In the summer of 1971, one of the most famous experiments in the history of psychology was conducted in a 35-foot section of the Stanford Psychology Department basement. Eighteen healthy young men were randomly assigned—by the flip of a coin—to play the role of a prisoner or guard for a one-to-two–week experiment on prison life. The Palo Alto police “arrested” the prisoners in their homes and charged them with robbery. They were fingerprinted, strip searched and given new identities.

In the meantime, the guards were provided with wooden batons, uniforms and mirrored sunglasses. Within days, the guards began to employ psychological tactics to control and humiliate the prisoners, and several guards began to exhibit genuine sadistic tendencies. The sanitary conditions swiftly declined when the guards forced the prisoners to urinate and defecate in a bucket placed in their cell—and refused to let the prisoners empty it.

The cruelty of the guards eventually led one prisoner to “act crazy,” including screaming, cursing and begging to be released. When another prisoner went on a hunger strike to protest the abuse, he was locked in a dark closet for “solitary confinement.” The abuse of prisoners continued to escalate until the experiment had to be brought to an immediate halt several days before the study was scheduled to end.

Within weeks, these shocking findings had hit the press and the study quickly became a staple of social science classes around the country. The young professor in change of the experiment, Philip Zimbardo, became a star witness in congressional testimony on issues ranging from American prison riots to the torture of prisoners at Abu Ghraib.

The lesson, reflected in countless lectures, textbooks and several feature length movies about the study seemed clear: people slip naturally into social roles—affirming the power of the situation. Alongside the work on obedience to authorities by Stanley Milgram, these conclusions have played a key role in shaping public understanding of the psychology of tyranny and evil for the past half century.

But brand-new evidence from the archives has cast this classic study in a fresh light. Nearly a half century after the original experiment, the Stanford library made a range of materials from the experiment available for analysis. As a result, scholars were finally able to listen to audio tapes of the prisoners receiving instructions from the experimenters. These materials suggest a shockingly different explanation for the vicious behavior in the original experiment. 

As we explain in a forthcoming paper in the American Psychologist, this new evidence makes it very clear that merely assigning people to play the role of a prison guards did not lead them to engage in cruelty naturally, of their own accord. Instead, the tapes provide clear evidence that the experimenters in charge of the “prison” used psychological tactics to persuade reluctant guards to adopt an aggressive style in their interactions with the helpless prisoners. 

In one recording, the prison “superintendent”—a member of the research team—meets with one of the guards who was reluctant to engage in cruelty towards the prisoners. The guard says We noticed this morning that you weren’t really lending a hand … but we want to get you active and involved because the guards have to know every guard is going to be what we call a tough guard.” The guard repeatedly resists the pressure to engage in the harsh treatment of the prisoners.

This is where we noticed new evidence that the superintendent in the prison used a form of "identity leadership.” Specifically, he encouraged the guard to see himself as sharing the same mission as the experimenters—cultivating a shared in-group identity (and us versus them). For instance, the warden used collective pronouns 57 times during the short conversation (or once every 30 words) to cultivate a sense of shared identity with the guard—centered around toughness towards the prisoners. This is the same language that politicians who win elections are far more likely to use (using them once every 79 words) than losing candidates (using them only once every 136 words).

The superintendent then framed toughness towards prisoners as necessary for the achievement of a virtuous shared goal to help revolutionize the prison system. He says “We’re not trying to do this just because we’re sadists,” and then supplies a rationale for harming the prisoners: “If you need an excuse, and I think most of us do really, it is so we can learn what happens in a total institution.… And we want to know about them. So that we can, we can get on the media and, um, and, and into press with it. And, and, and say “Now look at what, what this really about.”

But you do not need to take our word for it. The audio tape we quoted can be accessed directly online. For the first time, people can be flies on the wall to learn what happened behind the scenes in one of the most famous experiments in psychology.

Another new paper by Jared Bartels found that Zimbardo’s instructions to the guards may have had a similar impact. The day before the experiment began, the research team held an orientation for the guards in which they communicated expectations for hostility towards the prisoners. Bartels showed a completely new set of participants this same script almost 50 years later and found that people expected hostile and oppressive behaviour from Zimbardo and assumed it was expected from themselves, as guards. Thus, the language of the leadership team clearly sanctioned abuse among the guards.

Together, this new evidence suggests that adopting a role or entering a system of brutality is not sufficient to produce cruel behavior. These findings converge with recent work showing how identity leadership was central to Milgram’s famous studies on obedience to authority and to the dynamics of tyranny and cruelty outside the lab.

Indeed, once one appreciates the role of identity leadership in both the Stanford Prison Experiment and Milgram’s obedience studies, the underlying lesson comes into focus. This is how leaders throughout history have validated cruelty to minorities or immigrants—by signalling that these are out-group members and that the ends justify the means. The actions and language of leaders matter.

Thankfully, this isn’t the whole story. Our analysis also highlights striking evidence that some guards were actively resisting engaging in brutality—fighting back against the pressures from leadership. From Oskar Schindler to Rosa Parks, history is brimming with people who have bucked oppressive authority figures and abusive systems. It is not only time for a radical rethink of one of psychology’s most influential experiments, but also for recognition that our most famous studies on cruelty, conformity and obedience have critical lessons to teach us about leadership and resistance.