فصل 09

PART III

READING OTHERS

9

THE WOMAN WHO KNEW TOO MUCH

Her father had an explosive temper, and as a child she was always terrified that he might be about to erupt. So Katrina, as I’ll call her, learned to be hypervigilant, straining to sense the small cues—a rise in his tone of voice, the lowering of his eyebrows into a glower—that signaled he was heading toward another rampage. That emotional radar grew more sensitive as Katrina grew older. In graduate school, for example, just by reading their body language she realized that a fellow student had secretly slept with a professor. She saw how their bodies synchronized in a subtle dance. “They would shift together, move in unison,” Katrina told me. “When she wiggled, he wiggled. When I saw they were intimately attuned at the body level, like lovers, I had the thought, Oh, creepy . . . “Lovers don’t know they’re doing it, but you both become super-responsive to each other at a primal level,” she added. Only months later did the student confide the clandestine affair to Katrina, who adds, “Their affair had stopped, but their bodies were still together.” Whenever she’s with someone, Katrina says, “I’m hyperaware of dozens of streams of information people don’t usually sense—things like the lift of an eyebrow, the movement of a hand. It’s disruptive—I know way too much and it kills me. I’m overly aware.” What Katrina senses—and sometimes spills into the open—not only upsets other people; it can throw her off, too. “I came late to a meeting and made everyone wait. They were all being perfectly friendly in what they said—but what they were telling me with their bodies was not. I could see by their postures and the way they would not meet my eyes that everyone there was angry. I felt a rush of sadness and a lump in my throat. The meeting didn’t go great. “I’m always seeing things I’m not supposed to—and it’s a problem,” she added. “I poke into private stuff without meaning to. For a long time I didn’t realize I do not have to share every telling thing I know.” After getting feedback from people on her team that she was being too intrusive, Katrina began working with an executive coach. “The coach told me I have a problem leaking emotional cues—when I pick up this stuff I’m not supposed to notice, I react in a way that makes people think I’m angry all the time. So now I have to be careful about that, too.” People like Katrina are social sensitives, keenly attuned to the most minimal emotional signals, with an almost uncanny knack for reading cues so subtle that other people miss them. A slight dilation of your iris, lift of your eyebrow, or shift of your body is all they need to know how you feel. This means trouble if, like Katrina, they can’t handle such data well. But these same talents can make us socially astute, sensing when not to broach a touchy topic, when someone needs to be alone, or when people would welcome words of comfort. A trained eye for the subtle cue offers advantage in many life arenas. Take top players in sports like squash and tennis, who can sense where an opponent’s serve will land by noting subtle shifts in his posture as he positions himself to hit the ball. Many of baseball’s great hitters, like Hank Aaron, would watch films over and over of the pitchers they would face in their next game, to spot telling cues that revealed which pitch would come next. Justine Cassell, director of the Human-Computer Interaction Institute at Carnegie Mellon University, applies a similar well-trained empathy in the service of science. “Observing people was a game we played in our family,” Cassell told me. That childhood propensity was refined when as a graduate student she spent hundreds of hours studying hand movements in videos of people describing a cartoon they had just seen. Working with thirty-frames-per-second slices of the video, she’d annotate a hand’s shape as it changed, as well as the stream of shifts in its orientation, placement in space, and trajectory of movement. And to check her accuracy, she’d then work back from her notes to see if she could precisely reproduce the movement of the hand. Cassell more recently has done similar work with tiny movements of the facial muscles, with eye gaze, eyebrow raises, and head nods, all scored second by second and checked. She’s done that for hundreds of hours—and does it to this day with grad students in her lab at Carnegie Mellon. “Gestures always occur just before the most emphasized part of what you’re saying,” Cassell tells me. “One reason why some politicians may look insincere is that they have been taught to make particular gestures, but have not been taught the correct timing, and so when they produce those gestures after the word, they give us the sense that something fake is going on.” The timing of the gesture interprets its meaning. If your timing is off, a positive statement can have negative impact. Cassell gives this example: “If you say, ‘She’s a great candidate for the job’ and raise your eyebrows, nod, and emphasize the word great all at the same time, you send a very positive emotional message. But if as you say the same sentence your head nod and eyebrow raise come in the short silence after great, then it shifts the emotional meaning to sarcasm—you’re really saying she’s not all that great.” Such readings of meta-messages in nonverbal channels occur to us instantly, unconsciously, and automatically. “We cannot not make meaning of what someone tells us,” says Cassell, whether in words or just gestures, or both together. Everything we attend to in another person generates meaning at an unconscious level, and our bottom-up circuitry constantly reads it. In one study, listeners remembered having “heard” information they only saw in gesture. For example, somebody who heard “He comes out the bottom of the pipe” but saw the speaker’s hand formed into a fist and bouncing up and down said that he had heard “and then goes down stairs.”1 Cassell’s work makes visible what typically whizzes by us in microseconds. Our automatic circuitry gets the message, but our top-down awareness misses almost all of it. These hidden messages have powerful impacts. Marital researchers have long known, for instance, that if one of the partners repeatedly makes fleeting facial expressions for disgust or contempt during conflicts, the odds are great against that couple staying together.2 In psychotherapy, if the therapist and client move in synch with one another, there are likely to be better therapeutic outcomes.3 While Cassell was a professor at MIT’s Media Lab, one way she deployed this extremely precise analysis of how we express ourselves was in developing a system that guides professional animators in the art of nonverbal behavior. The system—called BEAT—allows animators to type in a segment of dialogue and get back an automatically animated cartoon person with the right gestures, head and eye movement, and posture, which they can then tweak for artistic value.4 Getting the “feel” just right of a virtual actor’s remarks, tone of voice, and gestures seems to demand a top-down grasp of bottom-up processes. These days Cassell is building similarly animated cartoons where, she says, images of children “act as virtual peers to elementary school students, using social skills to build rapport, and then using that rapport to facilitate learning.” When we met over coffee while on a break at a conference, Cassell explained how those hundreds of hours of parsing nonverbal messages have fine-tuned her sensitivity. “Now I automatically track this when I’m with anyone,” she told me—which, I confess, made me a bit self-conscious (even more so when I realized she probably noticed that, too).