A theory dating back more than 100 years suggested that there is an optimal amount of stress for good performance, based on a study that had nothing to do with humans.

Researchers ran mice through a maze, using electric shocks to punish wrong choices. On harder tasks, moderate shocks produced the fastest learning, while stronger shocks made it worse. Over the years, this got reinterpreted into claims about punishment, reward, motivation and stress in humans.

But the original study was about how animals learn to avoid fear and pain, not about optimal learning at all. Once a mouse learns to avoid a shock, escape itself triggers dopamine release; in other words, the escape – not the discovery – is the reward.