Think Again

The problem it starts from

In August 1949 a crew of smokejumpers parachuted into Mann Gulch in Montana to fight what looked like a routine fire. The wind changed and the fire raced uphill toward them. The foreman, Wagner Dodge, did something no one had been trained to do. He stopped running, lit a fire in the grass in front of him, and lay down in the burned patch it left. He survived. Most of his crew, who did not follow him, did not. Dodge survived because he was able to drop, in seconds, a response he had learned well and replace it with a new one.

Adam Grant opens Think Again, published in 2021, with that story. The organisational theorist Karl Weick had studied Mann Gulch in 1993 as a case of how a group’s shared sense of what is happening can collapse in a crisis, leaving people to fall back on what they have always done. Grant uses it to introduce his subject: how hard it is to let go of what we know, and how much can depend on doing so.

The practices

Notice which role you are playing. Grant borrows a framework from the psychologist Philip Tetlock, who described people in arguments as acting like preachers, prosecutors or politicians. A preacher defends a sacred belief, a prosecutor attacks the other side’s case, and a politician campaigns for approval. None of them is trying to find out what is true. Grant adds a fourth role, the scientist, who treats beliefs as hypotheses and looks for evidence that would show them wrong. The practice is to notice which of the first three modes one has slipped into and to step back into the fourth.

Hold confident humility. The goal is not doubt about one’s ability but doubt about one’s current knowledge. Grant draws on research into the Dunning-Kruger effect, which describes how people with little knowledge in a domain often overestimate it, and on the observation that the most capable people are often quickest to admit uncertainty. Confidence in one’s ability to learn, combined with humility about what one already knows, makes rethinking possible.

Enjoy being wrong. Grant describes people who take visible pleasure in discovering their mistakes, on the reasoning that every error found is one less to make. The habit depends on separating identity from opinions. People who define themselves by their beliefs experience every correction as a threat; people who define themselves by their values, such as curiosity or learning, can change beliefs without losing themselves.

Build a challenge network. The people closest to us tend to agree with us. Grant recommends deliberately gathering people who will disagree, and rewarding them when they do, as opposed to a support network that only reassures.

Argue to learn. When trying to change someone else’s mind, the instinct is to bring more and stronger arguments. Grant reviews research on skilled negotiators and debaters suggesting that fewer, better reasons work better, as does acknowledging common ground and asking genuine questions. He devotes a chapter to motivational interviewing, a method developed by the psychologists William Miller and Stephen Rollnick for helping people change behaviour by drawing out their own reasons rather than supplying them.

Make organisations safe for rethinking. In the final part of the book Grant turns to teams and institutions. Drawing on Amy Edmondson’s research on psychological safety, he argues that a culture of rethinking needs both safety, so that people can admit errors, and accountability for process, so that decisions are judged by how they were made as well as how they turned out. His examples include the loss of the Space Shuttle Columbia, where engineers’ concerns were not heard.

The idea underneath

The book’s central idea is old. Karl Popper argued that knowledge grows by conjecture and refutation, and that a theory is scientific only if it says what evidence would show it false. Grant applies that view to individual minds. Everyone holds theories about the world, their work and themselves, and the difference between people who learn and people who do not lies largely in whether they treat those theories as open to refutation.

The psychological obstacle is what Leon Festinger called cognitive dissonance: the discomfort of holding evidence that conflicts with a belief one is committed to, which people often relieve by dismissing the evidence. The site’s idea page Cognitive Dissonance and Bad Faith explores it. Grant’s practices are, in effect, ways of lowering the cost of dissonance so that the evidence can win.

There is a moral dimension too. Grant treats rethinking as a form of respect, for the truth and for the people with whom one disagrees. Treating one’s own views as provisional makes it possible to treat other people’s views seriously.

Since it was published

The book appeared early in 2021, during a period of intense polarisation, and much of its reception focused on its relevance to political argument. The research it draws on continues to develop. Some findings in this broad area, such as the strength of the Dunning-Kruger effect, have been debated by researchers who argue that part of the pattern reflects statistical artefacts. Grant’s broader claim, that the habits of scientific thinking can be learned and are valuable outside science, rests on more than any single study, and it fits the evidence from Tetlock’s work on the forecasters who update their views most often.

A first step

Write down three beliefs you hold about your work, your industry or your customers that would change what you do if they were wrong. For each, ask what evidence would convince you that it is false, and when you last looked for it. Then name one person who disagrees with you about one of them and ask them to make their case, with no aim except to understand it.