Classic Group Psychology: Conformity, Obedience, and an Overrated "Situation Myth"

In 1951 Solomon Asch asked people to judge which of two lines matched in length — the answer was obvious. But when a room of "peers" (confederates) unanimously gave a clearly wrong answer, about 75% of real subjects conformed at least once, going along with the error on 37% of the critical trials. People will deny what they plainly see, just to align with the group.

Social influence comes in two kinds (Deutsch & Gerard, 1955): informational influence — using others as cues when uncertain — and normative influence — going along out of fear of rejection. Asch's answer was unmistakable, so the driver was mostly the latter: not genuine belief, but not wanting to be the odd one out. Berns's (2005) fMRI work goes further, suggesting conformity can alter early perceptual processing in the brain, not just verbal agreement — pressure can seep into perception itself.

Reflection: Return to the last time you stayed silent or agreed in a group — were you genuinely unsure (informational), or afraid of being isolated (normative)?

In 1961-63 at Yale, Stanley Milgram had subjects, under the orders of a lab-coated authority, deliver what they believed were escalating and ultimately lethal 450V shocks to a "learner" next door. About 65% of ordinary people went all the way to the maximum. They were not evil; most were visibly distressed, trembling, protesting — and pressed on anyway.

Milgram's explanation was the agentic state: people hand responsibility up to authority and feel they are merely "instruments." But modern reinterpretation (Haslam & Reicher) is sharper: this isn't "blind obedience" but engaged followership — the more subjects identified with the experimenter's "scientific cause" and saw him as one of their own, the more they obeyed. Crucially: situational variables decide almost everything.

Reflection: Think of a decision you made "on orders" — if you had to explain it face-to-face to the people it affected, would you still do it? If not, your conscience probably never agreed.

Philip Zimbardo's 1971 prison simulation has long been told as: "good college students, once given a guard identity, spontaneously turned cruel" — situation utterly overpowering human nature. But archives and recordings released since 2018 show this widely repeated story is seriously methodologically unsound.

A properly controlled redo reaches the opposite conclusion: Reicher & Haslam's BBC Prison Study (2002) found that people do not automatically adopt assigned roles — whether a group becomes oppressive depends on whether a shared identity can form, which is in turn strongly shaped by leadership and narrative. In short: cruelty must be authorized and encouraged; it does not happen automatically just from putting on a uniform.

The real lesson of all three experiments isn't "humans are evil" but that we systematically overweight character and underweight situation — the fundamental attribution error (Lee Ross, 1977). But beware the post-2000 overcorrection: casting people as pure puppets of the situation is equally wrong.

Of others we say "that's just who he is"; of ourselves, "the circumstances were special." The asymmetry comes from attentional focus: looking at others, we see the person; looking at ourselves, we see the situation. The modern synthesis (Haslam, Reicher): situations are powerful but not mechanically determining — identity, leadership framing, and active choice mediate between situation and behavior. Arendt's "banality of evil" has also been revised: Stangneth (2014), drawing on Eichmann's own writings, shows he was no thoughtless cog but a committed ideologue. Neither pure character nor pure situation.