The Better Angels of Our Nature: Why Violence Has Declined

by Steven Pinker

  Moral norms, even when ineffable, can sometimes be effective brakes on violent behavior. In the modern West, as we have seen, the avoidance of some kinds of violence, such as mercy-killing an abandoned child, retaliating for an insult, and declaring war on another developed state, consist not in weighing the moral issues, empathizing with the targets, or restraining an impulse, but in not having the violent act as a live option in the mind at all. The act is not considered and avoided; it is unthinkable or laughable.

  The combination of radical cultural differences in which behaviors are moralized with moral dumbfounding in our own culture may create the impression that norms and taboos are arbitrary—that there may be a culture out there somewhere in which it is immoral to utter a sentence with an even number of words or to deny that the ocean is boiling hot. But the anthropologist Richard Shweder and several of his students and collaborators have found that moral norms across the world cluster around a small number of themes.170 The intuitions that we in the modern West tend to think of as the core of morality—fairness, justice, the protection of individuals, and the prevention of harm—are just one of several spheres of concern that may attach themselves to the cognitive and emotional paraphernalia of moralization. Even a glance at ancient religions like Judaism, Islam, and Hinduism reminds us that they moralize a slew of other concerns, such as loyalty, respect, obedience, asceticism, and the regulation of bodily functions like eating, sex, and menstruation.

  Shweder organized the world’s moral concerns in a threefold way.171 Autonomy, the ethic we recognize in the modern West, assumes that the social world is composed of individuals and that the purpose of morality is to allow them to exercise their choices and to protect them from harm. The ethic of Community, in contrast, sees the social world as a collection of tribes, clans, families, institutions, guilds, and other coalitions, and equates morality with duty, respect, loyalty, and interdependence. The ethic of Divinity posits that the world is composed of a divine essence, portions of which are housed in bodies, and that the purpose of morality is to protect this spirit from degradation and contamination. If a body is merely a container for the soul, which ultimately belongs to or is part of a god, then people do not have the right to do what they want with their bodies. They are obligated to avoid polluting them by refraining from unclean forms of sex, food, and other physical pleasures. The ethic of Divinity lies behind the moralization of disgust and the valorization of purity and asceticism.

  Haidt took Shweder’s trichotomy and cleaved two of the ethics in two, yielding a total of five concerns that he called moral foundations.172 Community was bifurcated into In-group Loyalty and Authority/Respect, and Autonomy was sundered into Fairness/Reciprocity (the morality behind reciprocal altruism) and Harm/Care (the cultivation of kindness and compassion, and the inhibition of cruelty and aggression). Haidt also gave Divinity the more secular label Purity/Sanctity. In addition to these adjustments, Haidt beefed up the case that the moral foundations are universal by showing that all five spheres may be found in the moral intuitions of secular Westerners. In his dumbfounding scenarios, for example, Purity/Sanctity underlay the participants’ revulsion to incest, bestiality, and the eating of a family pet. Authority/ Respect commanded them to visit a mother’s grave. And In-group Loyalty prohibited them from desecrating an American flag.

  The system I find most useful was developed by the anthropologist Alan Fiske. It proposes that moralization comes out of four relational models, each a distinct way in which people conceive of their relationships.173 The theory aims to explain how people in a given society apportion resources, where their moral obsessions came from in our evolutionary history, how morality varies across societies, and how people can compartmentalize their morality and protect it with taboos. The relational models line up with the classifications of Shweder and Haidt more or less as shown in the table on page 626.

  The first model, Communal Sharing (Communality for short), combines In-group Loyalty with Purity/Sanctity. When people adopt the mindset of Communality, they freely share resources within the group, keeping no tabs on who gives or takes how much. They conceptualize the group as “one flesh,” unified by a common essence, which must be safeguarded against contamination. They reinforce the intuition of unity with rituals of bonding and merging such as bodily contact, commensal meals, synchronized movement, chanting or praying in unison, shared emotional experiences, common bodily ornamentation or mutilation, and the mingling of bodily fluids in nursing, sex, and blood rituals. They also rationalize it with myths of shared ancestry, descent from a patriarch, rootedness in a territory, or relatedness to a totemic animal. Communality evolved from maternal care, kin selection, and mutualism, and it may be implemented in the brain, at least in part, by the oxytocin system.

  Fiske’s second relational model, Authority Ranking, is a linear hierarchy defined by dominance, status, age, gender, size, strength, wealth, or precedence. It entitles superiors to take what they want and to receive tribute from inferiors, and to command their obedience and loyalty. It also obligates them to a paternalistic, pastoral, or noblesse oblige responsibility to protect those under them. Presumably it evolved from primate dominance hierarchies, and it may be implemented, in part, by testosterone-sensitive circuits in the brain.

  Equality Matching embraces tit-for-tat reciprocity and other schemes to divide resources equitably, such as turn-taking, coin-flipping, matching contributions, division into equal portions, and verbal formulas like eeny-meenyminey-moe. Few animals engage in clear-cut reciprocity, though chimpanzees have a rudimentary sense of fairness, at least when it comes to themselves being shortchanged. The neural bases of Equality Matching embrace the parts of the brain that register intentions, cheating, conflict, perspective-taking, and calculation, which include the insula, orbital cortex, cingulate cortex, dorsolateral prefrontal cortex, parietal cortex, and temporoparietal junction. Equality Matching is the basis of our sense of fairness and our intuitive economics, and it binds us as neighbors, colleagues, acquaintances, and trading partners rather than as bosom buddies or brothers-in-arms. Many traditional tribes engaged in the ritual exchange of useless gifts, a bit like our Christmas fruitcakes, solely to cement Equality Matching relationships.174

  (Readers who are comparing and contrasting the taxonomies may wonder why Haidt’s category of Harm/Care is adjacent to Fairness and aligned with Fiske’s Equality Matching, rather than with more touchy-feely relationships like Community or Sanctity. The reason is that Haidt measures Harm/Care by asking people about the treatment of a generic “someone” rather than the friends and relatives that are the standard beneficiaries of caring. The responses to these questions align perfectly with the responses to his questions about Fairness, and that is no coincidence.175 Recall that the logic of reciprocal altruism, which implements our sense of fairness, is to be “nice” by cooperating on the first move, by not defecting unless defected on, and by conferring a large benefit to a needy stranger when one can do so at a relatively small cost to oneself. When care and harm are extended outside intimate circles, then they are simply a part of the logic of fairness.)176

  Fiske’s final relational model is Market Pricing: the system of currency, prices, rents, salaries, benefits, interest, credit, and derivatives that powers a modern economy. Market Pricing depends on numbers, mathematical formulas, accounting, digital transfers, and the language of formal contracts. Unlike the other three relational models, Market Pricing is nowhere near universal, since it depends on literacy, numeracy, and other recently invented information technologies. The logic of Market Pricing remains cognitively unnatural as well, as we saw in the widespread resistance to interest and profits until the modern era. One can line up the models, Fiske notes, along a scale that more or less reflects their order of emergence in evolution, child development, and history: Communal Sharing > Authority Ranking > Equality Matching > Market Pricing.

  Market Pricing, it seems to me, is specific neither to markets nor to pricing. It reall
y should be lumped with other examples of formal social organization that have been honed over the centuries as a good way for millions of people to manage their affairs in a technologically advanced society, but which may not occur spontaneously to untutored minds.177 One of these institutions is the political apparatus of democracy, where power is assigned not to a strongman (Authority) but to representatives who are selected by a formal voting procedure and whose prerogatives are delineated by a system of laws. Another is a corporation, university, or nonprofit organization. The people who work in them aren’t free to hire their friends and relations (Communality) or to dole out spoils as favors (Equality Matching), but are hemmed in by fiduciary duties and regulations. My emendation of Fiske’s theory does not come out of the blue. Fiske notes that one of his intellectual inspirations for Market Pricing was the sociologist Max Weber’s concept of a “rational-legal” (as opposed to traditional and charismatic) mode of social legitimation—a system of norms that is worked out by reason and implemented by formal rules.178 Accordingly, I will sometimes refer to this relational model using the more general term Rational-Legal.

  For all their differences in lumping and splitting, the theories of Shweder, Haidt, and Fiske agree on how the moral sense works. No society defines everyday virtue and wrongdoing by the Golden Rule or the Categorical Imperative. Instead, morality consists in respecting or violating one of the relational models (or ethics or foundations): betraying, exploiting, or subverting a coalition; contaminating oneself or one’s community; defying or insulting a legitimate authority; harming someone without provocation; taking a benefit without paying the cost; peculating funds or abusing prerogatives.

  The point of these taxonomies is not to pigeonhole entire societies but to provide a grammar for social norms.179 The grammar should reveal common patterns beneath the differences among cultures and periods (including the decline of violence), and should predict people’s response to infractions of the reigning norms, including their perverse genius for moral compartmentalization.

  Some social norms are merely solutions to coordination games, such as driving on the right-hand side of the road, using paper currency, and speaking the ambient language.180 But most norms have moral content. Each moralized norm is a compartment containing a relational model, one or more social roles (parent, child, teacher, student, husband, wife, supervisor, employee, customer, neighbor, stranger), a context (home, street, school, workplace), and a resource (food, money, land, housing, time, advice, sex, labor). To be a socially competent member of a culture is to have assimilated a large set of these norms.

  Take friendship. Couples who are close friends operate mainly on the model of Communal Sharing. They freely share food at a dinner party, and they do each other favors without keeping score. But they may also recognize special circumstances that call for some other relational model. They may work together on a task in which one is an expert and gives orders to the other (Authority Ranking), split the cost of gas on a trip (Equality Matching), or transact the sale of a car at its blue book value (Market Pricing).

  Infractions of a relational model are moralized as straightforwardly wrong. Within the Communal Sharing model that usually governs a friendship, it is wrong for one person to stint on sharing. Within the special case of Equality Matching of gas on a trip, an infraction consists of failing to pay for one’s share. Equality Matching, with its assumption of a continuing reciprocal relationship, allows for loose accounting, as when the ranchers of Shasta County compensated each other for damage with roughly equivalent favors and agreed to lump it when a small act of damage went uncompensated.181 Market Pricing and other Rational-Legal models are less forgiving. A diner who leaves an expensive restaurant without paying cannot count on the owner to let him make it up in the long run, or simply to lump it. The owner is more likely to call the police.

  When a person violates the terms of a relational model he or she has tacitly agreed to, the violator is seen as a parasite or cheater and becomes a target of moralistic anger. But when a person applies one relational model to a resource ordinarily governed by another, a different psychology comes into play. That person doesn’t violate the rules so much as he or she doesn’t “get” them. The reaction can range from puzzlement, embarrassment, and awkwardness to shock, offense, and rage.182 Imagine, for example, a diner thanking a restaurateur for an enjoyable experience and offering to have him over for dinner at some point in the future (treating a Market Pricing interaction as if it were governed by Communal Sharing). Conversely, imagine the reaction at a dinner party (Communal Sharing) if a guest pulled out his wallet and offered to pay the host for the meal (Market Pricing), or if the host asked a guest to wash the pots while the host relaxed in front of the television (Equality Matching). Likewise, imagine that the guest offered to sell his car to the host, and then drove a hard bargain on the price, or the host suggested that the couples swap partners for a half-hour of sex before everyone went home for the evening.

  The emotional response to a relational mismatch depends on whether it is accidental or deliberate, which model is substituted for which, and the nature of the resource. The psychologist Philip Tetlock has suggested that the psychology of taboo—a reaction of outrage to certain thoughts being aired—comes into play with resources that are deemed sacred.183 A sacred value is one that may not be traded off against anything else. Sacred resources are usually governed by the primal models of Communality and Authority, and they trigger the taboo reaction when someone treats them with the more advanced models of Equality Matching or Market Pricing. If someone offered to purchase your child (suddenly thrusting a Communal Sharing relationship under the light of Market Pricing), you would not ask how much he was offering but would be offended at the very idea. The same is true for an offer to buy a personal gift or family heirloom that has been bestowed upon you, or to pay you for betraying a friend, a spouse, or your country. Tetlock found that when students were asked their opinion on the pros and cons of an open market for sacred resources like votes, military service, jury duty, body organs, or babies put up for adoption, most of them did not articulate a good case against the practice (such as that the poor might sell their organs out of desperation) but expressed outrage at being asked. Typical “arguments” were “This is degrading, dehumanizing, and unacceptable” and “What kind of people are we becoming?”

  The psychology of taboo is not completely irrational.184 To maintain precious relationships, it is not enough for us to say and do the right thing. We also have to show that our heart is in the right place, that we don’t weigh the costs and benefits of selling out those who trust us. When you are faced with an indecent proposal, anything less than an indignant refusal would betray the awful truth that you don’t understand what it means to be a genuine parent or spouse or citizen. And that understanding consists of having absorbed a cultural norm that assigns a sacred value to a primal relational model.

  In an old joke, a man asks a woman if she would sleep with him for a million dollars, and she says she would consider it. He then asks if she would sleep with him for a hundred dollars, and she replies, “What kind of a woman do you think I am?” He answers, “We’ve already established that. We’re just haggling over the price.” To understand the joke is to appreciate that most sacred values are in fact pseudo-sacred. People can be induced to compromise on them if the tradeoff is obfuscated, spin-doctored, or reframed.185 (The joke uses the landmark figure “a million dollars” because it reframes a mere exchange of money into a life-transforming opportunity, namely becoming a millionaire.) When life insurance was first introduced, people were outraged at the very idea of assigning a dollar value to a human life, and of allowing wives to bet that their husbands would die, both of which are technically accurate descriptions of what life insurance does.186 The insurance industry mounted advertising campaigns that reframed the product as an act of responsibility and decency on the part of the husband, who would simply be carrying out his duty to his family during a period
in which he happened not to be alive.

  Tetlock distinguishes three kinds of tradeoffs. Routine tradeoffs are those that fall within a single relational model, such as choosing to be with one friend rather than another, or to purchase one car rather than another. Taboo tradeoffs pit a sacred value in one model against a secular value in another, such as selling out a friend, a loved one, an organ, or oneself for barter or cash. Tragic tradeoffs pit sacred values against each other, as in deciding which of two needy transplant patients should receive an organ, or the ultimate tragic tradeoff, Sophie’s choice between the lives of her two children. The art of politics, Tetlock points out, is in large part the ability to reframe taboo tradeoffs as tragic tradeoffs (or, when one is in the opposition, to do the opposite). A politician who wants to reform Social Security has to reframe it from “breaking our faith with senior citizens” (his opponent’s framing) to “lifting the burden on hardworking wage-earners” or “no longer scrimping in the education of our children.” Keeping troops in Afghanistan is reframed from “putting the lives of our soldiers in danger” to “guaranteeing our nation’s commitment to freedom” or “winning the war on terror.” The reframing of sacred values, as we will see, may be an overlooked tactic in the psychology of peacemaking.

  The new theories of the moral sense, then, have helped explain moralized emotions, moral compartmentalization, and taboo. Now let’s apply them to differences in moralization across cultures and, crucially, over the course of history.

  Many assignments of a relational model to a set of social roles feel natural to people in all societies and may be rooted in our biology. They include the Communal Sharing among family members, an Authority Ranking within the family that makes people respect their elders, and the exchange of bulk commodities and routine favors under Equality Matching. But other kinds of assignment of a relational model to a resource and a set of social roles can differ radically across time and culture.187