Why Smart People Resist Changing Their Minds

In 1956, Leon Festinger published a book called When Prophecy Fails. He had been studying a small UFO cult whose leader had predicted that the world would end on a specific date. Festinger and his colleagues infiltrated the group as participant-observers, curious about what would happen on the morning after the predicted apocalypse failed to occur. The natural expectation was that the believers would feel embarrassed and either leave the cult or quietly distance themselves from the prediction.

Almost none of them did. Instead, the prediction was reinterpreted: the cult's faith had saved the world; the prediction had been correct on a deeper, spiritual plane; the failure was a test. Several previously private members became more public proselytizers, more committed than before. Festinger's interpretation, later formalized as the theory of cognitive dissonance, was that human beings are sometimes more committed to their beliefs after disconfirming evidence than before.

The theory was new and controversial. It is now part of the basic vocabulary of social psychology. It also describes a pattern that almost everyone, on reflection, can identify in themselves and other people.

The puzzle that interests this essay is a related one. Why is the pattern often worse, not better, in people who are unusually intelligent?

The intuition is that intelligent people, having a sharper grip on evidence, should update faster when the evidence changes. The evidence is the opposite. Studies of motivated reasoning consistently find that more intelligent participants are not better at arriving at the correct answer when the correct answer threatens a belief they care about. They are better at generating reasons why the correct answer is wrong.

This is sometimes called the intelligence trap. The trap is not intelligence itself but the use to which intelligence is put. The mind has two related capacities: a capacity for finding the truth, and a capacity for defending whatever the mind already believes. In most people most of the time, these capacities compete. A more intelligent mind is better at both. If the second capacity is in charge, more intelligence makes the mind more resistant to correction, not less.

The mechanism is identity. Beliefs do not float free in the mind; they are attached, often deeply, to the person's sense of who they are. The doctor who has been giving certain advice for twenty years has staked something larger than an opinion on it. The activist who has built a public identity around a position will find it materially expensive to abandon. The intellectual who has written books on a subject is, in a real sense, the books. To change the belief is to dismantle a piece of the self that depended on it. The mind has strong protections against this.

The discipline that works against this is unusual and uncomfortable. It involves, first, separating the belief from the believer — treating beliefs as instruments that you use rather than possessions that define you. Second, it involves pre-committing to what evidence would, in fact, change your mind. The act of saying in advance what would cause you to update is itself diagnostic. If you cannot name such evidence, you do not really hold the belief on grounds; you hold it on commitment.

The Bayesian framing, popular in some intellectual circles, is a useful approximation of this discipline. You hold a probability on a belief, not a flag. New evidence adjusts the probability. The adjustment is mechanical, not personal. In practice almost no one operates this cleanly. But the framing is a useful reminder that minds at their best are updating instruments, not static positions.

The cost of being wrong about a particular belief is almost always small. The cost of being unteachable — of becoming a mind that cannot update, because each update is a wound to identity — is enormous. It is the difference between continuing to grow and slowly fossilizing inside the views one happened to acquire before the age of forty.