“I was just following orders.” People were chilled when they heard Adolf Eichmann, lieutenant colonel for the Nazi Party, explain his role in the Holocaust. But the phenomenon of people doing unspeakable acts is not limited to Eichmann, and certainly has not remained there. In offices, people have covered up for the CEO; guards abuse prisoners.

For years, the predominant explanation for this phenomenon lied in obedience. Robert Milgram conducted the first series of experiments on this idea, where participants were asked to play the role of “teacher” and administer tasks to an actor, the “learner” on the other side of a partition. If the actor failed the task, the participants were told to give a shock to the actor. Though the actor was not actually being shocked, and participants could not see the person, they could hear the voice through their partition – cries of agony and pain and pleas to stop. Milgram found that the majority of people continued administering lethal shocks at beyond a human threshold for pain, because the experimenter, a man in a white coat – ostensibly a doctor, in a position of authority – told them to.

People were understandably horrified by Milgram’s conclusion. But when scientists have recreated the experiments, with varying conditions, they have not always had the same result. Researchers from the United Kingdom and Scotland sought to answer the question of why.

The researchers theorized that social identification may have played a larger hand than obedience. In studies in which the conclusions were the same as Milgram’s, participants identified more with the experimenter in Milgram’s original experiment. In studies in which the conclusions contrasted with Milgram’s, participants identified more with the learner.

The study authors had two different sample groups – 32 expert psychologists from Great Britain and Australia – and 96 first-year psychology students. They were all given a short summary of the Milgram experiment, and given a list of 15 variants. From the list, the respondents were asked to define whether the variants would make participants identify in greater numbers with the experimenter or with the learner.

The results confirmed researchers’ hypothesis – increased identification with the experimenter would lead to greater obedience, while increased identification with the learner led to decreased obedience. That explains that in cases like Eichmann, people are not merely following orders. It also explains why people are more obedient under certain conditions more than others.

The results of the study were published in Perspectives of Psychological Science.