Blink

ONE

The Theory of Thin Slices: How a Little Bit of Knowledge Goes a Long Way

Some years ago, a young couple came to the University of Washington to visit the laboratory of a psychologist named John Gottman. They were in their twenties, blond and blue-eyed with stylishly tousled haircuts and funky glasses. Later, some of the people who worked in the lab would say they were the kind of couple that is easy to like—intelligent and attractive and funny in a droll, ironic kind of way—and that much is immediately obvious from the videotape Gottman made of their visit. The husband, whom I'll call Bill, had an endearingly playful manner. His wife, Susan, had a sharp, deadpan wit.

They were led into a small room on the second floor of the nondescript two-story building that housed Gottman's operations, and they sat down about five feet apart on two office chairs mounted on raised platforms. They both had electrodes and sensors clipped to their fingers and ears, which measured things like their heart rate, how much they were sweating, and the temperature of their skin. Under their chairs, a "jiggle-o-meter" on the platform measured how much each of them moved around. Two video cameras, one aimed at each person, recorded everything they said and did. For fifteen minutes, they were left alone with the cameras rolling, with instructions to discuss any topic from their marriage that had become a point of contention. For Bill and Sue it was their dog. They lived in a small apartment and had just gotten a very large puppy. Bill didn't like the dog; Sue did. For fifteen minutes, they discussed what they ought to do about it.

The videotape of Bill and Sue's discussion seems, at least at first, to be a random sample of a very ordinary kind of conversation that couples have all the time. No one gets angry. There are no scenes, no breakdowns, no epiphanies. "I'm just not a dog person" is how Bill starts things off, in a perfectly reasonable tone of voice. He complains a little bit—but about the dog, not about Susan. She complains, too, but there are also moments when they simply forget that they are supposed to be arguing. When the subject of whether the dog smells comes up, for example, Bill and Sue banter back and forth happily, both with a half smile on their lips.

Sue: Sweetie! She's not smelly...

Bill: Did you smell her today?

Sue: I smelled her. She smelled good. I petted her, and my hands didn't stink or feel oily. Your hands have never smelled oily.

Bill: Yes, sir.

Sue: I've never let my dog get oily.

Bill: Yes, sir. She's a dog.

Sue: My dog has never gotten oily. You'd better be careful.

Bill: No, you'd better be careful.

Sue: No, you'd better be careful....Don't call my dog oily, boy.

1. The Love Lab

How much do you think can be learned about Sue and Bill's marriage by watching that fifteen-minute videotape? Can we tell if their relationship is healthy or unhealthy? I suspect that most of us would say that Bill and Sue's dog talk doesn't tell us much. It's much too short. Marriages are buffeted by more important things, like money and sex and children and jobs and in-laws, in constantly changing combinations. Sometimes couples are very happy together. Some days they fight. Sometimes they feel as though they could almost kill each other, but then they go on vacation and come back sounding like newlyweds. In order to "know" a couple, we feel as though we have to observe them over many weeks and months and see them in every state—happy, tired, angry, irritated, delighted, having a nervous breakdown, and so on—and not just in the relaxed and chatty mode that Bill and Sue seemed to be in. To make an accurate prediction about something as serious as the future of a marriage—indeed, to make a prediction of any sort—it seems that we would have to gather a lot of information and in as many different contexts as possible.

But John Gottman has proven that we don't have to do that at all. Since the 1980s, Gottman has brought more than three thousand married couples—just like Bill and Sue—into that small room in his "love lab" near the University of Washington campus. Each couple has been videotaped, and the results have been analyzed according to something Gottman dubbed SPAFF (for specific affect), a coding system that has twenty separate categories corresponding to every conceivable emotion that a married couple might express during a conversation. Disgust, for example, is 1, contempt is 2, anger is 7, defensiveness is 10, whining is 11, sadness is 12, stonewalling is 13, neutral is 14, and so on. Gottman has taught his staff how to read every emotional nuance in people's facial expressions and how to interpret seemingly ambiguous bits of dialogue. When they watch a marriage videotape, they assign a SPAFF code to every second of the couple's interaction, so that a fifteen-minute conflict discussion ends up being translated into a row of eighteen hundred numbers—nine hundred for the husband and nine hundred for the wife. The notation "7, 7, 14, 10, 11, 11," for instance, means that in one six-second stretch, one member of the couple was briefly angry, then neutral, had a moment of defensiveness, and then began whining. Then the data from the electrodes and sensors is factored in, so that the coders know, for example, when the husband's or the wife's heart was pounding or when his or her temperature was rising or when either of them was jiggling in his or her seat, and all of that information is fed into a complex equation.

On the basis of those calculations, Gottman has proven something remarkable. If he analyzes an hour of a husband and wife talking, he can predict with 95 percent accuracy whether that couple will still be married fifteen years later. If he watches a couple for fifteen minutes, his success rate is around 90 percent. Recently, a professor who works with Gottman named Sybil Carrère, who was playing around with some of the videotapes, trying to design a new study, discovered that if they looked at only three minutes of a couple talking, they could still predict with fairly impressive accuracy who was going to get divorced and who was going to make it. The truth of a marriage can be understood in a much shorter time than anyone ever imagined.

John Gottman is a middle-aged man with owl-like eyes, silvery hair, and a neatly trimmed beard. He is short and very charming, and when he talks about something that excites him—which is nearly all the time—his eyes light up and open even wider. During the Vietnam War, he was a conscientious objector, and there is still something of the '60s hippie about him, like the Mao cap he sometimes wears over his braided yarmulke. He is a psychologist by training, but he also studied mathematics at MIT, and the rigor and precision of mathematics clearly moves him as much as anything else. When I met Gottman, he had just published his most ambitious book, a dense five-hundred-page treatise called The Mathematics of Divorce, and he attempted to give me a sense of his argument, scribbling equations and impromptu graphs on a paper napkin until my head began to swim.

Gottman may seem to be an odd example in a book about the thoughts and decisions that bubble up from our unconscious. There's nothing instinctive about his approach. He's not making snap judgments. He's sitting down with his computer and painstakingly analyzing videotapes, second by second. His work is a classic example of conscious and deliberate thinking. But Gottman, it turns out, can teach us a great deal about a critical part of rapid cognition known as thin-slicing. "Thin-slicing" refers to the ability of our unconscious to find patterns in situations and behavior based on very narrow slices of experience. When Evelyn Harrison looked at the kouros and blurted out, "I'm sorry to hear that," she was thin-slicing; so were the Iowa gamblers when they had a stress reaction to the red decks after just ten cards.

Thin-slicing is part of what makes the unconscious so dazzling. But it's also what we find most problematic about rapid cognition. How is it possible to gather the necessary information for a sophisticated judgment in such a short time? The answer is that when our unconscious engages in thin-slicing, what we are doing is an automated, accelerated unconscious version of what Gottman does with his videotapes and equations. Can a marriage really be understood in one sitting? Yes it can, and so can lots of other seemingly complex situations. What Gottman has done is to show us how.

2. Marriage and Morse Code

I watched the videotape of Bill and Sue with Amber Tabares, a graduate student in Gottman's lab who is a trained SPAFF coder. We sat in the same room that Bill and Sue used, watching their interaction on a monitor. The conversation began with Bill. He liked their old dog, he said. He just didn't like their new dog. He didn't speak angrily or with any hostility. It seemed like he genuinely just wanted to explain his feelings.

If we listened closely, Tabares pointed out, it was clear that Bill was being very defensive. In the language of SPAFF, he was cross-complaining and engaging in "yes-but" tactics—appearing to agree but then taking it back. Bill was coded as defensive, as it turned out, for forty of the first sixty-six seconds of their conversation. As for Sue, while Bill was talking, on more than one occasion she rolled her eyes very quickly, which is a classic sign of contempt. Bill then began to talk about his objection to the pen where the dog lives. Sue replied by closing her eyes and then assuming a patronizing lecturing voice. Bill went on to say that he didn't want a fence in the living room. Sue said, "I don't want to argue about that," and rolled her eyes—another indication of contempt. "Look at that," Tabares said. "More contempt. We've barely started and we've seen him be defensive for almost the whole time, and she has rolled her eyes several times."

At no time as the conversation continued did either of them show any overt signs of hostility. Only subtle things popped up for a second or two, prompting Tabares to stop the tape and point them out. Some couples, when they fight, fight. But these two were a lot less obvious. Bill complained that the dog cut into their social life, since they always had to come home early for fear of what the dog might do to their apartment. Sue responded that that wasn't true, arguing, "If she's going to chew anything, she's going to do it in the first fifteen minutes that we're gone." Bill seemed to agree with that. He nodded lightly and said, "Yeah, I know," and then added, "I'm not saying it's rational. I just don't want to have a dog."

Tabares pointed at the videotape. "He started out with 'Yeah, I know.' But it's a yes-but. Even though he started to validate her, he went on to say that he didn't like the dog. He's really being defensive. I kept thinking, He's so nice. He's doing all this validation. But then I realized he was doing the yes-but. It's easy to be fooled by them."

Bill went on: "I'm getting way better. You've got to admit it. I'm better this week than last week, and the week before and the week before."

Tabares jumped in again. "In one study, we were watching newlyweds, and what often happened with the couples who ended up in divorce is that when one partner would ask for credit, the other spouse wouldn't give it. And with the happier couples, the spouse would hear it and say, 'You're right.' That stood out. When you nod and say 'uh-huh' or 'yeah,' you are doing that as a sign of support, and here she never does it, not once in the entire session, which none of us had realized until we did the coding.

"It's weird," she went on. "You don't get the sense that they are an unhappy couple when they come in. And when they were finished, they were instructed to watch their own discussion, and they thought the whole thing was hilarious. They seem fine, in a way. But I don't know. They haven't been married that long. They're still in the glowy phase. But the fact is that she's completely inflexible. They are arguing about dogs, but it's really about how whenever they have a disagreement, she's completely inflexible. It's one of those things that could cause a lot of long-term harm. I wonder if they'll hit the seven-year wall. Is there enough positive emotion there? Because what seems positive isn't actually positive at all."

What was Tabares looking for in the couple? On a technical level, she was measuring the amount of positive and negative emotion, because one of Gottman's findings is that for a marriage to survive, the ratio of positive to negative emotion in a given encounter has to be at least five to one. On a simpler level, though, what Tabares was looking for in that short discussion was a pattern in Bill and Sue's marriage, because a central argument in Gottman's work is that all marriages have a distinctive pattern, a kind of marital DNA, that surfaces in any kind of meaningful interaction. This is why Gottman asks couples to tell the story of how they met, because he has found that when a husband and wife recount the most important episode in their relationship, that pattern shows up right away.

"It's so easy to tell," Gottman says. "I just looked at this tape yesterday. The woman says, 'We met at a ski weekend, and he was there with a bunch of his friends, and I kind of liked him and we made a date to be together. But then he drank too much, and he went home and went to sleep, and I was waiting for him for three hours. I woke him up, and I said I don't appreciate being treated this way. You're really not a nice person. And he said, yeah, hey, I really had a lot to drink.'" There was a troubling pattern in their first interaction, and the sad truth was that that pattern persisted throughout their relationship. "It's not that hard," Gottman went on. "When I first started doing these interviews, I thought maybe we were getting these people on a crappy day. But the prediction levels are just so high, and if you do it again, you get the same pattern over and over again."

One way to understand what Gottman is saying about marriages is to use the analogy of what people in the world of Morse code call a fist. Morse code is made up of dots and dashes, each of which has its own prescribed length. But no one ever replicates those prescribed lengths perfectly. When operators send a message—particularly using the old manual machines known as the straight key or the bug—they vary the spacing or stretch out the dots and dashes or combine dots and dashes and spaces in a particular rhythm. Morse code is like speech. Everyone has a different voice.

In the Second World War, the British assembled thousands of so-called interceptors—mostly women—whose job it was to tune in every day and night to the radio broadcasts of the various divisions of the German military. The Germans were, of course, broadcasting in code, so—at least in the early part of the war—the British couldn't understand what was being said. But that didn't necessarily matter, because before long, just by listening to the cadence of the transmission, the interceptors began to pick up on the individual fists of the German operators, and by doing so, they knew something nearly as important, which was who was doing the sending. "If you listened to the same call signs over a certain period, you would begin to recognize that there were, say, three or four different operators in that unit, working on a shift system, each with his own characteristics," says Nigel West, a British military historian. "And invariably, quite apart from the text, there would be the preambles, and the illicit exchanges. How are you today? How's the girlfriend? What's the weather like in Munich? So you fill out a little card, on which you write down all that kind of information, and pretty soon you have a kind of relationship with that person."

The interceptors came up with descriptions of the fists and styles of the operators they were following. They assigned them names and assembled elaborate profiles of their personalities. After they identified the person who was sending the message, the interceptors would then locate their signal. So now they knew something more. They knew who was where. West goes on: "The interceptors had such a good handle on the transmitting characteristics of the German radio operators that they could literally follow them around Europe—wherever they were. That was extraordinarily valuable in constructing an order of battle, which is a diagram of what the individual military units in the field are doing and what their location is. If a particular radio operator was with a particular unit and transmitting from Florence, and then three weeks later you recognized that same operator, only this time he was in Linz, then you could assume that that particular unit had moved from northern Italy to the eastern front. Or you would know that a particular operator was with a tank repair unit and he always came up on the air every day at twelve o'clock. But now, after a big battle, he's coming up at twelve, four in the afternoon, and seven in the evening, so you can assume that unit has a lot of work going on. And in a moment of crisis, when someone very high up asks, 'Can you really be absolutely certain that this particular Luftwaffe Fliegerkorps [German air force squadron] is outside of Tobruk and not in Italy?' you can answer, 'Yes, that was Oscar, we are absolutely sure.'"

The key thing about fists is that they emerge naturally. Radio operators don't deliberately try to sound distinctive. They simply end up sounding distinctive, because some part of their personality appears to express itself automatically and unconsciously in the way they work the Morse code keys. The other thing about a fist is that it reveals itself in even the smallest sample of Morse code. We have to listen to only a few characters to pick out an individual's pattern. It doesn't change or disappear for stretches or show up only in certain words or phrases. That's why the British interceptors could listen to just a few bursts and say, with absolute certainty, "It's Oscar, which means that yes, his unit is now definitely outside of Tobruk." An operator's fist is stable.

What Gottman is saying is that a relationship between two people has a fist as well: a distinctive signature that arises naturally and automatically. That is why a marriage can be read and decoded so easily, because some key part of human activity—whether it is something as simple as pounding out a Morse code message or as complex as being married to someone—has an identifiable and stable pattern. Predicting divorce, like tracking Morse Code operators, is pattern recognition.

"People are in one of two states in a relationship," Gottman went on. "The first is what I call positive sentiment override, where positive emotion overrides irritability. It's like a buffer. Their spouse will do something bad, and they'll say, 'Oh, he's just in a crummy mood.' Or they can be in negative sentiment override, so that even a relatively neutral thing that a partner says gets perceived as negative. In the negative sentiment override state, people draw lasting conclusions about each other. If their spouse does something positive, it's a selfish person doing a positive thing. It's really hard to change those states, and those states determine whether when one party tries to repair things, the other party sees that as repair or hostile manipulation. For example, I'm talking with my wife, and she says, 'Will you shut up and let me finish?' In positive sentiment override, I say, 'Sorry, go ahead.' I'm not very happy, but I recognize the repair. In negative sentiment override, I say, 'To hell with you, I'm not getting a chance to finish either. You're such a bitch, you remind me of your mother.'"

As he was talking, Gottman drew a graph on a piece of paper that looked a lot like a chart of the ups and downs of the stock market over the course of a typical day. What he does, he explains, is track the ups and downs of a couple's level of positive and negative emotion, and he's found that it doesn't take very long to figure out which way the line on the graph is going. "Some go up, some go down," he says. "But once they start going down, toward negative emotion, ninety-four percent will continue going down. They start on a bad course and they can't correct it. I don't think of this as just a slice in time. It's an indication of how they view their whole relationship."

3. The Importance of Contempt

Let's dig a little deeper into the secret of Gottman's success rate. Gottman has discovered that marriages have distinctive signatures, and we can find that signature by collecting very detailed emotional information from the interaction of a couple. But there's something else that is very interesting about Gottman's system, and that is the way in which he manages to simplify the task of prediction. I hadn't realized how much of an issue this was until I tried thin-slicing couples myself. I got one of Gottman's tapes, which had on it ten three-minute clips of different couples talking. Half the couples, I was told, split up at some point in the fifteen years after their discussion was filmed. Half were still together. Could I guess which was which? I was pretty confident I could. But I was wrong. I was terrible at it. I answered five correctly, which is to say that I would have done just as well by flipping a coin.

My difficulty arose from the fact that the clips were utterly overwhelming. The husband would say something guarded. The wife would respond quietly. Some fleeting emotion would flash across her face. He would start to say something and then stop. She would scowl. He would laugh. Someone would mutter something. Someone would frown. I would rewind the tape and look at it again, and I would get still more information. I'd see a little trace of a smile, or I'd pick up on a slight change in tone. It was all too much. In my head, I was frantically trying to determine the ratios of positive emotion to negative emotion. But what counted as positive, and what counted as negative? I knew from Susan and Bill that a lot of what looked positive was actually negative. And I also knew that there were no fewer than twenty separate emotional states on the SPAFF chart. Have you ever tried to keep track of twenty different emotions simultaneously? Now, granted, I'm not a marriage counselor. But that same tape has been given to almost two hundred marital therapists, marital researchers, pastoral counselors, and graduate students in clinical psychology, as well as newlyweds, people who were recently divorced, and people who have been happily married for a long time—in other words, almost two hundred people who know a good deal more about marriage than I do—and none of them was any better than I was. The group as a whole guessed right 53.8 percent of the time, which is just above chance. The fact that there was a pattern didn't much matter. There were so many other things going on so quickly in those three minutes that we couldn't find the pattern.

Gottman, however, doesn't have this problem. He's gotten so good at thin-slicing marriages that he says he can be in a restaurant and eavesdrop on the couple one table over and get a pretty good sense of whether they need to start thinking about hiring lawyers and dividing up custody of the children. How does he do it? He has figured out that he doesn't need to pay attention to everything that happens. I was overwhelmed by the task of counting negativity, because everywhere I looked, I saw negative emotions. Gottman is far more selective. He has found that he can find out much of what he needs to know just by focusing on what he calls the Four Horsemen: defensiveness, stonewalling, criticism, and contempt. Even within the Four Horsemen, in fact, there is one emotion that he considers the most important of all: contempt. If Gottman observes one or both partners in a marriage showing contempt toward the other, he considers it the single most important sign that the marriage is in trouble.

"You would think that criticism would be the worst," Gottman says, "because criticism is a global condemnation of a person's character. Yet contempt is qualitatively different from criticism. With criticism I might say to my wife, 'You never listen, you are really selfish and insensitive.' Well, she's going to respond defensively to that. That's not very good for our problem solving and interaction. But if I speak from a superior plane, that's far more damaging, and contempt is any statement made from a higher level. A lot of the time it's an insult: 'You are a bitch. You're scum.' It's trying to put that person on a lower plane than you. It's hierarchical."

Gottman has found, in fact, that the presence of contempt in a marriage can even predict such things as how many colds a husband or a wife gets; in other words, having someone you love express contempt toward you is so stressful that it begins to affect the functioning of your immune system. "Contempt is closely related to disgust, and what disgust and contempt are about is completely rejecting and excluding someone from the community. The big gender difference with negative emotions is that women are more critical, and men are more likely to stonewall. We find that women start talking about a problem, the men get irritated and turn away, and the women get more critical, and it becomes a circle. But there isn't any gender difference when it comes to contempt. Not at all." Contempt is special. If you can measure contempt, then all of a sudden you don't need to know every detail of the couple's relationship.

I think that this is the way that our unconscious works. When we leap to a decision or have a hunch, our unconscious is doing what John Gottman does. It's sifting through the situation in front of us, throwing out all that is irrelevant while we zero in on what really matters. And the truth is that our unconscious is really good at this, to the point where thin-slicing often delivers a better answer than more deliberate and exhaustive ways of thinking.

4. The Secrets of the Bedroom

Imagine that you are considering me for a job. You've seen my résumé and think I have the necessary credentials. But you want to know whether I am the right fit for your organization. Am I a hard worker? Am I honest? Am I open to new ideas? In order to answer those questions about my personality, your boss gives you two options. The first is to meet with me twice a week for a year—to have lunch or dinner or go to a movie with me—to the point where you become one of my closest friends. (Your boss is quite demanding.) The second option is to drop by my house when I'm not there and spend half an hour or so looking around. Which would you choose?

The seemingly obvious answer is that you should take the first option: the thick slice. The more time you spend with me and the more information you gather, the better off you are. Right? I hope by now that you are at least a little bit skeptical of that approach. Sure enough, as the psychologist Samuel Gosling has shown, judging people's personalities is a really good example of how surprisingly effective thin-slicing can be.

Gosling began his experiment by doing a personality workup on eighty college students. For this, he used what is called the Big Five Inventory, a highly respected, multi-item questionnaire that measures people across five dimensions:

1. Extraversion. Are you sociable or retiring? Fun-loving or reserved?

2. Agreeableness. Are you trusting or suspicious? Helpful or uncooperative?