Aristotle’s Doctrine of the Mean Reconsidered
Michael Pakaluk
“In the Steps of the Ancients,” St. Andrews University, Nov. 5-6, 2011
1. Aristotle’s doctrine of the mean—of betweenness and intermediacy—is commonly understood to
be about states of character, on the one hand, and actions and emotions, on the other. Virtues are
states of character which are located between two vices, one marked by deficiency and the other by
excess; and virtuous actions and emotional responses fall between alternative ways of acting and
feeling which would be misguided through being either deficient or excessive. On this common
understanding, a state of character, or hexis, is an enduring quality of the soul, a disposition, which
may be actualized in actions and emotions, but which need not be—a person, for instance, who fell
asleep for the rest of his life would possess all of his states of character, but these would not be
manifested. It is obscure how a quality of the soul may be located between other such qualities,
especially when these qualities are exclusive, and a person can have only one at a time. Hence on
this common understanding a virtue’s being “between” two vices is understood simply as its being
for, its being realized in, emotions and actions which are “between” vicious ways of acting.
Naturally enough the doctrine of the mean is then taken to have significance only if it can help to
account for why virtuous actions are indeed those that we ought to do. But as actions and
emotions do not seem to be right because they are moderate, and as we often go astray because we
get things wrong in a way that cannot be construed as wrong in degree, Aristotle’s doctrine of the
mean can begin to look like either a pious hold over from Plato and Greek tradition, or even worse,
as Rosalind Hursthouse has recently concluded, a “whacky” pseudo-scientific principle which
Aristotle regrettably imported into his ethics from his natural scientific investigations, and which is
best discarded.
2. I shall propose an alternative interpretation of the doctrine of the mean, the key idea of which is
that a hexis for Aristotle, in his discussions of virtues and vices, is best understood not as a quality of
the soul, or even as a “disposition,” but rather in the sense of a “habit”, and that a hexis in this
sense, although admittedly said by Aristotle to be “in the soul”, is just as public and observable as an
emotion, which--with its characteristic onset, progress, and decline--Aristotle also describes as being
“in the soul.” The doctrine of the mean as I construe it is true in an obvious and commonplace sort
of way, and ends up being crucial for the interpretation of the Ethics as a whole.
3. The interpretation of the doctrine of the mean which I am rejecting is often explained with respect
to two words that Aristotle uses, which may both be rendered as “mean” or “intermediate”, namely,
mesotēs and meson. It is held that, although Aristotle sometimes uses these words interchangeably,
he typically reserves mesotēs for a state of character which is intermediate, and meson for an
action or emotional response which is intermediate. So I wish to examine first the meaning of
mesotēs and then that of meson.
4. Aristotle introduces mesotēs in the following passage in Nic Eth II.2, which is the first of two
statements of his of the doctrine of the mean. I parse the text so that its structure becomes clear.
Two theses are advanced, and, for each, Aristotle first gives an illustration or analogue from the
condition of the body, and then he applies each thesis to the two virtues of courage and
moderation:1
[from Nic Eth II.2]
(Thesis) [1104a11] The first thing to observe is that these sorts of things are naturally
such that they are ruined by deficiency and excess
(Analogue)—since visible things should be used as evidence for invisible things— just as
we see in the case of strength and health. [1104a15] Physical training if either excessive
or deficient ruins strength. So too when portions of food or drink are too large or too
small they ruin health, but suitable portions produce, increase, and preserve health.
(Ethical claim) The same thing holds, then, with moderation and courage, and the other
virtues.
(Courage) [1104a20] Anyone who flees from and fears everything, and faces
nothing, becomes a coward; anyone who basically fears nothing but approaches
everything becomes rash.
(Moderation) So too anyone who takes his fill of every pleasure and abstains
from none becomes self-indulgent, and anyone who flees from every pleasure,
as rustics do, becomes a kind of insensient person.[1104a25]
(Restatement of thesis) Hence moderation and courage are ruined by excess (huperbolē)
and deficiency (elleipsis), but they are preserved by a mesotēs.
(Thesis) Yet not only does their origin and increase, and ruin, result from the same
things and by means of the same things, but also their operation will consist in the same
things.
(Analogue) After all, among the other, more evident cases, this sort of thing holds.
Strength, for example, [1104a30] results from taking much nourishment and enduring
many exertions, and the person who is especially able to do these things is the strong
man.
(Ethical claim) The same thing holds with the virtues also.
(Moderation) From abstaining from pleasures we become moderate, and
[1104a35] once we have become so, we are especially able to abstain from
them.
(Courage) So too[1104b1]with courage, since through becoming accustomed to
looking upon fearful things with contempt and standing firm in the face of them,
1
See the Appendix for the Greek text.
we become courageous, and once we have become so we are especially able to
stand firm in the face of fearful things.
5. I offer some comments on this passage.
a. First, the term “these sorts of things” (toiauta) at the opening is important, as it indicates
what Aristotle takes to be the scope of the doctrine of the mean. What are “these sorts of
things”? Clearly, for Aristotle, at least, conditions of the body and of the soul. But
probably the scope is greater. The passage is immediately preceded, you may recall, by one
of those methodological passages in the Ethics where Aristotle insists that “accounts
concerned with actions” can be stated only in outline, and that in particular cases the
practitioners themselves have to look at what the occasion requires, and then he refers to
medicine and piloting a ship as examples. So “these sorts of things” are perhaps not simply
conditions of body or soul but rather anything which falls under the care of some expertise
or skill. If so, then Aristotle is conceiving of the doctrine of the mean to apply to any
expertise which achieves a goal through some kind of ordering which gets adapted to
particular circumstances.
b. Second, just as Aristotle is concerned with plans or patterns of physical training which
increase or ruin strength, so in the passage, the terms huperbolē (“excess”) and elleipsis
(”deficiency”) mean patterns of actions, or ways of life, and a mesotēs is a pattern of action
or way of life which is intermediate between these. His claim as regards the mesotēs is that
the pattern of action which leads to the acquisition of the virtue is the same pattern of
action that represents the operation of the virtue. The identity of the pattern before and
after acquisition of the virtue, of course, leads to the puzzle considered in chapter 4,
namely, of how ever we can do the actions of a virtue prior to our having acquired the
virtue; and Aristotle’s answer in that chapter, anticipated here, is that the acquisition of a
virtue is a matter of doing the same sorts of actions from a different motive (choosing them,
for their own sake) and doing them habitually—which in the present passage is analogized
as a kind of strength in doing them.
c. Third, Aristotle’s predominant interest in the passage is with how extremes ruin good
condition. The extremes are taken to be obvious, both in the patterns of action they display
and in the tendencies that lead to them, and the intermediate is picked out relative to these
extremes. “These sorts of things are naturally such that they are ruined by deficiency and
excess,” he says, and “moderation and courage are ruined by excess and deficiency.” The
evident lesson of the passage is that acquiring or holding onto an ethical virtue is a matter of
staying clear of the corrupting extremes.
d. Fourth, Aristotle’s characterization of the tendencies to the extreme, “flees from and fears
everything ... faces nothing... takes his fill of every pleasure”, looks like a caricature, but
rather it seems an idealization, and an important idealization, of what Aristotle elsewhere
refers to as “living by one’s emotions” (to kata pathos zēn). What he has in mind is
someone for whom the basis for his action or omission is no more than that that is what he
felt like or did not feel like doing. In physical training, to the extent that one’s motive is like
that, to that extent one had departed from one’s regimen and is no longer “in training”. For
that time, unless by accident what one feels like happens to coincide with what one’s
discipline would require, one is moving in the direction of becoming more flabby or
becoming more worn down; also, unless the spell of doing what one feels like was allowed
in advance and circumscribed by the regimen, there is no reason why it should ever end or
not reintroduce itself as the basis for one’s actions the next day. So to say that a person is
not someone who “takes his fill of every pleasure and abstains from none” is not to say
something trivial, that there is at least one pleasure which he abstain from, but rather to say
that he follows some kind of rule or discipline in enjoying pleasures, and, insofar as he is
pursuing virtue, he is never not under a discipline. We reach the Socratic conclusion that to
acquire or practice a virtue is as if to be perpetually in training.
e. Finally, it seems that some notion of prohairesis (choice) is implicit in the passage: which is
important because commentators object that a virtue is defined later in chapter 6 as a
prohairetic hexis—a habit of making choices—and yet Aristotle says nothing explicitly about
prohairesis before then. To see this, consider that a lot of highly deliberate and structured
behavior may be the expression of someone’s simply acting on fear: as in Aristotle’s
example of the soldier who faces the enemy on the battlefield because he is aware that if he
runs away his commanding officer is standing behind the lines ready to kill him. Likewise by
“abstains from no pleasure” Aristotle cannot mean something like fasting from lunch to
enjoy all the more an expected extravagant dinner—the sort of thing that Plato called the
“shadow image” (skiagaphia) of virtue. So Aristotle needs to postulate some standard or
basis for action—something gained in action— for which pleasure and safety are definitively
traded away or exchanged, which is of course the kalon. As he says later, epithumia is not
set in opposition to epithumia: some other principle of action must be posited. And so we
see here incipiently the view, which gets developed later, that a virtue sets down, or is set
upon, the kalon in some domain and represents an habitual dedication to the kalon as to a
goal or telos, and that phronesis works out exactly how that goal is to be achieved in the
circumstances (as it considers ta pros to telos), and that a virtuous person habitually chooses
that action so deliberated, for its own sake, taking pleasure in that very action which is
chosen.
6. In the remainder of the Ethics, Aristotle’s usage of mesotēs is consistent with his usage here.
Typically the mesotēs cannot be identified with the virtue, and by that term Aristotle does not mean
a quality of the soul. For example, when later Aristotle again says that the mesotēs preserves the
virtue (τῆς δὲ μεσότητος σῳζούσης, 1106b12), or that it is characteristic of the virtue (τῆς δ' ἀρετῆς
ἡ μεσότης, 1106b34), or that not every action or emotion “admits of” a mesotēs (οὐ πᾶσα δ'
ἐπιδέχεται πρᾶξις οὐδὲ πᾶν πάθος τὴν μεσότητα 1107a9ff), or that there is a mesotēs “in” an
emotion or “in” receipts and expenditures of money (εἰσὶ δὲ καὶ ἐν τοῖς παθήμασι καὶ περὶ τὰ πάθη
μεσότητες 1108a31, ἐν λήψει καὶ δόσει χρημάτων μεσότης ἔστι, 1125b4), then the mesotēs is one
thing, clearly, a pattern or range of action, and the quality of the soul something else. And
Aristotle’s famous remark that justice has a mean different from other virtues simply cannot be
construed on the common interpretation: “the virtue of justice is a mesotēs,” he says, “not in the
same way as the other virtues, but because it is of a meson”—but that is just what any mesotēs is
supposed to be, on the common view, a state of character which is in a mean because it is of, or
related to, actions and emotions which are in a mean. In the small number of passages in which he
most looks like he is identifying a virtue with a mesotēs, still he qualifies in some way. In the famous
definition, a virtue is a prohairetic hexis which “consists in a mesotēs” (ἐν μεσότητι οὖσα); a virtue is
“a kind of mesotēs” (μεσότης τις, 1106b27). In one of his summaries he states that he has argued
“that a virtue is a mesotēs and that it is a hexis” (ὅτι μεσότητές εἰσιν καὶ ὅτι ἕξεις, 1114b27), as if
these were different claims—as they would be, if a mesotēs is a pattern of action, and to say that
the pattern is carried out habitually is something else.
7. As mentioned, it is no more strange that Aristotle refer to a pattern of action as a virtue than that
we refer to someone’s slamming a door with a certain kind of red face as anger. The point is
stronger if we think of the virtue, according to the literal sense of hexis, as a “holding” or “keeping”
to a certain pattern of action: so when Aristotle says that, because the striving or orexis for honor
admits of more and less, the hexis itself of striving for an intermediate amount is praised, even
though it is a mesotēs which lacks its own name (ἐπαινεῖται δ' οὖν ἡ ἕξις αὕτη, μεσότης οὖσα περὶ
τιμὴν ἀνώνυμος, 1125b20-21), there is no need to take this assertion to mean that something
besides the consistent striving is praised, the quality which causes it, rather than someone’s
“holding” or “keeping” to that way of life in a variety of circumstances. Indeed, Aristotle sometimes
refers to a virtue as the entirety of action of a certain type over a person’s life, a holon, as in his
discussion of the voluntariness of courage and moderation in relation to particular actions of those
virtues, ta kath’ hekasta (1119a27-32).
8. I now pass to the meaning of meson, which is a term not used until the other discussion of the
mean, in II.6. The passage which has received the bulk of scholarly attention is where Aristotle
explains what he means by “a mean relative to us”:2
In everything which is continuous and divisible, it is possible to take that which is
more, that which less, and that which is equal, and, these either with respect to the
thing itself (kat’ auto to pragma), or relative to us (pros hēmas). That which is equal
is a kind of meson between going too far and falling short.
By “a meson of the thing” I mean that keeps away by an equal amount from each
of the extremes, and this is precisely one thing and the same thing for everyone, but
by “a meson relative to us” I mean what neither takes too much nor falls short. This
is not one thing, and it is not the same thing for everyone.
For instance, suppose ten are many and two are few. Then six are the meson for
those who are taking “with respect to the thing,” since the amount by which six
exceeds, and that by which it is exceeded, are equal. This is the meson by arithmetic
proportion.
But that which is a meson relative to us is not to be taken in that way. It is not
that ,if eating ten ounces is a lot and eating two is a little, then the trainer will
prescribe six—since presumably even that is a lot for the person who will be taking
it, or a little: for Milo it is a little, but for someone who has only begun training it is a
lot; similarly for running and wrestling.
2
See the Appendix for the Greek text.
It is in that way, then, that every knowledgeable expert avoids going too far and
falling short; he seeks the meson, and chooses this, not that of the thing, but that
relative to us.
9. Two important points about the framing of this passage need to be emphasized. First, this passage
purports to be a more careful treatment of the earlier account in ch. 2 of ethical virtue. Immediately
prior to the quoted passage Aristotle says that “we have already discussed” (ἤδη μὲν εἰρήκαμεν) the
matter, but that he is going to offer the remarks that follow on the assumption that “we wish to
understand the particular sort of nature” of an ethical virtue (ἐὰν θεωρήσωμεν ποία τίς ἐστι ἡ
φύσις αὐτῆς , 1106a24-6), and his remarks, he says, will make this even more evident (ἔτι δὲ καὶ ὧδ'
ἔσται φανερόν). So the ch. 6 account of the mean is intended to be fuller or more precise or more
revealing version of the account which we have examined from ch. 2, and so presumably it needs to
be understood in that spirit.
10. But, second, it is a mistake to focus, as most commentators do, on the passage just given from ch 6,
as though that were alone or primarily giving the fuller account of ethical virtue. That passage
merely attempts to clarify what a “meson relative to us” is, but it does not intend to clarify what that
phrase means or implies specifically for ethical virtues. To see that this is so, the larger structure of
ch. 6 needs to be considered. Aristotle begins the chapter by saying that every virtue puts a thing in
good condition and makes it so that it carries out its work well, as, for instance, the virtue of an eye
makes it a healthy eye and also makes it so that it can see well. His concern in the chapter, he says,
is to explain how an ethical virtue makes it so that a person carries out his distinctive work well, -presumably, his ergon of living in accordance with logos. It is at this point that Aristotle explains
what a mean “relative to us” is. He then says, immediately after the quoted passage:
If every expertise in this way completes its work well (τὸ ἔργον εὖ ἐπιτελει), looking to
the meson and shaping its products by that [standard] .. and a virtue is even more
precise and better than any skill (just as nature is), then a virtue would be the sort of
thing that characteristically aims at and hits the meson (I mean, ethical virtue), (1106b816).
The passage about Milo and the meson relative to us is meant to explain something that holds of
types of expertise generally, as the basis for an a fortiori argument that ethical virtue has that
character to an even greater degree. It follows that Aristotle’s observations about a meson relative
to us have to be understood in a way that is completely general for any type of expertise. He is not
offering comments about ethical virtue in particular; that application comes later. It is pointless, for
instance, to wonder, as some commentators do, whether Milo or the trainer represents ethical
virtue.
11. Immediately we see that something like a distinction between what might be called an “active
meson” and a “passive meson” needs to be made. An active meson would pertain to some
determination by a skill or expertise, which could go wrong in something like the manner of a poor
calculation, as to excess and defect, and which would issue in a directive; and a passive meson
would pertain to the implementation or following of such a directive, and it would go wrong in the
manner of a poor execution, as to excess or defect. For example, consider a master stonemason
giving directions to an apprentice as to laying a foundation of stones for a house. The master
stonemason’s calculations as to the appropriate length, direction, and thickness of the foundation
are a matter of finding the appropriate proportion or fit of the foundation to the house under
construction, and these determinations could go wrong by way of excess or defect. In that sense
his determinations are a matter of finding a meson. But they are an active meson, as they issue in
directives which the apprentice needs to execute. The apprentice’s execution follows a meson, as in
execution he can depart from the stonemason’s directives by going too far or not far enough in
various directions, and this is a passive mean as he does not himself determine the dimensions of
the foundation but rather his job is to correspond accurately to what the stonemason says. As
Aristotle distinguishes “the part of the soul that possesses a logos” and that which does not possess
it but “shares” in it by “heeding” and “obeying it”, and each part has its own virtue, we should
expect that the way in which phronēsis “makes it so that someone does his ergon well” is by helping
someone find an active meson, that is, putting affairs in his life in a proper proportion to the goal he
is trying to achieve, and the way in which ethical virtues do so is by making it so that he can carry
out accurately the directives of phronēsis. The latter is presumably what Aristotle means when he
says that an ethical virtue is stochastikē tou mesou, that is, as rendered above, “the sort of thing
that characteristically aims at and hits the meson”, that is, the meson set by phronēsis.
12. The passage about the meson relative to us is best interpreted if, for the moment, we put aside the
question of what “relative to us” means, and why Aristotle used that phrase, but consider rather
what Aristotle is intent on denying. The main lesson seems to be that it does not suffice, for any
expertise, to determine what is to be done merely by adopting the aim of avoiding evident
extremes. Finding an arithmetic mean would be the obvious and most precise way of avoiding
extremes, but such a procedure does not yield a determination which is helpful in practice, and
certainly not a determination which would count as expert: the stonemason would not be regarded
as skilled if he determined the thickness of the foundation wall, say, by picking a dimension midway
between recognizably overbuilt and recognizably inadequate foundations. Note that the mesotēs
discussed in ch. 2 was that sort of thing, merely a middle ground between obvious extremes. So
applying the lesson to ethical virtue we get something like: it does not suffice, to have one of those
virtues, merely to avoid evident extremes—being good is not a matter of not being bad—but rather
highly refined considerations are required, which control every dimension, just as a skilled craftsman
controls every dimension in the production of his handiwork. That is why Aristotle concludes his
explanation of the meson relative to us by saying, as we have seen, “It is in that way, then, that
every knowledgeable expert avoids going too far and falling short; he seeks the meson, and chooses
this, not that of the thing, but that relative to us”: an expert does not simply avoid the extremes but
rather seeks the meson.
13. What Aristotle means by meson is simply what the dimension of an action should be like, when
every dimension is under control. That is why, immediately after introducing the notion of meson,
he almost stops using it entirely, and shifts instead to using various lists of what commentators have
called “parameters” of an action or emotion: toward whom it should be directed, under what
circumstances, for what cause, for how long and how often it should, in the degree and manner in
which it should be done, and so on. It is not that, as Rosalind Hursthouse has suggested, Aristotle in
practice finds the concept of a “middle” or “intermediate” unhelpful for accounting for right action
or emotion, but rather that meson was introduced in the first place to provide a bridge to this more
refined account of what has to be attained by an ethical virtue for it to be appropriately refined and
“skillful.”
14. But the second lesson we are meant to take from the passage about a meson relative to us is that, in
the matters that Aristotle is concerned about, namely, objects of various types of expertise, the
extremes are not actually “far” extremes, allowing a large middle ground, but “near” extremes,
falling immediately to either side of the correct determination—as the land falls away sharply on
either side of the Aonach Eagach. These near extremes are a consequence of the standard and do
not serve to define it antecedently. That is why Aristotle says the following in one of his summaries:
That ethical virtue is a mesotēs (understand, “a
middle ground”), and how it is, and that it is a
mesotēs between two vices, of excess and of
deficiency, and that it is that sort of thing on
account of its being stochastikē of the meson in
emotions and actions, has been sufficiently
discussed (1109a20-24)
Ὅτι μὲν οὖν ἐστὶν ἡ ἀρετὴ ἡ ἠθικὴ μεσότης, καὶ
πῶς,καὶ ὅτι μεσότης δύο κακιῶν, τῆς μὲν καθ'
ὑπερβολὴν τῆς δὲ κατ' ἔλλειψιν, καὶ ὅτι τοιαύτη
ἐστὶ διὰ τὸ στοχαστικὴ τοῦ μέσου εἶναι τοῦ ἐν τοῖς
πάθεσι καὶ ἐν ταῖς πράξεσιν, ἱκανῶς εἴρηται.
15. It follows that the relationship between mesotēs and meson is something like that between rough
tuning and fine tuning. The mesotēs is a name for the obvious middle ground between evidently
extreme, vicious ways of life, which can be seen by extrapolation to be the result of tendencies we
have which, if treated as sufficient for deciding our action, would amount to living “by emotion”; the
term “meson” is a determinable and stands for the whatever counts towards the complete and
refined determination of action and emotion, as proportionate to some larger goal or context, as
this is judged to be so by phronēsis. That is how I suggest the following should be understood:
That is why as regards its substance, and the
account of what is it for it to exist, a virtue is a ,
but as regards what is best and good, it is a
pinnacle. (1107a6-8)
διὸ κατὰ μὲν τὴν οὐσίαν καὶ τὸν λόγον τὸν τὸ
τί ἦν εἶναι λέγοντα μεσότης ἐστὶν ἡ ἀρετή, κατὰ δὲ
τὸ ἄριστον καὶ τὸ εὖ ἀκρότης.
That is to say, if we want to mark out just what it is for a virtue to exist at all—what “conserves” or
“preserves” it, Aristotle elsewhere says, it is enough to say that it is the middle ground between
extremes—as Aristotle does say when initially distinguishing virtues from corresponding vices in the
“chart” of virtues and vices in II.7, and perhaps, too, to say this is to say how a virtue “puts someone
in good condition” and makes him a good rather than a bad human being, but if we want to speak to
its excellent functioning and how a virtue makes it so that someone carries out his distinctive work
well, then it has to be described as the reliable seeking and attaining of that which is meson. The
famous definition of ethical virtue should be interpreted similarly:
A virtue then is a habit of choosing which exists
in its occupying a middle ground—[but] relative
to us, as that is marked out by logos and as a
person with phronēsis would mark it out.
Ἔστιν ἄρα ἡ ἀρετὴ ἕξις προαιρετική, ἐν μεσότητι
οὖσα τῇ πρὸς ἡμᾶς, ὡρισμένῃ λόγῳ καὶ ὡς ἂν ὁ
φρόνιμος ὁρίσειεν.
Without stretching, we see that the definition expresses two thoughts, the second refining the first.
16. Nonetheless, we might still want to know why Aristotle uses “relative to us” as the label for the
meson that interests him. Clearly “relative to us” does not mean “relative to particular cases”, since
that construction is too general, as there is nothing about something’s being adapted to particular
cases that so far makes it count as a meson, or as a special sort of meson—indeed, the arithmetic
mean needs to be applied correctly to the particular case. The phrase also cannot mean “relative to
human purposes”, as Lesley Brown has suggested, at least not in the sense in which she meant it,
that is, “relative to the basic constraints and needs of human nature”—because, as we have seen,
Aristotole understands the meson relative to us as being at stake in any skill or expertise, not simply
ethics. However, the phrase might mean “relative to human purposes” in the trivial sense in which
human purposes come into play whenever someone is adapting some stage in a process or part of a
whole to that process or whole, so that the whole thing comes out well—as human art and action
always involve human purposes.
17. But I rather think the phrase comes from Aristotle’s concern to explain how a dimensionality of
more and less may arise immediately around a standard which may not itself have the same
dimensionality, or any dimensionality at all, and his view is that a more or less, excess and defect,
are the consequence of the attempt to meet a standard in the presence of inclination. Consider the
following example. You are looking at a golf green from above, perhaps from one of those camera
cranes lifted high above a golf tournaments, and you see a white golf ball on the green, in the area
between the hole and the front of the green. The ball is not in the hole; clearly, the golfer who shot
it missed getting it into the hole; and to that extent, he failed at what he was trying to do. But did he
fail by hitting it too far or too short? You cannot tell. The ball might have been hit from either the
front of the green, and fell short, or from the back, and went too long; and similarly for right and for
left. If we consider only at the ball on the green, then the mistake of the ball’s not getting in the
hole has no dimensionality to it: it is not too much, too little, too far right, or too far left, anything.
A dimension can be attributed to the mistake only after some direction of striving is brought to the
circumstance, and we know where the golfer was hitting the ball from and where he was aiming. So
Aristotle uses the term “pros hēmas”, because this phrase or something like it is what he standardly
uses to indicate a relationality in the thing that is imputed only because of its relationship to us, as
when in the Physics he holds apparently that the directions of up and down are in the cosmos itself,
because, because of the tendencies of falling and rising bodies, but right and left are not in the
cosmos but attributable to it pros hēmas, that is, as a consequence of our having a right and left
(208b13-22). Relationality in the thing by nature is “always the same” (aei to auto, b15-16) and is
the same for everyone (malista pasi, 212a23); relationality pros hēmas differs, of course, depending
upon how we orientate ourselves first to it. Because Aristotle makes the same point in the Physics
about mathematical objects—“Though they have no real place, they nevertheless, in respect of their
position relatively to us (pros hēmas), have a right and left as attributes ascribed to them only in
consequence of their relative position, not having by nature (phusei) these various characteristics.”
(208b22-25)3—when he looks in the Ethics for a phrase for imputed dimensionality in contrast to a
3
δηλοῖ δὲ καὶ τὰ μαθηματικά· οὐκ ὄντα γὰρ ἐν τόπῳ ὅμως κατὰ τὴν θέσιν τὴν πρὸς ἡμᾶς ἔχει δεξιὰ καὶ ἀριστερὰ
ὡς τὰ μόνον λεγόμενα διὰ θέσιν, οὐκ ἔχοντα φύσει τούτων ἕκαστον, Ross translation above.
mathematical meson, he uses the same phrase. For an analogous use elsewhere in the Ethics
consider that when Aristotle argues that sometimes a vice is more contrary to the virtue than the
opposite vice, he says that contrariety can result either from a difference in the actions themselves
(ex autou tou pragmatos, 1109a6), as running away in battle is more opposed to courage than
rushing forward, or it can result from our greater inclination to one vice rather than the other, in
which case the contrariety “derives from us ourselves” (ex hemōn autōn, a12-13), as self-indulgence
is more opposed to moderation than insentience, because we are highly drawn to such pleasures,
even though superficially the way of acting of a self-indulgent person may look more like
moderation than the way of acting of an insentient rustic.
18. So tying up these threads we get something like the following picture. The vices are as it were
extremes set “by nature” as they represent extrapolations of living “from” emotions and inclinations
that we all have as being human; they are easily identifiable and agreed upon by all, which is why we
find that words are available naming them all, except those such as insensibility for which the
contributing inclination is weak; and similarly the middle ground in between them, the mesotēs, is
obvious by contrast –“not living in the manner of either vice”—and, I take it, importantly, one does
not need phronēsis to identify the mesotēs. Yet to pick out the mesotēs is only very roughly to mark
out the area of the virtue; it represents, as it were, only that some kind of rule or discipline has been
adopted. It marks out where the virtue must be operative in order to exist at all. Nothing like
knowledge or expertise would involve achieving its result in only that rough sort of way, so the
operation of an ethical virtue must be more precise, not, indeed, by its attempting to calculate more
precisely, and foolishly, an intermediate point equidistant from each of the vices, but rather simply
in its manner of looking for, and wanting to follow, indications of logos with respect to very open
dimension affecting an inclination, desire, or action, which can then be counted, too, as a meson,
because of the imputation of a dimensionality as a result of inclinations we bring to bear in our
following the indications of logos. Avoiding the far extremes, the vices, is hardly relevant to
someone who is practicing the virtues, but rather such a person will be concerned with fine
discriminations, choices, and preferences, which likewise can be faulted as being too much or too
little.
19. Criticisms of the doctrine of the mean seem to fall flat with respect to this interpretation. The
doctrine of the mean would obviously not represent any kind of importing into ethics of a “whacky”
pseudo-scientific theory of equilibrium. It is not intended to account for our judgments about what
should be done but presupposes the possibility of such judgments. (Most of what we look for in
accounting for rightness and wrongness would be for Aristotle a matter of justice, which, as we have
seen, he says is a meson in a distinctive way.) It is not an empty doctrine, but rather highly
interesting, because, for instance, it the way, in which it holds that the vices are obvious but fine
points of virtuous action are not, it points a way out of the circularity problem that seems to beset
phronēsis; and its notion that an ethical virtue heeds logos by being stochastic, by looking for a
rational standard that presents itself subjectively as a meson between more and less, where that
appearance is a consequence of our inclinations and biases, is highly sophisticated and apparently
invites self-reflection and self-knowledge, as essential to virtue.
20. The doctrine of the mean, moreover, ends up being important for the interpretation of the
Nicomachean Ethics as a whole, in the following way. The main interpretative problem of the Ethics,
it is thought, is the task of squaring what are believe to be Aristotle’s fairly conventional accounts of
the virtues in books II-V, with his vaunting of philosophical contemplation in book X. The force of
the problem perhaps arises largely from the sense that the virtues are given and prior, and have
their own shape and demands, and then contemplation looks like something that has to fit in with
or trump them. Let’s call that outlook “Aristotle as David Hume.” But suppose Aristotle is not like
Hume but is really more like Socrates in the Phaedo, and he is convinced that devotion to
philosophical contemplation has a reforming effect on all of the virtues? Well, we see something
like a systematic reform of the virtues undertaken in book X, in the sense that the domains in which
these virtues operate become limited in importance. I do not mean merely that the ethical virtues
are diminished on the grounds that the gods do not practice them, or because they are bound up
with the whole of human nature and therefore engage the body and the emotions, not simply nous,
but I mean, rather, on the one hand we are told that no limit is to be set to philosophical
contemplation, that we should attempt as much as possible to act as an immortal through practicing
it, but then, we are told, that such a person will not need many material goods, and bodily pleasures
will pale in comparison for him, and he will not need to strive for great actions, or even influence
and power, which courage needs. And then we may think of how already in the accounts of the
various virtues their significance was being crafted and adapted: greatness of soul as taking virtue to
not political office to be great; readiness to die in battle as being something which a virtuous and
eudaimōn person will not enthusiastically seek, liberality as being a virtue which when really
practiced is almost guaranteed to make someone poor, and so on. And then for this purpose, of
accounting for how the virtues should be adapted to an end or horos such as philosophical
contemplation, Aristotle’s doctrine of the mean is admirably suited, since it allows a philosopher to
say that he avoids the same extremes as anyone else but that his austerity, poverty and obscurity
represent a suitably different meson in what remains a middle ground: just as Milo needs to take so
much more that to others his portion would appear almost gluttonous, the philosopher assigns
himself scandalously less.4
4
I wish to express my debt to Dan Graham and James Siebach at BYU, and Henry Richardson at Georgetown, for
helpful comments on earlier drafts of this paper.
APPENDIX
(from Nic Eth II.2)
πρῶτον οὖν τοῦτο θεωρητέον, ὅτι
τὰ τοιαῦτα πέφυκεν ὑπ' ἐνδείας καὶ ὑπερβολῆς φθείρεσθαι, (δεῖ γὰρ ὑπὲρ τῶν ἀφανῶν τοῖς φανεροῖς μαρτυρίοις
χρῆσθαι) ὥσπερ ἐπὶ τῆς ἰσχύος καὶ τῆς ὑγιείας ὁρῶμεν·
1104a.15
τά τε γὰρ ὑπερβάλλοντα γυμνάσια καὶ τὰ ἐλλείποντα
φθείρει τὴν ἰσχύν, ὁμοίως δὲ καὶ τὰ ποτὰ καὶ τὰ σιτία
πλείω καὶ ἐλάττω γινόμενα φθείρει τὴν ὑγίειαν, τὰ δὲ
σύμμετρα καὶ ποιεῖ καὶ αὔξει καὶ σῴζει. οὕτως οὖν καὶ
ἐπὶ σωφροσύνης καὶ ἀνδρείας ἔχει καὶ τῶν ἄλλων ἀρετῶν.
1104a.20
ὅ τε γὰρ πάντα φεύγων καὶ φοβούμενος καὶ μηδὲν ὑπομένων δειλὸς γίνεται, ὅ τε μηδὲν ὅλως φοβούμενος ἀλλὰ
πρὸς πάντα βαδίζων θρασύς· ὁμοίως δὲ καὶ ὁ μὲν πάσης ἡδονῆς ἀπολαύων καὶ μηδεμιᾶς ἀπεχόμενος ἀκόλαστος, ὁ δὲ πᾶσαν φεύγων, ὥσπερ οἱ ἄγροικοι, ἀναίσθητός
1104a.25
τις· φθείρεται δὴ σωφροσύνη καὶ ἡ ἀνδρεία ὑπὸ τῆς
ὑπερβολῆς καὶ τῆς ἐλλείψεως, ὑπὸ δὲ τῆς μεσότητος σῴζεται.
ἀλλ' οὐ μόνον αἱ γενέσεις καὶ αὐξήσεις καὶ αἱ
φθοραὶ ἐκ τῶν αὐτῶν καὶ ὑπὸ τῶν αὐτῶν γίνονται, ἀλλὰ
καὶ αἱ ἐνέργειαι ἐν τοῖς αὐτοῖς ἔσονται· καὶ γὰρ ἐπὶ τῶν
1104a.30
ἄλλων τῶν φανερωτέρων οὕτως ἔχει, οἷον ἐπὶ τῆς ἰσχύος·
γίνεται γὰρ ἐκ τοῦ πολλὴν τροφὴν λαμβάνειν καὶ πολλοὺς πόνους ὑπομένειν, καὶ μάλιστα ἂν δύναιτ' αὐτὰ ποιεῖν
ὁ ἰσχυρός. οὕτω δ' ἔχει καὶ ἐπὶ τῶν ἀρετῶν· ἔκ τε γὰρ
τοῦ ἀπέχεσθαι τῶν ἡδονῶν γινόμεθα σώφρονες, καὶ γενό1104a.35
μενοι μάλιστα δυνάμεθα ἀπέχεσθαι αὐτῶν· ὁμοίως δὲ
1104b.1
καὶ ἐπὶ τῆς ἀνδρείας· ἐθιζόμενοι γὰρ καταφρονεῖν τῶν φοβερῶν καὶ ὑπομένειν αὐτὰ γινόμεθα ἀνδρεῖοι, καὶ γενόμενοι μάλιστα δυνησόμεθα ὑπομένειν τὰ φοβερά.
(from Nic Eth II.6)
ἐν παντὶ δὴ συνεχεῖ καὶ διαιρετῷ ἔστι
λαβεῖν τὸ μὲν πλεῖον τὸ δ' ἔλαττον τὸ δ' ἴσον, καὶ ταῦτα
ἢ κατ' αὐτὸ τὸ πρᾶγμα ἢ πρὸς ἡμᾶς· τὸ δ' ἴσον μέσον
τι ὑπερβολῆς καὶ ἐλλείψεως. λέγω δὲ τοῦ μὲν πράγμα1106a.30
τος μέσον τὸ ἴσον ἀπέχον ἀφ' ἑκατέρου τῶν ἄκρων, ὅπερ
ἐστὶν ἓν καὶ τὸ αὐτὸ πᾶσιν, πρὸς ἡμᾶς δὲ ὃ μήτε πλεονάζει μήτε ἐλλείπει· τοῦτο δ' οὐχ ἕν, οὐδὲ ταὐτὸν πᾶσιν.
οἷον εἰ τὰ δέκα πολλὰ τὰ δὲ δύο ὀλίγα, τὰ ἓξ μέσα
λαμβάνουσι κατὰ τὸ πρᾶγμα· ἴσῳ γὰρ ὑπερέχει τε καὶ
1106a.35
ὑπερέχεται· τοῦτο δὲ μέσον ἐστὶ κατὰ τὴν ἀριθμητικὴν
ἀναλογίαν. τὸ δὲ πρὸς ἡμᾶς οὐχ οὕτω ληπτέον· οὐ γὰρ
1106b.1
εἴ τῳ δέκα μναῖ φαγεῖν πολὺ δύο δὲ ὀλίγον, ὁ ἀλείπτης ἓξ μνᾶς προστάξει· ἔστι γὰρ ἴσως καὶ τοῦτο πολὺ
τῷ ληψομένῳ ἢ ὀλίγον· Μίλωνι μὲν γὰρ ὀλίγον, τῷ δὲ
ἀρχομένῳ τῶν γυμνασίων πολύ. ὁμοίως ἐπὶ δρόμου καὶ
1106b.5
πάλης.
οὕτω δὴ πᾶς ἐπιστήμων τὴν ὑπερβολὴν μὲν καὶ
τὴν ἔλλειψιν φεύγει, τὸ δὲ μέσον ζητεῖ καὶ τοῦθ' αἱρεῖται, μέσον δὲ οὐ τὸ τοῦ πράγματος ἀλλὰ τὸ πρὸς ἡμᾶς.