Doctrine of the Mean

What is the doctrine of the mean?

Aristotle's doctrine of the mean holds that every moral virtue is a disposition to choose the intermediate "mean" between two vices of excess and deficiency. We can feel confidence, anger, or pity either too much or too little, but the virtuous person (in both feelings and actions) aims at the mean. Virtue, on Aristotle's definition, is "concerned with choice, lying in a mean, that is, the mean relative to us, this being determined by the rational procedures by which a wise man (phronimos) would determine it." Courage, for instance, is a mean between rashness and cowardice; temperance, a mean between self-indulgence and insensitivity.

Is this an arithmetic midpoint?

The mean is not an arithmetic midpoint. Such a mechanical reading would be intuitively unsatisfying, since we often regard certain virtues as more closely aligned with one extreme than the other - generosity, for example, looks more admirable when it inclines toward lavishness than toward stinginess. Aristotle is concerned with a normative meson: what neither goes too far nor falls short, but is the appropriate amount given the particularities of the situation.

Is it a one-dimensional spectrum?

Nor is the mean located on a single one-dimensional spectrum. Using the example of bravery, Aristotle does not hold that fear and confidence are opposites on one axis; they are distinct variables that vary independently. One can therefore fail the mean in several ways: too much fear and too little confidence (the coward), too little fear and too much confidence (the rash), or too little of both (the insensible).

Is there a singular mean for each emotion across situations?

There is also no singular mean for each emotion across occasions. The person of excellent character feels only mild annoyance at a trivial slight but is enraged by torture, and on each occasion his reaction hits the mean. This person is more praiseworthy than someone who expresses an intermediate level of anger to both trivial and grave provocations.

To have one's emotions in a mean is to feel each emotion "at such times, on such matters, toward such people, for such reasons, and in such ways as are proper." Getting angry or spending money is easy; doing so toward the right person, in the right amount, at the right time, for the right end, and in the right way is "rare" and "praiseworthy" (II.9.2). Succeeding in this challenge is what merits a man of excellent character.

How does Aristotle attempt to show that the doctrine is true?

Aristotle's attempt to show that the doctrine is true takes the form of running through a quite large set of specific excellences of character and showing that the doctrine of the mean fits them all.

How can virtue be in a mean, if as far as the best and the good are concerned it is an extremity?

While virtue is formally defined as a mean between the vices of excess and deficiency, Aristotle famously argues that "in respect of its highest good and its care, it is an extremity” (II.6). This means, for example, that while courage sits between rashness and cowardice, it represents a vertical peak of excellence.

Aristotle’s metaphor of the virtuous person aiming at a target helps illuminate the extremity claim. While the bullseye is a mean in the sense that it is surrounded by potential misses (high, low, left, right), hitting it is the extreme of success.

What is the wrong interpretation of ‘relative to us’?

Recall that Aristotle defines virtue in NE II.6 as a “state concerned with choice, lying in a mean relative to us.” One of the most contentious aspects of Aristotle’s doctrine of the mean is his use of this phrase “relative to us.”

Some scholars, such as Broadie, favour an agent-centered reading on which the mean is relative to the individual’s own psychological starting point and capacities - much like Milo the Wrestler’s diet is relative to his specific caloric needs.

This reading draws support from Aristotle's accompanying trainer analogy (II.6.7), where Milo needs far more food (say ten pounds) than a novice gymnast (say two pounds). Hence the “right” amount for each would not be the same. It would be bad for the trainer to feed them both six pounds of food; six would be too little for Milo the athlete but too much for the beginner in gymnastics.

Applied to moral character, this suggests that virtues are agent-relative, with the correct level of each emotion differing from person to person. The worry is that virtue then becomes overly subjective and loses its authority.

What is the better interpretation of ‘relative to us’?

This is a misinterpretation, however, according to Brown, and stems from a misplaced focus on the athletes. Aristotle’s analogy is not suggesting that each individual has a different mean for each excellence (two pounds versus ten). Instead, Milo and the gymnast actually represent different situations and the trainer, who is deciding how to feed them, represents the moral agent. The important point is that any trainer of excellence would have prescribed ten pounds of food to Milo and two pounds to the gymnast.

The “mean relative to us” is therefore whatever is right for human beings in each particular situation: the “us” refers to human beings in general, not each individual agent. This interpretated is supported by Aristotle’s broader use of ‘us’ for ‘human beings’ throughout the NE (e.g. heroic excellence is “beyond us”) and in the Physics, “relative to us” regularly means “relative to the human viewpoint.”

So what is the proper view of ‘relative to us’?

In saying that we aim at the “mean relative to us”, Aristotle is not implying that different people of virtue hit different standards just because they are different individuals. The mean is relative to us as human beings - given our shared nature and ends - and relative to the particular situation at hand, as determined by reason i.e. by that reason which the phronimos would determine it. This does not change depending on the person making the decision.

Strongest evidence for this reading:

The strongest evidence for this view comes from Aristotle's own criteria for hitting the mean: acting "at the right time, about the right things, towards the right people, for the right end, and in the right way." None of these criteria mentions the agent's internal feelings or natural temperament; all point to external, situational variables. For a soldier on a battlefield, there is an objective mean of courage required by the situation, regardless of whether he is naturally cowardly or naturally brave.

Aristotle's doctrine therefore maintains the objectivity he clearly intended. Right action is sensitive to context but not arbitrary or purely individual, and there is no sign in any of Aristotle's writings that the virtues differ from agent to agent or that any excellence admits multiple correct forms. Where many varieties of a virtue are discussed - courage being the clearest case - Aristotle makes clear that only one is true courage.

Why does the doctrine appear useless to some?

Hans Kelsen famously dismissed the doctrine as a tautological or empty formula. If virtue is defined as hitting the mean, and the mean is defined as "what the person of practical wisdom (phronimos) would do," then the advice "act according to the mean" simply translates to "do what the right person does."

Bernard Williams similarly dismissed it as little more than common sense or vacuous advice. The doctrine imposes no concrete rule, leaves the notion of the mean extremely vague, and leaves it to each individual to work out the mean for each situation using experience and reason. Hence, to some, it appears useless.

How is it still action guiding?

Hursthouse, however, argues that the doctrine is action-guiding through its generation of v-rules (virtue rules). While Aristotle does not provide a single master-rule, he provides a rich vocabulary of thick moral terms: the doctrine tells us to "be courageous" and "avoid being rash." These carry significant prescriptive weight.

Instead of aiming for bravery - an unknown abstraction - it is much more accessible to aim for the mean between rashness and cowardice, two concepts that are immediately recognisable and universally rejected.

How else is it action-guiding?

Further, the doctrine guides action by directing the agent's attention to the specific dimensions of a situation - the "right time," the "right person," the "right reason" - that must be satisfied for an action to be noble.

Curzer develops this point further, arguing that Aristotle's dimensions of the mean provide an action-guiding checklist for the agent: if you are angry, the doctrine forces you to ask whether you are angry at the right person, for the right reason, in the right way. Its importance lies in its ability to direct attention to the multi-axial nature of a situation.

Who rejects we should search for an action-guiding rule at all?

McDowell goes further and rejects the assumption that we should search for an action-guiding rule at all. Aristotle is famously explicit that his ethical inquiry does not aim at the same precision as the theoretical sciences, insisting instead that the subject matter of human conduct only allows for an account delivered "in outline." This is due to the inherent variability of "the fine and the just," which are too contingent upon circumstance to be reduced to universal laws.

This is not a structural defect but a deliberate feature of Aristotle's agent-centered framework: because the goal of the Nicomachean Ethics is not merely "to know" what virtue is but "to become good," it is no failure of the doctrine if it is not strictly action-guiding in a legalistic sense, since it was never intended as a decision-making algorithm.

The importance of the doctrine lies in its acknowledgment that ethics is uncodifiable: it does not provide a map that anyone can read but describes the kind of thing a virtuous person perceives. It illuminates the sensitivity of the phronimos and hence explains why ethics requires practical wisdom rather than a rule book, since rightness is a target that requires perception, not a formula.

How does our natural temperament matter?

Whilst the end is objective, the means to becoming a virtuous individual remains agent-relative.

In II.9.4 Aristotle prescribes that “we must examine what we ourselves drift into easily.” This is because different people have different natural tendencies towards different goals.” Some people are naturally quick-tempered, and others may be so because of their upbringing. On the other hand, others may be very slow to anger and find it very easy to deal with many anger-provoking situations.

Different hot-tempered individuals will have their tempers flare in different settings and not others, and we can “come to know our own tendencies from the pleasure or pain that arises in us” in given situations. Thus, different agents have different emotional constitutions – which we can work out by examining our reactions to certain situations – that will prevent us, or make it more difficult, to hit the target of the mean.

What must we do to address our nature?

Aristotle therefore suggests that we “we must drag ourselves off in the contrary direction” in order to reach the intermediate condition like “straightening bent wood” (II.9.5).

Each of us will, for example, in trying to act or react courageously, have to make adjustments for our different character traits. 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, must – when uncertain about the mean and in particular situations - aim at what is timid to avoid being blown further towards rashness.

This is how Aristotle reconciles his commitment to an objective mean with his evident awareness that individual psychology shapes the practical task of becoming virtuous: the mean is not relativised to the agent, but the route to it is.

How is ethical life like a skill?

Furthermore, a point Aristotle makes in II.1 is that “none of the virtues of character arise in us naturally” but we acquire them “by having first performed virtuous actions.” For example, “abstaining from pleasures makes us become temperate, and once we have become temperate we are most capable of abstaining from pleasures. (II.2.9).

In the same way, the more an individual aims at the mean and, subsequently, acts in accordance with virtue, the more closely they resemble the person of practical wisdom (phronimos) and hence more able they are to correctly identify the mean. It is in sense that ethical life is a skill or habit that one acquires rather than a calculus.

Hitting the target of the mean is “a matter of active, engaged participation in a complex situation”, in the words of Peter Losin, and there is no procedure we can go through which will enable us to fix in advance the location of the mean. Only through skilled guesswork and the experimental use of reason, in a range of situations, will an individual become a virtuous one.

Are there certain things for which there is no mean?

Aristotle does not assert universal application of the doctrine of the mean. Instead, the very nature of certain types of action and feelings makes them by definition vices, and thus not open to the trichotomy of deficiency, mean and excess.

The doctrine of the mean only applies to those feelings and actions that can be good or bad by varying degrees whereas some feelings, such as spite, shamelessness and envy, or actions, such as adultery, murder or theft, are never virtuous in any degree. They themselves, “rather than their excess or deficiencies, are base” (II.6.18). It is impossible to do them well; by committing murder, for instance, at the right time in the right way with the right victim.

What is the definitional defence?

One defense makes a definitional point, that some terms have wrongness built into their meaning. "Adultery," "theft," and "murder" are not neutral descriptions of action-types that can be done well or badly; they pick out, respectively, wrongful sex with another's spouse, wrongful taking, and wrongful killing.

The neutral descriptive equivalents (killing, taking, sexual intercourse) do admit means: there's a right amount of killing in war, a right time to take what one needs, a right partner for sex. Aristotle's point is that once you've reached for the morally loaded term, the wrongness is already presupposed. To ask for the mean of adultery is like asking for the mean of injustice.

What is the substantive defence?

A second, more substantive defence holds that some emotions and actions are constitutively vicious in their object or aim. The wrongness here is not merely linguistic but lies in the intentional structure of the act or feeling itself.

Spite is not an excess of some neutral underlying capacity that could be moderated; it is pleasure directed at a wrongful object - namely, others' undeserved suffering. Envy is pain directed at a wrongful object: others' deserved good fortune. The corruption lies in the relation between the feeling and what it is aimed at, not in any quantitative overshoot. There is no "moderate amount of taking pleasure in a stranger's suffering" that would be virtuous, because the trouble is not the quantity of pleasure but the fact that the pleasure is taken in that object at all.

This extends naturally to the action cases. Adultery is sexual activity directed at a wrongful object (another's spouse); theft is acquisition directed at a wrongful object (another's rightful property); murder is killing directed at a wrongful object (an innocent human life).

Why is the substantive reading superior?

This substantive reading has two advantages over the linguistic defence.

First, it explains why Greek ethical vocabulary came to encode wrongness in these places.

Second, it gives a unified account of both halves of Aristotle's list, actions and emotions, rather than the two halves requiring separate explanations.

Pakaluk emphasises a related point for the emotion cases: these are not continua with a virtuous middle, because the zero point - feeling no pleasure at others' suffering - is not a deficiency but the appropriate response. The mean structure requires that both extremes be vicious; here, only one direction is, because the underlying intentional orientation is corrupt from the start.

Does Aristotle’s list appear arbitrary?

A further worry is that Aristotle's list looks arbitrary. Why adultery but not, say, lying, when Aristotle is clear elsewhere (NE IV.7) that truthfulness is a virtue and falsehood "in itself mean and culpable"? Why murder but not cruelty to slaves? The selection looks less like the output of a principled criterion and more like a list of acts that fourth-century Athenian society had already encoded as paradigmatic wrongs.

The charitable response is that Aristotle is offering examples rather than an exhaustive list. But if the category is what matters and the specific list is incidental, Aristotle's failure to articulate the criterion for membership becomes more conspicuous, not less.

Why are moral dilemmas worrying?

The deepest worry concerns moral dilemmas. The doctrine of the mean is rightly praised for its sensitivity to circumstance: virtuous action requires perceiving the morally salient features of a situation and responding with the right feeling, toward the right person, at the right time, in the right way. Aristotle insists that ethics is not a matter of fixed rules but of perception (aisthesis) and practical wisdom (phronesis) attuned to particulars. But the no-mean claim sits awkwardly with this.

If theft, adultery, and murder are wrong simpliciter, then Aristotle is committed to a class of exceptionless prohibitions that operate independently of the perceptual machinery he develops elsewhere.

Could there really never be a case where theft is the right thing to do - stealing bread to feed a starving child, lying to the murderer at the door, killing in self-defence? Each of these looks like a paradigm case for phronesis: the practically wise person perceives that what would normally be wrong is, in this circumstance, the right thing — or even required.

What is the natural defence?

Aristotle's most natural defence is the redescription move: the bread-stealer is not really committing theft, the killer in self-defence is not really committing murder.

But if the no-mean claim is preserved only by allowing redescription whenever circumstances justify the act, the claim becomes empty: it tells us only that wrongful taking is wrong and wrongful killing is wrong, leaving the substantive question - when is taking wrongful, when is killing wrongful - to the very circumstantial reasoning the no-mean claim seemed to exclude.

Better response:

The more promising response is to acknowledge tragic conflict without permitting the wrongful acts: the agent who steals bread for the starving child does wrong, but does the lesser wrong, and should regret what she has done while accepting that she had to do it.

This preserves the no-mean claim while allowing for moral complexity, but at the cost of admitting that virtue does not guarantee right action - the virtuous agent in a tragic dilemma will inevitably do wrong.

Why does Aristotle even discuss these cases?

This is more a solution to an objection someone might raise “is there a mean of adultery.” If we’re talking about adultery, we’re already talking about something that’s already wrong.