Social Intelligence: The New Science of Human Relationships

by Daniel Goleman

  Studies in Sweden of workers at different levels and in the United Kingdom among civil servants show that people in the lower positions in an organization are four times more likely to develop cardiovascular disease than are those in the top rungs, who don’t have to put up with the whims of bosses such as themselves.14 Workers who feel unfairly criticized, or whose boss will not listen to their problems, have a rate of coronary heart disease 30 percent higher than those who feel treated fairly.15

  In rigid hierarchies bosses tend to be authoritarian: they more freely express contempt for their subordinates, who in turn naturally feel a messy mix of hostility, fear, and insecurity.16 Insults, which can be routine with such authoritarian managers, serve to reaffirm the boss’s power while leaving their subordinates feeling helpless and vulnerable.17 And because their salary and very job security depend on the boss, workers tend to obsess over their interactions, reading even mildly negative exchanges as ominous. Indeed, across the board, just about any conversation with someone of higher status at work elevates a person’s blood pressure more than does a similar conversation with a coworker.18

  Take how one handles an insult. In a relationship among peers, an affront can be challenged, an apology asked for. But when the insult comes from someone who holds all the power, subordinates (perhaps wisely) suppress their anger, responding with a resigned tolerance. But that very passivity—with the insult going unchallenged—tacitly confers permission to a superior to continue in that vein.

  People who respond to insults with silence experience significant hikes in blood pressure. As the demeaning messages continue over time, the person holding back feels increasingly powerless, anxious, and ultimately depressed—all of which, if prolonged over long periods, markedly increases the likelihood of cardiovascular disease.19

  In one study a hundred men and women wore devices that took readings of their blood pressure whenever they interacted with someone.20 When they were with family or enjoyable friends, their blood pressure fell; these interactions were pleasant and soothing. When they were with someone who was troublesome, there was a rise. But the biggest jump came while they were with people they felt ambivalent about: an overbearing parent, a volatile romantic partner, a competitive friend. A mercurial boss looms as the archetype, but this dynamic operates in all our relationships.

  We try to steer clear of people we find unpleasant, but many unavoidable people in our lives fall into this “mixed” category: sometimes they make us feel good, and other times terrible. Ambivalent relationships put an emotional demand on us; each interaction is unpredictable, perhaps potentially explosive, and so requires a heightened vigilance and effort.

  Medical science has pinpointed a biological mechanism that directly links a toxic relationship to heart disease. Volunteers for an experiment on stress had to defend themselves against a false accusation that they had shoplifted.21 As they talked, their immune and cardiovascular systems mobilized in a potentially deadly combination. The immune system secreted T lymphocytes, while the walls of blood vessels emitted a substance that binds to those T cells, setting in motion the formation of artery-clogging plaque on the endothelium.22

  Most surprising medically was that even relatively minor upsets seemed to trigger this mechanism. Presumably, this distress-to-endothelium chain reaction would put us at risk for heart disease if such stressful encounters become routine events in our daily lives.

  THE CAUSAL CHAIN

  It’s all very well to find a general correlation between stressful relationships and poor health, and to identify a pathway or two in a possible causal chain. But despite the occasional studies suggesting biological mechanisms, skeptics often argue that very different factors may be at play. For instance, if a difficult relationship leads someone to drink or smoke too much or to sleep poorly, that might be a more immediate cause of ill health. So researchers have continued to look for a distinct biological link—one clearly separable from these other reasons.

  Enter Sheldon Cohen, a psychologist at Carnegie Mellon University who has intentionally given colds to hundreds of people.23 Not that Cohen has a malicious streak—it’s all in the interest of science. Under meticulously controlled conditions, he systematically exposes volunteers to a rhinovirus that causes the common cold. About a third of people exposed to the virus develop the full panoply of symptoms, while the rest walk away with nary a sniffle. The controlled conditions allow him to determine why.

  His methods are exacting. Cohen’s experimental volunteers are quarantined for twenty-four hours before they are exposed, to be sure they have not picked up a cold elsewhere. For the next five days (and for $800) the volunteers are housed in a special unit with other volunteers, all of whom are kept at least three feet from one another, lest they reinfect someone.

  During those five days their nasal secretions are tested for technical indicators of colds (like the total weight of their mucus) as well as the presence of the specific rhinovirus, and their blood samples are tested for antibodies. This way Cohen takes the measure of the cold with a precision that goes far beyond counting runny noses and sneezes.

  We know that low levels of vitamin C, smoking, and sleeping poorly all increase the likelihood of infection. The question is, can a stressful relationship be added to that list? Cohen’s answer: definitely.

  Cohen assigns precise numerical values to the factors that make one person come down with a cold while another stays healthy. Those with an ongoing personal conflict were 2.5 times as likely as the others to get a cold, putting rocky relationships in the same causal range as vitamin C deficiency and poor sleep. (Smoking, the most damaging unhealthy habit, made people three times more likely to succumb.) Conflicts that lasted a month or longer boosted susceptibility, but an occasional argument presented no health hazard.24

  While perpetual arguments are bad for our health, isolating ourselves is worse. Compared to those with a rich web of social connections, those with the fewest close relationships were 4.2 times more likely to come down with the cold, making loneliness riskier than smoking.

  The more we socialize, the less susceptible to colds we become. This idea seems counterintuitive: don’t we increase the likelihood of being exposed to a cold virus the more people we interact with? Sure. But vibrant social connections boost our good moods and limit our negative ones, suppressing cortisol and enhancing immune function under stress.25 Relationships themselves seem to protect us from the risk of exposure to the very cold virus they pose.

  THE PERCEPTION OF MALICE

  Elysa Yanowitz is not alone in the indignities she suffered at work. A woman working at a pharmaceutical company sends me this e-mail: “I’m having personality clashes with my boss, who is not a very nice person. For the first time in my professional career, my confidence is shaken—and since she’s friends with all of the upper hierarchy at my company, I feel I have no recourse. The whole thing is making me physically ill from the stress.”

  Is she merely imagining the link between her toxic boss and her physical illness? Perhaps.

  On the other hand, her plight fits well with findings from an analysis of 208 studies involving 6,153 individuals who were subjected to stressors ranging from loud, obnoxious noises to confrontations with equally obnoxious people.26 Of all the sorts of stress, the worst by far was when someone was the target of harsh criticism and was helpless to do anything about it—like both Yanowitz and the pharmaceutical employee who clashed with her boss.

  Why this should be so has been revealed by Margaret Kemeny, an expert in behavioral medicine at the University of California Medical School in San Francisco, who analyzed hundreds of stress studies with her colleague Sally Dickerson. Threats or challenges, Kemeny told me, are most stressful “when you have an audience and feel you are being judged.”

  Stress reactions in all the studies were gauged by the rise in a person’s level of cortisol.27 The largest spikes in cortisol occurred when the source of stress was interpersonal—for example, when someone sat in ju
dgment on a volunteer who had to subtract the number 17 from 1,242 aloud then continue to subtract 17 from the resulting number, as rapidly as possible. When a person performed such an onerous task while being judged on how they did, the effect on cortisol was about three times greater than when the stress was comparable but impersonal.28

  Imagine, for example, that you’re being interviewed for a job. As you talk about the talents and expertise that you believe qualify you, something unsettling happens. You see the interviewer responding to you with a stony, unsmiling expression, coolly making notes on a clipboard. Then, to make things worse, the interviewer makes critical remarks, belittling your skills.

  That was the nerve-wracking predicament of volunteers for a devilish measure of social stress, all of whom were actually in the midst of applying for a job and had come for a practice interview. But the “practice sessions” were actually a stress test. Developed by researchers in Germany, this experimental ordeal has been used in labs around the world because it produces powerful data. Kemeny’s lab routinely has used a variation of the test to assess the biological punch of social stress.

  Dickerson and Kemeny argue that being evaluated threatens the “social self,” the ways we see ourselves through others’ eyes. This sense of our social value and status—and so our very self-worth—comes from the cumulative messages we get from others about how they perceive us. Such threats to our standing in the eyes of others are remarkably potent biologically, almost as powerful as those to our very survival. After all, the unconscious equation goes, if we are judged to be undesirable, we may not only be shamed, but suffer complete rejection.29

  An interviewer’s unnerving, hostile reaction reliably triggers the HPA axis to produce some of the highest levels of cortisol of any laboratory stress simulation ever tested. The social stress test hikes cortisol much more than does that classic lab ordeal, in which volunteers do increasingly difficult math problems under intense time pressure against annoying background noise, with a noxious buzzer signaling wrong answers—but without the presence of someone making nasty judgments.30 Impersonal ordeals are soon forgotten, but judgmental scrutiny delivers a particularly strong—and lingering—dose of shame.31

  Astonishingly, a symbolic judge existing only in the mind delivers just as large a dose of angst. A virtual audience can affect the HPA system as powerfully as a live one, Kemeny explains, because “the moment you think of something, you create an internal representation, which in turn acts on the brain” much as would the reality it represents.

  Feeling helpless adds to the stress. In the cortisol studies analyzed by Dickerson and Kemeny, threats were perceived as all the worse when they were beyond the person’s ability to do anything about them. When a threat persists no matter what efforts we might make, the cortisol rise magnifies. This parallels the predicament of someone, for example, who finds himself the target of vicious prejudice—or those two beleaguered women whose bosses turned against them. Relationships that are continually critical, rejecting, or harassing keep the HPA axis in constant overdrive.

  When the source of stress seems impersonal, like an obnoxious auto alarm we are helpless to stop, our most basic need for acceptance and belonging goes unthreatened. Kemeny found that for such impersonal stress, the body got over its inevitable jump in cortisol within forty minutes or so. But if the cause was a negative social judgment, cortisol stayed high 50 percent longer, taking an hour or more to return to normal.

  Brain imaging studies suggest which parts of the brain may react so strongly to such perception of malice. Recall from Chapter 5 the computer simulation in Jonathan Cohen’s lab at Princeton, where volunteers in an MRI scanner played the Ultimatum Game. The game requires two partners to divide some money, with one making offers of a split that the other can accept or reject.

  When a volunteer felt that the other person had made them an unfair offer, their brain showed activity in the anterior insula, which is known to activate during feelings of anger and disgust. Fittingly, they showed signs of bitterness and were more likely to reject not just this offer but the next one too, whatever it might be. Yet when they believed that the other “partner” in the game was just a computer program, their insula stayed quiet, no matter how seemingly one-sided the offer. The social brain makes a crucial distinction between accidental and intentional harm, and it reacts more strongly if it seems malevolent.

  This finding may solve a puzzle for clinicians attempting to understand post-traumatic stress disorder (PTSD): why catastrophes of similar intensity more often lead to lasting suffering if the person feels their trauma was purposely inflicted by someone else rather than a random act of Nature. Hurricanes, earthquakes, and other natural catastrophes leave many fewer victims of PTSD than do malicious acts like rape and physical abuse. The aftereffects of trauma, like all stress, are worse the more personally targeted the victim feels.

  THE CLASS OF ’57

  It was in 1957 that Elvis Presley broke into America’s national consciousness by appearing on the Sunday night Ed Sullivan Show, then the most-watched hour on television. The American economy was in the midst of a lengthy postwar boom, Dwight D. Eisenhower was president, cars had grotesque tail fins, and teens socialized at heavily chaperoned school dances called “sock hops.”

  In that year researchers from the University of Wisconsin began to study about ten thousand graduating high school seniors, nearly a third of those in the entire state. These teens were later reinter-viewed as they reached age forty, and again as they reached their mid-fifties. Then, as they approached sixty-five, a group of the grads were recruited for follow-up research by Richard Davidson at the University of Wisconsin and brought to the W. M. Keck Laboratory for Functional Brain Imaging and Behavior. Using measures far more sophisticated than anything available in 1957, Davidson undertook to correlate their social history, brain activity, and immune function.

  The quality of the grads’ relationships over the course of their lives had been discerned in the earlier interviews. Now it was compared to the wear and tear on their bodies. The grads were assessed on the chronic activity of systems that fluctuate as they handle stress, including blood pressure, cholesterol, and levels for cortisol and other stress hormones. The sum of these and similar measures predicts not just the likelihood of cardiovascular disease but also late-life declines in mental and physical functioning. A very high total score predicts an earlier death.32 The researchers found that relationships mattered: there was a strong association between having a high-risk physical profile and an unfavorable cumulative emotional tone in the grads’ most important life relationships.33

  Take, for example, the anonymous Class of 1957 graduate I’ll call Jane. Her relationship life had been tough, a litany of disappointments. Both Jane’s parents were alcoholics, and she saw little of her father during most of her childhood. He molested her while she was in high school. As an adult, she was extremely fearful of people, alternately angry at and anxious with those closest to her. Though Jane married, she soon divorced, and her scant social life offered her little solace. In the medical survey for Davidson’s study, she had nine out of twenty-two common medical symptoms.

  On the other hand, Jill, one of Jane’s high school classmates, was the picture of a full and rich relationship history. Although Jill’s father had died when she was just nine, she felt her mother had been extremely caring. Jill was close to her husband and her four sons, and she felt her family life was extremely satisfying. So too was her active social life, filled as it was with many close friends and confidants. And in her sixties Jill reported having problems with only three of the twenty-two symptoms.

  Again, correlation is not causation. To demonstrate a causal link between relationship quality and health includes identifying the specific biological mechanisms at work. Here the Class of 1957 provided some telling clues, based on Davidson’s tests of brain activity.

  Jill, the woman with the caring mother, satisfying relationships, and very few medical complaints
in her sixties, was the member of the Class of ’57 with the greatest activity in her left prefrontal cortex relative to the right. That brain activity pattern, Davidson has found, suggests that Jill’s days were filled with mostly pleasant moods.

  Jane, whose parents had been alcoholics and who was divorced with many medical problems in her sixties, had the opposite brain pattern. She had the highest activity in her right prefrontal area relative to the left of anyone studied in her class. That pattern suggests that Jane more often reacted to life with intense distress and recovered slowly from her emotional setbacks.

  The left prefrontal area, as Davidson had found in earlier research, regulates a cascade of circuitry in lower brain areas that determine our recovery time from distress—that is, our resilience.34 The more of this left prefrontal activity (relative to the right side), the better we are at developing cognitive strategies for emotional regulation and the faster our emotional recovery. That in turn determines how quickly cortisol returns to normal.

  Resilient health depends in part on how well the high road has learned to manage the low.

  Davidson’s earlier study went one step further. His research group discovered that activity in the same left prefrontal area correlated highly with the ability of a person’s immune system to respond to a flu shot. Those with the highest activation had immune systems that mobilized flu antibodies three times more than did others.35 Davidson believes that these differences are clinically significant—in other words, that those with high left prefrontal activity are less likely to get the flu if exposed to the virus.

  Davidson sees in such data a window on the anatomy of resilience. A soundly secure relationship history, he theorizes, gives people the inner resources to bounce back from emotional setbacks and losses—as was the case with Jill, the woman who lost her father at age nine but whose mother was so loving.