The Better Angels of Our Nature: Why Violence Has Declined

by Steven Pinker

  Since the only commodity at stake in contests of dominance is information, once the point has been made about who’s the boss, the violence can come to an end without setting off rounds of vendetta. The primatologist Frans de Waal discovered that in most primate species, after two animals have fought, they will reconcile.105 They may touch hands, kiss, embrace, and in the case of bonobos, have sex. This makes one wonder why they bother to fight in the first place if they were just going to make up afterward, and why they make up afterward if they had reason to fight. The reason is that reconciliation occurs only among primates whose long-term interests are bound together. The ties that bind may be genetic relatedness, collective defense against predators, being in cahoots against a third party, or, in an experiment, getting fed only if they work together.106 The overlap of interests is not perfect, so they still have reason to fight for dominance or retaliation within the group, but it is not zero, so they cannot afford to smack each other around indefinitely, let alone kill each other. Among primates whose interests are not bound up in any of these ways, adversaries are unforgiving, and violence is likelier to escalate. Chimpanzees, for example, reconcile after a fight within their community, but they never reconcile after a battle or raid with members of a different community.107 As we shall see in the next chapter, reconciliation among humans is also governed by the perception of common interests.

  The metaphor of competitive distance urination suggests that the gender with the equipment best suited to competing is the gender that is most likely to participate in contests of dominance. Though in many primate species, including humans, both sexes jockey for preeminence, usually against members of their own sex, it seems to loom larger in the minds of men than in the minds of women, taking on a mystical status as a priceless commodity worth almost any sacrifice. Surveys of personal values in men and women find that the men assign a lopsided value to professional status compared to all the other pleasures of life.108 They take greater risks, and they show more confidence and more overconfidence.109 Most labor economists consider these sex differences to be a contributor to the gender gap in earnings and professional success.110

  And men are, of course, by far the more violent sex. Though the exact ratios vary, in every society it is the males more than the females who play-fight, bully, fight for real, carry weapons, enjoy violent entertainment, fantasize about killing, kill for real, rape, start wars, and fight in wars.111 Not only is the direction of the sex difference universal, but the first domino is almost certainly biological. The difference is found in most other primates, emerges in toddlerhood, and may be seen in boys who (because of anomalous genitalia) are secretly raised as girls.112

  We have already seen why the sex difference evolved: mammalian males can reproduce more quickly than females, so they compete for sexual opportunities, while females tilt their priorities toward ensuring the survival of themselves and their offspring. Men have more to gain in violent competition, and also less to lose, because fatherless children are more likely to survive than motherless ones. That does not mean that women avoid violence altogether—Chuck Berry speculated that Venus de Milo lost both her arms in a wrestling match over a brown-eyed handsome man—but they find it less appealing. Women’s competitive tactics consist in less physically perilous relational aggression such as gossip and ostracism.113

  In theory, violent competition for mates and violent competition for dominance needn’t go together. One doesn’t have to invoke dominance to explain why Genghis Khan inseminated so many women that his Y chromosome is common in Central Asia today; it’s enough to observe that he killed the women’s fathers and husbands. But given that social primates regulate violence by deferring to dominant individuals, dominance and mating success in practice went hand in hand during most of our species’ history. In nonstate societies, dominant men have more wives, more girlfriends, and more affairs with other men’s wives.114 In the six earliest empires, the correlation between status and mating success can be quantified precisely. Laura Betzig found that emperors often had thousands of wives and concubines, princes had hundreds, noblemen had dozens, upper-class men had up to a dozen, and middle-class men had three or four.115 (It follows mathematically that many lower-class men had none—and thus a strong incentive to fight their way out of the lower class.) Recently, with the advent of reliable contraception and the demographic transition, the correlation has been weakened. But wealth, power, and professional success still increase a man’s sex appeal, and the most visible clue to physical dominance—height—still gives a man an edge in economic, political, and romantic competition.116

  Whereas instrumental violence deploys the seeking and calculating parts of the brain, dominance deploys the system that Panksepp calls Intermale Aggression. It really should be called Intrasexual Competition, because it is found in women too, and the human habit of male parental investment means that women as well as men have an evolutionary incentive to compete for mates. Still, at least one part of the circuit, a nucleus in the anterior preoptic portion of the hypothalamus, is twice as large in men as it is in women.117 And the entire system is studded with receptors for testosterone, which is about five to ten times more plentiful in the bloodstream of men than of women. The hypothalamus, recall, controls the pituitary gland, which can secrete a hormone that tells the testes or the adrenal glands to produce more testosterone.

  Though testosterone is often identified in the popular imagination as the cause of male pugnacity—“the substance that drives men to behave with quintessential guyness, to posture, push, yelp, belch, punch and play air-guitar,” as the journalist Natalie Angier put it—biologists have been nervous about blaming it for male aggression itself.118 Raising testosterone undoubtedly makes most birds and mammals more obstreperous, and lowering it makes them less so, as the owner of any neutered dog or cat is aware. But in humans the effects are less easily measured, for a number of boring biochemical reasons, and they are less directly tied to aggression, for an interesting psychological reason.

  Testosterone, according to scientists’ best guess, does not make men more aggressive across the board, but prepares them for a challenge of dominance.119 In chimpanzees, testosterone goes up in the presence of a sexually receptive female, and it is correlated with the male’s dominance rank, which in turn is correlated with his aggressiveness. In men, testosterone levels rise in the presence of an attractive female and in anticipation of competition with other men, such as in sports. Once a match has begun, testosterone rises even more, and when the match has been decided, testosterone continues to rise in the winner but not in the loser. Men who have higher levels of testosterone play more aggressively, have angrier faces during competition, smile less often, and have firmer handshakes. In experiments they are more likely to lock their gaze onto an angry face, and to perceive a neutral face as angry. It’s not just fun and games that pump up the hormone: recall that the southern men who were insulted in Richard Nisbett’s experiment on the psychology of honor responded with a rise in testosterone, and that they looked angrier, shook hands more firmly, and walked out of the lab with more of a swagger. At the tail end of the belligerence spectrum, prisoners with higher levels of testosterone have been found to commit more acts of violence.

  Testosterone rises in adolescence and young adulthood, and declines in middle age. It also declines when men get married, have children, and spend time with their children. The hormone, then, is an internal regulator of the fundamental tradeoff between parenting effort and mating effort, where mating effort consists both in wooing the opposite sex and in fending off rivals of the same sex.120 Testosterone may be the knob that turns men into dads or cads.

  The rise and fall of testosterone over the life span correlates, more or less, with the rise and fall of male pugnacity. Incidentally, the first law of violence—it’s something that young men do—is easier to document than to explain. Though it’s clear why men should have evolved to be more violent than women, it’s not so clear why young men shoul
d be more violent than old men. After all, young men have more years ahead of them, so when they take up a violent challenge, they are gambling with a greater proportion of their unlived lives. On mathematical grounds one might expect the opposite: that as men’s days are numbered, they can afford to become increasingly reckless, and a really old man might go on one last spree of rape and murder until a SWAT team cuts him down.121 One reason this does not happen is that men always have the option of investing in their children, grandchildren, nieces, and nephews, so older men, who are physically weaker but socially and economically stronger, have more to gain in providing for and protecting their families than in siring more offspring.122 The other is that dominance in humans is a matter of reputation, which can be self-sustaining with a long payout. Everyone loves a winner, and nothing succeeds like success. So it is in the earliest rounds of competition that the reputational stakes are highest.

  Testosterone, then, prepares men (and to a lesser extent women) for contests of dominance. It doesn’t cause violence directly, because many kinds of violence have nothing to do with dominance, and because many contests of dominance are settled by displays and brinkmanship rather than violence itself. But to the extent that the problem of violence is a problem of young, unmarried, lawless men competing for dominance, whether directly or on behalf of a leader, then violence really is a problem of there being too much testosterone in the world.

  The socially constructed nature of dominance can help explain which individuals are most likely to take risks to defend it. Perhaps the most extraordinary popular delusion about violence of the past quarter-century is that it is caused by low self-esteem. That theory has been endorsed by dozens of prominent experts, has inspired school programs designed to get kids to feel better about themselves, and in the late 1980s led the California legislature to form a Task Force to Promote Self-Esteem. Yet Baumeister has shown that the theory could not be more spectacularly, hilariously, achingly wrong. Violence is a problem not of too little self-esteem but of too much, particularly when it is unearned.123 Self-esteem can be measured, and surveys show that it is the psychopaths, street toughs, bullies, abusive husbands, serial rapists, and hatecrime perpetrators who are off the scale. Diana Scully interviewed many rapists in their prison cells who bragged to her that they were “multitalented superachievers.”124 Psychopaths and other violent people are narcissistic: they think well of themselves not in proportion to their accomplishments but out of a congenital sense of entitlement. When reality intrudes, as it inevitably will, they treat the bad news as a personal affront, and its bearer, who is endangering their fragile reputation, as a malicious slanderer.

  Violence-prone personality traits are even more consequential when they infect political rulers, because their hang-ups can affect hundreds of millions of people rather than just the unlucky few who live with them or cross their paths. Unimaginable amounts of suffering have been caused by tyrants who callously presided over the immiseration of their peoples or launched destructive wars of conquest. In chapters 5 and 6 we saw that the tail-thickening wars and dekamegamurders of the 20th century can be attributed in part to the personalities of just three men. Tin-pot tyrants like Saddam Hussein, Mobutu Sese Seko, Moammar Khaddafi, Robert Mugabe, Idi Amin, Jean-Bédel Bokassa, and Kim Jong-il have immiserated their people on a scale that is smaller but still tragic.

  The study of the psychology of political leaders, to be sure, has a deservedly poor reputation. It’s impossible to test the object of investigation directly, and all too tempting to pathologize people who are morally contemptible. Psychohistory also has a legacy of fanciful psychoanalytic conjectures about what made Hitler Hitler: he had a Jewish grandfather, he had only one testicle, he was a repressed homosexual, he was asexual, he was a sexual fetishist. As the journalist Ron Rosenbaum wrote in Explaining Hitler, “The search for Hitler has apprehended not one coherent, consensus image of Hitler but rather many different Hitlers, competing Hitlers, conflicting embodiments of competing visions. Hitlers who might not recognize each other well enough to say ‘Heil’ if they came face to face in Hell.”125

  For all that, the more modest field of personality classification, which pigeonholes rather than explains people, has something to say about the psychology of modern tyrants. The Diagnostic and Statistical Manual of Mental Disorders (DSM) of the American Psychiatric Association defines narcissistic personality disorder as “a pervasive pattern of grandiosity, need for admiration, and a lack of empathy.”126 Like all psychiatric diagnoses, narcissism is a fuzzy category, and overlaps with psychopathy (“a pervasive pattern of disregard for, and violation of, the rights of others”) and with borderline personality disorder (“instability in mood; black and white thinking; chaotic and unstable interpersonal relationships, self-image, identity, and behavior”). But the trio of symptoms at narcissism’s core—grandiosity, need for admiration, and lack of empathy—fits tyrants to a T.127 It is most obvious in their vainglorious monuments, hagiographic iconography, and obsequious mass rallies. And with armies and police forces at their disposal, narcissistic rulers leave their mark in more than statuary; they can authorize vast outlays of violence. As with garden-variety bullies and toughs, the unearned self-regard of tyrants is eternally vulnerable to being popped, so any opposition to their rule is treated not as a criticism but as a heinous crime. At the same time, their lack of empathy imposes no brake on the punishment they mete out to real or imagined opponents. Nor does its allow any consideration of the human costs of another of their DSM symptoms: their “fantasies of unlimited success, power, brilliance, beauty, or ideal love,” which may be realized in rapacious conquest, pharaonic construction projects, or utopian master plans. And we have already seen what overconfidence can do in the waging of war.

  All leaders, of course, must have a generous dose of confidence to have become leaders, and in this age of psychology, pundits often diagnose leaders they don’t like with narcissistic personality disorder. But it’s important not to trivialize the distinction between a politician with good teeth and the psychopaths who run their countries into the ground and take large parts of the world with them. Among the pacifying features of democracies is that their leadership-selection procedure penalizes an utter lack of empathy, and their checks and balances limit the damage that a grandiose leader can do. Even within autocracies, the personality of a leader—a Gorbachev as opposed to a Stalin—can have an enormous impact on the statistics of violence.

  The damage done by the drive for dominance can be multiplied in a second way. This multiplier depends on a feature of the social mind that can be introduced with a benign anecdote. Every December my heart is warmed by a local tradition: the province of Nova Scotia sends a towering spruce to the city of Boston for its Christmas tree in gratitude for the humanitarian aid that Boston organizations extended to Haligonians after the horrendous 1917 explosion of a munitions-laden ship in Halifax harbor. As a Canadian expatriate in New England, I get to feel good twice: once in gratitude for the generous help that had been extended to my fellow Canucks, once in appreciation for the thoughtful gift that is being returned to my Boston brethren. Yet this entire ritual is, when you think about it, rather odd. I was not a party to either act of generosity, and so I neither earned nor should express gratitude. The people who find, fell, and send the tree never met the original victims and helpers; nor did the people who erect and decorate it. For all I know, not a single person touched by the tragedy is alive today. Yet we all feel the emotions that would be appropriate to a transaction of sympathy and gratitude between a pair of individual people. Everyone’s mind contains a representation called “Nova Scotia” and a representation called “Boston” which are granted the full suite of moral emotions and valuations, and individual men and women act out their roles in the social behavior that follows from them.

  A part of an individual’s personal identity is melded with the identity of the groups that he or she affiliates with.128 Each group occupies a slot in their minds that is
very much like the slot occupied by an individual person, complete with beliefs, desires, and praiseworthy or blameworthy traits. This social identity appears to be an adaptation to the reality of groups in the welfare of individuals. Our fitness depends not just on our own fortunes but on the fortunes of the bands, villages, and tribes we find ourselves in, which are bound together by real or fictive kinship, networks of reciprocity, and a commitment to public goods, including group defense. Within the group, some people help to police the provision of public goods by punishing any parasite who doesn’t contribute a fair share, and they are rewarded by the group’s esteem. These and other contributions to the group’s welfare are psychologically implemented by a partial loss of boundaries between the group and the self. On behalf of our group, we can feel sympathetic, grateful, angry, guilty, trustful, or mistrustful with regard to some other group, and we spread these emotions over the members of that group regardless of what they have done as individuals to deserve them.

  Loyalty to groups in competition, such as sports teams or political parties, encourages us to play out our instinct for dominance vicariously. Jerry Seinfeld once remarked that today’s athletes churn through the rosters of sports teams so rapidly that a fan can no longer support a group of players. He is reduced to rooting for their team logo and uniforms: “You are standing and cheering and yelling for your clothes to beat the clothes from another city.” But stand and cheer we do: the mood of a sports fan rises and falls with the fortunes of his team.129 The loss of boundaries can literally be assayed in the biochemistry lab. Men’s testosterone level rises when their team defeats a rival in a game, just as it rises when they personally defeat a rival in a wrestling match or in singles tennis.130 It also rises or falls when a favored political candidate wins or loses an election.131