The changing of the guard

A British study that revisited the 'Stanford prison experiment' for reality TV has outraged the US academic who inspired the work, says Harriet Swain

For viewers, it was the unfolding power struggle between "prisoners" and "guards" that made The Experiment gripping television. But with a team of academics at its core, the programme, broadcast by the BBC in 2001, was more than just another "reality" show. A host of conflicts raged behind the scenes -between the needs of science and the demands of entertainment, between opposing European and US psychological theories and between the British academics who designed the televised study and the American professor whose work inspired it.

The Experiment explored a fundamental question about human nature: what makes ordinary people ill-treat other ordinary people? Four years on, the clash of ideas that it engendered is set to come to a head when the results of the research carried out during the programme are published in peer-reviewed journals.

In the next few months, papers based on the study co-written by Alex Haslam, professor of social psychology at Exeter University, and Steve Reicher, professor of psychology at St Andrews University, will appear in the British Journal of Social Psychology, The Journal of Applied Psychology and Leadership Quarterly . There will also be an article in Scientific American Mind and chapters in two books - a triumph for a study commissioned as a television programme.

But appearing alongside one of the Haslam and Reicher articles will be a scathing commentary written by Philip Zimbardo, professor emeritus at Stanford University. He is the man behind the infamous 1971 "Stanford prison experiment" whose exploration of human behaviour during incarceration led to the BBC programme.

Zimbardo accuses Haslam and Reicher of producing a "scientifically irresponsible" study, "made for TV" and unsuitable for acceptance "by the social psychological community in Britain, the US, or anywhere except in media psychology". In turn, Haslam and Reicher have accused Zimbardo of trying to halt debate on a subject with far-reaching implications. "The findings of the study really challenge his work," Haslam says. "And he doesn't like it."

The two studies, separated by three decades, reach rather different conclusions. The results of the Stanford experiment were chilling. It divided student volunteers into prisoners and guards and kept them in a prison-like setting with Zimbardo as superintendent. The study was stopped after six days as the guards' behaviour descended into sadism and several participants showed signs of mental disturbance. Zimbardo argued that it demonstrated "the inherently pathological characteristics of the prison situation" and that for ethical reasons it should never be repeated.

Haslam argues that this meant it would be impossible to question Zimbardo's findings. He and Reicher felt it was important to challenge a message that was "dangerous and problematic politically and socially", especially as the study was later used to explain extreme behaviour such as that of terrorists. The two academics say that the BBC study employed enough safeguards to make such an attempt possible.

For their study, 15 men were recruited, screened and then randomly divided into two groups. Prisoners had their heads shaved and were given orange boiler suits. Guards were told they were chosen for their superior qualities and were given better food and conditions. They were instructed to draft rules and punishments to ensure that the institution ran smoothly.

There was daily psychometric testing of both groups, and saliva swabs were taken to measure levels of cortisol, an indicator of stress.

From the start, the guards were troubled by exercising power and did not identify as a group. The prisoners, on the other hand, identified with each other. This led to a shift in power and the collapse of the prisoner-guard system. Both groups then tried to come together and to operate as a commune. But they failed to organise themselves and by the end were about to set up the kind of draconian system they had tried to replace.

While Zimbardo contended that groups and power always lead to tyranny, Reicher and Haslam conclude that "it is the breakdown of groups and powerlessness that creates the conditions under which tyranny can triumph".

They warn of the danger of using the Stanford prison experiment to "justify and normalise" actions such as the abuse of prisoners at Abu Ghraib.

Instead, their findings point to ways in which oppression can be overcome.

Reicher says: "Zimbardo's is a depressing study that says that if you put people in a tyrannical role, they will act in tyrannical ways. Our sense is that isn't always true. People sometimes submit to tyranny, but sometimes they rebel."

So while groups can behave badly, they can also change things. Reicher and Haslam say their study is not only about beliefs and values but about organising a society in which those values can be established.

"We don't believe Zimbardo's work was wrong or his arguments were rubbish," Haslam says. "We think they were limited and there were failures of interpretation, and our data supports that alternative analysis."

Zimbardo strongly disagrees. In his peer review for the British Journal of Social Psychology , he criticised the researchers for failing to recreate the sense of a real prison and for intervening too often in the research setting. He also accused them of creating an "artificial subtext of European social identity theory pitted against American role theory".

Above all, he criticised the study because of the involvement of TV. He argued that the visible cameras and microphones and the participants'

awareness that their actions would be screened nationwide probably altered the behaviour of participants.

Zimbardo insists that he is no malcontent and says that some of the participants also had reservations about the study. But his was a lone voice among the reviewers. The other three supported publication and Brian Parkinson, the journal's editor, notes: "The behaviour shown on TV programmes is often real social behaviour that is worthy of social psychological attention."

He adds: "I would certainly never rule out using this kind of data solely on the basis of its being televised."

Haslam and Reicher say the physiological and psychometric data prove that those under scrutiny were not play-acting. And, they say, surveillance is increasingly part of our lives, and humans frequently have to consider how their actions may be seen. They add that they never intended to recreate exactly either Zimbardo's experiment or a real prison. Rather, they wanted to represent a broad class of institutions in which one group of people has power over another.

Finally, they claim that whatever else it has achieved, the project has reopened the debate about tyranny and brought to an end the unquestioned dominance of Zimbardo's work. For Zimbardo, who returns to the subject of his experiment in a book out next year, his interpretation of man's inhumanity to man prevails.