Behave: The Biology of Humans at Our Best and Worst

by Robert M. Sapolsky

  Amid some evidence for single-gene selection (an obscure phenomenon called intragenomic conflict, which we won’t go into), most people who vote for the importance of gene(s) over phenotype view single-gene selfishness as a bit of a sideshow and vote for the genome level of selection being most important.

  Meanwhile, there’s the view that phenotype trumps genotype, something championed by Ernst Mayr, Stephen Jay Gould, and others. The core of their argument is that it’s phenotypes rather than genotypes that are selected for. As Gould wrote, “No matter how much power Dawkins wishes to assign to genes, there is one thing he cannot give them—direct visibility to natural selection.” In that view, genes and the frequencies of their variants are merely the record of what arose from phenotypic selection.43

  Dawkins introduced a great metaphor: a cake recipe is a genotype, and how the cake tastes is the phenotype.* Genotype chauvinists emphasize that the recipe is what is passed on, the sequence of words that make for a stable replicator. But people select for taste, not recipe, say the phenotypists, and taste reflects more than just the recipe—after all, there are recipe/environment interactions where bakers differ in their skill levels, cakes bake differently at various altitudes, etc. The recipe-versus-taste question can be framed practically: Your cake company isn’t selling enough cakes. Do you change the recipe or the baker?

  Can’t we all get along? There’s the obvious bleeding-heart answer, namely that there’s room for a range of views and mechanisms in our rainbow-colored tent of evolutionary diversity. Different circumstances bring different levels of selection to the forefront. Sometimes the most informative level is the single gene, sometimes the genome, sometimes a single phenotypic trait, sometimes the collection of all the organism’s phenotypic traits.44 We’ve just arrived at the reasonable idea of multilevel selection.

  The Resurrection of Group Selection

  Hooray, progress. Sometimes it makes the most sense to pay attention to the recipe, sometimes to the baking process; the recipe is what is replicated, the taste what is chosen.

  But there’s another level. Sometimes cake sales can be changed most consequentially by altering something other than recipe or taste—advertisements, packaging, or the perception of whether the cake is a staple or a luxury. Sometimes sales are changed by tying the product to a particular audience—think of products that advertise fair-trade practices, the Nation of Islam’s Your Black Muslim Bakery, or the Christian fundamentalist ideology of Chick-fil-A restaurants. And in those cases recipe and taste can both be trumped by ideology in purchasing decisions.

  This is where neo–group selection fits into multilevel selection—the idea that some heritable traits may be maladaptive for the individual but adaptive for a group. This has cooperation and prosociality written all over it, straight out of the analysis of Tit for Tat–ers finding one another in a sea of Always Defect–ers. Stated more formally, it’s when A dominates B but a group of Bs dominates a group of As.

  Here’s a great example of neo–group selectionism: As a poultry farmer, you want your groups of chickens to lay as many eggs as possible. Take the most prolific egg layer in each group, forming them into a group of superstar chickens who, presumably, will be hugely productive. Instead, egg production is miniscule.45

  Why was each superstar the egg queen in her original group? Because she would aggressively peck subordinates enough to stress them into reduced fertility. Put all these mean ones together, and a group of subordinated chickens will outproduce them.

  This is a world away from “animals behave for the good of the species.” Instead, this is the circumstance of a genetically influenced trait that, while adaptive on an individual level, emerges as maladaptive when shared by a group and where there is competition between groups (e.g., for an ecological niche).

  There’s been considerable resistance to neo–group selectionism. Part of it is visceral, often pronounced among the old guard—“Great, we’ve finally confiscated all the Wild Kingdom videos, and now we’re back to playing Whac-A-Mole with group selection sentimentality?” But the more fundamental resistance is from people who distinguish bad old group selection from neo–group selection, accept that the latter can occur, but think it’s very rare.

  Maybe so, across the animal kingdom. But neo–group selection plays out with great frequency and consequence in humans. Groups compete for hunting grounds, pastures, water sources. Cultures magnify the intensity of between-group selection and lessen within-group selection with ethnocentrism, religious intolerance, race-based politics, and so on. The economist Samuel Bowles, of the Santa Fe Institute, emphasizes how intergroup conflict like war is the driving force for intragroup cooperation (“parochial altruism”); he refers to intergroup conflict as “altruism’s midwife.”46

  Most in the field now both accept multilevel selection and see room for instances of neo–group selection, especially in humans. Much of this reemergence is the work of two scientists. The first is David Sloan Wilson of the State University of New York at Binghamton, who spent decades pushing for neo–group selection (although he sees it not really as “neo” but rather as old-style group selection finally getting some scientific rigor), generally being dismissed, and arguing his case with research of his own, studies ranging from fish sociality to the evolution of religion. He slowly convinced some people, most importantly the second scientist, Edward O. Wilson of Harvard (no relation). E. O. Wilson is arguably the most important naturalist of the last half of the twentieth century, an architect of the sociobiology synthesis along with a number of other fields, a biology god. E. O. Wilson had long dismissed David Sloan Wilson’s ideas. And then a few years back, the octogenarian E. O. Wilson did something extraordinary—he decided he was wrong. And then he published a key paper with the other Wilson—“Rethinking the Theoretical Foundation of Sociobiology.” My respect for these two, both as people and as scientists, is enormous.47

  Thus something resembling détente has occurred among the advocates for the importance of differing levels of selection. Our three-legged chair of individual selection, kin selection, and reciprocal altruism seems more stable with four legs.

  AND US

  Where do humans fit into all this? Our behavior closely matches the predictions of these evolutionary models. Until you look more closely.48

  Let’s start by clearing up some misconceptions. First, we are not descended from chimps. Or from any extant animal. We and chimps share a common ancestor from roughly five million years ago (and genomics show that chimps have been as busy evolving since then as we have).49

  And there are misconceptions as to which ape is our “closest relative.” In my experience, someone who is fond of duck hunting and country music usually votes chimp, but if you eat organic food and know about oxytocin, it’s bonobo. The reality is that we’re equally related to both, sharing roughly 98 to 99 percent of our DNA with each. Svante Pääbo of the Max Planck Institutes in Germany has shown that 1.6 percent of the human genome is more related to bonobos than to chimps; 1.7 percent more to chimps than to bonobos.*50 Despite the combination of some of our most fervent wishes and excuses, we’re neither bonobos nor chimps.

  On to how the conceptual building blocks of behavioral evolution apply to humans.

  Promiscuous Tournament or Monogamous Pair-Bonded?

  I can’t resist starting with an irresistible question—so, are we a pair-bonded or tournament species?51

  Western civilization doesn’t give a clear answer. We praise stable, devoted relationships yet are titillated, tempted, and succumb to alternatives at a high rate. Once divorces are legalized, a large percentage of marriages end in them, yet a smaller percentage of married people get divorced—i.e., the high divorce rate arises from serial divorcers.

  Anthropology doesn’t help either. Most cultures have allowed polygyny. But within such cultures most people are (socially) monogamous. But most of those men would presumably be polygamous
if they could buy more wives.

  What about human sexual dimorphism? Men are roughly 10 percent taller and 20 percent heavier than women, need 20 percent more calories, and have life spans 6 percent shorter—more dimorphic than monogamous species, less than polygamous ones. Likewise with subtle secondary sexual characteristics like canine length, where men average slightly longer canines than women. Moreover, compared with, say, monogamous gibbons, human males have proportionately bigger testes and higher sperm counts . . . but pale in comparison with polygamous chimps. And back to imprinted genes, reflecting intersexual genetic competition, which are numerous in tournament species and nonexistent in pair-bonders. What about humans? Some such genes, but not many.

  Measure after measure, it’s the same. We aren’t classically monogamous or polygamous. As everyone from poets to divorce attorneys can attest, we are by nature profoundly confused—mildly polygynous, floating somewhere in between.*

  Individual Selection

  At first pass we seem like a great example of a species where the driving force on behavior is maximizing reproductive success, where a person can be an egg’s way of making another egg, where selfish genes triumph. Just consider the traditional perk of powerful men: being polygamous. Pharaoh Ramses II, incongruously now associated with a brand of condoms, had 160 children and probably couldn’t tell any of them from Moses. Within half a century of his death in 1953, Ibn Saud, the founder of the Saudi dynasty, had more than three thousand descendants. Genetic studies suggest that around sixteen million people today are descended from Genghis Khan. And in recent decades more than one hundred children each were fathered by King Sobhuza II of Swaziland, Ibn Saud’s son King Saud, the dictator Jean-Bédel Bokassa of the Central African Republic, plus various fundamentalist Mormon leaders.52

  The human male drive to maximize reproductive success is shown by a key fact—the most common cause of individual human violence is male-male competition for direct or indirect reproductive access to females. And then there is the dizzyingly common male violence against females for coercive sex or as a response to rejection.

  So plenty of human behaviors would make sense to a baboon or elephant seal. But that’s only half the story. Despite Ramses, Ibn Saud, and Bokassa, numerous people forgo reproducing, often because of theology or ideology. And an entire sect—the United Society of Believers in Christ’s Second Appearing, aka the Shakers, will soon be extinct because of its adherents’ celibacy. And finally, the supposed selfishness of human genes driving individual selection must accommodate individuals sacrificing themselves for strangers.

  Earlier in the chapter I presented competitive infanticide as stark evidence of the importance of individual selection. Does anything like that occur in humans? The psychologists Martin Daly and (the late) Margo Wilson of McMaster University in Canada looked at patterns of child abuse and made a striking observation—a child is far more likely to be abused or killed by a stepparent than by a parent. This is readily framed as parallel to competitive infanticide.53

  This finding, termed the “Cinderella effect,” while embraced by human sociobiologists, has also been robustly criticized. Some charge that socioeconomic status was not sufficiently controlled for (homes with a stepparent, rather than two biological parents, generally have less income and more economic stress, known causes of displacement aggression). Others think there’s a detection bias—the same degree of abuse is more likely to be identified by authorities when it’s committed by a stepparent. And the finding has been independently replicated in some but not all studies. I think the jury is still out on the subject.

  Kin Selection

  Where do humans fit in terms of kin selection? We’ve already seen examples that fit well—e.g., the fraternal polyandry in Tibet, the weirdness of women liking the smell of their male cousins, the universality of nepotism.

  Moreover, humans are obsessed with kin relations in culture after culture, with elaborate systems of kinship terms (just go into a store and look at the Hallmark cards organized by kinship category—for a sister, a brother, an uncle, and so on). And in contrast to other primates who leave their natal group around adolescence, when humans in traditional society marry someone from another group and go live with them, they maintain contact with their family of origin.54

  Moreover, from New Guinea highlanders to the Hatfields and McCoys, feuds and vendettas occur along clan lines of relatedness. We typically bequeath our money and land among our descendants rather than among strangers. From ancient Egypt to North Korea and on to the Kennedys and Bushes, we have dynastic rule. How’s this for a display of human kin selection: Subjects were given a scenario of a bus hurtling toward a human and a nondescript dog, and they could only save one. Whom would they pick? It depended on degree of relatedness, as one progressed from sibling (1 percent chose the dog over the sibling) to grandparent (2 percent) to distant cousin (16 percent) to foreigner (26 percent).55

  As another measure of the importance of kinship in human interactions, people can’t be compelled to testify in court against a first-degree relative in many countries and American states. And when humans have damage to the (emotional) vmPFC, they become so unemotionally utilitarian that they would choose to harm family members in order to save strangers.56

  There’s a fascinating historical example of how wrong it feels when someone chooses strangers over kin. This is the story of Pavlik Morozov, a boy in Stalin’s Soviet Union.57 Young Pavlik, according to the official story, was a model citizen, an ardent flag-waving patriot. In 1932 he chose the state over his kin, denouncing his father (for supposed black marketeering), who was promptly arrested and executed. Soon afterward the boy was killed, allegedly by relatives who felt more strongly about kin selection than he did.

  The regime’s propagandists embraced the story. Statues of the young martyr to the revolution were erected. Poems and songs were written; schools were named for him. An opera was composed, a hagiographic movie made.

  As the story emerged, Stalin was told about the boy. And what was the response of the man most benefiting from such fealty to the state? Was it “If only all my citizens were that righteous; this lad gives me hope for our future”? No. According to historian Vejas Liulevicius of the University of Tennessee, when told about Pavlik, Stalin snorted derisively and said, “What a little pig, to have done such a thing to his own family.” And then he turned the propagandists loose.

  Thus even Stalin was of the same opinion as most mammals: something’s wrong with that kid. Human social interactions are profoundly organized around kin selection; with the rare exception of a Pavlik Morozov, blood is thicker than water.

  Naturally, until you look more closely.

  For starters, yes, across cultures we are obsessed with kinship terms, but the terms often don’t overlap with actual biological relatedness.

  We certainly have clan vendettas, but we also have wars where combatants on opposing sides have higher degrees of relatedness than do fighters on the same side. Brothers fought on opposing sides in the Battle of Gettysburg.58

  Relatives and their armies battle over royal succession; the cousins George V of England, Nicholas II of Russia, and Wilhelm II of Germany happily oversaw and sponsored World War I. And intrafamily individual violence occurs (although at extremely low rates when corrected for amount of time spent together). There’s patricide, often an act of revenge for a long history of abuse, and fratricide. Rarely due to conflicts over issues of economic or reproductive importance—stolen birthrights of biblical proportion, someone sleeping with their sibling’s spouse—fratricide is most often about long-standing irritants and disagreements that just happen to boil over into lethality (in early May 2016, for example, a Florida man was charged with second-degree murder in the killing of his brother—during a dispute over a cheeseburger). And then there is the hideous commonness of honor killings in parts of the world, as we’ve seen.59

  The most puzzling cases of
intrafamily violence, in terms of kin selection, are of parents killing children, a phenomenon most commonly arising from combined homicide/suicide, profound mental illness, or abuse that unintentionally proves fatal.*60 And then there are cases where a mother kills an unwanted child who is viewed as a hindrance—parent/offspring conflict flecked with the spittle of madness.61

  While we bequeath money to our descendants, we also give charitably to strangers on the other side of the planet (thank you, Bill and Melinda Gates) and adopt orphans from other continents. (Sure, as we’ll see in a later chapter, being charitable is tinged with self-interest, and most people who adopt kids do so because they cannot have biological offspring—but the occurrence of either act violates strict kin selection.) And in the primogeniture system of land inheritance, birth order trumps degree of relatedness.

  Thus we have textbook examples of kin selection, but also dramatic exceptions.

  Why do humans have such marked deviations from kin selection? I think this often reflects how humans go about recognizing relatives. We don’t do it with certainty, by innate recognition of MHC-derived pheromones, the way rodents do (despite our being able to distinguish degrees of relatedness to some extent by smell). Nor do we do it by imprinting on sensory cues, deciding, “This person is my mother because I remember that her voice was the loudest when I was a fetus.”

  Instead we do kin recognition cognitively, by thinking about it. But crucially, not always rationally—as a general rule, we treat people like relatives when they feel like relatives.

  One fascinating example is the Westermarck effect, demonstrated by marriage patterns among people raised in the Israeli kibbutz system.62 Communal child rearing is central to the ethos of the traditional socialist agricultural kibbutz approach. Children know who their parents are and interact with them a few hours a day. But otherwise they live, learn, play, eat, and sleep with the cohort of kids their age in communal quarters staffed by nurses and teachers.