The Mirror That Thinks It Is a Mind
How rooms full of intelligent people talk themselves into decisions none of them would make alone.
In 1979, ten thousand people pressed toward a set of doors at Riverfront Coliseum in Cincinnati. They had come to hear The Who. Most of the seating was general admission, and a rumour rippled through the packed forecourt that the concert had already begun. The crowd surged. At the front, near the still-locked doors, the pressure became unbearable. People fell. Others, pushed from behind, could not stop themselves walking over them. Eleven died, most of them by compressive asphyxia, crushed while standing upright.1
What is unsettling about the story is not cruelty but its absence. Nobody in that forecourt wanted to hurt anyone. Each individual wanted a small, reasonable thing: to get inside, to find a seat, to not be left behind. And yet the sum of those reasonable wants produced something monstrous. The crowd made a decision that no single person within it would ever have chosen. This is the uncomfortable puzzle at the heart of social psychology. A group of intelligent, well-meaning people can arrive at a conclusion dumber, more dangerous, or more deluded than any of them would reach alone.
Most of us have felt a milder version of the same force. You are in a meeting. The room is nodding. A decision is forming, and something about it feels wrong to you, but the words to explain why will not come, and the moment to speak is passing, and then it has passed, and you nod too. Nothing was decided by argument. It was decided by momentum. The question this essay asks is a simple one with a long and strange history: why does gathering more brains sometimes produce a worse answer?
A doctor watching a mob
The first serious attempt to explain crowd behaviour came from a French physician named Gustave Le Bon. In 1895 he published The Crowd: A Study of the Popular Mind, a book that argued something genuinely new for its time.2 A crowd, Le Bon claimed, is not simply a collection of individuals. It is a distinct psychological entity with its own character, and that character is almost always inferior to the people who compose it.
Inside a crowd, Le Bon wrote, the conscious personality dissolves. Feelings and ideas take a single direction. The individual becomes suggestible, impulsive, and susceptible to contagion, like a cell absorbed into a larger organism. His most quoted line captures the tone: “By the mere fact that he forms part of an organised crowd, a man descends several rungs in the ladder of civilisation.”
Le Bon’s work has aged badly in most respects. It is soaked in the class anxieties and racial prejudices of late nineteenth-century Europe. He treated the masses as a mob to be feared and manipulated, and his ideas were later admired by propagandists who took his advice about steering crowds rather too literally. As science, his book is closer to nervous editorial than to evidence.
And yet he asked the right question. Why do sensible people abandon their own judgment when they gather? Le Bon located the problem in the roaring street, in the revolutionary throng and the lynch mob. He was looking in the wrong place. The real problem, later researchers would discover, was quieter and far closer to home. It was not the mob outside. It was the committee in the room next door: the jury, the boardroom, the panel of experts. Ordinary groups of thoughtful people, bending steadily toward error, and never noticing the bend.
Three lines and a room of liars
The first hard evidence arrived through an experiment about something as trivial as the length of a line. In 1951, the Polish-American psychologist Solomon Asch designed a test so simple it seemed pointless.3 He showed volunteers a card with a single reference line, then a second card with three comparison lines of obviously different lengths. The task was to say which of the three matched the reference. The correct answer was never in any doubt. Alone, people got it right virtually every time.
The trick was the setup. Each volunteer sat in a group of seven or eight others who appeared to be fellow participants but were in fact confederates, actors instructed by Asch. On certain rounds, every confederate confidently named the same wrong line. The real subject, seated near the end, then had to answer after hearing a chorus of unanimous, obvious error.
The results have unsettled psychologists ever since. Around three-quarters of subjects conformed to the wrong answer on at least one trial. Across all the critical trials, subjects went along with the incorrect majority roughly a third of the time. These were not people who could not see the lines. When interviewed afterward, most reported that they had known the true answer perfectly well. Some had begun to doubt their own eyes. Others simply did not want to stand out, to be the lone dissenter, the odd one, the difficult person.
Asch himself found the outcome sobering rather than triumphant. “That we have found the tendency to conformity in our society so strong that reasonably intelligent and well-meaning young people are willing to call white black is a matter of concern,” he wrote.3 The line test stripped the problem to its bones. When belonging and accuracy pulled in opposite directions, a large fraction of ordinary people chose belonging. And this was over lines, where nothing was at stake, where being wrong cost nothing but the mild discomfort of standing apart.
Now raise the stakes. Imagine the disagreement is not about lines but about whether to invade a country.
The disaster planned by brilliant men
In April 1961, a force of roughly fourteen hundred Cuban exiles, trained and armed by the United States, landed at the Bay of Pigs on Cuba’s southern coast. The plan was to spark an uprising against Fidel Castro. Instead the invaders were pinned down within hours and overwhelmed within three days. It was a total, humiliating failure, and it had been approved by one of the most impressive collections of talent ever assembled in an American administration: President Kennedy and a cabinet of celebrated intellects, war heroes, and scholars.
The psychologist Irving Janis spent years trying to understand how such capable minds had endorsed such a flawed plan. In his 1972 book Victims of Groupthink, he gave the phenomenon its enduring name.4 Groupthink, in Janis’s definition, is a mode of thinking that people fall into when they are deeply embedded in a cohesive group, when the members’ striving for unanimity overrides their motivation to realistically appraise alternatives.
The mechanism is almost gentle. In a close, high-status group, harmony feels good and important. Raising an objection begins to feel like an act of aggression, a breach of the group’s warmth. So doubts go unspoken. Each member, seeing no one else object, assumes the silence around them signals agreement, and censors themselves further. An illusion of unanimity forms, in which everyone privately harbours reservations and everyone publicly appears certain. Kennedy’s advisers later admitted they had felt uneasy but had said nothing, unwilling to be the skunk at the garden party.
Janis identified eight symptoms of groupthink, and reading them today feels like reading a diagnosis of half the institutions we know.4 An illusion of invulnerability breeding excessive optimism. Collective rationalising away of warnings. Unquestioned belief in the group’s own morality. Stereotyped views of outsiders and rivals. Direct pressure on any member who dissents. Self-censorship of doubts. A shared illusion of unanimity. And the emergence of self-appointed mindguards, members who shield the group from inconvenient information.
The cruel twist is that the trap tightens as the group improves. The more cohesive, the more accomplished, the more mutually respectful the members are, the more each one wants to preserve the group’s good feeling, and the less anyone wants to be the one who spoils it. Talent and camaraderie, the very things that ought to produce good decisions, quietly conspire to produce bad ones.
The heat of the room
Conformity and groupthink explain how a group can converge on error. But there is a second, stranger force at work, and it does the opposite of what most people expect. We tend to assume that when people talk something over, their views moderate. Extremes cancel out. Discussion cools the temperature. The research shows the reverse.
In the 1960s and 1970s, the French social psychologist Serge Moscovici and his collaborators demonstrated that group discussion does not average opinions toward the middle. It pushes them toward the extreme.5 The finding, now called group polarisation, is remarkably robust. Take a room of people who lean cautiously in one direction. Let them discuss the matter among themselves. When they emerge, they lean more strongly in that same direction than any of them did when they walked in. Cautious groups grow more cautious. Bold groups grow reckless.
Why does agreement intensify rather than soothe? Part of the answer is informational. When you argue your position, others offer supporting reasons you had not thought of, and you come away with more ammunition than you brought. Part of it is social comparison. Once you realise the group leans your way, you feel safe to lean harder, even to compete for the position of the most committed member. Every voice that echoes your view becomes fresh evidence that your instinct was correct all along. Discussion, in other words, does not test opinions against reality. It rewards them for being loud.
This is why online communities tend to drift toward their sharpest edges rather than settle into a reasonable centre. A forum of people who mildly distrust something can, through nothing but mutual conversation, become a forum of people who are certain. And it is why financial markets lurch from euphoria to terror with such violence. Prices are supposed to reflect cool calculation. In practice they reflect a crowd watching itself.
The classic illustration is the Dutch tulip mania of the 1630s. Speculation in rare tulip bulbs reached such a pitch that, at the peak in early 1637, single bulbs of prized varieties reportedly changed hands for sums greater than the price of a comfortable house.6 Then, almost overnight, the crowd changed its mind. Buyers vanished, prices collapsed, and fortunes built on the assumption that tulips would keep rising evaporated. What is striking is that nobody inside the bubble felt foolish while it was inflating. Each buyer was simply agreeing with everyone else, and everyone else was buying, and the agreement itself looked like proof. The madness was invisible from the inside because it was made entirely of ordinary, individually sensible acts of imitation.
The weight of an ox
Here the story takes a turn. If crowds conform, converge on error, and inflame each other into extremes, the tidy conclusion would be that crowds are simply dumber than individuals and should be avoided. That conclusion is wrong. The same species that stampedes and panics and blows soap-bubble fortunes is also capable, under the right conditions, of startling collective accuracy.
The most famous demonstration came from an unlikely source. In 1906, the Victorian polymath Francis Galton attended a country fair in Plymouth, England, where visitors were invited to guess the weight of a slaughtered and dressed ox. Nearly eight hundred people paid to enter estimates. Galton, who had a low opinion of the average person’s judgment and expected to prove the crowd hopelessly wrong, collected the tickets and ran the numbers.7 The individual guesses were scattered wildly. But their median came to 1,207 pounds. The ox’s actual weight was 1,198 pounds. The crowd, taken as a whole, had erred by less than one percent, outperforming most of the cattle experts present.
Galton’s ox has become the founding parable of what the journalist James Surowiecki later called The Wisdom of Crowds.8 Under the right conditions, aggregating many independent estimates cancels out individual errors and converges on something close to the truth. The errors are not identical. Some guess too high, some too low, and when you pool them, the mistakes tend to offset one another, leaving the signal behind.
So we have two crowds that appear to contradict each other. One tramples people at a stadium door and drives tulip prices to insanity. The other guesses the weight of an ox to within a fraction of a percent. They are not different species. They are the same crowd under different conditions, and the condition that separates wisdom from madness has a name.
Independence, and the courage of the awkward question
The hinge is independence. Galton’s crowd was wise because each visitor wrote down a private estimate without knowing what anyone else had guessed. Their errors were uncorrelated, so they cancelled. The stampede and the bubble were disasters because the opposite was true. Each person was watching, copying, and reacting to everyone else, so their errors did not cancel. They compounded. A crowd of independent minds averages toward truth. A crowd of imitating minds amplifies whatever it happens to be feeling.
This is the practical lesson buried under a century of research, and it is more hopeful than Le Bon’s gloom. The intelligence of a group is not fixed by the intelligence of its members. It is set by how the group is structured. The antidotes to groupthink, the ones Janis himself recommended, all serve to protect independence: genuine diversity of background and perspective, the deliberate cultivation of dissent, and leaders who withhold their own opinions until others have spoken so that the room does not simply converge on the boss.4
Good decision-making, it turns out, is largely a matter of engineering. Assign someone the explicit role of devil’s advocate, so that objection becomes a duty rather than a betrayal. Invite outside experts who have no stake in the group’s harmony. Break large groups into smaller ones that deliberate separately before comparing conclusions. And, perhaps most powerfully, vote privately before you discuss. When each person commits to a judgment in silence, before the room’s mood has formed, you preserve the very independence that makes crowds wise. Discuss first, and you get polarisation. Estimate first, then discuss, and you get something closer to the ox.
There is a personal version of this discipline too. When you sense a decision hardening around you, before you have decided your own mind, ask whether your growing certainty is built on evidence or merely on the sound of agreement. The nodding heads in a room are not data. They are a mirror.
That, in the end, is the truest thing to say about a crowd. It is not a mind, though it can convincingly pretend to be one. It has no wisdom of its own and no madness of its own. It reflects and magnifies whatever the people inside it bring, and it magnifies most fiercely the moment they stop thinking for themselves and start watching each other. The wisdom of crowds and the folly of crowds are not opposites. They are the same reflective surface, tilted toward truth or toward disaster by a single variable. So the next time everyone around you agrees, and the agreement feels warm and effortless and safe, that is precisely the moment to do the ungrateful thing. Ask the quiet, uncomfortable question. The room may need it more than it knows.
Sources
- Johnson, N. R., “Panic at ‘The Who Concert Stampede’: An Empirical Assessment,” Social Problems, 1987. — https://academic.oup.com/socpro/article-abstract/34/4/362/1620633
- Le Bon, Gustave, The Crowd: A Study of the Popular Mind, 1895. — https://www.gutenberg.org/ebooks/445
- Asch, S. E., “Opinions and Social Pressure,” Scientific American, 1955. — https://www.jstor.org/stable/24943779
- Janis, Irving L., Victims of Groupthink: A Psychological Study of Foreign-Policy Decisions and Fiascoes, Houghton Mifflin, 1972. — https://archive.org/details/victimsofgroupth0000jani
- Moscovici, S. and Zavalloni, M., “The Group as a Polarizer of Attitudes,” Journal of Personality and Social Psychology, 1969. — https://psycnet.apa.org/record/1969-15467-001
- Goldgar, Anne, Tulipmania: Money, Honor, and Knowledge in the Dutch Golden Age, University of Chicago Press, 2007. — https://press.uchicago.edu/ucp/books/book/chicago/T/bo5209816.html
- Galton, Francis, “Vox Populi,” Nature, 1907. — https://www.nature.com/articles/075450a0
- Surowiecki, James, The Wisdom of Crowds, Doubleday, 2004. — https://en.wikipedia.org/wiki/The_Wisdom_of_Crowds
Related reading
