The Better Angels of Our Nature: Why Violence Has Declined

by Steven Pinker

  The participants then read the partner’s feedback on their essay, which was scathing. They also were shown his or her taste preferences, which indicated a dislike of spicy foods. And then they were asked to prepare a snack for the partner from a bag of chips and a container of hot sauce prominently labeled SPICY. The amount they applied was measured by weighing the container after they left. The participants were also asked to rate their moods, including anger. The ones whose self-control had been depleted by having to forgo the donut didn’t get mad, but they did get even. They applied 62 percent more hot sauce to the chip of their insulting partner, presumably because they could not resist the impulse for revenge. Will-depleted subjects were also more likely to torment a critic by leaning on a button to blast him with an air horn every time he made an error in a computer game.

  Another study tapped aggressive fantasies by asking participants to imagine standing at a bar with a beloved girlfriend when a rival shows up and begins to flirt, to her visible enjoyment. (In the scenario given to women, it was a boyfriend who was chatted up by a rival woman.) The participant imagines confronting the rival, who responds by shoving him into the bar. Close at hand is a beer bottle. The participant was asked: “How likely would you be to smash the bottle on the person’s head? Indicate your response on a scale from -100 (not at all likely) to 100 (extremely likely).” The participants who were low in self-control, if their wills were ready and rested, indicated they probably would not retaliate. But if their wills had been depleted, they indicated they probably would.

  If we combine (1) Baumeister’s experiments, which found that reducing self-control in the lab can increase tendencies toward impulsive sex and violence; (2) the correlations across individuals between low self-control on the one hand and childhood misconduct, dissolute behavior, and crime on the other; (3) the neuroimaging studies that showed correlations between frontal lobe activity and self-control; and (4) the neuroimaging studies showing correlations between impulsive violence and impaired frontal lobe function, then we get an empirical picture that supports Elias’s conjecture that violence may be caused by weakness in an overarching neural mechanism of self-control.

  The picture is still incomplete. The existence of a trait that is stable in an individual over a span of decades, and that can be depleted over a span of minutes, cannot explain how a society can change over a span of centuries. We still need to show that whatever level of self-control a person is born with, he or she has ways to boost it. There is no paradox in the possibility that self-control can both be heritable across individuals and rise over time. That is exactly what happened with stature: genes make some of us taller than others, but over the centuries everyone got taller.117

  For as long as people have reflected on self-control, they have reflected on ways to enhance it. Odysseus had his sailors tie him to the mast and plug their ears with wax so he could hear the alluring song of the sirens without steering his ship onto the rocks. Techniques in which the present self handicaps the future self are sometimes called Odyssean or Ulyssean in his honor. There are hundreds of examples.118 We avoid shopping on an empty stomach. We throw out the brownies or the cigarettes or the booze at a time when we aren’t craving them, to foil ourselves at a time when we are. We put our alarm clock on the other side of the bedroom so we don’t turn it off and fall back asleep. We authorize our employers to invest a part of every paycheck for our retirement. We refrain from buying a magazine or a book or a gadget that will divert our attention from a work project until it is complete. We hand over money to a company like Stickk.com that returns it a fraction at a time if we meet certain goals, or donates it to a political organization we detest if we don’t. We make a public resolution to change, so our reputation takes a hit if we don’t.

  As we saw in chapter 3, one way the early modern Europeans used Odyssean self-control was to keep sharp knives out of reach at the dinner table. The familiar sign in the saloons in old Westerns—“Check your guns at the door”—served the same purpose, as do gun control laws and disarmament agreements today. Another tactic is to keep oneself away from trouble, such as avoiding the place where an aggrieved rival is known to hang out. Brawlers who allow themselves to be pulled apart by bystanders avail themselves of a similar tactic, with the added bonus that they have not conceded weakness or cowardice by disengaging.

  Other strategies of self-control are mental rather than physical. Walter Mischel showed that even four-year-olds can wait out a long interval for a double helping of marshmallows if they cover the alluring marshmallow in front of them, look away from it, distract themselves by singing, or even reframe it in their minds as a puffy white cloud rather than a sweet tasty food.119 An equivalent in the case of violence may be the cognitive reframing of an insult from a devastating blow to one’s reputation to an ineffectual gesture or a reflection on the immaturity of the insulter. Such reframing lies behind advice like “Don’t take it personally,” dismissals like “He’s just blowing smoke,” “He’s only a kid,” and “Take it from whom it comes,” and proverbs like “Sticks and stones can break my bones but words can never hurt me.”

  Martin Daly and Margo Wilson, invoking the economists’ theory of optimal interest rates and the biologists’ theory of optimal foraging, have suggested a third way in which self-control can be manipulated. They propose that organisms are equipped with an internal variable, like an adjustable interest rate, that governs how steeply they discount the future.120 The setting of the variable is twiddled according to the stability or instability of their environment and an estimate of how long they will live. It doesn’t pay to save for tomorrow if tomorrow will never come, or if your world is so chaotic that you have no confidence you would get your savings back. In a quantitative comparison of neighborhoods in a major city, Daly and Wilson found that the shorter the expected life span (from all causes other than violence), the higher the rate of violent crime. The correlation supports the hypothesis that, holding age constant, people are more reckless when they have fewer years of unlived life at risk. A rational adjustment of one’s discounting rate in response to the uncertainty of the environment could create a vicious circle, since your own recklessness then figures into the discounting rate of everyone else. The Matthew Effect, in which everything seems to go right in some societies and wrong in others, could be a consequence of environmental uncertainty and psychological recklessness feeding on each other.

  A fourth way people in a society might boost their self-control is by improving their nutrition, sobriety, and health. The frontal lobe is a big slab of metabolically demanding tissue with an outsize appetite for glucose and other nutrients. Pushing the metaphor of self-control as physical effort even further, Baumeister found that people’s blood glucose level plummets as their ego is depleted by an attention-consuming or a willpower-demanding task.121 And if they replenish their glucose level by drinking a glass of sugar-sweetened lemonade (but not a glass of aspartame-sweetened lemonade), they avoid the usual slump in the follow-up task. It is not implausible to suppose that real-world conditions that impair the frontal lobes—low blood sugar, drunkenness, druggedness, parasite load, and deficiencies of vitamins and minerals—could sap the self-control of people in an impoverished society and leave them more prone to impulsive violence. Several placebo-controlled studies have suggested that providing prisoners with dietary supplements can reduce their rate of impulsive violence.122

  Baumeister stretched the metaphor even farther. If willpower is like a muscle that fatigues with use, drains the body of energy, and can be revived by a sugary pick-me-up, can it also be bulked up by exercise? Can people develop their raw strength of will by repeatedly flexing their determination and resolve? The metaphor shouldn’t be taken too literally—it’s unlikely that the frontal lobes literally gain tissue like bulging biceps—but it’s possible that the neural connections between the cortex and limbic system may be strengthened with practice. It’s also possible that people can learn strategies of self-control, enj
oy the feeling of mastery over their impulses, and transfer their newfound tricks of discipline from one part of their behavioral repertoire to another.

  Baumeister and other psychologists tested the exercise metaphor by having participants undertake regimens of self-control for several weeks or months before taking part in one of their ego depletion studies.123 In various studies the regimens required them to keep track of every piece of food they ate; enroll in programs of physical exercise, money management, or study skills; use their nonpreferred hand for everyday tasks like brushing their teeth and using a computer mouse; and one that really gave the students’ self-control a workout: avoid curse words, speak in complete sentences, and not begin sentences with I. After several weeks of this cross-training, the students indeed turned out to be more resistant to ego depletion tasks in the lab, and they also showed greater self-control in their lives. They smoked fewer cigarettes, drank less alcohol, ate less junk food, spent less money, watched less television, studied more, and washed the dishes more often rather than leaving them in the sink. Score another point for Elias’s conjecture that self-control in life’s little routines can become second nature and be generalized to the rest of one’s comportment.

  In addition to being modulated by Ulyssean constraints, cognitive reframing, an adjustable internal discount rate, improvements in nutrition, and the equivalent of muscle gain with exercise, self-control might be at the mercy of whims in fashion.124 In some eras, self-control defines the paragon of a decent person: a grown-up, a person of dignity, a lady or a gentleman, a mensch. In others it is jeered at as uptight, prudish, stuffy, straitlaced, puritanical. Certainly the crime-prone 1960s were the recent era that most glorified the relaxation of self-control: Do your own thing, Let it all hang out, If it feels good do it, Take a walk on the wild side. The premium on self-indulgence is on full display in concert films from the decade, which show rock musicians working so hard to outdo each other in impulsiveness that it looks as if they put a lot of planning and effort into their spontaneity.

  Could these six pathways to self-control proliferate among the members of a society and come to define its global character? That would be the final domino in the chain of explanation that makes up the theory of the Civilizing Process. The exogenous first domino is a change in law enforcement and opportunities for economic cooperation that objectively tilt the payoffs so that a deferral of gratification, in particular, an avoidance of impulsive violence, pays off in the long run. The knock-on effect is a strengthening of people’s self-control muscles that allow them (among other things) to inhibit their violent impulses, above and beyond what is strictly necessary to avoid being caught and punished. The process could even feed on itself in a positive feedback loop, “positive” in both the engineering and the human-values sense. In a society in which other people control their aggression, a person has less of a need to cultivate a hair trigger for retaliation, which in turn takes a bit of pressure off everyone else, and so on.

  One way to bridge the gap between psychology and history is to look for changes in a society-wide index of self-control. As we have seen, an interest rate is just such an index, because it reveals how much compensation people demand for deferring consumption from the present to the future. To be sure, an interest rate is partly determined by objective factors like inflation, expected income growth, and the risk that the investment will never be returned. But it partly reflects the purely psychological preference for instant over delayed gratification. According to one economist, a six-year-old who prefers to eat one marshmallow now rather than two marshmallows a few minutes from now is in effect demanding an interest rate of 3 percent a day, or 150 percent a month.125

  Gregory Clark, the economic historian we met in chapter 4, has estimated the interest rates that Englishmen demanded (in the form of rents on land and houses) from 1170 to 2000, the millennium over which the Civilizing Process took place. Before 1800, he argues, there was no inflation to speak of, incomes were flat, and the risk that an owner would lose his property was low and constant. If so, the effective interest rate was an estimate of the degree to which people favored their present selves over their future selves.

  FIGURE 9–1. Implicit interest rates in England, 1170–2000

  Source: Graph from Clark, 2007a, p. 33.

  Figure 9–1 shows that during the centuries in which homicide plummeted in England, the effective interest rate also plummeted, from more than 10 percent to around 2 percent. Other European societies showed a similar transition. The correlation does not, of course, prove causation, but it is consistent with Elias’s claim that the decline of violence from medieval to modern Europe was part of a broader trend toward self-control and an orientation to the future.

  What about more direct measures of a society’s aggregate self-control? An annual interest rate is still quite distant from the momentary exercises of forbearance that suppress violent impulses in everyday life. Though there are dangers in essentializing a society by assigning it character traits that really should apply to individuals (remember the so-called fierce people), there may be a grain of truth in the impression that some cultures are marked by more self-control in everyday life than others. Friedrich Nietzsche distinguished between Apollonian and Dionysian cultures, named after the Greek gods of light and wine, and the distinction was used by the anthropologist Ruth Benedict in her classic 1934 ethnography collection Patterns of Culture. Apollonian cultures are said to be thinking, self-controlled, rational, logical, and ordered; Dionysian cultures are said to be feeling, passionate, instinctual, irrational, and chaotic. Few anthropologists invoke the dichotomy today, but a quantitative analysis of the world’s cultures by the sociologist Geert Hofstede has rediscovered the distinction in the pattern of survey responses among the middle-class citizens of more than a hundred countries.

  According to Hofstede’s data, countries differ along six dimensions.126 One of them is Long-Term versus Short-Term Orientation: “Long-term oriented societies foster pragmatic virtues oriented towards future rewards, in particular saving, persistence, and adapting to changing circumstances. Short-term oriented societies foster virtues related to the past and present such as national pride, respect for tradition, preservation of ‘face,’ and fulfilling social obligations.” Another dimension is Indulgence versus Restraint: “Indulgence stands for a society that allows relatively free gratification of basic and natural human drives related to enjoying life and having fun. Restraint stands for a society that suppresses gratification of needs and regulates it by means of strict social norms.” Both, of course, are conceptually related to the faculty of self-control, and not surprisingly, they are correlated with each other (with a coefficient of 0.45 across 110 countries). Elias would have predicted that both of these national traits should correlate with the countries’ homicide rates, and the prediction turns out to be true. The citizens of countries with more of a long-term orientation commit fewer homicides, as do the citizens of countries that emphasize restraint over indulgence.127

  So the theory of the Civilizing Process, like the theory of the expanding circle, has found support in experiments and datasets that are far from its field of origin. Psychology, neuroscience, and economics have confirmed Elias’s speculation that humans are equipped with a faculty of self-control that regulates both violent and nonviolent impulses, that can be strengthened and generalized over the lifetime of an individual, and that can vary in strength across societies and historical periods.

  So far I have not mentioned yet another explanation for the long-term growth of self-control: that it is a process of evolution in the biologist’s sense. Before turning to the last two of our better angels, morality and reason, I need to spend a few pages on this vexed question.

  RECENT BIOLOGICAL EVOLUTION?

  Many people casually use the word evolution to refer both to cultural change (that is, history) and to biological change (that is, a shift in the frequencies of genes across generations). Cultural and biological evolution c
an certainly interact. For instance, when tribes in Europe and Africa adopted the practice of keeping livestock for milk, they evolved genetic changes that allowed them to digest lactose in adulthood.128 But the two processes are different. They can always be distinguished, in theory, by experiments in which babies from one society are given up for adoption and brought up in another. If biological evolution has taken place in response to the distinctive culture of either society, then the adopted children should differ, on average, from their native peers.

  A frequently asked question about declines in violence is whether they can be attributed to recent biological evolution. In a society that has undergone a Pacification or Civilizing Process, has the genetic makeup of the people changed in response, helping the process along and leaving them permanently less disposed to violence? Any such change would not, of course, be a Lamarckian absorption of the cultural trend into the genome, but a Darwinian response to the altered contingencies of survival and reproduction. The individuals who happen to be genetically suited to the changed culture would outreproduce their neighbors and contribute a larger share of genes to the next generation, gradually shifting the population’s genetic makeup.