A man who could not choose
In the 1980s the neurologist Antonio Damasio and his colleague Paul Eslinger, at the University of Iowa, examined a man whom Damasio later called Elliot. A tumor had been growing in the covering of Elliot’s brain, just behind the forehead, and to save his life surgeons removed it together with some brain tissue beneath, on both sides of the front of the brain: a region called the ventromedial prefrontal cortex. Elliot recovered. His intelligence tests came back in the superior range. His memory, his language, and his logic were fine. In every way that a doctor could measure, he could think.
Yet his life fell apart. He lost his job, his business failed, and his marriage ended. What had gone wrong, Damasio concluded, was that Elliot could not decide. Damasio tells of one patient with this kind of injury who, offered two possible dates for his next appointment, spent the better part of half an hour listing reasons for and against each. He could analyze without end, and he did, but nothing in him pointed toward one answer. Elliot himself spoke about his misfortunes in a flat, detached way, as if he were reporting on someone else’s.
This is a strange fact, because we normally picture reason and feeling as opponents, with reason as the sensible driver and feeling as the passenger who grabs the wheel. Elliot had all the reason and almost none of the grabbing, and he could not drive. The question this course asks is how people really make choices, and the first answer is already surprising: not by reason alone.
The rational actor, and the doubt
Economists have long worked with a simplified person, sometimes called homo economicus. This person knows what they want, weighs every option against it, and picks the best. Economists do not mean that real people are cold or always right; “rational” has a technical sense of consistent and calculating. It is a very useful simplification for understanding markets, and it is the picture that psychologists began to test.
The first serious doubt came from within economics. In 1955 and 1956 Herbert Simon, who worked in what we would now call both psychology and economics, argued that no real chooser has the time, information, or brainpower to optimize. People, he said, satisfice, a word he built from “satisfy” and “suffice”: they search until they find an option that is good enough, and stop. Simon called this “bounded rationality,” and in 1978 he won the Nobel Memorial Prize in Economic Sciences, awarded to him alone.
The next step came from two Israeli psychologists. Daniel Kahneman and Amos Tversky began working together in 1969 at the Hebrew University of Jerusalem, and over the next decade they ran dozens of small experiments in which people judged and chose in ways that no calculating machine would. Their 1974 paper in Science and their 1979 paper on “prospect theory” are among the most cited in the social sciences. Tversky died in 1996. In 2002 Kahneman received the Nobel Memorial Prize in Economic Sciences (shared with Vernon Smith) for integrating psychology into economics; the prize is not awarded after death, which is why Tversky’s name is not on it. Kahneman himself died in 2024. Their work is the spine of this course.
Feeling is part of deciding
Damasio’s explanation for Elliot is known as the somatic marker hypothesis. In this view, when we face a choice, the brain quickly reruns bits of past experience, and the body produces a faint echo of the feeling that went with them: a flutter, a tightening, a lift. These “markers” steer attention toward some options and away from others before we have reasoned anything out. Without them, every option looks equally flat, and a mind that can compute is stuck.
The test that made the idea famous is a card game, the Iowa Gambling Task, designed by Antoine Bechara, Damasio, and colleagues and published in 1994. The player starts with a $2,000 loan of play money and picks 100 cards, one at a time, from four decks. Two decks pay $100 a card and two pay $50, but the cards also carry penalties. Decks A and B are traps: they pay well but cost more than they pay, and over ten cards a player loses $250. Decks C and D pay less and cost less, and over ten cards a player gains $250. Nobody is told this.
Healthy players drift toward C and D after a few dozen cards, and they begin to avoid A and B even before they can say why. In a 1997 follow-up, they also showed small sweat responses on the palm just before reaching for a bad deck, a bodily “warning” ahead of any words. Patients like Elliot, with damage to the same region, kept drawing from A and B, and some of them chose badly even after they could state the rule.
Another view
What the task shows, that patients with ventromedial damage choose badly, has been reproduced. What it means is disputed. A 2004 paper by Tiago Maia and James McClelland found that with more sensitive questioning, healthy players knew which decks were good about as early as they began to choose them, so no unconscious “marker” was needed to explain the choices. Others have argued that the patients’ problem is mainly failing to switch when rewards change, and that healthy players’ choices are more varied than the textbook story. The safe conclusion is that emotion-related brain regions matter for choice. Exactly how the body’s signals do their work is still open.
A brain still being built
The same region sits in the prefrontal cortex, the front part of the brain, which is also the last to finish growing. Popular accounts say the brain “matures at 25,” and the claim sounds like science. It is a simplification. Brain-imaging studies of children and young adults show prefrontal gray matter thinning through the teenage years, and the wiring around it continues to change into the late twenties; a 2022 analysis of about 100,000 brain scans put the peak of total white-matter volume at around age 29. There is no single birthday at which the brain is “done,” and the real timetable differs from person to person and from one ability to another.
The legal facts are just as untidy. In the United States a person is a legal adult at 18 in most states, but cannot buy alcohol anywhere in the country until 21. China sets adulthood at 18, and Japan lowered its age of majority from 20 to 18 in 2022. In 2005 the US Supreme Court, in Roper v. Simmons, ruled it unconstitutional to execute people for crimes committed under 18, citing differences such as impulsiveness, vulnerability to peer pressure, and a character that is still forming. Societies draw their lines according to what they are trying to protect, not because the brain has a switch.
Fast and slow
If feeling and reasoning work together, how do they share the load? A useful picture, popularized by Kahneman’s 2011 book Thinking, Fast and Slow, divides thinking into two modes. The labels “System 1” and “System 2” were introduced by the psychologists Keith Stanovich and Richard West in 2000. System 1 is fast, automatic, and effortless: it tells you that 2 + 2 is 4, recognizes a face, and senses that the driver ahead is about to swerve. System 2 is slow, deliberate, and effortful: it works out 31 × 28, checks a contract, and holds a rule in mind against temptation. System 1 runs all day and suggests answers; System 2 is called in to check them, and often just signs off.
You can feel the difference with the questions below. Answer fast, and then watch what your mind did.
Fast answer, slow answer
Five short questions. Type the first answer that comes to mind, then read the working. The first three are the “Cognitive Reflection Test” of the economist Shane Frederick.
The timer is only for you and nothing is stored. If you already knew a puzzle, your System 2 had already memorized the answer, which is itself a point about learning.
The bat-and-ball puzzle is the most famous of the three. The tempting answer is 10 cents. It arrives at once, it feels right, and it is wrong: the ball costs 5 cents. In the data Kahneman reports from Frederick’s experiments, about half of Princeton students and 56 percent of University of Michigan students gave the 10-cent answer. Frederick’s test is not about arithmetic. These people can all subtract. It is about whether you check an answer that already feels finished.
Another classic shows how easily System 1 smooths over an error. “How many animals of each kind did Moses take on the ark?” In the original 1981 study by Thomas Erickson and Mark Mattson, about half of the readers (52 percent) answered “two,” even though they had been told that some of the questions contained errors and they did know the right name. The word “Moses” fits the sentence well enough, in sound and in subject, that the mind never stops to complain. When a name that clearly did not belong, such as Nixon, was swapped in, the illusion disappeared.
Fast thinking is not the villain here. It is how we speak, drive, and read a room: across ten languages, people begin to reply within a fraction of a second after the previous speaker finishes, which leaves no time for deliberation. The fast system is usually right in the situations it has seen many times, and it is worth a great deal. The trouble is that it does not announce which situations are unfamiliar. The next activity gives a small taste of the fast system’s grip.
Name the ink
Two short rounds. In the first, name the colour of the squares. In the second, name the colour the word is printed in, and ignore what the word says. Press the button, or the key shown on it.
A rough measure: your device’s timing and your reaction speed both add noise. Compare your own two numbers, not other people’s.
Someone answers “10 cents” to the bat-and-ball puzzle. What does that most likely show?
The frugal brain
Why would a mind prefer shortcuts? The psychologists Susan Fiske and Shelley Taylor gave the tendency a name in 1984: the cognitive miser. People will spend attention on a problem only as needed and will otherwise rely on whatever is easiest. A tidy explanation is often given: the brain is about 2 percent of a person’s weight and uses about 20 percent of the body’s energy, so it must economize.
The numbers are right, but the explanation needs a repair. A brain at rest burns almost as much energy as a brain working hard on a puzzle: the extra cost of concentrating is measured in a few percent. Most of the bill is for keeping the system running. So the miserliness is not mainly about calories. It is about a bottleneck: we can attend to very little at once, and slow thinking occupies it. A mind that deliberated over every sentence you hear would fall silent.
The same economy explains the myth that we use only 10 percent of our brains. The 2014 film Lucy built a story on it, in which a drug unlocks the other 90 percent. It is false. Brain scans show activity throughout the brain over a day, and damage to almost any part causes some loss. The brain is not underused. It is already running at a high baseline, and its organizing principle is to make the repeated cheap.
That is what habits do. In a diary study led by the psychologist Wendy Wood in 2002, American college students recorded what they were doing every hour. About 43 percent of what they wrote down turned out to be habits, meaning actions repeated almost daily in the same place, and about half of the things people did were done without thinking about them. (You will sometimes see the figure “60 percent.” It is not from this research.) A habit is a decision that has been moved out of System 2 and into System 1, and that is a gift, as long as the environment doesn’t change.
The senses economize in the same way. Walk into a room that smells strongly of fish and the first minute is overwhelming; soon you hardly notice. Psychologists call this adaptation, and the Chinese tradition noticed it long ago. A passage attributed to Confucius in the 孔子家语 (Kǒngzǐ jiāyǔ) says that to spend time with good people is like entering a room of orchids, and after a while you cannot smell them; to spend time with bad people is like entering a salted-fish shop, and after a while you cannot smell that either, and you are changed by it. The point of the saying is a warning about company, and the physiology is sound: the nose, like the mind, stops reporting what stays constant.
Closer to home
The Analects records a small joke on exactly this topic. A minister named Ji Wenzi was known to think three times before acting, 三思而后行 (sān sī ér hòu xíng). When Confucius heard, he said 再，斯可矣 (zài, sī kě yǐ): “Twice is enough” (Analects 5.20). Today the phrase sān sī ér hòu xíng is used as advice to be careful, nearly the opposite of what Confucius said. English has the same pair of proverbs: “Look before you leap” and “He who hesitates is lost.” Both are true, in different situations, and that is the point of this lesson: the skill is knowing which kind of problem you are facing. Many families, Chinese and American, add a third rule for the biggest decisions: sleep on it.
A picture, not a map
The fast-and-slow picture has been enormously useful, and it is also easy to overstate. Three cautions follow.
Another view
First, the two systems are metaphors. The psychologists Mark Melnikoff and John Bargh argued in 2018 that dividing thinking into two types, one automatic and one controlled, rests on assumptions that do not hold up, because real processes mix features from both. Kahneman’s own phrase “fictitious characters” invites that caution. Second, several famous findings that were used to illustrate the picture did not survive replication. “Ego depletion,” the claim that willpower is a muscle that tires and could be refueled with sugar, was tested in 2016 by 23 laboratories with 2,141 participants and found an effect of almost exactly zero; a second, larger study in 2021 (36 laboratories, 3,531 people) found essentially the same. The sugar part of the theory failed in several studies. “Social priming,” the idea that subtle cues change behavior, as when people who read words about old age supposedly walked more slowly, failed when researchers repeated the study with automated timing (Doyen and colleagues, 2012). In 2012 Kahneman wrote an open letter to priming researchers warning of a coming problem, and in 2017 he said he had put too much faith in small, underpowered studies. “Power posing,” the claim that standing like a superhero changes hormones and risk-taking, disappeared in a larger sample (Ranehill and colleagues, 2015), and one of its original authors said in 2016 that she no longer believed in the effects.
Third, the German psychologist Gerd Gigerenzer argues that the picture mistakes shortcuts for flaws. Many shortcuts, he says, are well adapted to the world they came from, and some beat elaborate calculation. We take up that debate in Lesson 3. For now, the fair summary is this: the experiments that Kahneman and Tversky designed to catch shortcuts out are solid and have been repeated many times. The story that wraps around them, with two characters and a muscle of willpower, is a useful sketch, not a map of the brain.
What to keep
- Reasoning alone cannot make choices. Patients with damage to the front of the brain, like Elliot, can think well and still cannot decide, because feeling helps single out an option.
- Thinking comes in a fast, automatic mode and a slow, effortful one. The fast mode is usually right and always on; the slow one checks, but only if it is called in.
- The mind economizes because attention is the scarce resource, not mainly because calories are. Habits, adaptation, and shortcuts are the result, and they are mostly gifts.
- The core experiments hold up; the broader stories (two “systems,” willpower as a fuel, priming) are metaphors, and several of the claims built on them failed replication.
Think it over
Your notes are saved in this browser.
Sources and further reading
- Antonio Damasio, Descartes’ Error: Emotion, Reason, and the Human Brain, Putnam, 1994; and Paul Eslinger and Antonio Damasio, “Severe disturbance of higher cognition after bilateral frontal lobe ablation: patient EVR,” Neurology 35 (1985).
- Antoine Bechara, Antonio Damasio, Hanna Damasio, and Steven Anderson, “Insensitivity to future consequences following damage to human prefrontal cortex,” Cognition 50 (1994); Tiago Maia and James McClelland, “A reexamination of the evidence for the somatic marker hypothesis,” PNAS 101 (2004).
- Herbert Simon, “A behavioral model of rational choice,” Quarterly Journal of Economics 69 (1955).
- Daniel Kahneman, Thinking, Fast and Slow, Farrar, Straus and Giroux, 2011; Keith Stanovich and Richard West, “Individual differences in reasoning,” Behavioral and Brain Sciences 23 (2000).
- Shane Frederick, “Cognitive reflection and decision making,” Journal of Economic Perspectives 19 (2005); Thomas Erickson and Mark Mattson, “From words to meaning: a semantic illusion,” Journal of Verbal Learning and Verbal Behavior 20 (1981).
- Wendy Wood, Jeffrey Quinn, and Deborah Kashy, “Habits in everyday life,” Journal of Personality and Social Psychology 83 (2002).
- Mark Melnikoff and John Bargh, “The mythical number two,” Trends in Cognitive Sciences 22 (2018); Martin Hagger and colleagues, “A multilab preregistered replication of the ego-depletion effect,” Perspectives on Psychological Science 11 (2016).
- Analects 5.20 and Kǒngzǐ jiāyǔ, chapter “Liù běn.”