful world even were there no observers of it” (in his “Kant and Stoic Ethics,” in Aristotle, Kant and the Stoics: Rethinking Happiness and Duty). Schneewind is right, of course, to insist that in Kant's account value is not independent of rational willing: a maxim is an act of rational willing, and it is the maxim, and the good will which is expressed in the maxim, which possess this value. For a comparison between Kant's conception of unconditional value and Moore's conception of intrinsic value, see my “Two Distinctions in Goodness” (CKE essay 9). For a more detailed account of the sense in which maxims have intrinsic value, see SN 3.3.5–3.3.6, pp. 107–12.
(9) Butler, in Sermon 11 of his Fifteen Sermons Preached at the Rolls Chapel (Sermon 4 in Stephen Darwall, ed., Joseph Butler: Five Sermons Preached at the Rolls Chapel and A Dissertation Upon the Nature of Virtue, pp. 46–57), and, following him, Hume, in the conclusion of the Enquiry Concerning the Principles of Morals (2E 9, 281–2) make arguments in favor of beneficent action which take this form. To be fair, neither of them thinks that this is the way to establish the moral value of beneficence; it is just a way to establish the harmony of beneficence and self‐interest. For further discussion, see SN 2.2.3, pp. 55–60, and Charlotte Brown, “Hume Against the Selfish Schools and the Monkish Virtues.”
(10) Kant has some tendency to exaggerate the active reflectiveness of human beings, and the possibility of an unreflective version of the motive of duty may not even have occurred to him. In the Critique of Practical Reason, for example, after explaining how a natural law can serve as a “type” of the moral law, he says, “Everyone does, in fact, appraise actions as morally good or evil by this rule. Thus one says: If everyone permitted himself to deceive when he believed it to be to his advantage . . . and if you belonged to such an order of things, would you be in it with the assent of your will?” (C2 5:69). Probably the most startling instance of this optimism occurs at Groundwork 4:450, where Kant suggests that even “the commonest understanding” draws a rough distinction between the sensible and intelligible worlds. But optimism about human reflectiveness is not the only thing at work here. Perhaps the closest thing to an unreflective version of the motive of duty in Kant's system is the inclination to honor; and I will explain why Kant didn't see this as an unreflective version of the motive of duty below.
(11) I thank Arata Hamawaki and Michael Hardimon for pressing me on this point, and for useful discussion of this argument in general.
(12) I describe this process of reflection and how it leads one to a recognition of the moral law in more detail in “Morality as Freedom,” CKE essay 7, pp. 27–31.
(13) There is an important similarity between this way of characterizing the difference between the naturally sympathetic person and the dutiful person and the way in which, according to Jennifer Whiting, Aristotle characterizes the difference between merely agathos (the merely good person) and the kaloskagathos (the noble and good person) in the Eudemian Ethics. The kaloskagathos, as Whiting characterizes him, is superior in his reflective understanding of the reasons for good actions and therefore chooses them for their own sake, rather than for the sake of external or natural goods (see her “Self‐Love and Authoritative Virtue: Prolegomenon to a Kantian Reading of Eudemian Ethics 8.3,” in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty). I would not say that the naturally sympathetic person chooses beneficence merely for the sake of the natural goods, but rather that he chooses it merely as a natural good. Whiting compares the Eudemian Ethics' agathos to a character in the Nicomachean Ethics, namely, the person who has merely natural as opposed to “authoritative” virtue; later in this essay I will compare this same character to the naturally sympathetic person.
(14) These remarks are based on the discussion in the Critique of Practical Reason's chapter on “The Incentives of Pure Practical Reason,” especially C2 5:71–76, with some supplementation from the discussion of how pleasure and pain are related to moral motivation at C2 5:116–118. For an interesting account of how the experience of respect for law is related to the workings of the principle of self‐love, see Allen W. Wood, “Self‐Love, Self‐Benevolence, and Self‐Conceit,” in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty.
(15) In his “Deliberation and Moral Development in Aristotle's Ethics,” John McDowell suggests that Aristotle sometimes overstates the extent to which actions that reveal virtue “issue from actual courses of thinking” (in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty, p. 25). I do not really mean to disagree with that claim here: McDowell and I would agree, I think, that the important point is that the actions be conceived as in some way good, where the good is the object of reason or thought. We might disagree somewhat about how articulate Aristotle expects his agents to be—and how articulate agents ought to be—in explaining why the action is good, and so about the extent to which they must be capable of providing something like a retrospective deliberation if asked to justify their actions. This however is a disagreement about a matter of degree. I am not certain what exactly McDowell has in mind when he criticizes those whom he supposes envisage a “straightforward” or “mechanical” application of principles; or at least I think it must be a misleading way to describe his worry (p. 21). The application of a principle by a thinking or conscious agent is always going to be perceptual rather than mechanical: perhaps there are places where the perceptual and the merely mechanical seem to run together, say for instance in the phototropic responses of plants, but this has nothing to do with the subject. The question, as I expect McDowell would agree, is surely about how much moral content perception must already have before we can begin to articulate, deliberate, and argue about the application of principles. I believe that even on the most algorithmic conception of the categorical imperative procedure the Kantian answer to that question could not be “none,” because an agent who views others as persons and things and actions as possible means and ends has already taken up what is broadly speaking an ethical perspective on the world. But this is not the place to pursue this point.
(16) See also the somewhat more detailed account of choice, and why incontinent action isn't chosen, in my “Aristotle's Function Argument,” Essay 4 in this volume, pp. 146–8. In the text above I say that the incontinent agent knows that his end is bad, but in “Aristotle's Function Argument,” I say that the incontinent agent knows that the performance of a certain act for the sake of a certain end is not a thing worth doing for its own sake. I now believe that is the correct account.
(17) I owe these references and some of what I say here about their implications to Terence Irwin, who discusses them in the notes to his translation of the Nicomachean Ethics. See especially the discussion of to kalon or as he renders it “the fine” at pp. 401–2. Julia Annas, in her “Aristotle and Kant on Morality and Practical Reasoning,” focuses on these same passages to support her claim that Aristotle like Kant draws a distinction between moral and non‐moral reasoning: “for the sake of the noble” is a distinctively moral reason (in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty, p. 241). While I am of course sympathetic to the comparison, I would prefer to phrase the conclusion in what is to some extent an opposite way: that both think there is only one kind of reason, although the considerations we use to identify a reason are complex. This is because I do not think that Aristotle would agree that you could ever really have a reason to do something base or ignoble, any more than Kant would agree that you could really have a reason to do something immoral. The person who fails to take nobility or obligation into account acts at best for an imperfect or incomplete reason, not for a different kind of reason. This is not merely a verbal fuss, for the question is whether we may avoid the Sidgwickean problems that arise when one acknowledges two distinct sources of normativity.
(18) Actually, there are two points to this skepticism. One point is the denial that we can formulate any reliable general rules to guide us morally: moral value belongs intrinsically to particular actions, and no set of general rules is sufficiently refined to pick them out. The other is the view that we must (therefore?) pick them out by means of perception. Now Kant has no reason to disagree with the first point. Kant certainly thinks that moral value belongs to particular actions, indeed that it is an intrinsic property of those actions. They do not inherit their value from any rules that are external to them. (See my “Kant's Analysis of Obligation: The Argument of Groundwork I,” CKE essay 2, pp. 60–2 for a discussion of this important point.) The categorical imperative test is a test on particular maxims, and any circumstance that is really relevant to the moral value of an action may properly be included in its maxim. Kant himself may have had some tendency to exaggerate the extent to which the categorical imperative's findings could be captured in a set of general rules, but nothing in the theory requires this. Now precisely because there is such a thing as the categorical imperative test, Kantians must deny that the failure of general rules leaves us no recourse but perception. But of course perception and judgment must at some level play a role, as anyone must agree (see note 15). So even this disagreement between Kant and Aristotle has been exaggerated. On the essential point, that moral value is an intrinsic property of particular actions in all of their rich particularity, Aristotle and Kant are in accord.
(19) SN 3.3.5, pp. 107–8.
(20) Where the action is done for its own sake, these will not be different.
(21) Now that I have made some fairly strong claims about Aristotle and Kant sharing a view of moral value, I want to wave my hands a little over the vexed question of categorizing ethical theories. People used to categorize theories as deontological or teleological; lately, we have started opposing deontology to consequentialism, and what seems to be a new category, “virtue ethics,” has come upon the scene, although it is unclear whether it is a rival theory or a rival view about what the direction of our attention should be. Certainly no one seems to have a very clear idea what deontology, consequentialism, and virtue ethics might be three theories of, but suppose we try saying that they are theories of what makes an action right. Consequentialism is the theory that what makes an action right is its consequences; deontology is the theory that the action's rightness is intrinsic, or consists in its conformity to a rule; and virtue ethics is the theory that what makes an action right is that it is the sort of action a good person would do. If “the action's rightness is intrinsic” means that the outward performance, the act, has intrinsic rightness, then perhaps only traditional rational intuitionists, like Clarke, Price, Ross, and Prichard, are deontologists, if anybody is. Kant and Aristotle, like Hume and Hutcheson, think that what makes an action right is that it is the sort of action a good person—for Kant an autonomous person and for Aristotle a person of practical wisdom—would choose. “Virtue ethics,” however, would be a rather wild misnomer for this view in their case, since one does not have to have the virtues in order to choose well—even Aristotle admits that a continent person may choose well. On the other hand, suppose we say that deontology is the view that what makes an action right is its conformity to a rule of reason. Then Kant and Aristotle, along with rationalistic consequentialists like Sidgwick, are deontologists, as opposed to Hume and Hutcheson. On this view, in fact, Aristotle and Kant must be categorized as both deontologists and virtue theorists, since they think that the good person acts in accordance with, or even is the source of, a rule or at least a direction of reason. Since this seems unhelpful, suppose we say that the categories do not represent three views about what makes an action right, but three views about what gives an action moral worth. The resulting view about consequentialism—that consequences give an action moral worth—seems insane, and I am sure no one holds it. Consequentialists, if they are going to employ the notion of moral worth at all, will have to hold that it is the intention to produce good consequences which constitutes moral worth, and then their view will be a species of so‐called “virtue ethics.” (Of course, I have already suggested in the text that consequentialists might better be thought of as not employing this notion, or even, in this sense, as doing ethics.) Traditional rational intuitionists will hold that it is the intention to do what is right that gives an action moral worth, so they will be virtue ethicists too. This is no good. Perhaps, then, we should return to the earlier distinction, between deontology and teleology? Deontologists are interested in rightness and rules; teleology thinks that ethics has something to do with value or the good. Fine. If teleology is the view that the moral value of an action consists in its promoting the good, then Aristotle is a deontologist, since he thinks moral actions embody the orthos logos and so are good in themselves. If teleology is meant to include the view that moral actions are themselves good, then Kant is a teleologist, since he holds this view. Do you find that these efforts to categorize theories fill your mind with darkness rather than light? That of course is the point. Well, then maybe we should drop that, and oppose theories like Aristotle's and Hume's, which are primarily concerned with the virtues of character, with theories like Kant's and Sidgwick's, which are primarily concerned with the rightness of actions? I won't even bother to object to these tendentious descriptions of the “primary concerns” of these philosophers, since it is easier to ask what possible reason we could have for opposing theories if they have different primary concerns. Well, it may be replied, the issue is a methodological one: in subtle and even unconscious ways, our theories are shaped by their primary concerns. Now there is a great deal in this, and it is worth being aware it of when we study another philosopher's theories. But it is not a ground for opposing or categorizing different kinds of theories. Nor can we choose our own methodology by deciding in advance in which subtle and unconscious ways we would like our theories to be shaped. See also Barbara Herman, “Leaving Deontology Behind,” in The Practice of Moral Judgment, pp. 208–40.
(22) At this point it is worth mentioning one apparent difference between the two philosophers. Earlier I mentioned that Aristotle supposes that adult human actions done “on the spur of the moment” are voluntary but not chosen. This raises a question: would Aristotle say that an act of impulsive sympathy was voluntary but not chosen? If so, he would not only deny Kant's view that it was done under the tacit or implicit principle of self‐love, but that it was done under any principle of volition. More generally, the point is that Kant seems to suppose that any adult human action is implicitly or tacitly done under some principle of volition, while Aristotle seems to think that merely voluntary action is still possible for adult human beings. Kant's view seems to be that the capacity for reflective choice, whether exercised or not, makes a difference to every action: adult human actions take place in the light, so to speak, of reflective thought, and can no longer be exactly like the actions of children and animals. Who is right? I believe that this question raises very complex issues about the third‐person attribution of mental states and conditions (belief, choice, etc.), and whether those attributions are moral or merely factual. Aristotle's view suggests that a merely voluntary action performed “on the spur of the moment” is not a proper subject of moral judgment, since the agent is just following nature, and it is choice, not the merely voluntary, that reveals character. But there is something to be said for Kant's view, for surely if an adult human being performed too many actions on the spur of the moment, and failed to sufficiently exercise the power of choice, we would make a negative moral judgment about him (perhaps the judgment that he lacks character). This shows that we do think that once the capacity to exercise choice is present, it makes a difference to every action (or at least to actions in general), just as Kant says. But Kant's decision to attribute a principle of volition to people who perform thoughtless actions is not a guess about their actual volitional states. It is a moral choice: a decision that adult human beings are to be held responsible for thoughtless actions, because they might have thought. (For further discussion, see my “Creating the Kingdom of Ends: Reciprocity and Responsibility in Personal Relations,” CKE essay 7.)
(23) I do not mean in this sentence to have said what the argument is; it is actually rather subtle and I am not taking it up here.
(24) At least this is why Aristotle says incontinence is a better state than intemperance (NE 7.8 1151a25). Although the merely naturally virtuous person and the continent person each lack an essential element of fully realized virtue, and this might seem to put them on a footing, still, the continent person can perform a noble act for the sake of its nobility, and the merely naturally virtuous person cannot do this essential thing.
(25) I think this shows that McDowell is wrong in characterizing the natural virtues as “mindless behavioral propensities” that merely “correspond” with the virtues (in “Deliberation and Moral Development in Aristotle's Ethics,” p. 20). But I think this conclusion is one that he should welcome as friendly to his reading of Aristotle. McDowell thinks that the result of habituation is a primitive form of practical wisdom. As I understand it, there is already a primitive form of practical wisdom built into the passions of the naturally virtuous person; the result of habituation is to refine it, and the result of intellectual training is to render it articulate. I think that this makes it clearer why habituation and intellectual training must proceed together (rather than habituation coming entirely first) and also avoids committing us to the somewhat implausible idea that habituation changes the ontological status of the passions altogether—transforming them from mere mechanical propensities into perceptions, as McDowell's view, as it stands, seems to imply.
(26) This point may be strengthened by the following consideration. What health does is preserve the form of the living body; what virtue does is preserve the form of the soul. The form of the human soul is that it is governed by reason: that is why reason is the human function (ergon). Healthy actions—those motivated by healthy appetites—tend to preserve health in the body. Virtuous actions, then, tend to preserve reason's government in the soul. This shouldn't sound wild—I mean, at least as an attribution to Aristotle—since it is the view explicitly advocated by Plato in Republic 4, at 443cff. I take it to be the view Aristotle is also espousing when he says “the intellect always chooses what is best for itself, and the good man obeys his intellect” (NE 9.8 1169a16–17). Of course making these claims plausible or even comprehensible is another matter. They are abstract because it seems so difficult to form a conception of how virtuous actions tend to preserve the rational form of the soul. I have discussed Aristotle's view on this question in “Aristotle on Function and Virtue” (Essay 5 in this volume); lately, it has seemed to me that Plato's attempt to explain it in Books 8 and 9 of the Republic may be more perspicuous. In any case, if virtue puts reason in a position to choose what is best for itself, then virtue does enable reason to be active rather than merely reactive in its relationship to the world. Choice choosing choice mimics the divine activity, the purest of all activities, thought thinking itself (Metaphysics 12.6–9). But I leave these extremely abstract thoughts for another occasion.
(27) The infamous passages in the second Critique in which Kant sounds so much like Bentham (C2 5:22–24) are actually an expression of this view. The point isn't that all we care about is our own pleasure. The point is that if our carings are just feelings, it doesn't really matter which ones we satisfy—that is, we have no reason, intrinsic to those carings themselves, to satisfy one rather than another.
(29) In Nicomachean Ethics 6 Aristotle seems to struggle to give a correct account of the respective relations of perception, practical wisdom, and nous (see, for instance, NE 6.7 1141b14–23; NE 6.8 1142a 23–30; NE 6.11 1143a 35–1143b6). I believe his view is that both practical wisdom and scientific wisdom are like perceptual states, both in the sense that you have a direct grasp of the first principles or reasons of things, and in the sense that that grasp somehow inhabits your actual perceptions of the particulars. To the person of practical wisdom and to the person of scientific wisdom, the world literally looks different from the way it does to those who perceive but do not yet understand. The person of scientific wisdom sees the essences of things unfolding in their activities; the person of practical wisdom sees the good, and opportunities to realize the good, in the circumstances in which she finds herself. Providing textual evidence for this view would be an immense undertaking, so for now I will just state that I think that's what he means.
(30) I defend the claim that pain is the perception of a reason at SN 4.3.4–4.3.12, pp. 147–56. In 4.3.5 I cite some other philosophers on the pains of pity in particular and it may be useful to repeat those citations here. Hutcheson says “If our sole Intention, in Compassion or Pity, was the Removal of our Pain, we should run away, shut our Eyes, divert our Thoughts from the miserable Object, to avoid the Pain of Compassion, which we seldom do: nay, we crowd about such Objects, and voluntarily expose our selves to Pain” (An Inquiry Concerning Moral Good and Evil, quoted in Selby‐Bigge, British Moralists, p. 93). The point is reiterated by Thomas Nagel: “Sympathy is not, in general, just a feeling of discomfort produced by the recognition of distress in others, which in turn motivates one to relieve their distress. Rather, it is the pained awareness of their distress as something to be relieved” (The Possibility of Altruism, p. 80 n.). Wittgenstein says “How am I filled with pity for this man? How does it come out what the object of my pity is? (Pity, one may say, is a form of conviction that someone else is in pain)” (Philosophical Investigations, section 287, p. 98).
(31) Some might think that the view suggested—that there is such a thing as perceiving a reason—implies a form of realism about reasons which is inconsistent with Kant's constructivist outlook. In The Sources of Normativity, I present a version of Kant's view that may be characterized as constructivist, and in 4.5.5 (p. 166) I explain the sense in which it can be harmonized with a form of realism. The view of pleasure and pain sketched at 4.3.1–4.3.12, pp. 147–56, is intended to show how the thesis that pleasure and pains are perceptions of reasons fits into that view.
(32) The type of work I have in mind is exactly that Barbara Herman undertakes in her paper “Making Room for Character,” in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty.
(33) This essay was originally written for the conference whose proceedings are published under the title Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty. I am grateful to Stephen Engstrom and Jennifer Whiting for organizing one of the best conferences I have ever attended. I also read it as the Gareth Evans Memorial Lecture at Oxford, and at the Universities of California at Irvine and Michigan. I am grateful to the audiences on those occasions for discussion.