Another of the thirty. This is the most famous experiment in psychology, and the archives say it did not happen the way you were taught.
August 1971. A Stanford basement is dressed as a prison, students are assigned at random to be guards or prisoners, and within days the guards turn cruel and the prisoners start breaking down. A study planned for two weeks was stopped after six. The lesson taught ever since is that decent people put in the wrong situation will do terrible things, because the situation wins. Zimbardo built The Lucifer Effect on that claim, and after Abu Ghraib he carried it into court and testified for the defense of one of the guards.
Then the archives opened. Thibault Le Texier went through them in 2018, Ben Blum reported on the same material, and what they found is that the guards were coached. The encouragement to be tough was systemic and it came from Zimbardo’s own team, particularly from the man playing warden, David Jaffe. There was no pre-specified protocol. No independent oversight. No control condition. And Zimbardo was not watching from behind glass; he cast himself as prison superintendent inside his own experiment, which means the person interpreting the results was also producing them.
So the famous demonstration that situations manufacture cruelty had people manufacturing the cruelty at the center of it. That is not a footnote to the study. It is the study.
The better evidence points somewhere more interesting than either version. When the BBC Prison Study re-ran the setup in 2002 with real oversight, people did not turn cruel simply by putting on a uniform. Where cruelty did emerge it followed leadership: someone actively building an us-and-them identity and giving harsh norms permission. Situations apply pressure. What that pressure turns into is decided by leaders, and by what a group agrees is acceptable.
Zimbardo’s own slogan outlives his experiment, which is a small irony worth keeping. There aren’t just bad apples, he said, there are bad barrels, and bad barrel makers. Drop the deterministic reading, because a room does not do it by itself, and you are left with something weaker and far more useful: institutions can be built so that cruelty is easy, and somebody usually has to authorize it. That puts the blame somewhere less comfortable than a basement full of role-playing students.
It is still taught in the strong version in nearly every introductory course, decades after the criticism landed. I find that harder to explain than the original result, and it is the pattern I went looking for in the other twenty-nine books. The question is not only whether the famous study was right. It is what happens to a finding once it becomes famous.
The 30 Greatest Psychology Books: What They Actually Argued, and Where They Go Wrong. Paperback and Kindle.