Talking to Strangers

by Malcolm Gladwell

  For example, when Scott Carmichael said to Ana Montes during their first meeting, “Look, Ana. I have reason to suspect that you might be involved in a counterintelligence influence operation,” she just sat there looking at him like a deer in the headlights. In retrospect, Carmichael believed that was a red flag. If she had been innocent, she would have said something—cried out, protested. But Montes? She “didn’t do a freaking thing except sit there.”

  In the moment, however, Carmichael missed that clue. Montes was uncovered only by chance, four years later. What Levine found is that we nearly always miss the crucial clues in the moment—and it puzzled him. Why? What happens at the moment someone tells a lie that specifically derails us? To find an answer, Levine went back to his tapes.

  Here is a snippet of another of the videos Levine showed me. It’s of a young woman—let’s call her Sally. Levine walked her through the straightforward questions without incident. Then came the crucial moment:

  Interviewer: Now, did any cheating occur when Rachel left the room?

  Sally: No.

  Interviewer: Are you telling me the truth?

  Sally: Yeah.

  Interviewer: When I interview your partner, I’m going to ask her the same question. What is she going to say?

  Sally pauses, looks uncertain.

  Sally: Probably…the same answer.

  Interviewer: Okay.

  The moment Levine asks the question “Did any cheating occur?” Sally’s arms and face begin to turn a bright red. Calling it an embarrassed blush doesn’t quite do it justice. Sally gives a whole new meaning to the expression “caught red-handed.” Then comes the critical question: What will your partner say? Blushing Sally can’t even come up with a convincing “She’ll agree with me.” She hems and haws and says, weakly, “Probably…the same answer.” Probably? Blushing Sally is lying, and everyone called in to judge the tape realizes she’s lying.

  Here’s the next tape Levine showed me. It’s of a woman who spent the entire interview obsessively playing with her hair. Let’s call her Nervous Nelly.

  Interviewer: Now, Rachel had to get called out of the room. Did any cheating occur when she was gone?

  Nervous Nelly: Actually my partner did want to look at the scores, and I said no—was like, “I want to see how many we got right”—because I don’t cheat. I think it’s wrong, so I didn’t. I told her no. I was like, “I don’t want to do that.” But she did say, “Well, we’ll just look at one.” I was like, “No, I don’t want to do that.” I don’t know if that was part of it or not, but no, we didn’t do that.

  Interviewer: OK, so are you telling me the truth about the cheating?

  Nervous Nelly: Yeah, we didn’t—she wanted…my partner honestly said, “We’ll just look at one.” I was like, “No, that’s not cool, I don’t want to do that.” The only thing I said was, “I’m surprised they left all the money in here.” I honestly don’t steal or cheat, I’m a good person like that. I was just kind of surprised, because normally when people leave money behind, you are going to take it—that’s just what everybody does. But no, we didn’t cheat. We didn’t steal anything.

  The twirling of the hair never stops. Nor do the halting, overly defensive, repetitive explanations, nor the fidgeting and the low-level agitation.

  Interviewer: OK, so when I call in your partner for an interview, what is she going to say to that question?

  Nervous Nelly: I think she’ll say that she wanted to look.

  Interviewer: OK.

  Nervous Nelly: If she says otherwise, then that’s not cool at all, because I said, “No, I don’t want to cheat at all.” She just said, “Why not just look at one?” She said, “Well, the answers are right there,” and I was like, “No, I’m not going to do that. That’s not who I am. It’s not what I do.”

  I was convinced Nervous Nelly was lying. You would conclude the same, if you saw her in action. Everybody thought Nervous Nelly was lying. But she wasn’t! When her partner reported back to Levine, he confirmed everything Nervous Nelly said.

  Levine found this pattern again and again. In one experiment, for instance, there was a group of interviewees whom 80 percent of the judges got wrong. And another group whom more than 80 percent got right.

  So what’s the explanation? Levine argues that this is the assumption of transparency in action. We tend to judge people’s honesty based on their demeanor. Well-spoken, confident people with a firm handshake who are friendly and engaging are seen as believable. Nervous, shifty, stammering, uncomfortable people who give windy, convoluted explanations aren’t. In a survey of attitudes toward deception conducted a few years ago, which involved thousands of people in fifty-eight countries around the world, 63 percent of those asked said the cue they most used to spot a liar was “gaze aversion.” We think liars in real life behave like liars would on Friends—telegraphing their internal states with squirming and darting eyes.

  This is—to put it mildly—nonsense. Liars don’t look away. But Levine’s point is that our stubborn belief in some set of nonverbal behaviors associated with deception explains the pattern he finds with his lying tapes. The people we all get right are the ones who match—whose level of truthfulness happens to correspond with the way they look. Blushing Sally matches. She acts like our stereotype of how a liar acts. And she also happens to be lying. That’s why we all get her right. In the Friends episode, when Monica finally breaks the news to her brother Ross about her relationship, she takes Ross’s hand and says, “I’m so sorry that you had to find out this way. I’m sorry. But it’s true, I love him too.” We believe her in that moment—that she is genuinely sorry and genuinely in love, because she’s perfectly matched. She’s being sincere and she looks sincere.

  When a liar acts like an honest person, though, or when an honest person acts like a liar, we’re flummoxed. Nervous Nelly is mismatched. She looks like she’s lying, but she’s not. She’s just nervous! In other words, human beings are not bad lie detectors. We are bad lie detectors in those situations when the person we’re judging is mismatched.

  At one point in his pursuit of Bernie Madoff, Harry Markopolos approached a seasoned financial journalist named Michael Ocrant. Markopolos persuaded Ocrant to take Madoff seriously as a potential fraud, to the point that Ocrant made an appointment to interview Madoff in person. But what happened?

  “It wasn’t so much his answers that impressed me, but rather it was his entire demeanor,” Ocrant said years later.

  It was almost impossible to sit there with him and believe he was a complete fraud. I remember thinking to myself, If [Markopolos’s team] is right and he’s running a Ponzi scheme, he’s either the best actor I’ve ever seen or a total sociopath. There wasn’t even a hint of guilt or shame or remorse. He was very low-key, almost as if he found the interview amusing. His attitude was sort of “Who in their right mind could doubt me? I can’t believe people care about this.”

  Madoff was mismatched. He was a liar with the demeanor of an honest man. And Ocrant—who knew, on an intellectual level, that something was not right—was so swayed by meeting Madoff that he dropped the story. Can you blame him? First there is default to truth, which gives the con artist a head start. But when you add mismatch to that, it’s not hard to understand why Madoff fooled so many for so long.

  And why did so many of the British politicians who met with Hitler misread him so badly? Because Hitler was mismatched as well. Remember Chamberlain’s remark about how Hitler greeted him with a double-handed handshake, which Chamberlain believed Hitler reserved for people he liked and trusted? For many of us, a warm and enthusiastic handshake does mean that we feel warm and enthusiastic about the person we’re meeting. But not Hitler. He’s the dishonest person who acts honest.1

  3.

  So what was Amanda Knox’s problem? She was mismatched. She’s the innocent person who acts guilty. She’s Nervous Nelly.

  Knox was—to those who did not know her—confusing. At the time of the crime she was tw
enty and beautiful, with high cheekbones and striking blue eyes. Her nickname was “Foxy Knoxy.” The tabloids got hold of a list she had made of all the men she’d slept with. She was the femme fatale—brazen and sexual. The day after her roommate’s brutal murder, she was spotted buying red underwear at a lingerie shop with her boyfriend.

  In fact, the “Foxy Knoxy” nickname had nothing to do with sex. It was bestowed on her at age thirteen by soccer teammates for the deft way she moved the ball up and down the field. She was buying red underwear a few days after her roommate’s murder because her house was a crime scene and she couldn’t get access to her clothes. She wasn’t a femme fatale.2 She was an immature young woman only a few years removed from an awkward and pimply adolescence. Brazen and sexual? Amanda Knox was actually a bit of a misfit.

  “I was the quirky kid who hung out with the sulky manga-readers, the ostracized gay kids, and the theater geeks,” she writes in her memoir, published in 2011 after she was finally released from an Italian prison.

  In high school she was the middle-class kid on financial aid, surrounded by well-heeled classmates. “I took Japanese and sang, loudly, in the halls while walking from one class to another. Since I didn’t really fit in, I acted like myself, which pretty much made sure I never did.”

  Matched people conform with our expectations. Their intentions are consistent with their behavior. The mismatched are confusing and unpredictable: “I’d do things that would embarrass most teenagers and adults—walking down the street like an Egyptian or an elephant—but that kids found fall-over hilarious.”

  Kercher’s murder changed the way Kercher’s circle of friends behaved. They wept quietly, hushed their voices, murmured their sympathies. Knox didn’t.

  Just listen to a handful of quotations that I’ve taken—at random—from the British journalist John Follain’s Death in Perugia. Believe me, there are more like this. Here is Follain describing what happened when Kercher’s friends met up with Knox and Sollecito at the police station the day after the murder.

  “Oh Amanda. I’m so sorry!” Sophie exclaimed, as she instinctively put her arms around her and gave her a bear hug.

  Amanda didn’t hug Sophie back. Instead, she stiffened, holding her arms down by her sides. Amanda said nothing.

  Surprised, Sophie let go of her after a couple of seconds and stepped back. There was no trace of emotion on Amanda’s face. Raffaele walked up to Amanda and took hold of her hand; the couple just stood there, ignoring Sophie and gazing at each other.

  Then:

  Amanda sat with her feet resting on Raffaele’s lap…the two caressed and kissed each other; sometimes they’d even laugh.

  How could Amanda act like that? Sophie asked herself. Doesn’t she care?

  Then:

  Most of Meredith’s friends were in tears or looked devastated, but Amanda and Raffaele made smacking noises with their lips when they kissed or sent kisses to each other.

  And then:

  “Let’s hope she didn’t suffer,” Natalie said.

  “What do you think? They cut her throat, Natalie. She fucking bled to death!” Amanda retorted.

  Amanda’s words chilled Natalie; she was surprised both by Amanda talking of several killers, and by the coldness of her tone. Natalie thought it was as if Meredith’s death didn’t concern her.

  In an interview with Knox, Diane Sawyer of ABC News brought up that last exchange in the police station, where Knox snapped at Kercher’s friend and said, “She fucking bled to death.”

  Knox: Yeah. I was angry. I was pacing, thinking about what Meredith must have been through.

  Sawyer: Sorry about that now?

  Knox: I wish I could’ve been more mature about it, yeah.

  In a situation that typically calls for a sympathetic response, Knox was loud and angry. The interview continues:

  Sawyer: You can see that this does not look like grief. Does not read as grief.

  The interview was conducted long after the miscarriage of justice in the Kercher case had become obvious. Knox had just been freed after spending four years in an Italian prison for the crime of not behaving the way we think people are supposed to behave after their roommate is murdered. Yet what does Diane Sawyer say to her? She scolds her for not behaving the way we think people are supposed to behave after their roommate is murdered.

  In the introduction to the interview, the news anchor says that Knox’s case remains controversial because, in part, “her pleas for innocence seemed to many people more cold and calculating than remorseful”—which is an even more bizarre thing to say, isn’t it? Why would we expect Knox to be remorseful? We expect remorse from the guilty. Knox didn’t do anything. But she’s still being criticized for being “cold and calculating.” At every turn, Knox cannot escape censure for her weirdness.

  Knox: I think everyone’s reaction to something horrible is different.

  She’s right! Why can’t someone be angry in response to a murder, rather than sad? If you were Amanda Knox’s friend, none of this would surprise you. You would have seen Knox walking down the street like an elephant. But with strangers, we’re intolerant of emotional responses that fall outside expectations.

  While waiting to be interviewed by police, four days after Kercher’s body was discovered, Knox decided to stretch. She’d been sitting, slumped, for hours. She touched her toes, held her arms over her head. The policeman on duty said to her, “You seem really flexible.”

  I replied, “I used to do a lot of yoga.” He said, “Can you show me? What else can you do?” I took a few steps toward the elevator and did a split. It felt good to know I still could. While I was on the floor, legs splayed, the elevator doors opened. Rita Ficarra, the cop who had reprimanded Raffaele and me about kissing the day before, stepped out. “What are you doing?” she demanded, her voice full of contempt.3

  The lead investigator in the case, Edgardo Giobbi, says he had doubts about Knox from the moment she walked with him through the crime scene. As she put on protective booties, she swiveled her hips and said, “Ta-dah.”

  “We were able to establish guilt,” Giobbi said, “by closely observing the suspect’s psychological and behavioral reaction during the interrogation. We don’t need to rely on other kinds of investigation.”

  The prosecutor in the case, Giuliano Mignini, brushed off the mounting criticisms of the way his office had handled the murder. Why did people focus so much on the botched DNA analysis? “Every piece of proof has aspects of uncertainty,” he said. The real issue was mismatched Amanda. “I have to remind you that her behavior was completely inexplicable. Totally irrational. There’s no doubt of this.”4

  From Bernard Madoff to Amanda Knox, we do not do well with the mismatched.

  4.

  The most disturbing of Tim Levine’s findings was when he showed his lying videotapes to a group of seasoned law-enforcement agents—people with fifteen years or more of interrogation experience. He had previously used as his judges students and adults from ordinary walks of life. They didn’t do well, but perhaps that’s to be expected. If you are a real-estate agent or a philosophy major, identifying deception in an interrogation isn’t necessarily something you do every day. But maybe, he thought, people whose job it was to do exactly the kind of thing he was measuring would be better.

  In one respect, they were. On “matched” senders, the seasoned interrogators were perfect. You or I would probably come in at 70 or 75 percent on that set of tapes. But everyone in Levine’s group of highly experienced experts got every matched sender right. On mismatched senders, however, their performance was abysmal: they got 20 percent right. And on the subcategory of sincere-acting liars, they came in at 14 percent—a score so low that it ought to give chills to anyone who ever gets hauled into an interrogation room with an FBI agent. When they are confronted with Blushing Sally—the easy case—they are flawless. But when it comes to the Amanda Knoxes and Bernie Madoffs of the world, they are hapless.

  This is distress
ing because we don’t need law-enforcement experts to help us with matched strangers. We’re all good at knowing when these kinds of people are misleading us or telling us the truth. We need help with mismatched strangers—the difficult cases. A trained interrogator ought to be adept at getting beneath the confusing signals of demeanor, at understanding that when Nervous Nelly overexplains and gets defensive, that’s who she is—someone who overexplains and gets defensive. The police officer ought to be the person who sees the quirky, inappropriate girl in a culture far different from her own say “Ta-dah” and realize that she’s just a quirky girl in a culture far different from her own. But that’s not what we get. Instead, the people charged with making determinations of innocence and guilt seem to be as bad as or even worse than the rest of us when it comes to the hardest cases.

  Is this part of the reason for wrongful convictions? Is the legal system constitutionally incapable of delivering justice to the mismatched? When a judge makes a bail decision and badly underperforms a computer, is this why? Are we sending perfectly harmless people to prison while they await trial simply because they don’t look right? We all accept the flaws and inaccuracies of institutional judgment when we believe that those mistakes are random. But Tim Levine’s research suggests that they aren’t random—that we have built a world that systematically discriminates against a class of people who, through no fault of their own, violate our ridiculous ideas about transparency. The Amanda Knox story deserves to be retold not because it was a once-in-a-lifetime crime saga—a beautiful woman, a picturesque Italian hilltop town, a gruesome murder. It deserves retelling because it happens all the time.