Introduction Lawrence Blum has a review of John Doris's book which is titled Lack of Character: Personality and Moral Behavior on Notre Dame Philosophical Reviews (The first chapter of Doris's book is available as a free download here). Doris’s book is an important contribution to moral psychology, drawing on work by social psychologists to undermine philosophical theories of character in general and the virtues in particular. In a nutshell, Doris’s aim is to show that experimental evidence establishes that human behavior is more determined by situation that by character and that there is very little empirical evidence for robust character traits, such as courage, temperance, or justice. Doris’s work is the most comprehensive philosophical version of this argument, but others, including Gilbert Harman and Owen Flanagan, have also explored these topics. For ease of reference, let’s loosely define situationalism as the view that situation rather than character is more causally potent in determining human behavior than is character. Similarly, let’s use the ugly neologism characterism to stand for the view that character traits do important explanatory work. What follows is an overview of some of the issues raised by the social psychology research on character and situation--but not a review of Doris's book.
Why Legal Theorists Should Care Why should legal theorists care about situationalism? Let’s simply list a variety of reasons for caring about the situationalism/characterism debate in the context of law:
--Rules of Evidence. Typically, the rules of evidence prevent the prosecution in a criminal trial from introducing evidence of bad character and particularly of prior criminal acts by the defendant. If situationalism is true, then these exclusionary rules might well be shown consistent with the accuracy goal of the evidence law.
--Theories of Punishment. Explaining the normative basis for punishment is a central topic for legal theory. Theories of criminal punishment interact in various ways with moral psychology. If situationalism is true, there may well be implications for our views about when and why punishment is appropriate.
--Morality and General Jurisprudence. More generally, our views about morality will affect our views about the law in a variety of ways—direct and indirect. One way to divide up the universe of moral theories suggests that they can be categorized as belonging to the following three groups:
++Consequentialist Theories, such as utilitarianism and welfarism, which hold that the rightness or wrongness of an action depends on the goodness or badness of the states of affairs to which the action leads.
++Deontological Theories, such as Kant’s theory or Scanlon’s contractarianism, which hold that the rightness or wrongness (permissibility or impermissibility) of an action depends on its conformity with a set of moral rules.
++Aretaic Theories, such as Aristotle’s theory or contemporary virtue ethics, which make the cultivation of human excellence, rather than the performance of right-actions or the production of good consequences, the central them for moral theory.
General jurisprudence is, to some extent, independent of disputes about moral theory, but legal theorists frequently see connections between normative questions about the content of the law and deeper issues about moral theory. Perhaps the most famous example in recent years in Louis Kaplow and Steven Shavell's famous (infamous?) book, Fairness versus Welfare--which argues that welfarism ( a variety of consequentialism) is the correct theory of political morality and that this has multitudinous implications for legal theory. Deontologists are likely to argue that the aim of law is to protect autonomy or liberty. Similarly aretaic theories of morality and politics are usually associated with the view that the aim of the law is to faciliate human flourishing in general and the development of virtue in particular.
How does this connect to the situationalist/characterist debate? Let me assert (with breathtaking oversimplification), that the correctness of situationalism (in a strong form) would be fatal to aretaic moral theories and damaging to deontological theories. I've only scratched the surface, but I hope I've said enough to convince you that legal theorists should care about Doris's book and the wider debate it represents.
Situationalism Situationalism is the view that situations rather than character do most of the work in explaining human behavior. Situationalism is rooted in social psychology, and there is a long and complex story to tell about its historical roots. Once again, I must be your indulgence for simplifying, but that story goes something like this. Psychologists had once posited a simple explanation for the fact that in a given situation, some people will lie and others will not, some will steal and others will not, some will cheat and others will not. That simply explanation was based on the idea of dispositional personality trait. Some people were liars, thiefs, and cheats. Others were honest. Studies were done, and they failed to provide strong confirmation for the simple, personality trait theory. One famous experiment took children and put them in a variety of situations that tested honesty--an opportunity to take some spare change left in a public place, an opportunity to cheat on a test, and so forth. Atlhough the experiments revealed intrasituational consistency--take the change on one occasion and then take it again on another occasion--they did not show strong intersituational consistency--test cheating was not strongly correlated with change taking.
Another line of research led to the identification of what is called the fundamental attribution error. This line of research investigates what we might call folk social psychology, i.e. what ordinary folks believed about the causes of human behavior. The research revealed that there is a strong tendency to attribute behavior to character traits rather than to situations.
Most legal theorists have at least a second-hand acquaintance with situationalism. One possible upshot of situationalism is that criminals are not "bad people;" rather, bad acts result when ordinary people are put in situations which elicit criminal behavior. Thus, situationalism provides support for consequentialist theories of punishment--crudely, left situationalists favor rehabilitation and right situationalists favor deterrence and incapacitation. But most situationalists would agree that neither retribution nor desert can provide the underlying justification for punishment. Of course, the broad sketch in this paragraph is vastly oversimplified.
Aretaic Moral Theory and Situationalism Aretaic moral theories are focused on human excellence; they are virtue-centered. They take the virtues as the primarly (but not the exlusive) locus of morality. We might use Aristotle's theory as an illustration, bearing in mind that philosophers debate the question whether Aristotle's theory is properly understood as aretaic in the sense that I have specified. Aristotle argued that eudaimonia (happiness) is the highest humanly achievable good, and that eudaimonia was constituted by a life lived in accord with the human excellence or virtues. Aristotle classified these in two groups. The moral virtues, such as courage, temperance, and good-temper (proates) were understood as connected with the emotions. Courage, for example, is related to fear. The courageous human is disposed to fear that is appropriate to the situation. Thus, courage can be defined as the mean between two vices (defects): cowardice (the disposition to too much fear for the situation) and rashness (the disposition to too little). The intellectual virtues were sophia (theoretical wisdom) and phronesis (practical wisdom). Obviously, in a single paragraph I haven't even come close to adequately summarizing Aristotle's view, but that will have to do for now.
How does situationalism pose a threat to aretaic moral theories? You, gentle reader, have undoubtedly run ahead of me by this point. Situationalism is really a family of views, with stronger and weaker forms. Let's define strong situationalism as the view that no humans have stable intersituational dispositional traits. If strong situationalism is true, then there is no such thing as a courageous human or a good-tempered human--the are no virtuous men or women. Likewise, it also follows from strong situationalism that there are no cowardly humans and no ill-tempered humans--there are no vicious women or men. If strong situationalism is true, then the "coward" is a social myth--not unlike the "witch." Of even greater interest to legal theory, there are no "bad guys" or "good guys" only bad situations and good situations. Virtue-centered moral theory would rest on a mistaken belief in virtues.
The Milgram Experiments I assume that almost everyone who reads Legal Theory Blog will be familiar with Milgram's famous experiments. Here is a brief description:
In response to a newspaper ad offering $4.50 for one hour's work, an individual turns up to take part in a Psychology experiment investigating memory and learning. He is introduced to a stern looking experimenter in a white coat and a rather pleasant and friendly co-subject. The experimenter explains that the experiment will look into the role of punishment in learning, and that one will be the "teacher" and one will be the "learner." Lots are drawn to determine roles, and it is decided that the individual who answered the ad will become the "teacher."
Your co-subject is taken to a room where he is strapped in a chair to prevent movement and an electrode is placed on his arm. Next, the "teacher" is taken to an adjoining room which contains a generator. The "teacher" is instructed to read a list of two word pairs and ask the "learner" to read them back. If the "learner" gets the answer correct, then they move on to the next word. If the answer is incorrect, the "teacher" is supposed to shock the "learner" starting at 15 volts.
The generator has 30 switches in 15 volt increments, each is labeled with a voltage ranging from 15 up to 450 volts. Each switch also has a rating, ranging from "slight shock" to "danger: severe shock". The final two switches are labeled "XXX". The "teacher" automatically is supposed to increase the shock each time the "learner" misses a word in the list. Although the "teacher" thought that he/she was administering shocks to the "learner", the "learner" is actually a student or an actor who is never actually harmed. (The drawing of lots was rigged, so that the actor would always end up as the "learner.")
At times, the worried "teachers" questioned the experimenter, asking who was responsible for any harmful effects resulting from shocking the learner at such a high level. Upon receiving the answer that the experimenter assumed full responsibility, teachers seemed to accept the response and continue shocking, even though some were obviously extremely uncomfortable in doing so.
Although Milgram's experiments were not designed to test the hypothesis that humans have character traits, they are nonetheless frequently cited in debates over this issue, because they seem to indicate that a certain kind of human excellence is very rare. (For Milgram's book, go here.) The fact that many or most of us would torture the innocent is take as providing some evidence for situationalism.
Assessing Situationalism
Does the social psychology research on situationalism pose a threat to the view that human beings have character traits? This is obviously a large question. My aim here is simply to point towards some relevant considerations. Let's begin with the way the ordinary people think about character.
Folk Psychology
Folk psychology is simply the beliefs about human psychology that are embedded in our culture. And the folk psychological view is that humans have character traits. Here is the way that S.J. Burrow put it in her review of Doris's book:
Character plays a large part in our assessment of others: Pat is decidedly argumentative, Tish is patient, Jerome is shy. Such assessment is useful as a heuristic for explaining and predicting the actions of others: Tish will successfully cope with delays in traffic while Jerome won’t easily start conversations with strangers.
Most (almost all?) readers of this blog ascribe character traits to their friends and colleagues. One would think, therefore, that the weight of the evidence is against strong situationalism. But this is where the fundamental attribution error comes into play. Social psychologists have convincing demonstrated that we humans have a consistent tendancy to overestimate the role of character in determining human behavior. So perhaps folk psychology and common sense are simply in error.
Problems with the Research
There are, however, a variety of problems with the research cited in support of situationalism. Here are a few:
--Failure to Define Situation. Lawrence Blum, in his review of Doris's book, notes that a very basic problem with much of the research--the failure to give any adequate definition of "situation." But if we do not know what counts as a situation, it is difficult to make sense of the hypothesis that situation rather than charcter explains behavior.
--Ambiguous Notion of Character Trait. In order to determine whether or not there are character traits, we would need to know what a charcter trait is. Aristotle, for example, has a theory of psychology of the virtues--which he theorizes are dispositions with respect to morally neutral emotions, such as fear. Much of the research that Doris investigates, however, simply plucks supposed character traits out of thin air. "Honesty," for example, is the character trait investigated by one prominent line of research, but Aristotle's theory of the virtues would not include "honesty" as a virtue--because it is not a mean with respect to an emotion.
--Children as Subjects. Some of the most interesting experiments that seem to show a lack of intersituational character traits were performed on children, but some theories of character development would not predict that children would have developed consistent characters.
--Information Impoverished Experimental Designs With respect to the fundamental attribution error, there is another, particularly interesting, problem with the research. Not unexpectedly, many of the FAE experiments are simple, one-shot, situations. You are asked to rate the skill of basketball players after seeing them play once, and with little information provided about the situational variables.
Theoretical Incompleteness
"Situationalism" is sometimes offered as a theoretical rival to "characterism," but as a theory of human behavior, "situationalism" is radically incomplete. It is obvious that human behavior is not wholly determined by situations. Indeed, the view that situations wholly determine human behavior is obviously false. Bricks, dogs, and wheelbarrels behave nothing like humans when placed in similar situations. Rather, sympathetically understood, situationalism is the view that human psychological makeup interacts with situations to produce behavior. Of course, stated in this way, situationalism does not compete with characterism. So the situationalist hypothesis might be that human psychological makeup is uniform, with no variation between individuals. But once again, this hypothesis is radically implausible, because it is manifestly the case that human behavior differs enormously, given relevantly similar situations. So the situationalist might hypothesize that these differences are essentially random, but, of course, this move is really an evasion of the issue, because it does not offer an explanation at all. As a theory, situationalism is incomplete--it really doesn't offer an explanation of human behavior that would compete with characterism.
Let me be clear. This very brief discussion is hardly adequate in any way. I certainly don't want you to be convinced on the basis of what I say here that situationalism is wrong. My goal is very modest. I only want to convey a sense of the issues.
The Fundamental Attribution Error Again
This brings me back to common sense and folk psychology. Does the "fundamental attribution error" really undermine our folk psychological belief in character traits? I can't do this question justice here, but I do have a strong hunch. I've read both primary and secondary accounds of much of the fundamental attribution error research. Most of the experiments involve situations in which the experimental subjects are given very little information about a situation, and then show a tendancy to attribute behavioral differences to character (or other traits of the subjects) rather than to situational variables. But our beliefs about human character were not formed in circumstances like thse. Our beliefs about character are formed in informationally rich environments in which we observe the same individuals and thier situations for hours, days, weeks, months, and years.
One experiment cited in favor of the fundamental attribution error illustrates the danger of overgeneralizing from this research. In this experiment, basketball players performed in two different situations--low light and normal light. Experimental subjects attributed the differences in performance to the quality of the players rather than the situation. This was an error. But from that error, should we draw the conclusion that differences in basketball performance between players can be explained by situational variables rather than by differences in player skills? Obviously not. Both situation and skill have explanatory roles. Indeed, situation and skill interact to produce player performance in basketball. But the fact that experimental subjects sometimes fail to understand the importance of a particular situational variable (lighting conditions) tells us almost nothing about the question whether basketball players have different ability levels that explain differences in their performance. Indeed, real basketball games are played under conditions that are carefully designed to eliminate most situational variables associated with playing conditions. Not all, of course. The home court advantage is part of the game, and I suspect that any knowledgable fan would be able to identify the influence of this variable. (It's pretty obvious in the statistics.)
Are there differences in basketball preformance that are caused by traits rather than situations? Of course! A randomly selected team of high school players will lose to a randomly selected team of NBA players almost 100% of the time, and this result will hold under a variety of situations, so long as both teams have an adequate incentive to win the game. (If you don't believe this hypo, how about a randomly selected team of law professors?) The NBA players are better. Their skills, abilities, and conditioning explain why they play better than the randomly selected high school players. Situational variables do almost none of the explanatory work. And this is true, even thought a cleverly designed experiment can elicit the so-called "fundamental attribution error" with respect to basketball performance. Likewise, the fact that the fundamental attribution error can be elicited with respect to character traits does not show that character traits do not exist.
Conclusion
So what is my conclusion? Frankly, I'm just not sure. My impression is that a good deal of the psychology research is simply not directed at the questions that are interesting to legal theorists. Having read quite a bit, I am quite sure that many of the researchers have committed a variety of errors in drawing conclusions from their research. However, I am also fairly sure that many of our folk psychological beliefs about character have not been confirmed by research. And it is not clear to me that the kind of research that would be required to test a really robust theory of character is feasible (or ethical). But I am quite sure of one thing. John Doris's book has raised important issues that philosophers and legal theorists need to ponder.
Links and Resources
Books
Working Papers, Articles, and Book Reviews
On the Internet
And finally, my own paper, Virtue Jurisprudence: A Virtue-Centered Theory of Judging.
