Behave: The Biology of Humans at Our Best and Worst

by Robert M. Sapolsky

  Importantly, neither type of selection necessarily selects for “the” most adaptive version of a trait, which replaces all others. There can be frequency-dependent selection, where the rarer version of two traits is preferable, or balanced selection, where multiple versions of traits are maintained in equilibrium.

  BEHAVIOR CAN BE SHAPED BY EVOLUTION

  Organisms are amazingly well adapted. A desert rodent has kidneys that excel at retaining water; a giraffe’s huge heart can pump blood up to its brain; elephants’ leg bones are strong enough to support an elephant. Well, yes—it has to work that way: desert rodents whose kidneys weren’t great at retaining water didn’t pass on copies of their genes. Thus there is a logic to evolution, where natural selection sculpts traits into adaptiveness.

  Importantly, natural selection works not only on anatomy and physiology but on behavior as well—in other words, behavior evolves, can be optimized by selection into being adaptive.

  Various branches of biology focus on the evolution of behavior. Probably best known is sociobiology, premised on social behavior being sculpted by evolution to be optimized, just as biomechanical optimization sculpts the size of a giraffe’s heart.4 Sociobiology emerged in the 1970s, eventually generating the offshoot evolutionary psychology—the study of the evolutionary optimization of psychological traits; as we’ll see, both have been plenty controversial. As a simplifying convenience, I’ll refer to people who study the evolution of social behavior as “sociobiologists.”

  THE DEMISE OF GROUP SELECTION

  We start by grappling with an entrenched misconception about the evolution of behavior. This is because Americans were taught about the subject in the 1960s by Marlin Perkins on the TV program Mutual of Omaha’s Wild Kingdom.

  It was great. Perkins would host. Jim, his sidekick, did dangerous things with snakes. And there were always seamless segues from the program to ads from Mutual of Omaha—“Just as lions mate for hours, you’ll want fire insurance for your home.”

  Unfortunately, Perkins espoused wildly wrong evolutionary thinking. Here’s how it looked on the program: It’s dawn on the savanna; there’s a herd of wildebeest on a river’s edge. The grass is greener on the other side, and everyone wants some, but the river teems with predatory crocodiles. The wildebeest are hemming and hawing in agitation when suddenly an elderly wildebeest pushes to the front, says, “I sacrifice myself for you, my children,” and leaps in. And while the crocs are busy with him, the other wildebeest cross the river.

  Why would the old wildebeest do that? Marlin Perkins would answer with patrician authority: because animals behave For the Good of the Species.

  Yes, behavior evolves by “group selection” for the good of the species. This idea was championed in the early 1960s by V. C. Wynne-Edwards, whose wrongness made him modern evolutionary biology’s Lamarck.*5

  Animals don’t behave for the good of the species. But what about that wildebeest? Look closely and you’ll see what really happens. Why did he wind up saving the day? Because he was old and weak. “Good of the species” my keister. They pushed the old guy in.

  Group selection was done in by theoretical and empirical studies showing patterns of behavior incompatible with it. Key work was done by two gods of evolutionary biology, George Williams of SUNY Stony Brook and Oxford’s Bill (“W.D.”) Hamilton.6 Consider “eusocial insects,” where most individuals are nonreproductive workers. Why forgo reproduction to aid the queen? Group selection, obviously. Hamilton showed that eusocial insects’ unique genetic system makes a colony of ants, bees, or termites a single superorganism; asking why worker ants forgo reproduction is like asking why your nose cells forgo reproduction. In other words, eusocial insects constitute a unique type of “group.” Williams then elaborated on how the more standard genetic system, in species from noneusocial insects to us, was incompatible with group selection. Animals don’t behave for the good of the species. They behave to maximize the number of copies of their genes passed into the next generation.*

  This is the cornerstone of sociobiology and was summarized in Dawkins’s famed sound bite that evolution is about “selfish genes.” Time to see its building blocks.

  INDIVIDUAL SELECTION

  Passing on lots of copies of one’s genes is accomplished most directly by maximizing reproduction. This is summarized by the aphorism “A chicken is an egg’s way of making another egg”—behavior is just an epiphenomenon, a means of getting copies of genes into the next generation.

  Individual selection fares better than group selection in explaining basic behaviors. A hyena bears down on some zebras. What would the nearest one do if she’s a group selectionist? Stand there, sacrificing herself for the group. In contrast, an individual selectionist zebra would run like hell. Zebras run like hell. Or consider hyenas that have just killed a zebra. Group selection mind-set—everyone calmly takes turns eating. Individual selection—frenzied free-for-all. Which is what occurs.

  But wait, says the group selectionist, wouldn’t the zebra species benefit if it is the fastest animals who survive and pass on those fast-running genes? Ditto for the group benefits of the fiercest hyena getting the most food.

  As more nuances of behavior are observed, clinging to group selection requires increasingly tortuous arguments. But one single observation devastates group selection.

  In 1977 the Harvard primatologist Sarah Blaffer Hrdy documented something remarkable—langur monkeys in the Mount Abu region of India kill one another.7 People already knew that some male primates kill one another, fighting for dominance—okay, makes sense, boys will be boys. But that’s not what Hrdy reported; male langurs were killing infants.

  Once people believed her careful documentation, there was an easy answer—since babies are cute and inhibit aggression, something pathological must be happening.8 Maybe the Abu langur population density was too high and everyone was starving, or male aggression was overflowing, or infanticidal males were zombies. Something certifiably abnormal.

  Hrdy eliminated these explanations and showed a telling pattern to the infanticide. Female langurs live in groups with a single resident breeding male. Elsewhere are all-male groups that intermittently drive out the resident male; after infighting, one male then drives out the rest. Here’s his new domain, consisting of females with the babies of the previous male. And crucially, the average tenure of a breeding male (about twenty-seven months) is shorter than the average interbirth interval. No females are ovulating, because they’re nursing infants; thus this new stud will be booted out himself before any females wean their kids and resume ovulating. All for nothing, none of his genes passed on.

  What, logically, should he do? Kill the infants. This decreases the reproductive success of the previous male and, thanks to the females ceasing to nurse, they start ovulating.*

  That’s the male perspective. What about the females? They’re also into maximizing copies of genes passed on. They fight the new male, protecting their infants. Females have also evolved the strategy of going into “pseudoestrus”—falsely appearing to be in heat. They mate with the male. And since males know squat about female langur biology, they fall for it—“Hey, I mated with her this morning and now she’s got an infant; I am one major stud.” They’ll often cease their infanticidal attacks.

  Despite initial skepticism, competitive infanticide has been documented in similar circumstances in 119 species, including lions, hippos, and chimps.9

  A variant occurs in hamsters; because males are nomadic, any infant a male encounters is unlikely to be his, and thus he attempts to kill it (remember that rule about never putting a pet male hamster in a cage with babies?). Another version occurs among wild horses and gelada baboons; a new male harasses pregnant females into miscarrying. Or suppose you’re a pregnant mouse and a new, infanticidal male has arrived. Once you give birth, your infants will be killed, wasting all the energy of pregnancy. Logical response? Cut your losses with
the “Bruce effect,” where pregnant females miscarry if they smell a new male.10

  Thus competitive infanticide occurs in numerous species (including among female chimps, who sometimes kill infants of unrelated females).11 None of this makes sense outside of gene-based individual selection.

  Individual selection is shown with heartbreaking clarity by mountain gorillas, my favorite primate.12 They’re highly endangered, hanging on in pockets of high-altitude rain forest on the borders of Uganda, Rwanda, and the Democratic Republic of the Congo. There are only about a thousand gorillas left, because of habitat degradation, disease caught from nearby humans, poaching, and spasms of warfare rolling across those borders. And also because mountain gorillas practice competitive infanticide. Logical for an individual intent on maximizing the copies of his genes in the next generation, but simultaneously pushing these wondrous animals toward extinction. This isn’t behaving for the good of the species.

  KIN SELECTION

  To understand the next foundational concept, reflect on what it means to be related to someone and to pass on copies of “your” genes.

  Suppose you have an identical twin, with the same genome as you. As a startling, irrefutable fact, in terms of the genes being passed on to the next generation, it doesn’t matter if you reproduce or sacrifice yourself so that your twin reproduces.

  What about a full sibling who isn’t an identical twin? Recall from chapter 8 that you’d share 50 percent of your genes with him.* Thus reproducing once and dying so that he reproduces twice are evolutionarily identical. Half sibling, 25 percent of genes in common, calculate accordingly. . . .

  The geneticist J. B. S. Haldane, who, when asked if he’d sacrifice his life for a brother, is credited to have quipped, “I’ll gladly lay down my life for two brothers or eight cousins.” You can leave copies of your genes in the next generation by reproducing, but also by helping relatives reproduce, especially closer relatives. Hamilton formalized this with an equation factoring in the costs and benefits of helping someone, weighted by their degree of relatedness to you. This is the essence of kin selection.* This explains the crucial fact that in countless species, whom you cooperate with, compete with, or mate with depends on their degree of relatedness to you.

  Mammals first encounter kin selection soon after birth, reflecting something monumentally obvious: females rarely nurse someone else’s infants. Next, among numerous primates the mother of a newborn and an adolescent female may commence a relationship fraught with pluses and minuses—the mother occasionally lets the adolescent care for her offspring. For the mother the plus is getting time to forage without baby on board; the minus is that the babysitter may be incompetent. For the adolescent the plus is getting mothering experience; the minus, the effort of child care. Lynn Fairbanks of UCLA has quantified the pluses and minuses of such “allomothering” (including that adolescents who practiced mothering have a better survival rate for their own kids). And who is a frequent “allomother”? The female’s kid sister.13

  An extension of allomothering is the cooperative breeding of New World monkeys like marmosets. In their social groups only one female breeds, while the others—typically younger relatives—help with child care.14

  The extent to which a male primate cares for infants reflects his certainty of paternity.15 Among marmosets, who form stable pair-bonds, males do most of the child care. In contrast, among baboons, where a female mates with multiple males during her estrus cycle, it’s only the likely fathers (i.e., males who mated on the female’s most fertile day, when she had her most conspicuous estrus swelling) who invest in the well-being of the child, aiding him in a fight.*

  Among many primates, how often you groom someone depends on how closely related they are to you. Among baboons, females spend their whole life in their natal troop (whereas males migrate to a new troop at puberty); as a result, adult females have complex cooperative kinship relations and inherit their dominance rank from their mother. Among chimps it’s the opposite; females leave home at puberty, and kin-based adult cooperation occurs only among males (for example, where groups of related males attack solitary males from neighboring groups). And among langurs, when a female defends her infant against a new male, she most often is helped by elderly female relatives.

  Moreover, primates understand kinship. Dorothy Cheney and Robert Seyfarth of the University of Pennsylvania, studying wild vervet monkeys, have shown that if animal A is crummy to animal B, afterward, B is more likely to be crummy to A’s relatives. And if A is lousy to B, B’s relatives are more likely to be crummy to A. Furthermore, if A is lousy to B, B’s relatives are more likely to be crummy to A’s relatives.16

  In beautiful “playback” experiments, Cheney and Seyfarth first recorded vocalizations from each vervet in a group. They’d place a speaker in some bushes, and when everyone was sitting around, they’d play a recording of some kid giving a distress call. And the females would all look at the kid’s mother—“Hey, that’s Madge’s kid. What’s she going to do?” (Note that this also shows that monkeys recognize voices.)

  In a study of wild baboons, Cheney and Seyfarth would wait for two unrelated females to sit near the bush with the speaker and then play one of three vocalizations: (a) sounds of the two females’ relatives fighting with each other; (b) a relative of one fighting with a third party; (c) two other random females fighting.17 If a female’s relative was involved in the fighting, she’d look toward the speaker longer than if there were no relatives involved. And if it was relatives of the two females fighting each other, the higher-ranking one would remind the subordinate of her place by supplanting her from her spot.

  Another playback study created some baboon virtual reality.18 Baboon A dominates baboon B. Thanks to cutting and splicing of recordings of vocalizations, baboon A is heard making a dominance vocalization, B making a subordination one. When this happens, no baboons looked at the bushes—A > B, boring status quo. But if baboon A is heard making a subordination vocalization after B makes a dominance one—a rank reversal—everyone orients to the bushes (“Did you hear what I just heard?”). Then a third scenario—a dominance reversal between two members of the same family. And no one looks, because it’s uninteresting. (“Families, they’re crazy. You should see mine—we have these huge dominance reversals and are hugging an hour later.”) Baboons “classify others simultaneously according to both individual rank and kinship.”

  Thus other primates contemplate kinship with remarkable sophistication, with kinship determining patterns of cooperation and competition.

  Nonprimates are also into kin selection. Consider this—sperm in a female’s vaginal tract can aggregate, allowing them to swim faster. Among a deer mouse species where females mate with multiple males, sperm aggregate only with sperm from the same individual or a close relative.19

  As behavioral examples, squirrels and prairie dogs give alarm vocalizations when spotting a predator. It’s risky, calling attention to the caller, and such altruism is more common when in the proximity of relatives. Social groups built around female relatives occur in numerous species (e.g., lion prides, where related females nurse one another’s cubs). Moreover, while prides typically contain a single breeding male, on those occasions when it’s two males, better than chance that they’re brothers. There is a striking similarity in humans. Most cultures have historically allowed polygyny, with monogamy as the rarer beast. Even rarer is polyandry—multiple men married to one woman. This occurs in northern India, Tibet, and Nepal, where the polyandry is “adelphic” (aka “fraternal”)—a woman marries all the brothers of one family, from the strapping young man to his infant brother.*20

  —

  A challenging implication of kin selection arises.

  Those hot cousins. If one accrues fitness benefits by helping relatives pass on copies of their genes, why not help them do that by mating with them? Yech; inbreeding produces decreased fertility and those genetic unpleasant
nesses in European royalty.*21 So the dangers of inbreeding counter the kin-selection advantages. Theoretical models suggest that the optimal balance is third-cousin matings. And indeed, numerous species prefer to mate with between a first and a third cousin.22

  This occurs in insects, lizards, and fish, where, on top of that, cousin-mating pairs invest more in the rearing of their offspring than do unrelated parents. A preference for cousin matings occurs in quail, frigate birds, and zebra finches, while among pair-bonded barn swallows and ground tits, females sneak out on their partner to mate with cousins. Similar preferences occur in some rodents (including the Malagasy giant jumping rat, a species that sounds disturbing even without cousins shacking up with each other).23

  And what about humans? Something similar. Women prefer the smell of moderately related over unrelated men. And in a study of 160 years of data concerning every couple in Iceland (which is a mecca for human geneticists, given its genetic and socioeconomic homogeneity), the highest reproductive success arose from third- and fourth-cousin marriages.24

  Recognizing Relatives?

  These findings concerning kin selection require animals to recognize degrees of relatedness. How do they do this?

  Some species have innate recognition. For example, place a mouse in an arena; at one end is an unrelated female, at the other, a full sister from a different litter, never encountered before. The mouse spends more time with the sister, suggesting genetically based kin recognition.

  How does this work? Rodents produce pheromonal odors with individual signatures, derived from genes called the major histocompatibility complex (MHC). This is a super variable gene cluster that produces unique proteins that form a signature for an individual. This was first studied by immunologists. What does the immune system do? It differentiates between you and invaders—“self” and “nonself”—and attacks the latter. All your cells carry your unique MHC-derived protein, and surveillance immune cells attack any cell lacking this protein password. And MHC-derived proteins also wind up in pheromones, producing a distinctive olfactory signature.