Behave: The Biology of Humans at Our Best and Worst

by Robert M. Sapolsky

  The evolutionary biologists were delighted to generate increasingly sophisticated models with the theoretical economists and theoretical diplomats and theoretical war strategists. The real question was whether animal behavior actually fits any of these models.

  One bizarre animal system suggests Tit for Tat enforcement of cooperation involving the black hamlet fish, which form stable pair-bonds.37 Nothing strange there. The fish can change sex (something that occurs in some fish species). As per usual, reproduction is more metabolically costly for the female than the male. So the fish in a pair take turns being the more expensive female. Say fish A and fish B have been doing their sex-change tango, and most recently A was the expensive female and B the cheap male. Suppose B cheats by staying male, forcing A to continue as female; A switches to male and stays that way until B regains his social conscience and becomes female.

  Another widely cited study suggested a Tit for Tat strategy among stickleback fish.38 The fish is in a tank, and on the other side of a glass partition is something scary—a bigger cichlid fish. The stickleback tentatively darts forward and back, investigating. Now put a mirror in its tank, perpendicular to the axis of the two fish. In other words, thanks to the mirror, there appears to be a second cichlid next to the first. Terrifying, except from out of nowhere there’s this mysterious second stickleback who checks out the second cichlid every time our hero checks out the first—“I have no idea who this guy is, but we’re an amazing, coordinated team.”

  Now convince the stickleback his partner is defecting. Angle the mirror so that the stickleback’s reflection is deflected backward. Now when the fish darts forward, his reflection does as well, but—that jerk!—it looks like he’s hanging back safely (lagging back even half a body length decreases the likelihood of a fish being predated). When the fish believes his partner is defecting, he stops darting forward.

  Greater complexity in Tit for Tat–ing is suggested by some animals having multiple roles in their social groups.39 Back to the playback technique with lions, where the roar of a strange male emanated from a speaker in the bushes (or from a life-sized model of a lion). Lions tentatively came forward to investigate, a risky proposition. Consistently, certain lions hung back. The toleration of these habitual scaredy-cats seemed to violate the demands of reciprocity, until it was recognized that such animals took the lead in other domains (e.g., in hunts). A similar punch line emerges concerning the Damaraland mole rat. The social groups of it and its relative, the naked mole rat, resemble those of social insects, with nonreproductive workers and a single breeding queen.* Researchers noted some workers who never worked and were considerably fatter than the rest. It turns out that they have two specialized jobs—during the rains, they dig through flooded, collapsed tunnels of the burrows, and when necessary, they disperse with the risky task of starting a new colony.

  I’m not convinced that a Tit for Tat reciprocity has been clearly demonstrated in other species. But evidence of its strict use would be hard for Martian zoologists to document in humans—after all, there are frequently pairs where one human does all the labor, the other doing nothing other than intermittently handing him some green pieces of paper. The point is that animals have systems of reciprocity with sensitivity to cheating.

  Gigantic Question #2: How Can Cooperation Ever Start?

  So a handful of Tit for Tat–ers can outcompete a mix of other strategies, including highly exploitative, uncooperative ones, losing the battles but winning the war. But what if there’s only one Tit for Tat–er in a population of ninety-nine Always Defect–ers? Tit for Tat doesn’t stand a chance. Always Defect–ers playing each other produces the second-worst outcome for each. But a Tit for Tat–er playing an Always Defect–er does worse, getting the sucker payoff that first round before becoming a de facto Always Defect–er. This raises the second great challenge for reciprocal altruism: forget which strategy is best at fostering cooperation—how do you ever start any type? Amid a sea of Always Defect–ers, the first black hamlet fish, mole rat, or Dictyostelium amoeba who, after reading Gandhi, Mandela, Axelrod, and Hamilton, takes the first altruistic step is screwed, lagging behind everyone else forever. One can practically hear the Always Defect amoebas chortling derisively.

  Let’s make it slightly easier for Tit for Tat to gain a foothold. Consider two Tit for Tat–ers amid ninety-eight Always Defect–ers. Both will crash and burn . . . unless they find each other and form a stable cooperative core, where the Always Defect–ers either must switch to Tit for Tat or go extinct. A nidus of cooperation crystallizes outward through the population.

  This is where green-beard effects help, conspicuous features of cooperators that help them recognize one another. Another mechanism is spatial, where the cooperative trait itself facilitates cooperators finding one another.

  Another route has been suggested for jump-starting reciprocal altruism. Occasionally a geographic event occurs (say, a land bridge disappears), isolating a subset of a population for generations. What happens in such a “founder population”? Inbreeding, fostering cooperation via kin selection. Eventually the land bridge reappears, the inbred cooperative founder population rejoins the main group, and cooperation propagates outward.*

  We return to the issue of starting cooperation in the final chapter.

  STANDING ON THREE LEGS

  We’ve now seen the three foundations of thinking about the evolution of behavior—individual selection, kin selection, and reciprocal altruism. Moreover, we’ve seen how these three concepts can explain otherwise puzzling behaviors. Some concern individual selection, with competitive infanticide as the canonical example. Other behaviors are most explicable with kin selection—why there’s male-male aggression between groups in only some primate species; why many species have hereditary ranking systems; why cousin matings are more frequent than one might expect. And some behaviors are all about reciprocal altruism. Why else would a vampire bat, aware of the vanquishing power of group selection, regurgitate blood for someone else’s kid?

  Let’s consider a few more examples.

  Pair-Bonding Versus Tournament Species

  Suppose you’ve discovered two new species of primates. Despite watching both for years, here’s all you know: In species A, male and females have similar body sizes, coloration, and musculature; in species B, males are far bigger and more muscular than females and have flashy, conspicuous facial coloration (jargon: species B is highly “sexually dimorphic”). We’ll now see how these two facts allow you to accurately predict a ton of things about these species.

  Male-female pairs of tamarins (top) and mandrills (bottom)

  First off, which species has dramatic, aggressive conflict among males for high dominance rank? Species B, where males have been selected evolutionarily for fighting skills and display. Species A males, in contrast, are minimally aggressive—that’s why males haven’t been selected for muscle.

  What about variability in male reproductive success? In one species 5 percent of the males do nearly all the mating; in the other, all males reproduce a few times. The former describes species B—that’s what all the rank competition is about—the latter, species A.

  Next, in one species, if a male mates with a female and she conceives, he’ll do a ton of child care. In contrast, no such male “parental investment” is seen in the other species. No-brainer: the former describes species A; the few species B males who father most of the kids sure aren’t doing child care.

  One species has a tendency to twin, the other not. Easy—the twinning is in species A, with two sets of parental hands available.

  How picky are males about whom they mate with? In species B, males mate with anyone, anywhere, anytime—it only costs the price of some sperm. In contrast, males of species A, with its rule of “You get her pregnant, you do child care,” are more selective. Related to that, which species forms stable pair-bonds? Species A, of course.

  After correcting fo
r body size, which species’ males have bigger testes and higher sperm count? It’s species B, ever prepared for mating, should the opportunity arise.

  What do females look for in a potential mate? Species B females get nothing from a male except genes, so they should be good ones. This helps explain the flamboyant secondary sexual characteristics of males—“If I can afford to waste all this energy on muscle plus these ridiculous neon antlers, I must be in great shape, with the sorts of genes you’d want in your kids.” In contrast, species A females look for stable, affiliative behavior and good parenting skills in males. This is seen in bird species with this pattern, where males display parenting expertise during courtship—symbolically feeding the female with worms, proof that he’d be a competent worm winner. Related to that, among bird versions of species A and B, in which is a female more likely to abandon her offspring, passing on more copies of her genes by breeding with another male? Species A, where you see “cuckoldry”—because the male is going to stick there, caring for the kids.

  Related to that, in species A, females compete aggressively to pair-bond with a particularly desirable (i.e., paternal) male. In contrast, species B females don’t need to compete, since all they get from males is sperm, and there’s enough to go around from desirable males.

  Remarkably, what we’ve described here is a broad and reliable dichotomy between two social systems, where A is a “pair-bonding” species, B a “tournament” species.*

  Pair-Bonded

  Tournament

  Male parental behavior

  Extensive

  Minimal

  Male mating pickiness

  High

  Low

  Variability in male reproductive success

  Low

  High

  Testes size, sperm count

  Small/low

  Large/high

  Levels of male-male aggression

  Low

  High

  Degree of sexual dimorphism in body weight, physiology, coloration, and life span

  Low

  High

  Females select for

  Parenting skill

  Good genes

  Rates of cuckoldry

  High

  Low

  Primates that pair-bond include South American monkeys like marmosets, tamarins, and owl monkeys, and among the apes, gibbons (with nonprimate examples including swans, jackals, beavers, and, of course, chapter 4’s prairie voles). Classic tournament species include baboons, mandrills, rhesus monkeys, vervets, and chimps (with nonprimate examples including gazelles, lions, sheep, peacocks, and elephant seals). Not all species fit perfectly into either extreme (stay tuned). Nonetheless, the point is the internal logic with which the traits of each of these types of species cluster, based on these evolutionary principles.

  Parent-Offspring Conflict

  Another feature of behavior turns kin selection on its head. The emphasis until now has been on the fact that relatives share many genes and evolutionary goals. Nonetheless, except for identical twins, just as pertinent is relatives not sharing all their genes or goals. Which can cause conflict.

  There’s parent-offspring conflict. One classic example is whether a female should give her child great nutrition, guaranteeing his survival, but at the cost of nutrition for her other children (either current or future). This is weaning conflict.40

  This causes endless primate tantrums.41 Some female baboon looks frazzled and cranky. Three steps behind is her toddler, making the most pitiful whimpering and whining sounds imaginable. Every few minutes the kid tries to nurse; Mom irritably pushes him away, even slaps him. More wailing. It’s parent-offspring weaning conflict; as long as Mom nurses, she’s unlikely to ovulate, curtailing her future reproductive potential. Baboon moms evolved to wean their kids at the age where they can feed themselves, and baboon kids evolved to try to delay that day. Interestingly, as females age, with decreasing likelihood of a future child, they become less forceful in weaning.*

  There’s also mother-fetus conflict. You’re a fetus with an evolutionary agenda. What do you want? Maximal nutrition from Mom, and who cares if that impacts her future reproductive potential? Meanwhile, Mom wants to balance current and future reproductive prospects. Remarkably, fetus and Mom have a metabolic struggle involving insulin, the pancreatic hormone secreted when blood glucose levels rise, which triggers glucose entry into target cells. The fetus releases a hormone that makes Mom’s cells unresponsive to insulin (i.e., “insulin resistant”), as well as an enzyme that degrades Mom’s insulin. Thus Mom absorbs less glucose from her bloodstream, leaving more for the fetus.*

  Intersexual Genetic Conflict

  In some species the fetus has an ally during maternal/fetal conflict—the father. Consider a species where males are migratory, mating with females and then moving on, never to be seen again. What’s a male’s opinion about maternal/fetal conflict? Make sure the fetus, i.e., his child, grabs as much nutrition as possible, even if that lessens Mom’s future reproductive potential—who cares, that won’t be his kid down the line. He’s more than just rooting for his fetus.

  This helps explain a mysterious, quirky feature of genetics. Normally a gene works the same way, regardless of which parent it comes from. But certain rare genes are “imprinted,” working differently, or only being activated, depending on the parent of origin. Their purpose was discovered in a creative synthesis by evolutionary biologist David Haig of Harvard. Paternal imprinted genes bias toward more fetal growth, while maternal imprinted genes counter this. For example, some paternal genes code for potent versions of growth factors, while the maternal genes code for growth factor receptors that are relatively unresponsive. A paternally derived gene expressed in the brain makes newborns more avid nursers; the maternally derived version counters this. It’s an arms race, with Dad genetically egging on his offspring toward more growth at the cost of the female’s future reproductive plans, and Mom genetically countering this with a more balanced reproductive strategy.*

  Tournament species, where males have minimal investment in a female’s future reproductive success, have numerous imprinted genes, while pair-bonders don’t.42 What about humans? Stay tuned.

  MULTILEVEL SELECTION

  So we’ve got individual selection, kin selection, and reciprocal altruism. And then what happened in recent years? Group selection reappeared, sneaking in the back door.

  “Neo–group selection” crashed a long-standing debate as to the “unit of selection.”

  Genotype Versus Phenotype, and the Most Meaningful Level of Selection

  To appreciate this, let’s contrast genotype and phenotype. Genotype = someone’s genetic makeup. Phenotype = the traits observable to the outside world produced by that genotype.*

  Suppose there’s a gene that influences whether your eyebrows come in two separate halves or form a continuous unibrow. You’ve noted that unibrow prevalence is decreasing in a population. Which is the more important level for understanding why—the gene variant or the eyebrow phenotype? We know after chapter 8 that genotype and phenotype are not synonymous, because of gene/environment interactions. Maybe some prenatal environmental effect silences one version of the gene but not the other. Maybe a subset of the population belongs to a religion where you must cover your eyebrows when around the opposite sex, and thus eyebrow phenotype is untouched by sexual selection.

  You’re a grad student re
searching unibrow decline, and you must choose whether to study things at the genotypic or phenotypic level. Genotypic: sequencing eyebrow gene variants, trying to understand their regulation. Phenotypic: examining, say, eyebrow appearance and mate choice, or whether unibrows absorb more heat from sunlight, thereby damaging the frontal cortex, producing inappropriate social behavior and decreased reproductive success.

  This was the debate—is evolution best understood by focusing on genotype or phenotype?

  The most visible proponent of the gene-centered view has long been Dawkins, with his iconic “selfish gene” meme—it is the gene that is passed to the next generation, the thing whose variants spread or decline over time. Moreover, a gene is a clear and distinctive sequence of letters, reductive and irrefutable, while phenotypic traits are much fuzzier and less distinct.

  This is the core of the concept of “a chicken is just an egg’s way of making another egg”—the organism is just a vehicle for the genome to be replicated in the next generation, and behavior is just this wispy epiphenomenon that facilitates the replication.

  This gene-centered view can be divided in two. One is that the genome (i.e., the collection of all the genes, regulatory elements, and so on) is the best level to think about things. The more radical view, held by Dawkins, is that the most appropriate level is that of individual genes—i.e., selfish genes, rather than selfish genomes.