Ekman, Friesen, and another colleague, Robert Levenson (who has also collaborated for years with John Gottman; psychology is a small world) decided to try to document this effect. They gathered a group of volunteers and hooked them up to monitors measuring their heart rate and body temperature—the physiological signals of such emotions as anger, sadness, and fear. Half of the volunteers were told to try to remember and relive a particularly stressful experience. The other half were simply shown how to create, on their faces, the expressions that corresponded to stressful emotions, such as anger, sadness, and fear. The second group, the people who were acting, showed the same physiological responses, the same heightened heart rate and body temperature, as the first group.

A few years later, a German team of psychologists conducted a similar study. They had a group of subjects look at cartoons, either while holding a pen between their lips—an action that made it impossible to contract either of the two major smiling muscles, the risorius and the zygomatic major—or while holding a pen clenched between their teeth, which had the opposite effect and forced them to smile. The people with the pen between their teeth found the cartoons much funnier. These findings may be hard to believe, because we take it as a given that first we experience an emotion, and then we may—or may not—express that emotion on our face. We think of the face as the residue of emotion. What this research showed, though, is that the process works in the opposite direction as well. Emotion can also start on the face. The face is not a secondary billboard for our internal feelings. It is an equal partner in the emotional process.

This critical point has enormous implications for the act of mind-reading. Early in his career, for example, Paul Ekman filmed forty psychiatric patients, including a woman named Mary, a forty-two-year-old housewife. She had attempted suicide three times, and she survived the last attempt—an overdose of pills—only because someone found her in time and rushed her to the hospital. Her grown children had left home, and her husband was inattentive, and she was depressed. When she first went to the hospital, she did nothing but sit and cry, but she seemed to respond well to therapy. After three weeks, she told her doctor that she was feeling much better and wanted a weekend pass to see her family. The doctor agreed, but just before Mary was to leave the hospital, she confessed that the real reason she wanted a weekend pass was to make another suicide attempt. Several years later, when a group of young psychiatrists asked Ekman how they could tell when suicidal patients were lying, he remembered the film taken of Mary and decided to see if it held the answer. If the face really was a reliable guide to emotion, he reasoned, shouldn’t he be able to look back at the film and see that Mary was lying when she said she was feeling better? Ekman and Friesen began to analyze the film for clues. They played it over and over for dozens of hours, examining in slow motion every gesture and expression. Finally, they saw what they were looking for: when Mary’s doctor asked her about her plans for the future, a look of utter despair flashed across her face so quickly that it was almost imperceptible.

Ekman calls that kind of fleeting look a micro expression, which is a very particular and critical kind of facial expression. Many facial expressions can be made voluntarily. If I’m trying to look stern as I give you a tongue-lashing, I’ll have no difficulty doing so, and you’ll have no difficulty interpreting my glare. But our faces are also governed by a separate, involuntary system that makes expressions that we have no conscious control over. Few of us, for instance, can voluntarily do A.U. one, the sadness sign. (A notable exception, Ekman points out, is Woody Allen, who uses his frontalis, pars medialis to create his trademark look of comic distress.) Yet we raise our inner eyebrows without thinking when we are unhappy. Watch a baby just as he or she starts to cry, and you’ll often see the frontalis, pars medialis shoot up as if it were on a string. Similarly, there is an expression that Ekman has dubbed the Duchenne smile, in honor of the nineteenth-century French neurologist Guillaume Duchenne, who first attempted to document with a camera the workings of the muscles of the face. If I were to ask you to smile, you would flex your zygomatic major. By contrast, if you were to smile spontaneously, in the presence of genuine emotion, you would not only flex your zygomatic but also tighten the orbicularis oculi, pars orbitalis, which is the muscle that encircles the eye. It is almost impossible to tighten the orbicularis oculi, pars orbitalis on demand, and it is equally difficult to stop it from tightening when we smile at something genuinely pleasurable. This kind of smile “does not obey the will,” Duchenne wrote. “Its absence unmasks the false friend.”

Whenever we experience a basic emotion, that emotion is automatically expressed by the muscles of the face. That response may linger on the face for just a fraction of a second or be detectable only if electrical sensors are attached to the face. But it’s always there. Silvan Tomkins once began a lecture by bellowing, “The face is like the penis!” What he meant was that the face has, to a large extent, a mind of its own. This doesn’t mean we have no control over our faces. We can use our voluntary muscular system to try to suppress those involuntary responses. But, often, some little part of that suppressed emotion—such as the sense that I’m really unhappy even if I deny it—leaks out. That’s what happened to Mary. Our voluntary expressive system is the way we intentionally signal our emotions. But our involuntary expressive system is in many ways even more important: it is the way we have been equipped by evolution to signal our authentic feelings.

“You must have had the experience where somebody comments on your expression and you didn’t know you were making it,” Ekman says. “Somebody asks you, ‘What are you getting upset about?’ or ‘Why are you smirking?’ You can hear your voice, but you can’t see your face. If we knew what was on our face, we would be better at concealing it. But that wouldn’t necessarily be a good thing. Imagine if there were a switch that all of us had, to turn off the expressions on our face at will. If babies had that switch, we wouldn’t know what they were feeling. They’d be in trouble. You could make an argument, if you wanted to, that the system evolved so that parents would be able to take care of kids. Or imagine if you were married to someone with a switch. It would be impossible. I don’t think mating and infatuation and friendships and closeness would occur if our faces didn’t work that way.”

Ekman slipped a tape from the O.J. Simpson trial into the VCR. It showed Kato Kaelin, Simpson’s shaggy-haired houseguest, being questioned by Marcia Clark, the lead prosecutor in the case. Kaelin sits in the witness box, with a vacant look on his face. Clark asks a hostile question. Kaelin leans forward and answers her softly. “Did you see that?” Ekman asked me. I saw nothing, just Kato being Kato—harmless and passive. Ekman stopped the tape, rewound it, and played it back in slow motion. On the screen, Kaelin moved forward to answer the question, and in that fraction of a second, his face was utterly transformed. His nose wrinkled, as he flexed his levator labii superioris alaeque nasi. His teeth were bared, his brows lowered. “It was almost totally A.U. nine,” Ekman said. “It’s disgust, with anger there as well, and the clue to that is that when your eyebrows go down, typically your eyes are not as open as they are here. The raised upper eyelid is a component of anger, not disgust. It’s very quick.” Ekman stopped the tape and played it again, peering at the screen. “You know, he looks like a snarling dog.”