There’s a fantastic scene at the end of Quentin Tarantino’s film Death Proof. Three young girls are being hunted by Kurt Russell, a psychopath with a stunt car and a foot fetish. We have watched him torture and kill his way through the film, and, as he wends his way towards this group, their naivete is used as a foil to his ever increasing menace. In the last three minutes they beat him to death with an iron pole. It’s a great ending.

Outside Hollywood, the idea of revenge is often pushed into the background, especially when explaining our motives. We call it “justice” or “righting a wrong” or “balancing the scales”, but these notions don’t do much for us biologically and don’t show up in an obvious way in the brain (it’s a grey area). Revenge, on the other hand, has a very clear neural signal – and that signal is pleasure.

Research by Tania Singer at University College London published in 2006 looked at the idea through a classic scenario – the prisoner’s dilemma. You and your partner in crime have been arrested. You are interrogated separately, and each of you is offered a deal. If you confess and tell on your partner you will be given a year’s sentence and your partner four. If you both stay quiet, you both get two years, and if you both confess, you’ll both get three.

Singer asked two “prisoners” to play out this scenario in front of an audience, which was left to form opinions of each of them. Then members of the audience were put inside fMRI machines while they witnessed each prisoner receiving electric shocks to his hands. As the audience members watched the experience of the prisoners, they showed increased activity in neural pain areas – the evidence of empathy – but this empathy was present only when watching one of the “silent partners”. If a “confessor” was punished, the activity died down considerably; the brain cared much less about the pain of those who had betrayed their partner.

There was another finding that Singer didn’t expect. Watching “bad” prisoners get punished gave members of the audience pleasure: there was activation in reward-related areas of the brain, such as the ventral striatum and the nucleus accumbens. (This was limited to male subjects.)

So, we get pleasure from revenge and, according to similar studies, we also get activity in the left prefrontal cortex of the brain, which relates to goal planning. Revenge seems to be not only a passing delight, but a craving, something we need, that we plan for. The desire increases when we are mistreated in front of others – we need to show people we’re not to be pushed around – and decreases when we’ve got more to lose by exacting it. In other words, the emotion has a logic to it.

Does it have an evolutionary function? Literature is littered with sayings about revenge and how it never pays, or is better expressed by forgiving the other person, or is exacted by “living well” (with gritted teeth). Yet our neurobiology feels otherwise. Which is right?

Well, let’s return to the prisoner’s dilemma. The dilemma itself is whether or not to trust your partner, but this issue figures only in the beginning. When the game is played over and over again, prisoners at first seek petty revenge on each other, punishing the other for betraying them the first time. But ultimately, subdued by punishment, the two settle on a compromise. Which seems to be the point. The possibility of revenge might well be vital for human interaction: a sweet necessity.