Author: Ian Tully
Word Count: 959
The past half century has seen a considerable revival of interest in virtue. Moral philosophers are no longer merely interested in assessing our conduct (e.g., ‘don’t lie!’) or the outcomes we bring about (e.g., more happiness is better!); many now agree that ethics needs to be concerned with our characters, too. What matters isn’t just what we do, but who we are. Yet in recent years this focus on character has been challenged by empirically-minded philosophers who argue that recent work in experimental psychology reveals that most people don’t possess the kinds of traits which would constitute virtues. This empirical challenge to virtue is the focus of this essay.
According to a large and well-confirmed body of research going back to the 1920s—which now goes under the label of ‘situationist personality psychology’ or ‘situationism’—“seemingly insubstantial situational factors have substantial effects on what people do” (Doris 2002: 28). To illustrate what is meant by ‘situational factors’ consider a classic situationist study, Isen and Levin’s (1972) ‘dime in the phone booth’ experiment. In this experiment, the difference between finding or not finding a dime in a phone booth made a substantial difference in whether individuals chose to help a woman pick up some papers she had just dropped: those who had found the dime were 84% more likely to help the woman out. In other words, a very minor difference in the situation (the presence or absence of a dime) played a significant role in determining behavior. Such findings are at odds with an intuitive picture of human psychology and behavior which focuses exclusively or primarily on an agent’s character in order to predict and explain his or her behavior. On this view, people behave in the ways that they do largely because of the kinds of people they are: because they are kind or generous or greedy. Situationism challenges this account by revealing the surprising extent to which our behavior is a function of external, situational factors.
Examples are legion. Being in a hurry significantly decreases the likelihood that passersby will stop to help an apparently distressed individual (Darley and Batson 1973). Ambient noises (Matthews and Cannon 1975), ambient smells (Baron 1997), and the presence or absence of other people (Latane and Darley 1970) also modulate helping behavior to a considerable extent. More troublingly, Stanley Milgram (1974) found that the polite but firm insistence of an experimenter could induce many participants to administer potentially lethal shocks to an experimental confederate in another room. It seems, in short, that whether or not people engage in morally praiseworthy or morally deplorable behavior is to a large extent determined by often very insubstantial features of the situations they find themselves in.
A number of philosophers (Doris 1998, 2002; Harman 1999) have concluded that this evidence spells trouble for traditional accounts of moral character. According to Aristotle, the virtuous agent acts from “firm and unchangeable character”; the agent sees what the situation calls for (i.e., honesty, compassion, courage) and performs that action (Aristotle 1984: 1105a27-b1). In Doris’s terms, the virtuous agent’s character is organized by “robust traits,” that is, traits which issue in behavior that exhibits “consistency across situations,” (Doris 2002: 23). The thought is that in a variety of different situations – i.e., regardless of whether or not he is in a hurry, or in the presence of a good-smelling bakery, or whatever – the, e.g., compassionate agent will, when faced with a circumstance which clearly calls for compassion, act compassionately. Yet the experimental evidence seems inconsistent with the average person possessing such global traits, given the situational variability it reveals. Of course, one presumably doesn’t always need to behave compassionately in order to merit the term compassion, but certain circumstances seem diagnostic or criterial: if I leave you drowning in a shallow pond, then (all else equal) I’m not compassionate—and likewise if I give you a lethal shock.
In sum, Doris, Harman, and others argue that the experimental evidence is inconsistent with most people possessing the kinds of robust traits implicated in traditional theories of moral character. But if that’s right, then most people lack virtues and vices. At most, what they have are merely “local traits,” for instance, “office party sociability” (Doris 66). Of course, this evidence is still consistent with a few people possessing global (i.e., non-local) courage, or compassion, or honesty. (Or, conversely, with a few people possessing global vices.) But it does seem to reveal that those global traits are very rare. For most of us, our psychologies are just not adequately described as constellations of such cross-situationally consistent character traits.
Of course, such pessimistic conclusions have been strongly contested, and the debate over the implications of situationist findings for virtue is ongoing.1 Defenders of virtue ethics have developed a number of replies, though space considerations prevent thorough discussion of them here. One prominent response notes that virtue – full virtue – is, and was always expected to be, rare. Thus, situationist findings merely confirm what we already knew. Yet such a response makes many uneasy, for it reinforces worries that virtue ethics is problematically elitist, advocating as a normative ideal something attainable by only a few (cf. Driver 2001: 54). Others note that virtue is not merely a matter of behavior – it also concerns what one thinks and feels. Thus, situationist findings are in an important sense incomplete.2 Critics also worry about the probative value of “one-time performance” data in demonstrating that one does or does not have a given dispositional property (Sreenivasan 2008: 603). After all, we often fail to behave compassionately when we are stressed or tired, but it seems too strong to claim that those occasional failures mean we aren’t compassionate: what is needed, it seems, are iterated trials.3 Finally, some critics (Snow 2010) have argued that contemporary psychology is not nearly so hostile to robust traits: in fact, understood in the right way, people do in fact possess such traits. Details of her proposal are unfortunately beyond the scope of this essay.
1For a very thorough overview, see Miller (2013).
2But see Doris (2006, 2009) for evidence that such features are also situationally susceptible.
3See Doris (2002: 71-76) for an anticipation and response to these concerns.
Isen, A. M. and Levin, P. F. (1972) “Effects of Feeling Good on Helping: Cookies and Kindness,” Journal of Personality and Social Psychology 21: 384-388.
Ian is a doctoral student at Washington University in St. Louis. He also holds an M.A. in philosophy from the University of Colorado at Boulder and a B.A. in philosophy from George Washington University. He is interested in ethics, metaethics, moral psychology, and the philosophy of mind. Aside from philosophy, he likes reading fiction, going to rock concerts, and hiking in the Ozarks.