20th World Congress of Philosophy Logo

Modern Philosophy

Does Hume Have an Ethics of Virtue? Some Observations on Character and Reasoning in Hume and Aristotle

Marcia L. Homiak
Occidental College

bluered.gif (1041 bytes)

ABSTRACT: I argue that Hume's ethics can be characterized as a virtue ethics, by which I mean a view according to which character has priority over action and the principles governing action: virtuous character guides and constrains practical deliberation. In a traditional utilitarian or Kantian ethics, character is subordinate to practical deliberation: virtue is needed only to motivate virtuous action. I begin by outlining this approach in Aristotle's ethics, then draw relevant parallels to Hume. I argue that virtuous character in Aristotle is understood in terms of "self-love." A true self-lover enjoys most the exercise of the characteristic human powers of judging, choosing, deciding and deliberating. A virtuous agent's self-love enables sizing up practical situations properly and exhibiting the virtue called for by the situation. But if an agent's character is defective, the practical situation will be misapprehended and responded to improperly. I argue that though Hume claims moral judgments are the product of sympathy, they are actually the result of a complex process of practical reflection and deliberation. Although Hume writes as though anyone can be a judicious spectator, there is reason to think that persons of calm temperament, who enjoy deliberation and have a facility for it, are more likely to perform the corrections in sentiments that may be necessary. If this is so, an agent's character has priority over his or her practical deliberations.

bluered.gif (1041 bytes)

I am interested in the general question of how to characterize Hume's ethics, in particular, in whether Hume can be seen to offer some version of a virtue ethics. Let me first explain what I take a virtue ethics to be. For a virtue ethics, the central question is: "What kind of person should I be?" rather than "What action ought I to perform?" A virtue ethics is agent-centered in the priority it gives to being a virtuous person (or to having a virtuous character) over doing the moral action. Yet this priority has proven difficult to articulate and defend, since it looks as though character traits can be specified in terms of principles defining the kinds of action in which traits are expressed. (1) If so, a virtuous person is simply someone who acts in accordance with particular principles. Individual virtues, such as courage, then become derivative: they are secondary to the defining and governing principles of action, which are determined by practical reason. (2) At best, individual virtues are needed only to motivate virtuous action, since practical reason alone might not, or cannot, do so. If virtues are derivative in the way described, then principle- or action-based theories can welcome them, since they do not threaten the primacy of practical reason. (3)

In most of my work in moral philosophy, I have argued that Aristotle's mature ethical writings provide a legitimate alternative to principle- or action-based ethical theories. (4) They do so, I believe, by making available to us an ethics of virtue in which character is not subordinate to practical reason. I think Aristotle's writings contain a notion of character that can be articulated clearly enough to indicate, first, how character is not reducible to action and to the principles governing action, and, second, how character can function in a way that goes beyond providing motivation for the deliberations of practical reason that issue in the principles governing action. Character can go beyond providing motivation if the non-rational emotions, attachments, and attitudes that partly constitute character guide and constrain the deliberations of agents so as to affect what they determine to be best to do. If character can do that, then practical reason does not tell agents what to do, independently of character.

In this paper I will operate with a notion of character, regrettably broadly understood, and will not attend to whether it is reducible to principles of action. Rather, I will offer some support for the second claim, that character, broadly understood, can affect both the content of practical knowledge and its availability to agents.

Now someone might think that though this strategy could constitute a workable approach to Aristotle, it is an altogether wrongheaded approach to Hume. For, first, Hume denies the existence of practical reason by denying that reason can rank ends (see the famous passage at Treatise 416); second, he claims to have defeated the moral rationalists' position, (5) according to which the rules of morality are conclusions of reason (T 455-457); and, third, he argues later in the Treatise that the rules of morality are extensions of our tendencies to sympathize (T 581-585). In other words, there seems to be no practical reason, whose subordinate or superior relation to anything else we are being asked to understand.

But there are two reasons to think Hume does not successfully defeat the moral rationalists: first, his argument against the moral rationalists seems to confuse questions of moral epistemology with questions of moral motivation. (6) Surely Hume cannot be right to claim, as he does at T 455-457, that from the fact that I am not moved to do the right thing, it follows that I cannot understand what the right thing to do is. And, second, Hume's efforts to explain how moral distinctions arise from our passions at T 581-585 seem to invoke the very kind of substantive, non-instrumental reasoning (7) he earlier, at T 413-418, dismissed.

I have, then, several aims in this discussion: first, to indicate the kind of non-instrumental reasoning I take Hume to use; second, to indicate how this type of non-instrumental reasoning may be subordinate to character; and, third, to suggest how this subordination of reasoning to character can strengthen Hume's position against the moral rationalists. (8) I do not know whether I am reading Hume correctly. But if I am, I will have given some reason to think Hume, too, may have a version of an ethics of virtue.

I want to explain more clearly what I mean by saying that practical reasoning is subordinate to character by outlining how I take this approach to work in Aristotle. (9) Then I will draw the relevant parallels to Hume.

I think the best Aristotelian text for these purposes is the discussion in EN VII.1-3 of incontinence (weakness of will or akrasia). Aristotle begins by asking whether Socrates is right to think the incontinent agent is ignorant, and his answer is interesting. He characterizes the incontinent as failing to know the "last premise" of a practical syllogism (EN 1147b9) and as failing to have "perceptual knowledge" (EN 1147b18), thus seeming to lend support to Socrates' position. Yet Aristotle's view is more complicated, since he also maintains that during incontinence appetite interrupts the agent's normal deliberative processes, with the result that the incontinent is ignorant in the ways just described.

The question for interpreters of Aristotle is how to understand this description of incontinence. Should one concentrate on the ignorance and construe incontinence as a failure of deliberation and reasoning, or should one emphasize the presence of the contrary desire and construe incontinence as a failure of character, in particular, a failure to develop the non-rational attitudes and desires constitutive of virtuous character?

If the incontinent's normal deliberative processes are interrupted, then he fails to know, at the time of action, something he ordinarily knows and which virtuous (and continent) persons do know. There are various possibilities here. The incontinent could fail to know some general principle of right action, some specific application(s) of such a principle, or some principle of rationality, such as a principle of prudence. Unfortunately, none of these possibilities serves adequately to distinguish the virtuous person from the incontinent. If the incontinent failed to know a general principle of right action, he would be more like the morally vicious person. And if he failed to know a specific application of a general principle or a principle of prudence, then the virtuous person, to whom the incontinent person fails to measure up, would seem to be less flexible than Aristotle's description of the good deliberator suggests. (EN 1151a29-b4)

But what if we emphasize the presence of the contrary desire and construe incontinence as a failure of character? Then the breakdown in reasoning is accounted for by that desire and not by the failure of the agent to deliberate correctly, and the breakdown is repaired by developing different desires. On this view, the agent's more deeply rooted desires (i.e., those that constitute his character) will, by influencing what he takes to be practically salient, affect the ways he sizes up practical situations. In other words, the agent's character will influence his assessment of the kind of situation he faces, which is then reflected in the minor premise of the practical syllogism he applies to the situation. The minor premise calls forth an appropriate major premise, and the agent acts (or intends to act).

First consider how this works in the case of the virtuous person or true self-lover (EN 1169a2-3). A true self-lover enjoys most the expression of his characteristic human abilities, which are the ordinary human abilities of choosing, deciding, discriminating, and judging (EN 1166a17ff., 117b18-19; cf. EN 1095b6). Because the pleasures of touch and taste are relatively lacking in rational activity, and thus are easy enough for grazing animals, slaves, and women to enjoy (EN 1095b19-20, EN 1177a8, Politics 1260a12-13), they are not that interesting to the virtuous person. So the virtuous person will not characterize the chocolate cake in front of him as especially yummy and not to be missed. On the other hand, someone who lacks true self-love and is therefore not virtuous, who pursues the easy pleasures because more difficult ones bring disappointment and self-doubt, will see opportunities for experiencing the easy pleasures particularly enticing. He will characterize the chocolate cake in front of him differently from the way a true self-lover will, and this different characterization will call forth different practical premises. The agent's weak or absent self-love in effect causes him to misperceive the situation he is in, so that he does not see it as one to which his generally held principles apply. This misperception is his failure "to know the last premise". (10)

Here character is understood as the enjoyments and aversions associated with self-love or its absence. These non-rational desires influence the agent's reasoning processes by affecting the construction of practical premises. As the example, I hope, makes clear, the relevant non-rational desires are not the specific desires associated with individual virtues and vices. Rather, they are the non-rational desires constitutive of either (i) the stable and secure self-confidence of virtue or (ii) the weak and unstable self-confidence of the non-virtuous conditions of continence, incontinence, or moral vice. Thus character is being understood in a general sense. If an agent has true self-love, she will size up practical situations properly and then exhibit the particular virtue called for by the situation so perceived. (11)

If this is a correct interpretation of the relationship between deliberation and the non-rational desires of character, then the desire to do what is virtuous will be continuous with the natural enjoyment we take in the development of our talents and abilities. I mean by this that the desire to do what is virtuous will be an extension, or development, of our quite ordinary desire to exercise our human abilities. Because the enjoyment we take in that exercise makes being virtuous possible, our natural enjoyments are neither inevitable competitors, nor mere cooperators, with a desire to be virtuous. Thus Aristotle can say at EN 1099a9-11 that a horse is pleasing to the lover of horses "in the same way" that justice is pleasing to the lover of justice.

I turn now to Hume. I begin with Hume's use of the point of view of the judicious spectator to explain how moral judgments are arrived at and especially how agreement in moral judgment is possible. (T 580-587)

Hume opens his discussion by acknowledging that if moral judgments arise from our passional nature, then there is no reason to suppose that the content of morals is the same for all moral agents. This problem is especially acute if moral judgments arise from the operation of sympathy, since our sympathies vary depending on how closely connected we are to the objects of our sympathies. To prevent the "contradictions" (T 581) in judgment that would arise from our actual sympathies, and to secure the agreement in moral judgments necessary for the stable functioning of society, Hume says we "loosen" ourselves from our "station" (T 583) and take up a "steady and general" (T 581) point of view. From this point of view, our judgments express the moral approvals and disapprovals we would have were the persons surveyed to have the good fortune to produce the beneficial effects they would produce under normal circumstances. We sympathize with the hypothetical pleasure and pain produced by the circumstances we imagine to take place, and our judgments are then presumably proportional to the total happiness brought about by the persons under consideration.

For Hume, only the point of view of the judicious spectator accounts for our agreement in moral judgments, and only our shared human tendency to respond sympathetically makes possible the deliberations we undertake from that point of view. Here sympathy does not operate in addition to a mechanism of moral judgments that is already in place, as it would if it were subordinate to that mechanism. Rather, it makes the mechanism possible by showing how these judgments arise from the natural tendencies of our passions. In at least one sense, then, Aristotle and Hume are similar: the operation of our shared human tendencies accounts for, and makes possible, the reasoning that leads to our recognition that an action or character is virtuous.

But this point of comparison might disappear. For one could with good reason claim that Hume fails in his effort to show that moral judgments are based in sympathy. One could say that he fails because when one takes up the point of view of the judicious spectator and loosens oneself from one's station, one leaves behind one's actual sympathies and imagines what pleasures and pains would be produced were the persons under consideration to realize the effects normally realized by their characters. One asks, "What pleasures and pains would I sympathize with were circumstances appropriately different?" In effect, there are no actual pleasures and pains to feel; the claimed pleasures and pains are really the agent's imagined response to a hypothetical situation. Thus one might conclude that the moral judgments of the agent are actually a product of reflection and deliberation, not of human passions at all. For if we perform the corrections properly, we will need to know how much happiness tends to be produced by what sorts of characters in what sorts of circumstances, and we can't know this without having remembered and reflected on past experience. Our knowledge of which characters are "fitted to be beneficial to society" (T 585) is the basis for the general rules we rely on in making moral judgments in cases of unrealized virtue.

If that is so, one might think that Hume has given to the moral rationalist, his avowed opponent, all that she needs to claim that moral distinctions are based on reason. For our judgment that an individual is virtuous or not is a matter of our applying a complex rule for measuring pleasures and pains, which we have developed from thought and reflection. But I think Hume has an interesting response to the moral rationalist here, that again ties him with Aristotle, and indicates how reason remains subordinate to passion.

Even if we assume that Hume's moral judgments are the result of a complex process of reflection and deliberation, that process, like any other activity in which an agent engages, must itself be motivated by a passion, according to Hume. Since the moral sentiments are themselves calm passions, and since calm passions predominate in someone of strength of mind, it might help to consider how someone develops strength of mind if we wish to learn what motivates the reflection that leads to moral judgment.

Although Hume's official view is that reflection cannot alter our system of final passions (cf. Enquiry, sec. 244), he allows that deliberation can guide and direct the passions. This happens, for example, when reason reveals to us a better means to achieve an end we already have. (T 416) More interesting for our purposes, however, is Hume's view of how deliberation can alter the intensity and force of our present passions by putting some in the background and bringing others to the foreground. Consider his example at T 425-426. He recalls the situation in the Athenian assembly where the question is whether to accept Themistocles' latest scheme to insure the public good of Athens. Because the efficacy of the scheme depends on its being kept secret, the Assembly orders Themistocles to reveal the plan only to Aristides, known for his honesty and discretion. Aristides reports that Themistocles' plan is extremely advantageous, but also extremely unjust. The Assembly deliberates and rejects Themistocles' plan, thereby acting against the public good. Hume's explanation here is that because the Athenians knew the plan under no particular instantiation (it was merely a general idea, vague and unspecific), their deliberations did not render the plan's advantage as intensely felt as it would have been had they been able to deliberate with more information.

The point of this example seems to be that if we deliberate carefully with appropriately detailed information, and if we use that information to make the vague specific, then deliberation can call up and make influential previously weak passions, perhaps even latent passions we were not aware we had. Although reasoning cannot change our system of final passions, it can, then, affect which of our passions direct and control our actual conduct. This is an important point, for it amounts to a concession that we can, through the process of rational reflection, ultimately alter our tendencies of temper and disposition. In effect, we can, over time, alter the sort of persons we are.

This point seems confirmed by Hume's explanation of strength of mind. In a person with strength of mind, the calm passions prevail over the violent passions. (T 418) Their prevalence is explained in such a person by having been "corroborated by reflection, and seconded by resolution". (T 437) Perhaps Hume means that when the calm passions move us to deliberate, and deliberation shifts the violent passions into the background, thereby bringing other, calmer, passions to the attention of the agent, the practice of deliberation reinforces and confirms the calm passions' influence on our conduct.

Hume helps us understand more clearly how reflection confirms the influence of the calm passions by reminding us that, as we practice reflection and deliberation, we gain a facility at deliberating. With facility, we gain an inclination to engage in further deliberation, since the facility is enjoyable. "By degrees, the repetition produces a facility, which is another very powerful principle of the human mind, and an infallible source of pleasure." (T 423) That we enjoy deliberating and want to do more of it is, then, a sign that the calm passions influence our conduct.

If I am right that Hume allows for strength of mind to be built up over time through a process of reflection that produces an enjoyment taken in that very reflection, then Hume appears to associate a difference in character, broadly understood, with an enjoyment or lack thereof in the activities of deliberating and reflecting. This is our second point of comparison with Aristotle.

But there is a final point of comparison that seems to me to be more important. If a person of calm temperament, in whom the calm passions predominate over her violent ones, has built up her dispositions over time through an enjoyment taken in deliberation itself, then such a person will be more likely to take up the point of view of the judicious spectator. That kind of deliberation must also be motivated by a passion, and since it is a complicated deliberation that requires loosening oneself from one's own perspective and passions, a person who enjoys deliberation and who has gained a facility in deliberating will be more likely to want to take up this point of view and to be able to perform the corrections in sentiments that may be necessary.

It seems to me to follow that a person with facility at deliberating will be more likely to reason properly from the point of view of the judicious spectator. Making the corrections will be easier for a person of calm temperament, of "temper and judgment" (T 472), since she is better able to distinguish between sentiments that are easily confused, especially the sentiments of interest and morals. A person of calm temperament preserves herself from the "illusions" (T 472) that affect so many others and therefore deliberates properly from the point of view of the judicious spectator. She, and people like her, know that "virtue in rags is still virtue". (T 584) Moreover, a person of calm temperament will be more likely to know how much happiness tends to be produced by what sorts of characters in what sorts of circumstances. Having experienced the pleasures of deliberation, she won't overlook them in taking the measurements of hypothetical pleasure and pain. Persons of different character, however, in whom the violent passions predominate, will not enjoy the subleties of deliberating from a steady and general point of view. Their character might prevent them from adopting the point of view of the judicious spectator, or they might find the necessary corrections psychologically impossible to make. They will not recognize that virtue in rags is still virtue.

The final point of comparison between Aristotle and Hume could, then, be stated thus: moral knowledge depends on character, in the sense that it is available only to persons of decent character. On the interpretation of Aristotle presented here, the virtuous person's true self-love (her love of the rational activity typical of human beings) enables her correctly to recognize and respond to practical situations. And on the reading of Hume I have given, the deliberative process that issues in moral distinctions is available to a person of calm passions, an agent of "temper and judgment", whose knowledge is not obscured by violent passions, and whose enjoyment of deliberation influences her assessment of hypothetical pleasures and pains. (Cf. T 534) I say this, despite Hume's various remarks that suggest any person can take up the moral point of view. (See, for example, T 499-500, 583, 585, 586.) For this (Aristotelian) position seems to me to strengthen Hume's arguments against the moral rationalists in two ways: first, it does not deviate from his intended strategy of showing how moral judgments and moral sentiments are natural extensions of our nature. Second, it avoids the problems of the argument he does give against the moral rationalists, since it does not depend on being able to show that reasoning is merely instrumental.

Does Hume have an ethics of virtue? These brief remarks cannot settle this question. But I hope they have provided us with good reason to think the question should be pursued.

bluered.gif (1041 bytes)


(1) For a discussion of this difficulty, see William Frankena, Ethics, 2nd ed., Englewood Cliffs, NJ, Prentice-Hall, 1973, p. 65.

(2) By "practical reason", I mean non-instrumental reasoning about the ends of action or about which actions are right or are to be done. Successful practical reasoning will, then, often involve the development of other, related capacities, such as the capacity to grasp relevant particulars and means. In this sense, both Aristotle and Kant employ practical reasoning, whereas Hume, on the canonical reading, does not.

(3) For example, although a Kantian might consider some empirical motives a hindrance to moral action, she could accept other empirical motives as morality's helpmate. The development of sympathetic feeling, for instance, can make following the moral law easier, since sympathy "is one of the impulses placed in us by nature for effecting what the representation of duty might not accomplish by itself." (p. 122, The Metaphysical Principles of Virtue, tr. James Ellington, Indianapolis, Bobbs-Merrill, 1964). Here a virtuous character is subordinate to the directives of practical reason.

(4) See, for example, "Virtue and Self-Love in Aristotle's Ethics", The Canadian Journal of Philosophy, XI (no. 4), 1981, 633-651; "The Pleasure of Virtue in Aristotle's Moral Theory", Pacific Philosophical Quarterly, 66 (nos. 1-2), 1985, 93-110; "Politics as Soul-Making: Aristotle on Becoming Good", Philosophia, 20 (nos. 1-2), 1990, 167-193; and "Aristotle on the Soul's Conflicts: Toward an Understanding of Virtue Ethics", in Reclaiming the History of Ethics: Essays for John Rawls, eds. A. Reath, B. Herman, and C. Korsgaard, Cambridge, Cambridge University Press, 1997, 7-35.

(5) Ralph Cudworth and Samuel Clarke seem to represent well the position Hume is attacking. See An Inquiry Concerning the Principles of Morals, ed. L. A. Selby-Bigge, 2nd ed., Oxford, Clarendon Press, 1972, p. 197n.

(6) See, for example, the criticisms leveled against Hume in Christine M. Korsgaard, "Scepticism About Practical Reason", The Journal of Philosophy, LXXXIII (no. 1), 1986, 5-25.

(7) In particular, the kind of non-instrumental reasoning that would generate what Rawls calls principle-based desires. See John Rawls, Political Liberalism, New York, Columbia University Press, 1993, pp. 82-85.

(8) Since I shall be concerned to explore the roles played by sympathy and calm temper in the deliberations of the judicious spectator, I have restricted my discussion to the Treatise of Human Nature (cited as "T"). I follow the text of L. A. Selby-Bigge, 2nd ed., rev., ed. P. H. Nidditch, Oxford, Clarendon Press, 1978.

(9) I shall be concerned only with the Nicomachean Ethics (cited as "EN"). I follow the translation of Terence Irwin, Hackett Publishing, Indianapolis, 1985.

(10) For a more detailed discussion of this interpretation of akrasia, see my "Aristotle on the Conflicts of the Soul: Toward an Understanding of Virtue Ethics", in A. Reath, B. Herman, and C. Korsgaard, eds. (note 4, above).

(11) I do not mean to deny here that the virtuous person engages in deliberation or that she has formed particular practical principles as a result of deliberation. Nor do I wish to deny that she deliberates properly, in contrast to her non-virtuous counterparts. I mean only to uncover the non-rational conditions that cause her and other agents' deliberations to be as they are.

bluered.gif (1041 bytes)


Back to the Top

20th World Congress of Philosophy Logo

Paideia logo design by Janet L. Olson.
All Rights Reserved


Back to the WCP Homepage