Put your hand in water at fourteen degrees Celsius for sixty seconds. It hurts. Later, with the other hand, do it again, but this time the cold sixty seconds are followed by thirty more in which the water warms slightly, just enough to notice. Seven minutes later you are told one trial will be repeated, and asked which. In the experiment Kahneman describes, among people who had felt the water warm, 80 percent chose the longer one.
Every second of the short trial is inside the long one, plus thirty seconds nobody needed. They knew which was longer. They chose more pain anyway, because the version with the gentler ending was the one they disliked less in memory, and memory was doing the choosing.
Kahneman explains this with two characters, not two people in your head. The experiencing self answers "how does it feel right now?" and lives in one present moment after another. The remembering self answers "how was it, on the whole?" and turns the stream into a story. The story is what you consult when you decide whether the year was a good one. One self lives the life. The other keeps the diary and, most of the time, makes the next decision.
The diary has two habits. It weighs the worst moment and the ending heavily, the peak-end rule, and barely registers how long anything lasted. He and the physician Don Redelmeier saw both in the early 1990s, before sedation for colonoscopies was routine, when 154 patients rated their pain every sixty seconds. One patient's procedure took about eight minutes and ended on a seven out of ten. Another's took about twenty-four minutes, hit the same peak, and ended near a one. If both used the scale the same way, the second suffered more by any duration-weighted count. The first remembered it as far worse.
That puts a doctor in an odd position. To spare pain, shorten the procedure. To spare a bad memory, add a gentler stretch at the end. Both are real goods, and a nicer memory cannot cancel minutes already endured. Kahneman asks only that the clinic say which one it is optimizing.
A listener whose record screeched at the end of a symphony said the noise had ruined the whole experience. It had ruined his memory of it. The forty minutes of music had already happened, and nothing could travel back and take them away.
Stretch this from minutes to years and it gets stranger. In a study by Ed Diener, people read the life of a fictional woman, Jen, who was extremely happy for thirty years or for sixty and then died instantly in an accident. Doubling her happy years did not make her life seem better. Give her five extra years that were pleasant but a little less happy, and her life was rated worse. The weaker ending lowered the whole. Kahneman saw the same reflex in an opera audience aching for Violetta's lover to reach her before she dies. A reunion of a few minutes seems to change the meaning of her entire life, while an extra good year earlier on would change nothing.
Holidays make the two selves fight openly. A quiet week at a familiar beach is better hour by hour. A grueling trek gives you the better story and the photographs, and taking the photograph can replace enjoying the view with designing a future memory. In a spring break study, students' wish to go back followed their rating after the trip, not the diary of the days. Kahneman asks what you would pay for a holiday if, at the end, every photograph were destroyed and a drug erased the memory. He did not study the answers formally, but his impression was that the price collapsed and some people would not go at all. "A great holiday" quietly means two different products.
If the story cannot be trusted to report the days, measure the days directly. Kahneman describes a method that asks people to rebuild the previous day as episodes and rate the feeling in each, summarized as the U-index, the share of waking time in which the strongest feeling was unpleasant. In one sample of a thousand American women, that index was high during the commute and at work, and low while socializing. Roughly half had no unpleasant episode all day; a small group carried most of the misery. French and American women spent similar time eating, but the French paid attention to the meal and enjoyed it more. The measure is not a smile count. Flow, the deep absorption in work, art or a game that a person does not want interrupted, counts as a good state in its own right. What fills your hours, who is with you, and what holds your attention mattered more than almost anything on a form.
Money divides the two selves neatly. In more than 450,000 American responses, everyday feeling rose with income until household income reached about 75,000 dollars in high-cost areas at the time, then stopped improving on average. Life evaluation kept rising after that. The dollar figure belongs to that study and era; the shape is the point. Money keeps buying a better verdict on your life long after it has stopped buying a better Tuesday.
The verdict is also hijacked by whatever you happen to be thinking about. Ask students about their dating life and then their happiness, and the answers move together; reverse the order and they barely relate. Students in California and the Midwest liked their climates very differently and reported the same life satisfaction, while both predicted Californians must be happier. Kahneman calls this the focusing illusion. Whatever you are attending to seems enormous. A new car fills your mind while you choose it and then disappears into the background, which is why things that keep reclaiming your attention, a weekly game, a lesson, a club, may hold their value better than purchases that stop being noticed.
Adaptation is mostly attention moving on. Asked how much of the time a paraplegic would be in a bad mood one month after the accident, people who knew one said about three quarters and people who did not said about seven tenths. A year on, the strangers still said 68 percent; the acquaintances said 41. They had watched work, friends, jokes and politics take attention back. Not everything lets go. Chronic pain, noise and severe depression keep seizing it and resist the softening.
So which record should count? Kahneman refuses to pick, and I think he is right. Lived hours need a non-negotiable place, because a policy that shortens commutes or treats pain can spare millions of unpleasant hours without moving any satisfaction ranking. Memory, meaning and chosen goals have a claim too. In a study following about 12,000 people from elite colleges in 1976, those who at seventeen or eighteen rated money as highly important earned more later and were made far happier by it. A person is more than a container of pleasant minutes. But the diary lies predictably. Keep both ledgers open, ask what an ordinary week will actually contain, and stop letting the last thirty seconds vote for the whole hour.
Three things to keep
Before a big change, ask how often you will actually notice it after month three.
Judge an experience by its hours, not only by its peak and its ending.
Keep two ledgers, lived time and remembered story, and admit when they disagree.