The entire eudaimonic tradition in philosophy — which includes Plato, Aristotle, the Stoics, and others in the west, as well as Confucianism in the east — is based on the notion of character, which my desktop dictionary defines as: “the mental and moral qualities distinctive to an individual.”
But what if there is no such thing as character? What if people don’t behave in certain ways because they are, say, courageous or cowardly, temperate or intemperate, but rather simply because they are responding in idiosyncratic ways to local situations?
Welcome to the situationist debate in moral philosophy. Which is a strange debate for a number of reasons, perhaps the main one being that psychologists themselves have repudiated their own situationist literature since the 1980s. Some philosophers, it seems, are almost half a century behind, and haven’t caught up yet.
There are several aspects to the debate about character in moral philosophy, covered in detail in a highly recommended essay by Christian Miller, the author of The Character Gap: How Good Are We? (four-part commentary here). In this post I will focus on what I regard as the main one: situationism.
As I said, the debate actually started in psychology, back in the 1960s and ‘70s, with participants describing it as “traumatic” and “intense,” involving “warfare” and “heated but futile battles.” In other words, yet another transitory phase in the history of psychological science.
Much later on, we have the onset of situationism in philosophy, mostly due to the work of Gilbert Harman and John Doris. These authors proposed the following two-pronged strategy against the existence of character traits:
(I) They argue that psychological studies demonstrate that people typically do not have global character traits.
(II) They conclude that this is a serious problem for Aristotelian and other kinds of virtue ethics.
To fix our ideas, Miller provides the example of temperance. If you are a temperate person, presumably that doesn’t just mean that you eat moderately in restaurants. It also means that you moderate yourself when it comes to drinking, sex, and in fact pretty much everything else. This is why the Stoic philosopher Musonius Rufus says that sitting at the table is an excellent way to exercise one’s temperance. Evidently, the assumption is that the trait then translates into situations that have nothing to do with meals. But is that really the case?
Concerning point (I), Harman and Doris argue that the empirical evidence simply does not support the existence of global character traits, meaning traits that are both consistent (across a diversity of eliciting conditions) and stable (across repeated instances of similar eliciting conditions). In other words: maybe I’ll temperate at tonight’s dinner. But that doesn’t mean I’ll also be temperate at dinner next week, or that I’ll be temperate at any other activity outside of dining.
One serious problem with Harman and Doris’ claims, however, is that they limited their survey of the empirical studies to a single character trait: compassion, for which they used the psychological literature on pro-social behavior. They simply assumed that what goes for compassion goes for any other global character trait. That is one whopping assumption, though.
Moreover, there are significant problems even if we stick to compassion. Three classes of experiments studied pro-social behavior in the 1960s and ‘70s: variations on the “dime in the booth” situation, the “lady in distress” experiments, and the famous experiments on obedience to authority conducted by Stanley Milgram.
In the “dime in the booth” scenario, some of the subjects “accidentally” found a dime when they went to use a telephone booth (remember those?), while other subjects did not. Participants were then given a chance to pick up some dropped papers. 88% of the dime people did, compared with only 4% of the control group.
Subjects in the “lady in distress” protocol heard a crash in another room, followed by a woman screaming, apparently in pain. Would they help? If they were by themselves they helped about 70% of the time. But if someone else were around, that proportion dropped to a mere 7%.
The Milgram experiments consisted in an alleged doctor asking subjects to progressively increase the voltage of a punishing electric discharge administered to people who were giving wrong answers to certain questions (the charge was not real, and the people in apparent pain were confederates of the experimenters, obviously). Under pressure from what they thought was an authority figure, 80% of participants went as high as 270 volts, while 65% went all the way to the 450 volts, a level that — had it actually been administered — would have been lethal.
The first thing to notice about all three experiments is that it is highly questionable whether they are actually measuring compassion or pro-sociality. Moreover, several of these experiments — particularly the Milgram one — have now been called into question, either because they turned out not to be replicable, or because the original design was flawed. For instance, psychologists John Sabini and Maury Silver have argued that: “People who must act in such circumstances are confused and inhibited by the anticipation of embarrassment, and that we argue is the lesson to be drawn from social psychological research” (quoted in Miller).
Such criticisms, undermining as they do part (I) of the situationist claim, would be enough to end the discussion here. But in fact there are noteworthy rebuttals against point (II) as well.
Harman claimed that “this sort of [Aristotelian] virtue ethics presupposes that there are character traits of the relevant sort, that people differ in what character traits they have, and these traits help to explain differences in the way people behave.” And Doris wrote: “Aristotelian virtue ethics, when construed as invoking a generally applicable descriptive psychology … [is] subject to damaging empirical criticism.” (Both quoted in Miller.)
But in fact virtue ethics does not claim that most people are virtuous, on the contrary. For the Stoics, the only truly virtuous people are the (very rare) sages. The rest of us just muddle through, ideally trying to do our best. Aristotle himself clearly disagrees with Harman and Doris’ understanding of virtue ethics:
“The many naturally obey fear, not shame; they avoid what is base because of the penalties, not because it is disgraceful. For since they live by their feelings, they pursue their proper pleasures and the sources of them, and avoid the opposed pains, and have not even a notion of what is fine and truly pleasant, since they have had no taste of it.” (Nicomachean Ethics, 1179b11–16).
Other philosophers coming to the defense of virtue ethics, like Rachana Kamtekar, have pointed out that the sort of character traits often studied in psychology have little to do with the moral traits that are of interest to virtue ethicists. For Aristotle, virtues are behavioral dispositions to act appropriately to a given situation, reflecting one’s best judgment. But if someone is not trained in reflecting and improving on their judgments, then of course they will not display “global character traits.”
Finally, I’d like to note that this is a stark example of the difference between psychology and philosophy. Psychology is a science, and so it is in the business of describing things (like people’s behaviors), not of making prescriptive judgments (about how people ought to behave). While philosophy and psychology clearly have much to benefit from each other, Aristotle, Confucius, the Stoics and others are interested in training us to behave in a certain way. The fact that we often don’t is not a problem with their prescriptions, it is a problem with our behavior. It would be like observing that most people don’t have proper aerobic capacity and muscle tone and conclude that your trainer is therefore mistaken about what it would take for you to get fit.