Archie Bland: However nice you are, you might push the button too

To the general public, it sounds preposterous: the mere presence of an attractive television host and a baying mob should not be enough to make you electrocute an innocent stranger. You're a normal person, not a lunatic, and you just wouldn't do it.

But however untrue it feels, psychologists know differently; indeed, they have done for years. There is nothing new about the quasi-experiment broadcast on French television. Instead, it is a new twist on one of the most famous studies in the history of social psychology: the Milgram experiment.

That this new version tweaks the formula to add a live audience and replaces a man with a white coat with a woman in a shimmering dress does not alter the fundamental fact: most of us are more obedient than we would like to believe. When teachers ask easily-led pupils if they would follow little Johnny off a cliff if he suggested it, the answer is probably yes. And every genocide in human history has relied on the terrifying possibility that your nice next-door neighbour might pick up a weapon if someone in authority tells him to – and that you might, too.

Until Stanley Milgram, though, no one had quite acknowledged that fact, for the obvious reason that it is anathema to some of our most basic beliefs about human compassion. Indeed, when Milgram polled his colleagues about the likely results of the study in advance, they estimated that about one per cent of participants would administer the maximum, fatal, shock. In fact, 65 per cent did.

There are real questions about the ethical basis for such a study, questions that can only be sharpened by transplanting the process to television. And while it may be true that television is a uniquely horrible (and relatively new) means of exerting social control, the French documentary was not subject to any of the ordinary constraints that experimental psychologists would face.

But for all of the valid objections to making a circus of such a fundamental and distressing human behaviour – to say nothing of the consequences for the participants – there is one powerful counterargument: simply by making such a phenomenon more widely known, you make it less likely to recur.

In the Milgram study, participants were prodded into submission by no more aggressive means than the simple statement: "You have no other choice, you must go on." Perhaps, if the study and its hideous implications were more widely known, we might at least pause before pushing the button.