I. Introduction
Aristotle says that ‘no human achievement has the stability of activities that express virtue, since these seem to be more enduring even than our knowledge of the sciences’.Footnote 1 Most moral philosophers have assumed that, on this issue at least, Aristotle should be trusted. Persons have characters, and those characters dispose them to good or evil deeds. Folk psychology concurs; in everyday conversation, we typically explain and predict actions on the basis of people's long-term personality traits – blissfully prized commodities in our fractured times. Similarly, virtue ethics, which for some time now has swept everything before it in ethics circles, focuses on the cultivation of moral virtues qua stable dispositions conducive to human flourishing. And in education circles, character education – a close cousin of virtue ethics and sharing many of the same assumptions – has become not so much the flavour of the month as the flavour of the past decade.
Dispositionism (or globalism with respect to character traits) is not an undisputed thesis, however. Social psychologists will have none of it. They consider dispositionism to be refutable by empirical evidence. In recent years, a number of philosophers, spearheaded by Gilbert Harman and John Doris,Footnote 2 have followed suit, launching a sustained attack on residues of dispositionism in ethics and moral education. This challenge has a rallying cry: situationism. According to situationism, there is no such thing as character in its etymological sense of an indelible mark impressed on an object. People have no robust traits; how they act varies with the situation. If this is true, all the suspects in the story – folk psychology, virtue ethics and character education – are guilty of the same ‘fundamental attribution error’.Footnote 3 The chief villain is, however, Aristotle: the prime instigator of the error. Character attributions rest on the tenets of an Aristotelian psychology that is some 2,500 years old and, from a scientific perspective, out of the ark.Footnote 4 The rejection of dispositionism will thus have particularly devastating ramifications for the neo-Aristotelian theory prominent in moral philosophy for the past quarter century.Footnote 5 Aristotle-inspired folk psychology finally goes down the same drain as Aristotelian cosmology did 500 years ago; virtue ethics will be shown to be going nowhere fast; and from an educational perspective, if there is no such thing as character, then character education is an illusion.Footnote 6
Judging from the number of rejoinders to the situationist challenge, situationism seems to have repelled more moral philosophers than it has attracted.Footnote 7 There is a steadily growing mountain of such rejoinders.Footnote 8 Arguably, this literature has now reached what qualitative researchers call a ‘saturation point’: the point at which the addition of new participants fails to provide new and significant information. I do not aspire to lift the saturation point in this paper by producing new objections to situationism. Rather, my aim is threefold. First, I propose what I consider to be a helpful classification of the existing objections under four main headings: ‘the methodological objection’, ‘the moral dilemma objection’, ‘the bullet-biting objection’ and ‘the anti-behaviouristic objection’ (see Sections III and IV). The last two of these objections draw essentially on Aristotelian sources. I argue, however, that the anti-behaviouristic objection in particular has so far deployed but a miniscule part of the available Aristotelian arsenal. My second aim is to resuscitate a more powerful Aristotelian version of this objection by fleshing it out through varied illustrative examples of human conduct – thus deepening, if not widening, the current discourse on situationism (see Section IV). My third and final aim is to explore some of the implications of such resuscitation for our understanding of the salience of character and for future studies of its nature (see Section V). Prior to all of this, I rehearse briefly, in Section II, some of the basic ingredients of the situationist challenge.
II. The Psychological Experiments and Their Alleged Implications
The psychological experiments that, first, social psychologists, and, later, moral philosophers have used as grist for the situationist mill are, I trust, familiar to most readers and require only the briefest of rehearsals here. Let me focus on the four experiments most commonly cited:
Honesty Experiment. Over 8,000 schoolchildren were placed in situations in which they could (a) cheat on artificially created tests (unrelated to schoolwork) by peeping or asking friends, (b) fake records or cheat at party games, (c) steal change left on a table and (d) lie about their conduct. The correlation across behaviour types (a)–(d) was only .227. Nor was any significant correlation found between the children's behaviour and their knowledge of the Ten Commandments or the Boy Scout Code. The researchers (Hartshorne and May) concluded that there is no such thing as cross-situational honesty, and, indeed, no such thing as character. This experiment, conducted in the 1920s, is by far the oldest of this type, and the one which has had the clearest practical repercussions: it influenced Kohlberg's moral-reasoning approach and gave succour to his dismissal of traditional character building as a ‘bag of virtues’ – thereby influencing the content of moral education for decades, until the renascence of character education in the 1990s.
Dime Experiment. In this experiment, performed in the 1970s, the subjects were adults making calls from public telephones in US shopping malls. Some of the subjects found a dime secretly left in the phone by the experimenter; others did not. As the subjects left the phone booths, a confederate of the experimenter ‘accidentally’ dropped a folder full of papers on the floor. As it turned out, most of the subjects who had found a dime stopped to help the confederate pick up the papers, but only one out of 25 participants who did not find a dime offered help. The experimenters (Isen and Leven) hypothesised that finding a dime led subjects to feel in a good mood, and that their mood – rather than any consistent character trait – prompted the helping behaviour.
Good Samaritans Experiment. When it comes to responding to the needs of a sick person, does it matter what religious views people hold, or whether they are preparing a talk on a religious or a secular theme? Not according to this experiment, performed in the 1970s. Whether seminary students believed that they were on their way to delivering a talk on the parable of the Good Samaritan or on a practical topic did not have a significant impact on their stopping to help a ‘sick’ confederate. However, the experimenters (Darley and Batson) found that whereas 63% of subjects who did not consider themselves to be in a hurry to reach their destination offered help, only 10% of hurried participants helped. Again, a situational factor seemed to be the crucial variable.
Milgram Experiment. The most famous of the four, this experiment, which took place in the 1960s, focused on a fictitious learning–memory test in which the ‘learners’ were strapped to chairs and supposedly given electric shocks by the ‘teachers’ each time they made a mistake on a learning task. The ‘teachers’ sat in an adjacent room and administered the shocks by pushing a button. All the subjects recruited for the experiment as ‘teachers’ knew beforehand what was considered to be a safe moderate, strong, intense and life-threatening voltage. The experimenter instructed the subjects, with increasingly forceful verbal prods, to give shocks of higher voltage after each mistake made by the learners. Contrary to prior predictions by their peers, all the subjects went at least to the ‘intense’ level, and two-thirds went all the way to the end of the (fictitious) shock series, ignoring cries of increased agony emanating from the learners. Only 2.5% of unprodded subjects went all the way, however. The experimenter (Milgram) saw this as an explanation of the apparently irrational obedience to authority shown by Nazi soldiers during World War II.
As mentioned, recent years have witnessed a burgeoning literature on the moral implications of those psychological experiments. Ethical situationists claim that people's behaviour is essentially situation-dependent rather than character-dependent. Undue freight has been heaped on the idea of a fixed component, character, which simply does not seem to be operative in the experiments; people are not as set in their ways as we used to think; and we have systematically underappreciated the salience of situational factors. The results of these experiments, then, are deemed at least sufficient to shake our previously imperturbable confidence in the existence of consistent cross-situational dispositions, and to call for ‘a certain redirection of our ethical attention’;Footnote 9at most even sufficient to eliminate the very idea of character, and to damn the entire fields of virtue ethics and character education.Footnote 10
In slightly more technical terms, situationists do not typically deny that people have dispositions to behaviours that are stable from this day to the next within a range of reasonably similar situations – the correlations between children's behaviours listed within each group (a)–(d) in the Honesty Experiment, for instance, was relatively high. What they do deny is that people possess dispositions which are robustly consistent between diverse situations; even more do they deny the unity-of-the-virtues thesis that one virtue entails all the others. In other words, whereas some stability remains (for ‘local’, ‘situation-specific’ dispositions), two other core elements of globalism about character traits, ‘robustness’ and ‘evaluative integrity’, are rejected.Footnote 11 The cobweb that most urgently needs to be blown away is said to be the ‘fundamental attribution error’, committed in equal measure by folk psychologists and academics:Footnote 12 the error of explaining and predicting people's behaviour via robust character traits.
In the psychological literature, a distinction tends to be made between (a) personality traits, such as being gloomy, giggly or talkative, that are purely descriptive and not necessarily reason-responsive or identity-conferring, and (b) character traits, such as the traditional virtues and vices, that are reason-responsive, penetrate deeper to the core of a person's self than do mere personality traits, and have to do with a person's moral worth. There is some confusion abroad as to whether or not personality traits and character traits are mutually exclusive categories, or whether the latter merely forms a sub-category of the former. What seems patent, however, is that although situationism is meant to hit primarily at the assumption of robust character traits, it does not leave mere personality traits untouched either. Staple textbooks in personality psychology teach us that personality traits are enduring states, forming broad or generalised patterns across a range of situations. Situationism obviously denies the existence of such traits; Harman, for instance, explicitly mentions ‘talkativeness’ as one of the supposedly imaginary global dispositions,Footnote 13 and in psychological circles, the experiments described above are more often invoked as part of an ongoing feud between personality psychologists and social psychologists on the existence of a firm and enduring individual personality, than they are as ammunition in a debate about special moral characteristics.
As noted, situationists direct their animadversions specifically against the apotheosis of Aristotle's virtue theory in contemporary moral philosophy and moral education. It is obviously not the naturalistic streak in Aristotelian ethics that they resent. Quite the contrary, ethical situationists gladly subscribe to Flanagan's principle of ‘minimal psychological realism’: about the need to construct moral theories in accordance with actual human capacities.Footnote 14 They are all for letting moral philosophy meet social psychology and sundry other empirical disciplines. What they object to are the specifics of the Aristotelian story, which they consider factually wrong. Aristotelian virtue theory and Aristotle-inspired character education are faulty not because they are anchored in psychology but because they are anchored in a flawed psychology. There is no clash of two cultures here – as when some psychologists have denigrated moral philosophy, lock, stock and barrel, for being the site of harmful ‘musterbation’. In Doris's own words, situationism ‘is not radically revisonary, generally problematizing ethical thought, but conservatively revisionary, undermining only particular – and dispensable – features of ethical thought associated with Aristotelian characterological psychology’.Footnote 15
Noticeably, in this regard, situationism does not involve a complete rupture with the idea of human dispositions. Although cross-situational dispositions, such as global compassion, go overboard, intra-situational dispositions, such as consistent ‘dime-finding, dropped-paper compassion’, remain.Footnote 16 Furthermore, situationists acknowledge the fact that many people appear to possess global dispositions by virtue of having been diligent enough in selecting and modifying the situations to which they could be exposed. For example, if I assume the role of a devoted husband and carefully avoid situations in which my devotion can be tested, then I may be able to deceive myself and others into considering it a global character trait.Footnote 17 If done self-deceptively, situationists do not recommend such self-inuring strategies, however. Indeed, much is made in the situationist literature of the way in which misguided attributions of global characterological traits to oneself or others produce deceit, disappointment, prejudice and mawkish hero-worship.
On the other hand, situationists strongly recommend deliberate (un-self-deceived, if you will) situation selection as a strategy in moral education. We should teach children to avoid situations in which they are likely to get into trouble, arrange social institutions such that outlets for temptations are limited, and manipulate our own social settings so that they become propitious to decent behaviour. I should not, for instance, accept the invitation of a flirtatious colleague when my spouse is out of town, in the false belief that I can control my impulses when it comes to the crunch.Footnote 18 Apart from this advice, situationists have little patience with what nowadays goes by the name of moral education. Take character education, which revolves around the notion of character building, and where character is understood to include a sense of personal integrity, enduring consistency and steadfastness of purpose.Footnote 19 The situationist response here is Harman's blunt and scathing remark that ‘there is no such thing as character building’.Footnote 20
III. Two Initial Objections
Let me first review two common objections to situationism that are not specifically Aristotelian: ‘the methodological objection’ and ‘the moral dilemma objection’.
According to the methodological objection, there is something individually wrong with the way in which each of the situationism-supporting experiments has been conducted and/or interpreted. Consider first the Honesty Experiment. The subjects in this experiment were children, and it is no news that children's characters are more malleable than those of adults. More saliently, it may well be that the to-be-researched dimension of ‘honesty versus dishonesty’ existed only in the heads of the experimenters, not in the heads of the children. Did the children necessarily consider lying, stealing and cheating as instances of a common underlying principle? The child psychologist William Damon has studied the method employed in this experiment in detail and his verdict is one of damning indictment. When examining children's behaviour, Damon points out, we must try to understand its significance within the context of the child's own world rather than cherrypicking our favourite adult conceptualisation. Children have their own social lives and social roles, and they may interpret interpersonal events differently than adults do. There is no good reason to think that the subjects in this experiment understood that copying obscure answers to bizarre tests or breaking silly rules in games was to be considered ‘cheating’, let alone ‘dishonesty’. Perhaps, if the children realised that they were being tested, they thought that the test was one of helpfulness to friends, loyalty and cooperation – in which case they would have being playing a straight bat and scored high. From the perspective of current child psychology, the experimenters committed pedestrian errors.Footnote 21
The Dime Experiment – to turn to that – seems to show that trivial matters, such as minor mood swings, may affect behaviour. But do they affect morally important behaviour, such as helping someone in dire need? The Dime Experiment demonstrates only that they affect morally insignificant behaviour which is, in any case, not part of one's daily grind: failing to pick up a stranger's dropped papers is hardly an important manifestation of moral failings.Footnote 22 Perhaps this experiment is most charitably passed over in silence.
The Milgram Experiment has generated considerably more nuanced methodological discussion. The unpreparedness of the subjects, the relentless pressure exerted on them by the experimenter, the fast pace of the experiment (which gave the subject's behaviour a knee-jerk quality), and the stepwise, slippery slope nature of the subjects' decisions have all been mentioned as possible mitigating factors. In their careful scrutiny of this experiment, Sabini and Silver conclude that the disturbance it may cause to our conception of character will at most be ‘local’, not ‘global’. It does reveal two specific weaknesses to which most people happen to be prone: the tendency to yield more or less unquestioningly to the commands of articulate, domineering ‘institutional experts’, and to act like Romans when in Rome in order to avoid embarrassment: to follow uncritically what other apparently reasonable people around them seem to be doing.Footnote 23 The uncomfortable facts that few people have a spoon long enough to sup with the devil, and that many can bear adversity but few contempt, do not undermine any folk psychology about character, as they are already part and parcel of such psychology.
These methodological qualms notwithstanding, it may be too optimistic to think that all psychological experiments which seem conducive to situationism can be shown to fall prey to some methodological errors. I have not seen any such objection urged against the Good Samaritans Experiment, for example, apart from the fact that it has not been repeated. A second objection, however, cuts deeper to the core of what is at stake here morally. According to the moral dilemma objection, the psychological experiments in question do not place subjects in our typical day-to-day choice situations in which the imperatives of a virtue compete with those of a vice or of a neutral state. Rather, they place subjects in situations in which they face the pressure of competing virtue imperatives. Such dilemma situations, requiring one to walk a tightrope between two virtues and not but contravene one or both of them, constitute that perilous terrain where virtue ethics encounters its severest trials. Sometimes the diminution of one virtue can coherently and with impunity be set off against the proper manifestation of another, as in the case where a temperate person decides to suspend temperance momentarily while accepting a huge slice of creamy cake from an elderly grandmother – for saying no would be cruel to the old lady. Other cases remain, however, in which choice is inherently tragic. Situationists focus on experiments relating to isolated and out-of-the-ordinary dilemma situations and conclude from them that people do not possess robust virtues. What they should be doing instead is gauging behavioural consistencies over extended periods involving everyday situations.
In the Honesty Experiment, for example, it may have been the children's pride, loyalty and helpfulness rather than their dishonesty that they pitted against honesty and which, in the end, eclipsed it.Footnote 24 In the Good Samaritans Experiment, attending to the virtue of appropriate punctuality is at least a mitigating concern.Footnote 25 And in the Milgram Experiment, one must bear in mind that cooperativeness in group enterprises and certain deference to appropriate authority are virtuous up to a point.Footnote 26 That the subjects took those virtues too far is, in retrospect, not to be doubted, but let us not forget that they were the hapless victims of an artificially created situation which was always likely to overstrain human nature or at least bring it close to tipping point. Moreover, taking a virtue too far is one thing; being positively vicious (here: cruel to peers) is another.
On a particularist reading of virtue ethics that is popular in the present day, moral theory does not provide us with an algorithm to adjudicate the imperatives of conflicting virtues in dilemma situations. There is no yardstick – no single currency – to sum up and codify the variously dimensioned vectors of those imperatives; instead we must rely on some sort of intuitive artistry in such cases. And it is possible that equally virtuous persons will make radically different choices, all equally good. Some virtue ethicists consider it to be a strength, rather than a weakness, of modern virtue ethics and ‘entirely to its credit’ that it does not furnish us with any general decision procedure to apply theory to particular cases.Footnote 27 It would be salutary to explore the relationship between situationism and particularism. Although the former is essentially a psychological thesis and the latter a moral one, one could well imagine an argument to the effect that, because of the inherent particularity of morality, the fact that people's moral decisions always turn out to be situation-specific may not be such a bad thing after all. This is not a line of thought that I will not pursue further here, however.
Some moral particularists have been eager to enlist Aristotle as their ally.Footnote 28 The trouble for them is that in Aristotle's view, phronesis adjudicates moral conflicts, and phronesis relies not only upon situational appreciation but also upon general moral truths. Thus, when Aristotle uses particular examples, he does not abandon generalisations and tell us to attend only to the particularities of the described situation; rather he describes the generalisations we should seek. Moreover, there is no inkling in Aristotle of the thesis that all the virtues are of equal or incommensurable standing; rather, the greatest virtues are necessarily those most useful to others – justice, courage, generosity and great-mindedness, for instance. Aristotle is well aware of the intractability of some moral dilemmas. However, he goes out of his way to ‘try to offer help’ in solving such dilemmas, while admitting that ‘is not easy to define [such] matters exactly’.Footnote 29 It may, therefore, be urged that insofar as the moral dilemma objection stems from a brand of modern virtue ethics that renounces moral generalism and embraces instead claims of moral particularism, it does not really constitute an Aristotelian response to situationism.
IV. The Two Aristotelian Objections
How should Aristotelians react to situationism? I think they should do so by dint of two objections: ‘the bullet-biting objection’ and ‘the anti-behaviouristic objection’.
The former objection is correctly anticipated by Doris as follows:
The fact that many people failed morally in the observed situations tells us little about the adequacy of Aristotelian descriptive psychology, since such disappointing demographics are exactly what the virtue theorist would expect. Indeed, a virtue-based approach can explain the situationist data: it is precisely because so few people are truly virtuous that we see the results that we do.Footnote 30
In the Nicomachean Ethics, Aristotle proposes a complicated, if somewhat unsystematically explicated, stage theory of moral development, ranging from the level of ‘the many’ (including children and other moral learners) through the levels of ‘the soft’, ‘the resistant’, ‘the incontinent’ and ‘the continent’, to that of the ‘fully virtuous’. Aristotle forthrightly acknowledges in a couple of places that ‘most people’ are placed between the levels of the incontinent and the continent.Footnote 31
Aristotle must be referring there to adult citizens; it would be out of line with his description of the level of ‘the many’ to hold that this level is not the one where most people (if you include children, labourers, etc.) are placed. In any case, most people cannot be counted upon to respond virtuously in morally tricky situations, for only a small minority has reached the level of full virtue. Far from being a reductio of Aristotelian characterological explanations, the results of the psychological experiment turn out to be exactly what coherent Aristotelians would expect. And far from pointing to the poverty of character building, the results underscore the need for sustained and intense education of that sort.Footnote 32 This objection is felicitously referred to ‘bullet-biting’. Not only does it embrace with ease the allegedly embarrassing facts thrown at it, it positively relishes the data from the experiments – which tend to show that 20–30% of people actually possess robustly virtuous traits – as (happily) indicating a bigger minority than could have been expected.Footnote 33
Doris latches onto the point made by Aristotle that virtuous persons will never behave basely.Footnote 34 However, Howard Curzer's studied reading of the relevant portions of Aristotle's corpus clearly brings to light that this point is an idealisation which Aristotle modifies in various ways. If he did not, we would be unable to explain the various passages which indicate that virtue comes in degrees; that full virtue is still inferior to god-like heroic virtue; and that virtuous people sometimes act wrongly, while remaining virtuous. Indeed, Curzer identifies at least seven distinct ways in which fully virtuous persons can, by Aristotle's lights, act out of character without being displaced from their superior level.Footnote 35 This observation further substantiates the bullet-biting objection: not only are most people insufficiently virtuous, even the fully virtuous can have tiny glitches in their characters.
I return to the bullet-biting objection in the final section, when I consider Doris's rejoinders, but it is now time to explore the second, and more profound Aristotelian objection: the anti-behaviouristic one. For a start, Harman describes an Aristotelian character trait as a ‘relatively long-term stable disposition to act in distinctive ways’: an honest person being a person disposed to act honestly, and so forth.Footnote 36 Many commentators have pointed out that this is a crude behaviouristic understanding, which has little to do with the modern virtue ethical conception of character, let alone the Aristotelian one. So even if social psychologists succeed in convincing us that people do not possess character traits qua robust behavioural dispositions, it does not mean that people do not possess character in the more nuanced Aristotelian sense, which is holistic and inclusive of judgement, emotion and manner, as well as action.Footnote 37 Some critics have fleshed out this objection by noting that, owing to their behaviouristic bias, situationists would be prone to confusing the virtuous with the continent and the vicious with the incontinent, although the actions or inactions of those persons would issue from radically different motivations.Footnote 38 I believe that the anti-behaviouristic objectors are on the right track. I also believe, however, that this objection has been seriously underdeveloped. After explaining the difference between the reactions of the virtuous, continent, incontinent and vicious in a given case, for instance, Goldie comments that this ‘just about completes the list’.Footnote 39 To do justice to the subtleties of Aristotle's character distinctions – and thus to appreciate the full power of the possible Aristotelian response – a more fine-grained analysis is needed.
Let us begin by considering the following scenario: A number of persons, who are on their way home from work and waiting at a bus stop, are approached by a scantily clad girl who asks them, with tears in her eyes, to give her 80p so that she can take the bus home. This girl does not look like a typical street person or druggie: There is certain awkwardness in her demeanour, indicating that she is not street-smart, and she is young – hardly a teenager yet. The way she looks indicates that she may just undergone some terrible experience. Let us now look at a number of variations in what could follow:
P1 does not care a whit about the girl's predicament. He has a rule about never giving to beggars or to charity. He considers it a waste of money. He shakes his head at the girl.
P2 likes to be seen and hailed by the greatest number as a generous person. He waits until more passengers come to the bus stop, then he ostentatiously hands the girl the 80p.
P3 feels incredibly sad to see the state this girl is in. Although he occupies a low-paid post and has barely enough money to make a living, he passionately thrusts a 50-pound note into the girl's palm.
P4 might have given the girl 80p if he had been in a good mood. However, after a bad day at work, he feels blue and says no.
P5 feels compassion towards the girl. However, just before handing her the money, he realises that he might possibly be seen to be accosting her. He immediately hesitates and decides to say no.
P6 often behaves generously towards strangers in need. However, he recalls his wife's stern complaints last night about his spending money too freely. After contemplating for a while, he decides not to give the girl 80p.
P7 would normally give money under such circumstances. However, he had been planning all day to try his luck at a slot machine on the way home. He only has a pound in cash on him, so he decides that giving 80p away is not a good idea this time.
P8 is having family problems and is in a bad mood. He does not feel a hint of compassion towards the girl. However, being a person of principles – one of which is behaving generously – he hands her the coins.
P9 feels compassion towards the girl. Without the need for any deliberation, he searches his pockets for coins and hands them to the girl with a warm glow of pleasure.
P10 is himself extremely poor. He does not know how to feed his family until the end of the month, and today he only has a pound left to buy bread for them. Yet he unhesitatingly hands the girl his one pound.
If this had been a psychological experiment of people's generosity, we can envisage the outcome: P2, P3, P8, P9 and P10 would be deemed generous and the rest ungenerous, based on their displayed actions or inactions. Their behaviour would possibly be compared to their behaviour in other, differently designed, experiments – and lo and behold: correlations would likely be low, and the notion of character infirmed once again. The Aristotelian response would be to refuse the very gambit offered by such behaviouristic measurements as an exercise in intellectual bullying. In contrast, consider the following analysis as an alternative:
P1 is consistently deficient in giving and thus ungenerous.Footnote 40 P2 is a stable giver, but he does not give for the right reasons. He is the kind of person who typically decides to do without generosity in order to practise charity. In plain terms, he possesses the character of ‘vanity’.Footnote 41 P3 is excessive in giving and thus ‘wasteful’ rather than generous.Footnote 42 P4 is at the level of ‘the many’. They ‘live by their [non-reason-informed] feelings’ and ‘have not even a notion of what is fine’ and hence ‘truly pleasant’.Footnote 43 P5 belongs to ‘the soft’. They have some grasp of the virtuous thing to do in given circumstances, but they fail to heed it if doing so is accompanied by any hint of pain.Footnote 44 P6 is ‘resistant’. The resistant possess only a limited degree of control against painful appetites, even when these go against morality.Footnote 45 P7 is ‘incontinent’. The incontinent have managed to overcome the thrust of the painful appetites that prevent many people from aiming at the good. They are easily overcome by counter-moral pleasant appetites, however: fail in many cases to abide by reason, ‘because of too much [enjoyment]’.Footnote 46 P8 is continent. The continent have managed to overcome permanently both painful and pleasant counter-moral appetites and are able to do the right thing. They are fully self-controlled and listen diligently to reason. Yet they are far from being virtuous; self-control is not the ideal state, because continent persons still have base appetites, and merely force themselves to act as they should.Footnote 47 P9 is truly generous, possessing the virtue in full measure. Full virtue is achieved only when one's appetites and emotions have both become reasonable and morally fitting – when they ‘share in reason’, in the strong sense of ‘agreeing with reason’. The virtuous persons' perceptions of moral salience silence considerations that remain active for the continent person.Footnote 48 P10 is also generous. It is ‘definitely proper to the generous person to exceed so much in giving that he leaves less for himself, since it is proper to a generous person not to look out for himself’.Footnote 49 Aristotle's point is that, for at least a number of people, some virtues require – as a matter of psychological fact – their own intermittent excess for them to continue to exist as virtues. Generosity is one of them: to be generous to a fault may require, so to speak, being at times faultily (namely excessively) generous.
On this Aristotelian analysis, five out of those ten persons possess ‘firm and unchanging’ states of character:Footnote 50 P1 is ungenerous (qua deficit); P2 is ungenerous (qua vanity); P3 is ungenerous (qua excess); P9 is generous full stop; and P10 is generous also because he hits the relevant medial target often enough, although he has a small – but psychologically excusable – glitch in his virtuous character. The rest of the persons do not possess ‘firm and unchanging’ character states. Their souls have not become stably responsive to reason, either right reason (leading to virtue) or wrong reason (leading to vice). Their personae are still too fickle and erratic – too easily swayed by non-reason-infused feelings – to constitute character. However, if all is well, they are progressing towards character.
It is particularly salutary to consider why, in this Aristotelian picture, three out of the five persons who actually give the girl the money she requests would not count as generous. Generous persons are, according to Aristotle, good users of riches; they give the proper amounts to the right people, at the right time and for the right reason. In general, they aim at what is fine in giving, and they take pleasure in it, just as they take pleasure in other virtuous activities. However, they do not give in order to take pleasure in the giving or in being seen by others as good givers, like the vainglorious P2 did; their pleasure simply supervenes upon and completes the virtuous activity. Nor do they carelessly throw away their own possessions or overburden themselves and their families, like the do-gooder P3 did; for this would make them less able to continue giving in the future. And in the case of P8, he lacks the right frame of mind to be considered truly generous. For P2, think of the slimy lawyer, Clamence, in Camus's story The Fall; for P3, think of the over-zealous David in Nick Hornby's How to Be Good; for P8, think of Kant's ‘person of moral worth’, on what used to be the canonical interpretation, as a person whose goodness would be compromised by a co-operating inclination – and you get the picture.Footnote 51
From an Aristotelian perspective, the trouble with a behaviouristic interpretation of the above story variations is not only that it does not capture who are the generous and the ungenerous persons involved; more significantly, it tells us next to nothing about which persons possess character and which persons do not.
V. Rejoinders and Implications
Doris anticipates both the bullet-biting objection and the anti-behaviouristic objection and tries to meet them. I think his rejoinders misfire.
Consider first Doris's response to the bullet-biting objection, which he refers to as ‘the argument from rarity’. (1) He doubts that ‘reflection on a few extraordinary individuals’ facilities ethically desirable behaviour. (2) He notes that character training is, on virtue-based accounts, typically about inculcation rather than ‘reflection on a rarefied ideal’. (3) He complains that if the virtues touted in virtue-based moral theories cannot be appealed to in the explanation and prediction of behaviour, those theories become too ‘empirically modest’ to retain their current appeal.Footnote 52
Running quickly through possible Aristotelian responses, they should, I submit, take the following form. (1) Although Aristotle would admit that full virtue is comparatively rare,Footnote 53 it is not the privileged province of ‘a few extraordinary individuals’. If up to 20–30% of people possess robust character traits – witness some of the staple psychological experiments – that is already a considerable subset of the population. Doris may be thinking here of heroic virtue rather than ordinary full virtue. (2) The claim that virtue educators are not concerned about reflection on ideals does not stick in Aristotle's case. He characterises a special emotional virtue, called ‘emulation’: distress at the apparent presence among others of things honoured and possible for a person to acquire, with the distress arising not from the fact that another has them, but that the emulator does not.Footnote 54 Aristotle insists that this is one of two virtues specific to young moral learners, the other being shame. Role-modelling on ideals is thus an essential Aristotelian strategy of moral education, along with habituation, just as it has become in character-education accounts of late. (3) If Aristotle's virtue theory fails to satisfy the demands of the behavioural sciences for predictive reliability, this may say more about the limits of predictivism as a model of social inquiry than the limits of his virtue theory. Incidentally, I do not think that the current appeal of such theory lies essentially in its predictive value. But, in any case, as I suggest at the end of this section, Aristotelianism may well be compatible with scientific methods for investigating character.
Doris refers to the anti-behaviouristic objection as an ‘intellectualist account’, according to which the virtuous person is typified by a ‘distinctive outlook’ – some goings on ‘within the head’ – rather than reliable overt behaviours. Doris cavils at such an account for two distinct reasons. First, he finds it morally strange (and not ‘the most inspiring epitaph’) to say of someone that ‘his ethical perceptions were unfailingly admirable, although he behaved only avaragely’. Second, he points out that the alleged outlook of virtuous persons may also turn out to exhibit situational variability, just as other capacities and dispositions do, and thus be tarred with the same brush as overt behaviours.Footnote 55
In order to respond to Doris, we need first to consider what is to be understood by the ‘distinctive outlook’ of a virtuous person. In Aristotle's account, it is empirically true that the wellbeing of human beings consists of the realisation of intellectual and moral virtues and in the fulfilment of other specifically human physical and mental capabilities. Each moral virtue constitutes a specific medial character state, flanked by the extremes of deficiency and excess. There is only one way, the medial way, to be ‘correct’: to be inclined to act or feel in the right way, towards the right people, at the right time. But there are a number of ways in which to be ‘bad’.Footnote 56 To talk about right actions and feelings is crucial here, for a distinctive feature of Aristotle's virtue theory is the assumption that emotional reactions may also constitute virtues. Emotions can, no less than actions, have an intermediate and best condition proper to virtue. The precise relationship between virtues of action and emotional virtues is not always entirely clear in Aristotle's texts. He often seems to suggest that there is a general emotional trait that corresponds to each moral virtue, yet his civil virtues of friendliness, truthfulness and wit in social intercourse seem to exist without a distinct emotional corollary. The typical Aristotelian virtue will nevertheless be a complex character state (hexis), at the core of which lies moral sensitivity, exhibited through emotional reactions, to goings-on in the world around the agent. Whether a felt emotion should be acted upon or not is always a separate question, however. The answer to that question must take various situational factors into account and be adjudicated through the intellectual virtue of phronesis. The description of a given virtue is not fully exhausted by characterising its underlying emotional sensitivity and the range of possible actions to which it can give rise. Virtuous persons also comport themselves in certain distinctive ways which reverberate through all their attitudes and conduct; what matters is not only what they feel and do but also the manner in which they feel and do it. Great-minded or magnanimous persons, for instance, exude an aura of proper dignity and moral superiority which distinguishes them from other virtuous (but non-magnanimous) persons.Footnote 57 This is why each hexis is truly a complex state, rather than a mere disposition to feel and act. In our little story earlier, the actors could, from a crude behaviouristic perspective, only give or not give the girl the money she requested. From an Aristotelian perspective, they could feel in a number of different ways about the giving or not giving and – although this was not clearly revealed in the thinly described variations – give or not give in a radically different manner.
It was the emotional factor which turned out to be the distinguishing one in the Aristotelian analysis of the ten persons. Is that an ‘intellectualist’ idiosyncrasy? I very much doubt that it is. Consider once again the Milgram Experiment. The participants were subjected to overtaxing pressures to which most of them succumbed. They did not behave well. Yet, at the same time, they displayed ‘striking reactions of emotional strain’ and afterwards often reported significant levels of regret and post-traumatic stress. When relieved of the pressure to toe the line, only 20% continued to administer maximum-level shocks.Footnote 58 Moreover, although no correlation was found between the subjects' decisions to quit or not quit pushing the button and their Kohlbergian stages of moral reasoning, a clear demarcation was evidenced when subjects were asked at the end of the experiments about the desire they had felt to quit. The stages of moral reasoning were clearly related to behavioural independence in feeling and judgement, although not, under those taxing circumstances, to independence in action.Footnote 59 It is no wonder, then, that the one factor in the famous Big Five Personality Trait ModelFootnote 60 that most clearly zooms in on moral characteristics (namely, agreeableness, as it is somewhat infelicitously called there) is measured by asking respondents to assess themselves in relation to their emotions – if they have soft hearts, sympathise with others' feelings, feel concern for others, and so forth – rather than to their actions. The personality psychologists who designed this model seem to have realised that the clearest distinguishing factor between personality types is emotion. To return to Aristotle once again, recall that the kind of moral education he describes in greatest detail is emotion education. For him, the process of affective sensitisation plays a decisive role in the gradual consolidation of moral character, thus forming an indispensable starting point of any formal or informal programme of character education.Footnote 61 To be sure, it would be empirically conceivable that emotions turn out to exhibit the same situational variations as actions – as was Doris's sceptical suggestion. The empirical evidence so far, however, seems to check that scepticism; emotional inclinations have more permanence and robustness than our actions do.Footnote 62
But that brings us to Doris's other complaint about the anti-behaviouristic objection. If emotional factors are really the touchstones for measuring moral character, could we not end up with the bizarre implication of someone's ethical perception being deemed unfailingly admirable, although he behaves only sub-optimally? Aristotle has an answer to this. He says that wellbeing is an ‘activity’ rather than a ‘state’; for if it were not, someone could enjoy it and yet ‘be asleep for his whole life, living the life of a plant’.Footnote 63 Furthermore, ‘Olympic prizes are not for the finest and strongest, but for contestants, since it is only these who win; so also in life [only] the fine and good people who act correctly win the prize’.Footnote 64 Contrary to much of contemporary virtue ethics, there is no presumption of the ‘primacy of character’ here; rather, the tree is known by its fruit. Being endowed with a good character is, for Aristotle, clearly not praiseworthy as such; what matters is how that good character is manifested through particular performances. One could even read him as saying that attributing ‘good character’ to a person who fails to exemplify it in his deeds is a logical mistake.
Are these claims incompatible with an emphasis on moral character as emotional sensitivity and moral perception? Indeed not. The point being made in the analysis of my ten scenarios was that only an understanding of people's motivational structure can truly tell us whether they behave virtuously or viciously; and, moreover, whether or not they possess character in the first place. The emotional reactions leading up to the decision to act (or not to act) and the reactive attitudes the person experiences after the decision has been made are indispensable data for us to evaluate the moral propriety of the act and of the agent.Footnote 65 We can be sure that the emotional make-up of the agent will enter the texture of his general thinking and action, for it is essentially action-guiding. But each action or inaction of our ten persons in Section IV had to be understood as that action or inaction guided by that emotional make-up; otherwise we could not know if it exhibited the virtue of generosity.
If the anti-behaviouristic objection succeeds in revealing the poverty of the standard psychological experiments on character, what implications does this have for the ideal measurement of character? Kamtekar suggests that social psychologists should engage in more painstaking research into the considerations that experimental subjects have in mind when making moral decisions: the inner mediating events.Footnote 66 Now, despite their methodological hazards (such as possible self-deceptions and other-deceptions), self-reports, such as those elicited by Big Five researchers, are the instrument closest to hand in measuring inner mediating events. We should not assume, however, that subjects have privileged access to their inner lives; peer reports may also tell us a great deal about a person's character. Additionally, Webber mentions measurements of emotion-revealing hormones through saliva tests: people may be able to hide their feelings of, say, insult – but such tests will betray them.Footnote 67 There is something quintessentially Aristotelian about this suggestion, given Aristotle's keen awareness of the physical, as well as the cognitive, components of emotion. Most important, a person's character will reveal itself gradually over an extended period under varied circumstances. Longitudinal studies will thus be preferable to single-case studies.
What now has been said suffices, I submit, to parry Doris's rejoinders to the two Aristotelian objections. More generally, I hope to have fleshed out a more substantive Aristotelian response to situationism than has so far been advanced in repeated attempts to deflect the situationist challenge. Aristotelian characterology is not at death's door. There is quite a lot of life in the old dog yet.
