In the summer of 1971, Palo Alto became the home for one of psychology's most influential and compelling experiments. Philip Zimbardo, a young professor at Stanford, wanted to know how situations influenced behavior. He chose about 20 healthy college-age men, randomly assigned them roles as prisoners or guards and created prison-like conditions in the basement of a university building. He was supposed to pay them $15 a day for two weeks of imprisonment and guarding. Within days, however, the experiment turned into a nightmare, and through the intervention of one of his colleagues (later to become his wife), the whole thing was called off. Social psychology would never be the same.

Once the apparently normal young men were assigned the roles as guards, they very quickly developed tyrannical and abusive strategies for controlling "their" prisoners. As they seemed to forget that the basement offices were only the scene of an experiment, the guards became intensely protective of their authority. Sadism reared its head as the prisoners sank into their new identities as powerless people trapped in the system. Some of the prisoners suffered severe, if temporary, mental anguish, as they seemed to forget they could break their contracts and leave. The roles of prisoner and guard overwhelmed personal identity.

Zimbardo's Stanford Prison Experiment was a landmark because it undermined our core belief in the integrity of the individual moral self. For the past few hundred years, the dominant view in Western cultures has been that the individual self is formed over time, and that its chief characteristics are expressed not only variously in everyday life but also in extreme situations. In other words, we have preferred to think that some people behave badly and others behave well because of who they really are, and though situations may constrain an individual, they don't totally remake him or her.

Zimbardo thought otherwise, and as a social psychologist in the late 1960s, he believed that situations determine how we act in a significant way. But even he was unprepared for the radical transformation that simple role-playing brought about. Even he, as the chief experimenter, fell into a role as "head of the prison" and lost perspective on what was going on. He became more concerned about whether his prisoners were going to be "liberated" by an attack from the outside than about the welfare of his volunteer subjects. Only after he called off the experiment did Zimbardo realize that he, too, had become a prisoner of the situation he had created for research.

"The Lucifer Effect" provides a very detailed account of the now familiar experiment, and it remains fascinating to see how quickly and powerfully the situation changed the participants' behavior. But why, after all these years, is Zimbardo returning to the experiment? Abu Ghraib. It has brought the experiment back into the limelight and the author back into the courts as an expert witness. Although he is clear in saying that situations do not excuse illegal behavior, Zimbardo testified on behalf of one of the defendants to underscore how a normal person could be led into the worst sorts of behavior by a situation that seemed to demand abuse.

Zimbardo's message to the tribunal and to his readers is clear: Even you might behave tyrannically and violently if you were in a situation that encouraged such behavior. The experiment "has emerged as a powerful illustration of the potentially toxic impact of bad systems and bad situations in making good people behave in pathological ways that are alien to their nature."

Zimbardo does not believe that those bad situations reveal that we aren't such good people after all. For all his knowledge of the power of social context to determine who we are and what we do, he retains an even deeper faith that most people really are good and could prevent the distortion of personality by avoiding poisonous roles and institutions. He does not think he discovered that our innate sadism and love of tyranny are uncovered by situations that allow us to behave violently. His is the more comforting view that we can maintain our moral standards as true expressions of who we really are by creating more consistently moral or just institutions.

Zimbardo's video testimony (he did not want to go to the very bad situation of Iraq) did not have much effect on the Military court. That seems to have been the real stimulus for writing "The Lucifer Effect." The social psychologist plays the role of prosecutor in the last sections of this very long book to indict those responsible for creating "the System" that really is responsible for the bad situation of Abu Ghraib.

But all the evidence he brings together is readily available elsewhere, and the psychologist goes through much familiar political territory to show that the Bush administration is ultimately responsible (is guilty) for creating the poisonous contexts in which "otherwise normal" men and women behaved so badly. We are repeatedly told that it's the barrel, not a few apples, that's rotten.

Finally, and rather embarrassingly, this great psychologist provides a "ten-step program to resist unwanted influences." Where do human "wants," the individual's desires, come from in the first place? Zimbardo hasn't a clue. The banality of his "program" does a disservice to the quality of his experimental work. We don't need to be told to declare, "I am Me, the best I can be." Who does he think will be helped by the "resisting influence guide" (no joke, it's a guide against influence) on the "Lucifer Effect" Web site?

Zimbardo's parting message is to be yourself and to celebrate heroes. But his real cultural contribution occurred when he himself got lost in an experiment in a Stanford basement more than 30 years ago. That's when he, and we, discovered that we don't know who we are, nor what we can become, when we find ourselves in a new role in a novel situation.

Michael Roth is the former president of the California College of the Arts. He is now president of Wesleyan University.