Aristotle's Doctrine of the Mean

(Originally appeared in History of Philosophy Quarterly 4/3, July 1987.)

Aristotle's doctrine of the mean is sometimes dismissed as an unhelpful and unfortunate mistake in what would otherwise be -- or perhaps, in spite of this lapse, still is -- a worthwhile enterprise. Bernard Williams, for example, clearly regards it thus:

Aristotle's...views on [virtue] are bound up with one of the most celebrated and least useful parts of his system, the doctrine of the Mean, according to which every virtue of character lies between two correlative faults or vices..., which consist respectively of the excess and the deficiency of something of which the virtue represents the right amount. The theory oscillates between an unhelpful analytical model (which Aristotle himself does not consistently follow) and a substantively depressing doctrine in favor of moderation. The doctrine of the Mean is better forgotten.[1]

Williams's remark strikes me as both unfair to Aristotle and, perhaps as a result, blind to certain ethical insights of which Aristotle is keenly aware. In this essay I shall offer a more charitable interpretation of the doctrine of the mean. In sections I-III I bring together various things Aristotle says in developing his view that virtue or excellence lies in the observance of a mean. In section IV I turn to the obvious fact that as I have interpreted it the doctrine of the mean does not provide detailed and unambiguous guidance to agents deliberating in particular situations. I suggest that it was not intended to provide such guidance, and argue that this does not mean that it is not a useful part of Aristotle's ethical theory worth the attention of moral philosophers.

I

Aristotle develops the doctrine of the mean in the course of his discussion of aretê, excellence or virtue, in Book II of the Nicomachean Ethics (see also Eudemian Ethics, Book II, chapters 3 and 5).[2] There he writes that

all excellence makes what has it good, and also enables it to perform its function well. For instance, the excellence of an eye makes the eye good and enables it to function well as an eye; having good eyes means being able to see well. Likewise, the excellence of a horse makes it a good horse, and so good at galloping, carrying its rider, and facing the enemy. If this is true in all cases, then, the excellence of a human being will be that disposition which makes him a good human being and which enables him to perform his function well. (1106a16-25)

The function or characteristic activity of human beings, Aristotle has argued in Book I, is "a way of living... consisting in the exercise of the psyche's capacities in accordance with reason, or at any rate not in opposition to reason"; a good person "exercises these capacities and performs these activities well." Excellence, then, is that condition which best suits us to perform those activities which are distinctively human. Hence the best life for a human being will involve "the active exercise of his psyche's capacities in accordance with excellence" (1098a12-18).

But where does the mean come in? Aristotle summarizes his account of excellence in Book II, chapter 6:

excellence... is a settled disposition determining choice, involving the observance of the mean relative to us, this being determined by reason, as the practically wise person would determine it. (1106b36-1107a2; cf. EE II.5, 1222a6-10)

But why should excellence or virtue involve the observance of a mean?

The notion of the mean, and that of the observance of the mean, would have been familiar to those who attended Aristotle's lectures. They were at the conceptual center of the most advanced and sophisticated science of the day, medicine. Aristotle's father was a physician, and medical concepts and examples played an important and widely-recognized role in the philosophizing of Aristotle's day. Health was believed to lie in a balance of powers, in a mixture so constituted that none of its constituent elements eclipsed the others. The author of the Hippocratic treatise On Breaths writes that "opposites are cures for opposites. Medicine is in fact addition and subtraction, subtraction of what is in excess, addition of what is wanting."[3] Aristotle himself expresses this view, e.g. in the Topics (139b21, 145b7-10). Proper balance or proportion makes for health, lack of it for disease (On the Generation of Animals 767a20-35; cf. Physics 246b3-20).

Aristotle imports this way of thinking into his account of ethical excellence or excellence of character. Bodily strength and health are destroyed by excess and deficiency. Too much food, or too much exercise, are bad for health, just as too little food or exercise are. The same holds in ethical matters. Here too excellence is

so constituted as to be destroyed by excess and deficiency ... (here we must explain what is invisible by means of visible illustrations). (1104a12-13)

Bodily health is a matter of observing a mean between extremes of excess and deficiency. Further, Aristotle says, this provides an apt visible illustration of an invisible truth about ethical health. Excellence of any kind, Aristotle says,

aims at the mean [tou mesou... stochastikê: I discuss the importance of this construction below]. Excellence of character is concerned with emotions and acts, in which there can be excess or deficiency or a mean. For example, one can be frightened or bold, feel desire or anger or pity, and experience pleasure and pain generally, either more or less than is right, and in both cases wrongly; while to have these feelings at the right time, on the right occasion, toward the right people, for the right purpose and in the right manner, is to feel the best amount of them, which is the mean amount -- and the best amount is of course the mark of excellence. Likewise, in acts there can be excess, deficiency and a mean.... Hence excellence is a mean state in the sense that it aims at the mean. (1106b15-29; cf. EE II.3, 1220b22-34)

In this important passage, to which I shall return shortly, we are invited to compare excellence of character -- or the person who has such excellence -- to a skilled archer able to hit a target. Aristotle begins the NE with this simile (1094a23-24), and he returns to it throughout. I shall argue that it can shed a good deal of light on the idea that virtue or excellence lies in a mean.

Missing the mark[4] is possible in a virtually indefinite number of ways. A person aiming at a target can miss to the right, to the left, above, below; a crooked shot can glance off the target, etc. To hit the mark one must land a shot within a relatively small, more or less precisely defined, area. Just so, Aristotle suggests, what is excellent and commendable to do is definite and limited. There is a correspondingly vast, relatively unlimited, area for wrongs and shots that miss the mark:

Missing the mark is possible in many ways (for badness is a form of the indefinite, to use Pythagorean terms, and goodness a form of the definite), while success can be had only one way (which is why it is easy to err and hard to succeed -- easy to miss the mark and hard to hit it). (1106b29-33)

Now while hitting the mark is in this sense a much more precise matter than missing it, there is still room for variation within the shots that hit the mark. More than one shot can hit the bulls-eye of a good-sized target, and all such hits are scored the same. And a shot need not hit the exact center of the bulls-eye to be an excellent one. In the same way, Aristotle's simile suggests, virtue rarely demands a single precisely determined act, or an emotional reaction of a particular intensity, duration, frequency, etc. It rather demands that one's acts or emotions fall somewhere within a more or less precisely delineated range.

For example, the person who flees from every danger is cowardly; the person who does not flee from anything is rash. What is courageous, then, falls somewhere between these extremes; courage is "preserved by the observance of the mean" (1104a26). The same is true of temperance; what is temperate lies in a mean between the extremes of excessive enjoyment of sensual pleasures and deficient enjoyment of such pleasures. Similar things, Aristotle thinks, can be said for each virtue. There are important differences among the dispositions Aristotle calls virtues, of course; but each virtue involves the observance of a mean between extremes. One extreme consists in some sort of excess; another in some sort of deficiency, though (as I shall argue) this way of talking can mislead. Our task in trying to be good is to find these means and avoid these opposed extremes.

The means in question are "relative to us." What are we to make of this? Aristotle explains:

By the mean of a thing I mean what is equally distant from either extreme, which is one and the same for everyone; by the mean relative to us what is neither too much nor too little, and this is not the same for everyone. For instance, if ten are many and two few, we take the mean of the the thing if we take six; since it exceeds and is exceeded by the same amount; this then is the mean according to arithmetic proportion. But we cannot arrive thus at the mean relative to us. Let ten lbs. of food be a large portion for someone and two lbs. a small portion; it does not follow that a trainer will prescribe six lbs., for maybe even this amount will be a large portion, or a small one, for the particular athlete who is to receive it.... In the same way then one with understanding in any matter avoids excess and deficiency, and searches out and chooses the mean -- the mean, that is, not of the thing itself but relative to us. (1106a29-b8; cf. EE II.5, 1222a23-37)

"The mean according to arithmetic proportion" is a point, a fixed and determinate amount. We cannot arrive at the mean relative to us by this method, for at least four reasons. First, the mean relative to us need not be equidistant from two opposed extremes the way an arithmetic mean is. Secondly, unlike an arithmetic mean, the mean relative to us is "of considerable range and not indivisible" (On Generation and Corruption 334b26-30); by this Aristotle means that it is not an extensionless point. Thirdly, as we have seen, Aristotle's target simile suggests that there is room for variation among shots all of which hit the target. What virtue or excellence demands is not a fixed and determinate act or emotional response on a particular occasion, but that our acts and emotions fall within a certain more or less precisely delineated range. Aristotle himself points out that in practical matters the arithmetic mean is not particularly useful (see, e.g., Topics 139b21, 149a35-b4; On the Heavens 312b2). Fourthly, each of us is different; the mean relative to us will consequently also be different, and cannot be determined without close attention to features of the persons to whom such means are relative and the circumstances in which those persons are placed. The importance of this will become clear when I turn in section II to discussing particular Aristotelian virtues.

Seen one way, then, the possibilities for error are indefinitely various. Any shot that misses the mark in any direction qualifies. There is a sense, then, in which the remark Aristotle quotes at 1106b35 -- "there is but one way to act nobly, many ways to act basely" -- is true. Seen another way, however, the recipe for such error is absolutely precise: any shot that lands anywhere beyond the fixed edge of the target counts. This comports well with what Aristotle says earlier about excellence of character, that there is nothing fixed and invariable about matters of excellent or virtuous conduct (1104a4-12); the excellent thing to do is anything which falls within a certain range. What is excellent depends upon circumstances, just as the appropriate amount of food or exercise does. It cannot be determined with arithmetic precision (1104a1-6).

There are, however, emotions and acts which are absolutely vicious and disgraceful and are so in ways that do not depend upon circumstances. Aristotle's examples are malice, shamelessness, and envy (emotions), and adultery, theft, and murder (acts) (see, e.g., 107a12-26; cf. EE 1221b18-26). There cannot be commendable or praiseworthy exercises of malice, shamelessness, and envy; nor can one deserve praise for committing adultery, theft, or murder.

So Aristotle is not saying that one's emotions should always be of moderate intensity, or that one's acts should always express moderate amounts of particular emotions. The view that one should be moderate in everything is (pace the opening passage from Williams) not a fair statement of Aristotle's doctrine of the mean. Some things -- the acts and feelings just mentioned -- should never be done or felt; other things should be done or felt with our whole being 1166a15-23; 1169a18-36).

II

Aristotle points out that a general account of the mean is not likely to be helpful without concrete examples (1107a28-30; cf. EE II.3, 1221b8-9). In the course of Books II, III, and IV of the NE he discusses many virtues and their corresponding vices, arguing that in each case the virtue involves the observance of a mean between extremes. For example, in discussing andreia, courage, in Book III Aristotle suggests that it "is the observance of the mean regarding fear and confidence" (1115a6; see, however, the entire passage: 1115a6-1116a3). Aristotle does not, as some commentators have suggested, think of fear as the opposite or absence of confidence, or of confidence as the opposite or absence of fear.[5] Rather these are two distinct variables which can vary independently of one another. There are therefore several ways one can fail to hit the mean with respect to these variables. One can on a given occasion display too much fear and too much confidence; we have no special name for this kind of person, but while he puts on a show of courage, he does not endure (1115b31-33). One can display too much fear and not enough confidence; this is the coward. One can display too little fear and too much confidence; this is the rash person. Lastly, one can display too little fear and not enough confidence; this person is crazy or insensible.

J.O. Urmson suggests that Aristotle has in effect presented us with two continua:[6]

Emotions and acts can fall anywhere on the first continuum, and anywhere on the second. The courageous person observes the mean regarding fear and confidence; he avoids the errors listed above. The mean with respect to the first variable need not correspond exactly with the mean with respect to the second, for the variables are independent of one another. And, further, there is no particular point on the continuum from cowardice to insensitive fearlessness to which his act must correspond; neither is there such a point on the continuum from lack of confidence to overconfidence. The courageous person hits the mark; as we have seen, Aristotle's target simile suggests that this does not imply that there is room for only one excellent or commendable shot. His act or reaction falls within a range of acceptable alternatives on each continuum. What is acceptable depends at least in part on the circumstances. But what circumstances? And in what ways does what is excellent depend on these circumstances?

"The same things are not fearful to everyone" (1115b7). Some people are by nature confident and assured of themselves. Others are not. One who is naturally bold or overconfident may find it easy to conquer fear of certain things. A naturally timid person may not. Some people fear certain things and situations more than other people do, and certain things and situations more than other things and situations. Because the mean is relative to the individual one cannot tell whether an individual deserves praise for being courageous unless one knows something about that person -- specifically, about that concerning which she is especially fearful or fearless, unconfident or overconfident.

This is fine as far as it goes, but it is clear that Aristotle does not regard courage as simply a matter of landing a shot within a certain range on these two continua. The courageous person also avoids fearing the wrong objects, fearing things in the wrong ways and at the wrong times; and similarly, we are told, regarding situations inspiring confidence (1115b17-18).[7] Fearing the wrong objects or situations is not simply being too fearful, or fearing too many things; nor is fearing things in the wrong ways and at the wrong times simply fearing them too much. Likewise, being confident in the wrong ways and at the wrong time may involve being too confident, but it need not. Evidently the two-continuum picture is too simple.[8]

Aristotle introduces a further complicating element to his account of excellence as a mean when he looks more closely at courage in Book III, chapter 8 (see 1116a17-1117a26). Not every disposition which enables one to overcome fear and lack of confidence is equally excellent and commendable. In particular, citizen's courage [politikê andreia], the courage born of experience [empeiria] or of spirit [thumos], the courage of the merely optimistic [hoi euelpides] or the ignorant [hoi agnoountes], all enable their possessors to overcome fear and lack of confidence. But none of these dispositions is true courage; none is a genuine excellence of character.

True courage -- unlike the five imposters Aristotle mentions -- is a disposition in which fears and confidences are balanced and mastered "for the sake of the noble" (tou kalou heneka, 1115b12-13; dia to kalon, 1117b31).[9] So not only can one fail to hit the mark by being too fearful or not fearful enough, too confident or not confident enough; one can miss the mark by fearing the wrong things, by fearing them in inappropriate ways or on wrong occasions; one can also miss the mark set by true courage by fearing the right things in the right ways and on the right occasions, but by not doing so (as we might put it) for the right reasons or in the right spirit. And this element in Aristotle's discussion resists unpacking by setting out continua.

Does this not show that Aristotle's language of excess and deficiency is too crude, that the model he suggests (the "unhelpful analytical model," as Williams puts it) is not apt? Or should we rather resist the quasi-quantitative analysis given the notions of excess and deficiency by the continuum model? Before turning to this matter it will be useful to consider another example of an Aristotelian excellence of character. Consider, then, what Aristotle says about praotês, even-temperedness, at 1125b27-1126a29. We have here at least five continua:

As in the case of courage, acts and feelings can fall anywhere on each of these continua. Each represents, in principle anyway, a distinct variable, and each varies independently of the other four. There are, then, any number of ways to miss the mark with respect to anger. One can display anger too frequently or not frequently enough, too mildly or too violently, for too short a time or for too long a time; one can feel anger toward people who have done nothing to make anger appropriate or fail to feel anger toward people who have done something to which anger is a proper response; one can feel anger at insignificant things or fail to feel anger at important wrongs.

This is sufficient to show that Williams's claim that according to Aristotle "every virtue... lies between two correlative faults or vices" rests on an oversimple view of the doctrine of the mean. But this picture, replete as it is with possibilities for error, still does not capture an important part of what Aristotle is saying. Getting angry at the wrong people (1126a14) is not primarily a matter of getting angry at too many people. Nor is getting angry on occasions when anger is uncalled for (1126a18-20) a simple matter of feeling anger too often. And not getting angry when one should get angry (1126a4-9) cannot fairly be characterized as simply getting angry on too few occasions, or as a simple matter of reacting too mildly. Once again the continuum model seems misleading. The errors Aristotle is talking about cannot be so easily characterized. Excess and deficiency, it seems, are not to be unpacked in the simple quantitative way the continuum model suggests.

True even-temperedness, like true courage and any other true excellence of character, is "for the sake of the noble." Of course this makes it possible to miss the mark in even more ways. It is possible, I suppose, to attend scrupulously to my liability to anger, taking care not to be too violently angered by situations, or angry at the wrong people, or for too long a time; if I do this simply to impress others with my self-mastery or from fear of being blamed by someone, this is not genuine Aristotelian even-temperedness. It is not done for the sake of the noble. Not only must my acts and reactions fall within the proper range on the continua set out above; they must do so for the right reasons, in the right spirit. Excellence of character demands that excellent states be sought and chosen for the sake of the noble.

As in the case of courage, we cannot tell whether a person deserves commendation for her temper unless we know something about her -- in particular, about what she is especially provoked by, what sorts of situations and people she is especially sensitive to, and so on. People differ widely in these respects. Some people are naturally quick-tempered; others are so as a product of upbringing. Some others are at the opposite extreme: nothing provokes their anger, and they spend their lives getting stepped on like doormats. A naturally slow-tempered person may find it easy to deal with some (not necessarily all) anger-provoking situations. A naturally hot-tempered one may not, and her hot temper may flare only in certain settings and not others.

III

We are now in a position to see why the simple quantitative model will not do as an account of Aristotle's doctrine of the mean. First, avoiding extremes is only one necessary condition for hitting a particular dispositional mean-state. It is not sufficient. The extremes must be avoided for the right reasons, for the sake of the noble. Secondly, how the extremes are best avoided is not as simple as the continuum model suggests. We do not effectively avoid the extremes simply by seeking moderation in everything. We do not avoid the extremes simply by aiming to land a shot within a certain range on (even several) one-dimensional continua, hard as that might be. What is excellent or commendable does typically lie within such a range, but its excellence or commendability consists of more than its place on various continua.

Here, I think, is where the target and archery similes are most useful. Aristotle tells us that excellence, like an archer, aims at a target. Now an archer trying to hit a target must take into account various things which would cause him to miss the mark. He must (since he cannot possibly hold his bow perfectly still) coordinate his release of the bowstring with the subtle movements of the bow. If there are strong crosswinds he must aim slightly into the wind, and the wind will blow his arrow onto the part of the target he wishes to hit. If he is aiming into the wind, he must aim high to compensate for the slowing effect of the wind. If he is aiming from the crest of a hill above the target, he must adjust for the effects of gravity. And so on. Hitting the mark involves being aware of, and adjusting for, factors like these.

Our emotional constitutions provide us with a set of these complicating factors which can cause us to miss the mark, and will do so if we do not compensate for them. Each of us will, e.g., in trying to act or react courageously, have to make adjustments for different crosswinds. If I am naturally timid, I may in some settings have to aim toward what is rash to overcome the effects of my timidity. A naturally confident person, on the other hand, would be blown in the opposite direction; she must in certain circumstances aim at what is timid to avoid being blown further toward the rash. Likewise, a naturally slow-tempered person may have to work very hard to get angry at certain things he is naturally prone to endure meekly. The hot-tempered or bitter person might have to aim at what is timid in order to counteract her tendency, under particular sorts of conditions, to fly off the handle at slight provocations. These are the things of which an equable temper is made.

This is one of the reasons why Aristotle says that particular excellences of character involve observing a mean relative to us. It is also why he says that the mean relative to us cannot be determined with arithmetic precision: where we should aim to hit the mean will vary a great deal depending on the kinds and directions of crosswinds, headwinds and tailwinds; their strength; whether they are constant or intermittent; whether or not there are gusts; whether there are variations in the terrain which might produce unusual pockets of turbulence. Hitting a target in conditions like these is not a matter of fixing one's sight unwaveringly on one particular point (the geometrical center of the bulls-eye); it involves close attention to, and adjustment for, a variety of factors which would otherwise make us miss the mark. Hitting the mark is a matter of active, engaged participation in a complex situation. How, and how much, and when, and in what ways we should adjust is not something that can be said prior to close attention to the circumstances of the situation. There is no procedure we can go through which will enable us to fix in advance the location of the mean. (It is worth noting that the verb stochazesthai, literally "to take aim," e.g. at a target, is used in the NE and some contemporary works of a kind of skilled guesswork, an experimental use of reason which is sensitive to the details of particular situations (see, e.g., 1106b15; 1109a30; 1126b29; 1127a6-8; 1128a6; 1129b15; 1141b13-16; cf. Politics 1266b28; 1324b7; Rhetoric 1395b10; cf. Plato, Gorgias 465A2; Philebus 55E-56A; Laws 635A2, 962D1-5; cf. On Ancient Medicine, chapter 9). Our word "stochastic" has some of these connotations, though unlike its Greek ancestor it suggests randomness.)

Hitting the mark set by particular virtues, as we have seen, requires acting or reacting for the sake of the noble. An archer who is good at hitting the bullseye of a target under difficult conditions can use his skill to miss the mark when he finds occasion to do so, e.g. when someone has paid him to throw a match. He can use his skill as well for unworthy purposes -- destruction of property, e.g., or paid assassinations. Genuine excellences of character cannot be bought off in this manner. The excellent person's marksmanship is for the sake of the noble. Facility in hitting the mark, however commendable and essential to excellence of character, is not sufficient. But then Aristotle's talk of excess and deficiency is not adequately unpacked in the way the continuum model suggests. Avoiding extremes of excess and deficiency is a necessary condition of true excellence of character, but is not by itself sufficient for such excellence. Excellence or virtue is not mere skill.

IV

All this makes it very hard to say in advance with any precision where the mean lies relative to us:

it is not easy to define [ou radion diorisai] how and with whom and on what grounds and how long one should be angry, and up to what point one does correctly in so doing and where error begins.... Now how far and in what way someone must overstep to be blameworthy is not easy to set out by principle [ou radion tôi logôi apodounai], since what matters here are the details of the case, and the judgment lies in perception [en têi aisthêsei hê krisis]. (1126a32-34, b2-4)

And it is no easy matter to hit the mean, as Aristotle insists in a number of places (1109a25-29, 1109b13-24; cf. EE II.5, 1222a11-b4). Still, Aristotle has some general advice to offer those who are aiming at, trying to observe, the mean:

What is necessary first in aiming at the mean is to avoid that extreme which is the more opposed to the mean.... Since of the two extremes one is a more serious error than the other, and since hitting the mean accurately is hard, the second-best thing... is to take the lesser of the evils. The best way to do this is as we said. We must also attend to what we ourselves are most prone to, for different people naturally incline to different faults.... We must drag ourselves away in the opposite direction, for if we stay away from error we will attain the mean. And we must beware especially of what is pleasurable; none of us is an unbiased judge when it comes to pleasant things... All this makes it apparent that it is the mean disposition in every case that is to be praised, but also that sometimes we must lean to the side of excess and sometimes to that of deficiency, for this is the easiest way of hitting the mean and of doing well. (1109a30-b27; cf. 1108b11-1109a19)

What Aristotle is saying here is this. To determine where the mean lies in a particular case, and what the observance of the mean demands, I must attend to the details of the case. Among these details are those concerning my own character. I must realize, and adjust for, the aptnesses I have to various sorts of errors, most noticeably those involving excesses and deficiencies. I must compensate for my tendencies to over- or under-react, my susceptibilities to certain things and situations, my prejudices and biases. This may require that I overcompensate, aiming at what (were I to land a shot there) would be wide of the mark. I must realize that certain settings bring out the worst in me, and try to avoid those settings, or (again) compensate for their tendency to bring out the worst in me. And I should be especially wary of aspects of situations which I find pleasurable: pleasure -- and the prospect of pleasure -- is likely to impair my judgment, and it makes it very hard to find, let alone consistently hit, the mean.

All this seems to me very sound advice. That it is procedural and schematic, not substantive and informative; that it is not precise; that it does not by itself provide me with detailed and unambiguous guidance in particular cases, would not have bothered Aristotle. All that can be offered at this level of generality is a sketch (1094b19-23). Detailed informative advice comes only after close attention to particular cases (the point, after all, of the archery simile); in aiming at the mark we must "look to what suits the occasion" (1104a10).[10]

In general and for the most part, however, human beings are more liable to certain excesses and deficiencies than to others (see, e.g. EE II.5, 1222a38-b4). It may be that most of us are more prone to err to the right-hand side of the continua above; there may be some feature of human nature which accounts for this. Few of us are naturally prone to be too liberal with our possessions; in most the tendency is to the opposing extreme (see e.g. 1121a17-29). In some cases, then, one extreme is "more opposed to" the mean than the others -- and these will have to be compensated for by most people trying to hit the mean.

Still, there is no general way, no algorithm or principle, to fix or define the mean in particular cases. What is necessary, Aristotle says in many places, is aisthêsis, perception or sensitivity (see, e.g. 1109b23; 1142a27; 1147a26; 1172a36). The details of particular situations, which are too fine for coarse-grained rules to capture, can be caught by careful perception. Aristotle makes this point at 1109b22-23 and 1126b4-5 by contrasting matters which can be defined or set out by principle [tôi logôi aphorisai, tôi logôi apodounai] and those in which the judgment lies with perception [en têi aisthêsei hê krisis]. Krisis is judgment or discernment of the sort that rests on balanced and careful and active appreciation of the particulars of the case. As an archer aiming at a target, the person aiming at the mean must be sensitive to a very complex situation, and must be able to anticipate and adjust to minute changes in that situation:

in the case of...all the virtues there is a certain mark to aim at, on which the person who has reason fixes his gaze, and increases or relaxes the tension accordingly... (1138b21-23)

Aristotle argues in the paragraphs following this passage that the person whose perception and discernment is most acute is the practically wise person. (This is why, in the account of excellence or virtue quoted above [1106b36-1107a2], it is in observance of a mean relative to us, determined by reason, as the practically wise person would determine it, that excellence consists.) The practically wise person has a knack for hitting the mean, hits it consistently in a wide variety of circumstances. She is the balanced person, the person who is ethically healthy and whose character and emotions and actions therefore exhibit "proper balance or proportion." Aristotle is not suggesting that we blindly defer to this person's judgments and opinions about where the mean lies. He does suggest, however, that the reactions, opinions and considered judgments of the practically wise person are important standards to which we may find it useful to appeal in deliberation. Still, in the situations we face the mark we are interested in hitting is a mean that is relative to us, not to the person of practical wisdom. Such a person may be good at hitting such a mark, but she cannot do it for us. She may be able to advise us; but it is up to us to hit the mark (1105b5-18).

V

I have argued that Aristotle's doctrine of the mean is not the simple (and false) platitude that we should seek everything "in moderation." Nor is it "an unhelpful analytical model" of the sort suggested by the continuum model discussed in sections II and III. Nor is it the simple-minded view that "every virtue... lies between two correlative faults or vices." And it cannot fairly be regarded as a rule or set of rules designed to tell us what, in particular cases, to do. Aristotle develops the notion of the mean, as we have seen, as part of his account of excellence or virtue. Excellence is preserved by the observance of the mean (1104a26). The best life for a human being, then, namely one which consists of "the active exercise of his psyche's capacities in conformity with excellence" (1098a16-18), consists in the observance of the mean. Hitting the mean is not so much a matter of hitting one particular point on a target as it is a matter of avoiding the variety of mistakes it is possible to make in a complex situation. Observing the mean -- and so virtue or excellence -- is primarily a matter of careful awareness and avoidance of errors. Excellence of character, like health, involves a balance of opposite tendencies to act and react, a capacity to respond in various ways when and as occasions demand. This is the crux of Aristotle's doctrine of the mean. Far from being, as Williams suggests, one of the "least useful parts of his system" it seems to me both central to that system and a helpful and illuminating piece of ethics.[11]

NOTES

1. Bernard Williams, Ethics and the Limits of Philosophy (London, 1985), p. 36. [Back to text]

2. All unattached references to Aristotle's works are to the Nicomachean Ethics (hereafter "NE"). I shall refer to the Eudemian Ethics as "EE." [Back to text]

3. On Breaths 1, translated by W.H.S. Jones, Hippocrates and the Fragments of Heracleitus (Cambridge MA, 1923), volume 2, p. 229. [Back to text]

4. Hamartanein: forms of this verb appear frequently in the New Testament, where it is usually translated "to sin"; see, e.g., Romans 3:23; 7:7-25. [Back to text]

5. See e.g. H.H. Joachim, who takes confidence [tharros] to be "the contrary of" fear [phobos] -- Aristotle: The Nicomachean Ethics (Oxford, 1951), p. 117. Rackham, in the Loeb translation of the NE, comments that "in using ta tharralea ['situations inspiring confidence'] as the opposite of ta phobera ['situations inspiring fear'] Aristotle follows Plato, Rep. 450E, Protag. 359C, Lach. 195B, etc." (p. 156, note c). Leaving Plato out of it, these are misreadings of Aristotle. He is not suggesting that fears (or situations inspiring them) are "the opposite of" confidences (or situations inspiring them). For a sounder view see D.J. Allan, The Philosophy of Aristotle (London, 1970), p. 129. [Back to text]

6. See J.O. Urmson, "Aristotle's Doctrine of the Mean," American Philosophical Quarterly 10 (1973): 223-30, p. 230. [Back to text]

7. Successfully observing the mean involves achieving what is right [dei] and avoiding what is wrong [mê dei, ou dei]: see 1104b20-26; 1106b18-24; 1107b27-28; 1109a26-30; 1115b15-19; 1115b34-35; 1118b22-27; 1119b16-18; 1120a10-11; 1120a25-26; 1120a32-34; 1120b3-4; 1120b20-24; 1120b29-31; 1121a20-26; 1121a29-31; 1122a32-34; 1125b31-35; 1126a9-11; 1126a14-15; 1126a32-34; 1126b5-7; 1128a1-2; 1142b27-28. [Back to text]

8. Rosalind Hursthouse makes this point against Urmson's way of construing the doctrine of the mean in her "A False Doctrine of the Mean," Proceedings of the Aristotelian Society 81 (1980-81): 57-72, at pp. 60-61. [Back to text]

9. These expressions are ubiquitous in the NE: for the former see, e.g., 1115b22, 1117a4-5, 1120a24, 1120a28, 1122b6-7, 1123a24-25; for the latter, 1116a11-12, 1116b2-3, 1116b31-32, 1117a16-17, 1117b7-9, 1119b16, 1121a1-2, 1121b1-2, 1121b4-5, 1121b9-10, 1126b32, 1127a4-5, 1168a33. [Back to text]

10. In the early pages of the NE Aristotle likens proper ethical procedure to both medicine and navigation (1104a3-10, e.g.; cf. 1097a11-14, 29-32; 1137b13-33; 1141a21-25, 31-34; 1180b7-28). That Aristotle finds archery, medical, and navigational similes illuminating of ethical matters is surely important, and says a lot about how Aristotle conceives of ethics, but I cannot discuss his use of these similes here. [Back to text]

11. I would like to thank Jon Moline for helpful discussion, and for his comments on drafts of this essay. [Back to text]