Courtship in the
Pleistocene
To judge a new theory of human evolution, it can be more important to forget one’s preconceptions than to learn a set of new facts and ideas. Most of our images of human evolution come from popular culture. Film, television, cartoons, and advertising have filled our heads with a lot of colorful nonsense about prehistory. If the image in your mind is of cave-men clubbing cave-women unconscious and dragging them off, you may not grant sexual choice much significance in human evolution. This chapter aims to confront these preconceptions, inquiring how our ancestors did and did not form sexual relationships.
Popular culture images of prehistory are divided by market segmentation according to consumer age group, and by sexual content ratings. There is a children’s G-rated version of prehistory that eliminates all sex and most violence, where neither sexual selection nor natural selection have much force. Playmobil toy sets include multi-ethnic cave-men happily living alongside dinosaurs, hunting lions, and living in jungles. The Flintstones cartoons depicted a prehistory of capitalist affluence, suburban family values, and chaste monogamy. In these Gardens of Eden there is no hint of reproductive competition, the engine of evolution.
Then there is a “Parental Guidance” prehistory, with a bit more violence and a few coy allusions to romance. Our PG version of prehistory is usually compiled from Planet of the Apes films, television cartoons about time-traveling teenagers, school trips to natural history museums, and summer camp experiences with the odd broken bone or stinging insect. Since this version emphasizes adventure, danger, and survival, it makes more plausible the idea that our minds evolved for toolmaking, hunting, and warfare. The resulting theory of human evolution resembles the opening sequence of Stanley Kubrick’s 2001: A Space Odyssey, in which proto-human apes conquer their rivals by inventing bone clubs, which put us straight on the technological path to moon-going spacecraft. The PG version never shows how the proto-humans produced any offspring, so sexual selection remains invisible.
Adult versions of prehistory include sexual content, but almost always in the form of a prurient male fantasy where female choice is irrelevant. Please, forget the sexual favors Raquel Welch bestowed on the dinosaur-slaying cave-man in the film One Million Years B.C. Do not take seriously the scene in Quest for Fire in which a rough stranger visiting a more sophisticated tribe is invited to copulate with all of the tribe’s fertile women. Erase the memory of Daryl Hannah’s rape by Neanderthals in Clan of the Cave Bear. The torrid paleolithic romances of Jean Auel are good entertainment, as are the erotic daydreams that may float through the minds of college students during springtime physical anthropology courses. However, they are not good touchstones for judging a theory of mental evolution through sexual choice.
Most media portrayals of prehistory follow one of three strategies: eliminate sexual content entirely, show cave-women falling for adventure heroes who rescue them from peril, or offer a narcissistic sexual fantasy in which only the protagonist (usually male) exercises sexual choice. There seems to be no market for portrayals of our early ancestors exerting mutual choice. If we are to see all the genuine tensions and difficulties between the sexes, media producers assume we must be rewarded with a proper costume drama set in Imperial Rome or Regency England. After all, could Alan Rickman and Sigourney Weaver keep a straight face playing an intense romantic psychodrama set in Pleistocene Zaire, while wearing mangy furs, with ochre-smeared hair, and covered in ticks?
Maybe not, but a romantic psychodrama is just what we need to envision how sexual choice may have worked during human evolution. This is not a vain hope. In some ways we are better positioned to understand sexual selection than survival selection. The sexual challenges our ancestors faced were created by other members of their own species. Likewise today. If our thoughts and feelings about sexual relationships are not too different from those of our ancestors, then our sexual challenges must not be too different. We get infatuated, we fall in love, we feel ecstatic, jealous, or heartbroken, we grow bored with some partners, and, if lucky, we develop a companionable attachment to the sexual partners with whom we raise children. We are attracted to beautiful faces and bodies, but also to a good sense of humor, a kind personality, a keen intelligence, and a high social status. If these sexual tastes are part of human nature that evolved gradually, our ancestors must have felt similarly to some degree. We should not automatically project modern social arrangements back into prehistory, but it is probably valid to project our individual emotions on to our ancestors.
By contrast, it can be difficult to appreciate the survival challenges that shaped our mental adaptations. In the developed world, we drive around in cars, live in the same house for years, use money to buy food, work hard at specialized jobs, and go to hospitals when ill. Our ancestors had to walk everywhere, lived in makeshift shelters in dozens of different places every year, did little work other than foraging for food, and when they fell ill, they either recovered spontaneously or died. The economics of surviving have changed dramatically, while the romantic challenges of mating have remained rather similar.
Why are evolutionary psychologists so preoccupied with the Pleistocene? The Pleistocene was a geological epoch uniquely important in human evolution, because it included the evolution of all that is distinctively human. At the beginning of the Pleistocene, 1.6 million years ago, our ancestors were still relatively small-brained apes who walked upright and made just a few crude stone tools. They were almost certainly without language, music, art, or much creative intelligence. At the end of the Pleistocene, just 10,000 years ago, our ancestors were already modern humans, identical to us in bodily appearance, brain structure, and psychology. The evolution that shaped human nature all took place in the Pleistocene.
After the Pleistocene came the Holocene, occupying the last 10,000 years. The Holocene includes all of recorded history. During the Holocene, humans spread around the planet, invented agriculture, money, and civilization, and grew from populations of a few million to a few billion. The Holocene has been historically crucial but evolutionarily unimportant. Ten thousand years is only four hundred human generations, probably not enough time to evolve many new psychological adaptations. But it is plenty of time for runaway sexual selection to make populations diverge a bit in some aspects of body shape, facial appearance, and psychological traits. However, this book is not concerned with such relatively minor differences between populations. It is concerned with universal human mental abilities that our closest ape relatives do not share.
The Holocene changed patterns of human mating and reproduction dramatically. It saw the emergence of inherited wealth, arranged marriages, hierarchical societies, patriarchy, feminism, money, prostitution, monogamous marriage, harems, personal ads, telephones, contraception, and abortion. These make modern courtship rather different from Pleistocene courtship. But Pleistocene courtship is what drove sexual selection during the relevant period of human evolution, and human behavior in the Holocene still reflects our Pleistocene legacy.
Knowing that the human mind’s distinctive abilities evolved in the Pleistocene makes evolutionary psychology much easier. It means that all the ancestral environments that shaped the basic mental capacities of our species were physically contained within the African continent, since all pre-human ancestors lived in Africa, and humans spread out of Africa only towards the end of the Pleistocene. Our ancestors lived in areas of sub-Saharan Africa that contained mixtures of open savanna, scrub, and forest. Instead of caves or jungles, picture Africa’s broad, flat plains, with their baobabs and acacias, their wet and dry seasons, their hot days and cool nights, their plentiful hoofed herds and rare, emaciated predators, the incandescent sun, and millions of scrabbling insects.
A fairly coherent picture of Pleistocene life has emerged from anthropology, archeology, paleontology, primatology, and evolutionary psychology. Like other social primates, our hominid ancestors lived in small, mobile groups. Females and their children distributed themselves in relation to where the wild plant food grew, and clustered in groups for mutual protection against predators. Males distributed themselves in relation to where the females were. Many members of each group would have been blood relatives. Group membership may have varied daily and seasonally, according to opportunities for finding food and exploiting water sources.
Our ancestors would have known at least a hundred individuals very well by face and by personality. During their lifetimes they would have come into contact with several hundred or thousand members of the same local population. Almost all sexual partners would have been drawn from this larger tribal group, which, after language evolved, would probably have been identified by their shared dialect.
During the days, women would have gathered fruits, vegetables, tubers, berries, and nuts to feed themselves and their children. Men would have tried to show off by hunting game, usually unsuccessfully, returning home empty-handed to beg some yams from the more pragmatic womenfolk. Our ancestors probably did not have to work more than twenty or thirty hours a week to gather enough food to live. They did not have weekends or paid vacation time, but they probably had much more leisure time than we do.
There was intermittent danger from predators, parasites, and germs, but our ancestors would have become as accustomed to coping with those dangers as we are to crossing roads. Nature was not red in tooth and claw. Usually, it was really boring. Predators would have tended to kill the very young, the very ill, the very old, and the very foolish. Most illnesses would have been due to poor condition brought on by starvation or injury. Our ancestors did not spend all their time worrying about survival problems. They were among the longest-lived species on the planet, which implies that their daily risk of death was minuscule. Like most great apes, they probably spent their time worrying about social and sexual problems.
For most of evolution, our ancestors ranged across wide areas without being tied to a single home base or territory. They owned no more than they could carry, had no money, inherited no wealth, and could not store food today to insure against starvation next month. If individuals consistently appeared healthy, energetic, and well-fed, it was not because they were born rich. It must have been because they were good at foraging and good at making friends who took care of them during rough patches.
To understand how sexual selection may have operated in the Pleistocene, we have to ask how sexual relationships and sexual choice may have worked. We know that our hominid ancestors did not take each other out to restaurants and films, give each other engagement rings, or wear condoms. But what can we say about how they did select mates? We’ll start with a look at sexual choice in other primates, and then consider what was distinctive about sexual choice among our hominid ancestors.
In most primate species, the distribution of food in the environment determines the distribution of females, and the distribution of females determines the distribution of males. When food is so dispersed that females do best by foraging on their own, males disperse to pair up with the lone females. This gives rise to monogamous couples. It is a fairly rare pattern among primates, limited to gibbons, some lemurs, and some African and South American monkeys.
When food comes in patches large enough for several females to share, they tend to band together in small groups to find the food, and to protect each other against predators, unwanted males, and competing female groups. As long as the female band is not too large, a single male can exclude other males from sexual access to the band, which thus becomes “his.” This “harem system” of single-male polygyny is fairly common in primates, being found in hamadryas baboons, colobus monkeys, some langurs, and gorillas. The competition between males to guard the female groups creates very strong sexual selection pressures for male size, strength, aggressiveness, and large canine teeth.
When food comes in still larger patches, female groups can grow too large for any single male to defend them. The males must then form coalitions, resulting in a complex multi-male, multi-female group, as in some baboons, macaques, ring-tailed lemurs, howler monkeys, and chimpanzees. Our hominid ancestors probably lived in such groups, in which sexual selection gets more complicated. Sometimes, females in multi-male groups appear to use sperm-production ability as the main fitness indicator. A chimpanzee female might mate with every male in the group every time she becomes fertile. She lets their sperm fight it out in her reproductive tract, and the strongest swimmers with the best endurance will probably fertilize her egg.
In response to this sexual selection for good sperm, male chimpanzees have evolved large testicles, copious ejaculates, and high sperm counts. Female primates face a trade-off. They can select for the best-swimming sperm by mating very promiscuously, or they can select for the best courtship behavior by mating very selectively. Or they can do a little of both, selecting a small group of male lovers for their charm and then letting their sperm fight it out.
In species that do not get completely caught up in runaway sperm competition, females can favor various male behavioral traits. Multi-male groups obviously allow greater scope for females to choose between males. If they favor dominant males, males evolve through sexual selection to compete intensely for social status by individual force or by forming coalitions. If females favor kind males, males evolve through sexual selection to groom females, protect their offspring, and guard them from other males.
Given multi-male, multi-female primate groups, how does mate choice work? Female primates can exercise choice by joining groups that contain favored males, initiating sex with them during estrus, supporting them during conflicts, and developing long-term social relationships with them. Females can reject unfavored males by refusing to cooperate during copulation attempts, driving males away from the group, or leaving the group. But female mate choice criteria remain obscure for most primate species. In contrast to modern humans, female primates rarely favor males who can provide resources or paternal care of offspring. The sporadic male care that is observed, such as watching, carrying, and protecting infants, is better described as courtship effort than as paternal care. The male is unlikely to be the infant’s father, but is simply trying to mate with the infant’s mother by doing her a favor.
Primate researchers still know little about what traits are preferred by male and female primates. For example, we know less about female choice in other apes than we do about female choice in the Tungara frog, the guppy fish, or the African long-tailed widowbird. Nevertheless, three kinds of female preference have been reported in primates: preferences for high-ranking males capable of protecting females and offspring from other males; preferences for male “friends” that have groomed the female a lot and have been kind to her offspring; and preferences for new males from outside the group, perhaps to avoid genetic inbreeding. Each sort of preference could be explained in terms of female choice for good genes, or female choice for material and social benefits. Although male primates have evolved an astounding diversity of beards, tufts, and colorful hair styles, there has been very little research on female choice for male appearance. Also, there has been virtually no research on primate sexual choice for personality or intelligence. Female primates are sometimes reported to show “irrational” or “capricious” preferences that cannot be explained on the basis of male dominance, age, or group membership. Sometimes two primates just seem to like each other based on unknown features of appearance, behavior, or personality. Female primates might well be choosing males for their personalities and not just their status, but we do not know.
Most primates follow the general animal pattern of male sexual competition and female choosiness. But when the costs of male sexual competition and courtship are high, males also have incentives to be choosy. When male mate choice becomes important, sexual selection affects females as well as males. In monogamous marmosets and tamarins, females compete to form pairs with quality males and drive off competing females. In single-male harem systems, the dominant male’s sperm can become a limiting resource for female reproduction, and high-ranking females prevent low-ranking females from mating through aggression and harassment. In multi-male groups, females sometimes compete to form consortships and friendships with favored males. Such patterns of female competition suggest some degree of male mate choice. When the costs of sexual competition and courtship are high, males have an incentive to be choosy about how they spread their sexual effort among the available females. Males compete much more intensely for females who show signs of fertility such as sexual maturity, estrus swellings, and presence of offspring. Like females, some male primates also develop special friendships with particular sexual partners. It may not be romantic love, but, at least among some baboon pairs, it looks pretty similar.
Our closest ape relatives, the chimpanzees and the bonobos, live in multi-male, multi-female groups in which sexual choice is dynamic, intense, and complicated. Under these relentlessly social conditions, reproductive success came to depend on social intelligence rather than brute strength. Both sexes compete, both sexes have dominance hierarchies, and both sexes form alliances. Sexual relationships develop over weeks and years rather than minutes. Many primatologists and anthropologists believe that our earliest hominid ancestors probably lived under similar social and sexual conditions. Constant sociosexual strategizing in mixed-sex groups was the legacy of our ape-like ancestors. It was the starting point, not the outcome, of sexual choice in human evolution.
If we could look at the Earth through an extremely powerful telescope a million light-years away, we could see how our ancestors actually formed sexual relationships a million years ago. Until NASA approves that mission, we have to combine evidence from several less direct sources: the sexual behavior of other primates, the sexual behavior of modern humans who live as hunter-gatherers, the evidence for sexual selection in the human body and human behavior, and psychological findings on sexual behavior, sexual attraction, sexual jealousy, and sexual conflict. A number of good evolutionary psychology books already review this evidence, including David Buss’s The Evolution of Desire. A consensus is emerging about the key aspects of ancestral life, though there is still vigorous debate about many details.
Our ancestors probably had their first sexual experiences soon after reaching sexual maturity. They would pass through a sequence of relationships of varying durations over the course of a lifetime. Some relationships might have lasted no more than a few days. Given that it takes an average of three months of regular copulation before conception, very short-term partnerships would probably not produce a child. Longer-term relationships would have been much more evolutionarily important because they were much more likely to produce offspring. Indeed, in the absence of contraception the longer partnerships would almost inevitably produce a child every two or three years.
Most children were probably born to couples who stayed together only a few years. Exclusive lifelong monogamy was practically unknown. The more standard pattern would have been “serial monogamy”: a sequence of nearly exclusive sexual partnerships that were socially recognized and jealously defended. Relationships may have sometimes ended amicably, but perhaps more often one partner would reject or abandon the other, or one would happen to die. This is the pattern characteristic of most human hunter-gatherers, because they do not have the religious, legal, and property ties that reinforce ultra-long-term monogamous marriages in civilized societies.
Some desirable males were probably able to attract more than one regular sexual partner. Their polygyny opened the possibility of runaway sexual selection effects. But they were probably the exception. Much more common would have been the affairs and flings that bedevil ordinary sexual partnerships. For women, there were incentives to mate with males of higher fitness than their current partner. For men, there were incentives to mate with as many females as possible (if the current partner could stand it). Yet there were probably social pressures against such dalliances from jealous partners and their families. There is plenty of evidence from evolutionary psychology that men and women have physical, emotional, and mental adaptations for short-term liaisons and adulterous affairs. The different costs and benefits of such affairs for males and females explain most of the sex differences in human psychology. In particular, the higher incentives for males to attract large numbers of sexual partners through public displays of physical and mental fitness explain why males are so much more motivated to produce such displays.
Female mate choice was powerful in prehistory. Although sexual harassment of females by males was probably common, females could retaliate by soliciting assistance from female friends, male partners, and relatives. They would not have been jailed for killing a psychopathic stalker or an abusive boyfriend. Our female ancestors lost all visible signs of ovulation, so it would not have been possible for a would-be rapist to know when a woman was fertile. Concealed ovulation reduced the male incentives for rape, and it usually protected women from conceiving the offspring of rapists. From an evolutionary point of view, it guarded their power of sexual choice. Also, rapists would have been subject to vigilante justice by the male relatives of the victim. The power of clan members to enforce good sexual behavior is often overlooked in discussions of human evolution. Once language evolved, sexual gossip would have been a deterrent against illicit affairs, sexual harassment, and reputation-destroying rape accusations. Nevertheless, the prevalence of rape in human prehistory is still subject to intense debate. The higher the actual prevalence was, the less important female mate choice would have been, and the weaker my sexual choice theory would become.
Suppose that the level of fascination, happiness, and good humor that our ancestors felt in another individual’s company was a cue that they used to assess the individual’s mind and character. If an individual made you laugh, sparked your interest, told good stories, and made you feel well cared for, then you might have been more disposed to mate. Your pleasure in his or her presence would have been a pretty good indicator of his or her intelligence, kindness, creativity, and humor.
Now consider what happens in modern courtship. We take our dates to restaurants where we pay professional chefs to cook them great food, or to dance clubs where professional musicians excite their auditory systems, or to films where professional actors entertain them with vicarious adventures. The chefs, musicians, and actors do not actually get to have sex with our dates. They just get paid. We get the sex if the date goes well. Of course, we still have to talk in modern courtship, and we still have to look reasonably good. But the market economy shifts much of the courtship effort from us to professionals. To pay the professionals, we have to make money, which means getting a job. The better our education, the better our job, the more money we can make, and the better the vicarious courtship we can afford. Consumerism turns the tables on ancestral patterns of human courtship. It makes courtship a commodity that can be bought and sold.
During human evolution, though, one’s ability to make a good living did not automatically mean that one could buy a desired sexual partner good-quality entertainment. If you were a prehistoric hominid, you would have had to do the entertaining yourself. If you did not make a desired mate laugh, nobody would do it for you. And if they did, your date would probably run off with them instead of you.
The minds of our ancestors were relatively naked compared to ours. They did not spend twenty years in formal education ornamenting their memory with dead people’s ideas. They did not read daily newspapers so that they could recount human-interest stories. In courtship, they had to make up their ideas, stories, jokes, myths, songs, and philosophies as they went along. There was no masking a poor imagination with a good education, or a poor sense of rhythm with a good CD collection.
Perhaps even more importantly for long-term relationships, there was no television to keep your sexual partner amused after the first blush of romance faded. If they were bored in the relationship, there was no vicarious entertainment to be had. They either had to put up with your boring old self, or find a new lover. During the Holocene, when long-term monogamy thrived, people worked much harder and longer hours doing their planting, herding, trading, and career-climbing. There were fewer hours of leisure to fill, and more ways to fill them without talking to one another. Historically, humans did not begin to put up with lifelong marriage until they could no longer live off the land, property inheritance became the key to children’s survival, and couples had economic incentives to continue cooperating long after they were no longer on speaking terms. During prehistory, there were fewer economic incentives to stay together, fewer distracting entertainments to replace lost romance, and fewer ways to insulate oneself from new sexual opportunities.
Single mothers may have been the norm during most of human evolution, as they were during the previous 50 million years of primate evolution. As Sarah Blaffer Hrdy has argued in her book Mother Nature, human females have inherited a rich set of mental and physical adaptations fully sufficient to nurture their offspring with minimal assistance from males. Male help may have been a welcome luxury, but it was not a necessity.
Many Pleistocene mothers probably had boyfriends. But each woman’s boyfriend may not have been the father of any of her offspring. Or he may have been the father only of the most recent baby. Even so, his typical contribution to parenting is debatable. Males may have given some food to females and their offspring, and may have defended them from other men, but as we shall see, anthropologists now view much of this behavior more as courtship effort than paternal investment.
Viewed from the broad sweep of evolution, it is unlikely that male hominids did much direct fathering. In almost all mammals and all primates, females do almost all of the child care, with very little help from males. Males could never be sure which offspring really carried their genes, whereas females could be certain. This uncertainty about paternity leads most male mammals to invest much more in pursuing new sexual opportunities than in taking care of their putative offspring.
Like all other primates, the basic social unit among our ancestors was the mother and her children. Women clustered together for mutual help and protection. Male hominids, like males of other primate species, were probably marginal, admitted to the female group only on their forbearance. Herds of young bachelor males probably roamed around living their squalid, sexually frustrated lives, hoping they would eventually grow up enough for some group of women to take them in.
The traditional view that females needed males to protect them from predators has been challenged by an increased understanding of primate and hunter-gatherer behavior. To us, our sex differences in size and strength are salient. But to a large predator looking for an easy kill, female humans would have been only marginally less dangerous than males. Adult males may be more accurate at throwing things, but females tend to go around in larger groups while foraging, with many eyes and many hands to offer mutual vigilance and protection. An ancestral female would have been much safer in a group of a dozen sisters, aunts, and female friends than with a single male in a nuclear family. Female humans were among the largest primates ever to have evolved, and among the strongest omnivores in Africa. They did not necessarily need any help from boyfriends only 10 percent taller than themselves. Female hominids seem unlikely to have displayed the exaggerated physical vulnerability expected of women under patriarchy. When you picture ancestral females facing predators, do not imagine Marilyn Monroe whimpering and cowering. Imagine Steffi Graf brandishing a torch in place of a tennis racket.
The same group-protection effect would have guarded females against sexual predators. Ancestral women could protect one another from harassment and rape, just as other female primates do. From a female’s point of view, a strong male partner would be a mixed blessing. He could fend off unwanted attention from other males, but he could also beat you up if he got jealous or angry. Women consistently show preferences for tall, strong males in mate choice studies, but this may reflect a preference for good genes and high fitness, rather than a preference for a male capable of physical violence and intimidation that might get turned against her or her children.
Interviews with contemporary hunter-gatherer women by anthropologists such as Marjorie Shostak reveal that these women view many men as more trouble than they’re worth. If the men are hanging around, they usually eat more food than they provide, and demand more care than they give one’s children. If they have very high fitness, then their good genes, good sex, and good conversation might compensate for their messiness and lethargy. But if they are only average, their potential for sexual jealousy and violent irritability may render them a net cost rather than a benefit.
On the other hand, David Buss and other evolutionary psychologists have amassed considerable evidence that modern women generally favor tall, strong, healthy, and self-confident men, all else being equal. These traits may be favored because they would have correlated with good hunting abilities and protection abilities under ancestral conditions. However, as we shall see in the next chapter, many of these traits also reveal good genes—they are genetically heritable, and they work as effective fitness indicators. It is not yet clear whether the genetic or non-genetic benefits of such traits were more important to women. Mate choice mechanisms should evolve to capture both sorts of benefit whenever possible, so they may be difficult to disentangle.
There is still much debate about the importance of fathers in human evolution. Men show some signs of having been selected as good and helpful fathers, but our paternal instincts have not been well researched yet. Modern fathers form strong emotional attachments to their children, and this is probably an evolved propensity. A few of them even spend almost 20 percent as much time doing child care as their female partners do. Recent surveys show that Japanese fathers are starting to play with their children for almost seven minutes a day. That is a relatively high amount of paternal care compared to other male mammals. But to better understand the evolution of fathers, we need a closer look at how courtship may have overlapped with parenting.
Before contraception, our female ancestors would have produced their first child by around age 20, within a few years of reaching sexual maturity. (Female puberty probably happened several years later in prehistory than it does now, because the modern fat-rich diet artificially hastens puberty and increases teenage fertility.) Before legally imposed monogamous marriage, individuals probably passed through several sexual relationships during their reproductive years. These two patterns imply that most courtship during most of human evolution occurred between adults who already had children by previous relationships. Without nannies, nurseries, or schools, those children would have been hanging around their mothers almost all the time. (In the wild, no primate female ever grants parental custody of her children to their father after they split up.) Where there were women, there were usually already children. In modern Western societies we forget how parenting and courtship must have overlapped because we have children later in life, have very few of them, and exclude them from adult social life.
Female hominids must have juggled their courtship efforts with their mothering. Some of their courtship displays may have originated by turning normal motherly duties into better fitness indicators and entertainments. If they must tell stories to entertain their children, and if potential male mates are within earshot, they might as well make the stories appeal on both the child and the adult levels. If they must feed their children, and they want to attract a man, they might as well forage for something unusually tasty. Male mate choice almost never had the luxury of favoring a woman who did not yet have any children, who could spend all her time frolicking and canoodling. The important variable was not whether a female already had children, but whether she was a cheerful mother or a careworn mother, a beautiful mother or an ugly mother, an intelligent mother or a boring mother. Sexual competition between females was mostly sexual competition between mothers.
Moreover, mothers probably cared about the views of their children in choosing new sexual partners, so female choice must have intertwined with children’s choice. Kids who hated their mother’s new boyfriend might have destroyed his chances of sustaining a successful relationship. Mothers had good reasons to listen to their children’s likes and dislikes, because their children were the vehicles carrying their genes. The children were every mother’s paramount concern. A healthy child in hand was worth two male lovers in the bush. This put male hominids in an unusual position: their courtship had to appeal not only to mothers but to their children. This has a surprising implication. If children’s judgments influenced mate choice, then they influenced sexual selection, and children’s preferences indirectly shaped the evolution of adult male humans.
So, what did those hominid kids do to us? They did not make male humans as good at parenting as the average female mammal, but they made them better fathers than in almost any other male primate species. Men bring children food, make them toys, teach them things, and play with them. Their willingness to do this even for step-children could be viewed as a side-effect of a male adaptation for taking care of their own genetic offspring. But perhaps fatherly support and protection of step-children was the norm in the Pleistocene. If typical sexual relationships only lasted a few years, men were much more likely to be playing with some other guy’s children than their own. Many evolutionary psychologists have pointed out that what looks like paternal effort may actually have evolved through sexual choice as courtship effort. Men attracted women by pleasing their kids.
This is not to say that step-fathers are all sweetness and light. Evolutionary psychologists Martin Daly and Margo Wilson have found that men in every culture are about a hundred times more likely to beat and kill their step-children than their genetic children. There are clear evolutionary reasons for that. When male lions and langur monkeys mate with a new female, they routinely try to kill all of her existing offspring. Those offspring do not carry the males’ genes, so by killing them the males free the females to conceive their own offspring, who will carry their genes. The risk of infanticide by males is a big problem for many female primates. Yet is it much less of a worry for modern women. I want to highlight how kind most human step-fathers are compared with other male primate step-fathers. Not only do we consistently fail to kill our step-children like lions try to, we sometimes take reasonably good care of them. Surprisingly, human fathering instincts may have evolved through sexual selection for pleasing the existing children of potential female mates. Of course, where those existing children happen to be ours because we are still in a long-term sexual relationship, there are extra genetic incentives to be good fathers.
Mating among our ancestors was complicated, flexible, and strategic. When we talk about their “mating pattern,” this is just a generalization across a lot of individual strategic behavior. The individual sexual choices, not the aggregate mating pattern, drive sexual selection. To describe our ancestors as following mating patterns like “moderate polygamy” and “serial monogamy” is just a useful shorthand for identifying these sexual selection pressures.
For sexual choice to have any evolutionary effect, different individuals must produce different numbers of surviving offspring by virtue of their sexual attractiveness. How did the most attractive hominids leave more offspring? When we focus on the polygynous aspects of ancestral mating, it is easy to see. The most attractive males simply inseminate a larger proportion of females, and the least attractive males inseminate fewer. The next generation will inherit many genes from the most attractive males, and none from the least attractive. Polgyny raises the possibility of runaway sexual selection, which is driven mostly by differences in male reproductive success. Also, polygyny helps explain sex differences. The higher variation in reproductive success among males explains why male humans are so keen to show off, to dominate culture and politics, and to broadcast indicators of their fitness to any female who might listen. To the extent that our ancestors were polygynous, there were sexual selection pressures for males to display more intensely then females.
However, we should not assume that sexual selection requires polygyny. As Darwin appreciated, the sexual choices that lead to monogamous pairs can also be crucial. Is it possible that sexual selection can produce equal mental capacities for courtship in both sexes? How can the sexual choices that create monogamous couples possibly have any evolutionary effects? Sexual selection depends on differences in reproductive success, and at first glance monogamy looks as if it produces no such differences.
Suppose that sexual choice among our hominid ancestors worked as follows. Male and female hominids both tried to attract the best sexual partner they could. If they liked that partner’s company, they hung out a lot together, had a lot of sex, and produced a child. If they still liked each other after the baby arrived, they stayed together and produced another one. If they did not, they separated and looked for the best new partner they could find. Most hominids spent most of their lives in some kind of sexual relationship with somebody. Most sexual relationships longer than a few months produced at least one child.
To see how sexual selection can work even when everyone pairs up into couples, we need a thought experiment. Like all good thought experiments, it will be simplistic, unrealistic, and cartoon-like. But it will give us a surprising result. In this imaginary scenario, every hominid individual finds a sexual mate, every relationship is totally monogamous and permanent, and every relationship produces an identical number of babies. And yet, as long as sexual choice favors fitness indicators, sexual choice can still drive sexual selection by producing unequal numbers of grandchildren. Here’s how it works.
Imagine a tribe of hominids, half of them male and half female, all single, all just reaching sexual maturity at the same time. Some males have higher fitness than other males, and they advertise their higher fitness using fitness indicators such as vigorous dancing, intelligent conversing, or realistic cave-painting. Some females have higher fitness than other females, which they advertise through the same sorts of fitness indicator. Fitness is genetically heritable, so higher-fitness parents generally have higher-fitness offspring. The tribe has a tradition of strict monogamy and no infidelity. Every individual has to pick a partner once and stick with them until they die. Both sexes exercise mate choice, accepting and rejecting whomever they want.
What will happen? Each individual wants to attract the highest-fitness mate they can, because they want the best genes for their offspring. There will be a sorting process. Probably, the highest-fitness male will court the highest-fitness female first. If she is sensible, she will accept him, and they will pair off, leaving the rest of the tribe to sort themselves out. The second-highest-fitness male is disappointed. He wanted the highest-fitness female, but could not attract her. He must settle for the second-highest-fitness female. She is also disappointed, because she wanted the best male. But she settles for male number two, because she cannot do any better. Perhaps they fall in love, thanking their lucky stars that they did not end up with the cold and snooty number ones, or the repulsively inferior number threes. Now the third-highest-fitness male is doubly heart-broken. Golden female number one and silver female number two have both ignored him, leaving him to court bronze female number three. He can’t do any better, and neither can she, so they pair off. And so on. Eventually, the whole tribe sorts itself into mated pairs of roughly equal fitness.
The fitness matching does not result from any individual’s preference for a similarly ranked mate. Instead, it results from the interaction of everyone’s preferences during the sorting process. Everybody would prefer a higher-fitness mate rather than a same-fitness mate. But the opposite sex feels the same way too. For a male to mate above his fitness, a female would have to mate below her fitness. Her response to his offer will be “Dream on, loser.” Likewise for females trying to mate above their fitness. Individuals have no realistic hope of mating far above their own fitness level, or any willingness to mate below their fitness.
The result will be that mated pairs will correlate highly for fitness. If height correlates with fitness, they will be of similar height. If intelligence correlates with fitness, they will be similarly bright. If facial attractiveness correlates with fitness, they will be similarly beautiful. This is basically what we see in modern human couples: a fairly high degree of “assortative mating” for fitness indicators.
After the mated pairs start having sex, babies start arriving. To make this thought experiment challenging, let’s look at the situation where sexual selection seems weakest, and assume that every pair has exactly the same number of babies, say four babies per pair. During most of human evolution, probably only 50 percent of infants survived to sexual maturity, so two babies surviving out of four for every two parents will keep the population size stable. The question is, which mated pairs will contribute the most genes to future generations?
At first glance, it looks as if each pair should contribute the same number of genes, since they have the same number of babies. But we already know that mated pairs differ in their heritable fitness. That is what they were being choosy about when they were sorting themselves into pairs. So, the babies of higher-fitness couples will inherit higher-fitness genes. By definition, higher fitness leads to a better chance of surviving to sexual maturity. The offspring of male number one and female number one may have a very high chance of surviving. The offspring of the lowest-fitness male and the lowest-fitness female may only have a very low chance of surviving. By the time the babies’ generation grows up, there will be more surviving offspring of high-fitness parents than of low-fitness parents. In fact, the babies’ generation will have a higher average fitness than their parents’ generation did.
Evolution just happened. But did sexual selection happen? Things get a little complicated here, because there are two effects at work.
One effect of fitness matching is to increase the variation in fitness in the next generation. In fact, it creates the widest possible fitness differences between babies. Fitness matching by parents leads to fitness spreading among offspring. Consider the extremes of the fitness spread. The only way to produce a baby of the highest possible fitness given the parents available, would have been for the highest-fitness male to mate with the highest-fitness female. That is exactly what happened, through the mating market. And the only way to produce a baby of the lowest possible fitness would have been for the lowest-fitness male to mate with the lowest-fitness female. Again, that is exactly what happened. Fitness matching does not just increase the variation in fitness a little bit. It increases that variation as much as any mate choice process could, with or without monogamy.
The fitness-spreading effect is important because it creates a very tight link between sexual selection and natural selection. The power of natural selection is proportional to the fitness spread that is available in a population. Bigger fitness differences between babies lead to faster evolution. By creating the largest possible fitness spread, fitness matching gives natural selection the greatest diversity of raw material to work on. Psychologists Aaron and Steven Sloman emphasized the importance of this effect in an important paper they published in 1988.
From a genetic point of view, fitness matching concentrates harmful mutations from low-fitness parents in their low-fitness babies. When those babies die, they take a lot of harmful mutations with them. Fitness matching also concentrates helpful mutations (which are much rarer) in high-fitness babies. When those babies thrive at the expense of lower-fitness competitors, the helpful mutations increase their share of the gene pool. This is a heartlessly unromantic view of sexual selection’s effects, but evolution is heartless.
The fitness-spreading effect is interesting, but it doesn’t take us very far in understanding the evolution of the human mind. To do that, we have to ask how fitness matching affects the fitness indicators themselves. What follows is admittedly a subtle and speculative argument, but one I think is critical to understanding how sexual selection shaped the human mind.
In the above description of fitness matching, it was assumed that individuals could perceive each other’s fitness with perfect accuracy. But it is not that simple. Our hominid ancestors did not have portable DNA sequencing laboratories to measure the mutation load of every potential mate. They had to make do with fitness indicators such as sexual ornaments and courtship displays. By definition, fitness indicators have some correlation with fitness, but it is never a perfect correlation. The handicap principle keeps indicators relatively honest, but it cannot keep them perfectly honest, so there will always be a discrepancy between true fitness and apparent fitness. The evolution of fitness indicators is driven by this discrepancy.
Consider the mating market from female number two’s perspective. She is the second-highest-fitness female hominid in the tribe. She would love to get together with male number one and have his higher-fitness babies, who will survive better and attract better mates. But female number one stands in the way, seducing male number one with her high-fitness charms. (For the moment, we are still assuming strict monogamy and no adultery, so female number two cannot just have an affair with male number one.)
What can female number two do? She cannot raise her true heritable fitness, because on the African savanna she has no access to retroviral germ-line genetic engineering. But she could produce an appearance of higher fitness by allocating more energy to her fitness indicators. If she had a mutation that increased the quality of one of her fitness indicators, even at the expense of her other adaptations, she might look better than female number one. In fact, she would become female number one, in terms of apparent fitness. She could attract male number one, and produce high-fitness babies. She might produce the same number of babies she would have had with male number two, but now her babies have higher fitness, and are more likely to survive. Even though, according to our assumption, she has produced no more children than any other woman, she will produce more grandchildren who will carry her mutation. Her granddaughters and grandsons would inherit her propensity to allocate more energy to their sexual ornaments and courtship displays. If those displays included evolutionary novelties such as art, music, and language, sexual selection would improve their performance. This is how fitness matching can push fitness indicators to evolve. This is how sexual choice can drive sexual selection, even under strict monogamy.
Now, step back from female number two’s predicament and consider the general point. Here we have a hominid tribe that would make Puritans look sinful. They are perfectly monogamous, they have no adultery, and they all have exactly the same number of children. Yet even here, under the most impossible-looking conditions, sexual selection still works. It still favors more extreme, costlier, more impressive fitness indicators such as sexual ornaments and courtship displays. Sexual selection still works on fitness indicators because fitness still means something: some babies still survive better than others because they have higher fitness. Since fitness matching pays evolutionary dividends to those who have high apparent fitness, there are incentives for displaying the most extreme fitness indicators you can afford. The handicap principle will keep the fitness indicators within reasonably honest limits. It can keep low-fitness pretenders from displaying very high apparent fitness, but it cannot keep high-fitness competitors from escalating their sexual arms race. As long as there is some natural selection going on, fitness matching alone should suffice to drive sexual selection for indicators.
This fitness matching theory may sound speculative, but it is just a variation of Darwin’s theory of sexual selection in monogamous birds. Darwin faced the same problem: how to explain sexual ornaments that are equally extreme in both sexes in species that form monogamous pairs. He proposed a fitness matching process that relied on the fittest female birds arriving first at the best nesting sites in each breeding season, mating with the fittest male birds, and producing higher-fitness offspring who are more likely to survive. Sexual selection theorists such as Mark Kirkpatrick have shown that Darwin’s model can work as long as fitness remains heritable and sexual choice favors reliable fitness indicators. If fitness matching can explain ornamentation in monogamous birds, perhaps it can explain courtship abilities in relatively monogamous apes like us.
The pure fitness matching process would not produce any sex differences. All else being equal, males and females would evolve fitness indicators to precisely the same degree. This is because under strict monogamy they would have equal incentives for displaying their fitness and for selecting mates based on fitness. Fitness matching tends to promote sexual equality in the indicators it favors. This is one reason why it has the potential to be so important for human evolution. The sexual egalitarianism makes it an attractive model for explaining traits that are ornamental, costly, and sexually attractive, yet do not show the sex differences predicted by traditional models of sexual selection.
How many traits have these features predicted by the fitness matching model? Many traits in many species look ornamental and costly, show minimal sex differences, and probably influence mate choice. However, biologists since the 1930s have usually called such traits “species recognition markers.” They assumed, following the tradition of equating sexual selection with a mechanism for producing sex differences, that such traits simply advertise one’s species rather than one’s fitness. For the last fifty years, whenever a biologist noticed something that exists in both sexes, which would have been called a sexual ornament if it existed only in males, it was called a species recognition marker. If the marker was displayed vigorously by both sexes during mutual courtship, biologists would say that the animals are performing a “pair-bonding ritual.” This terminology obscured the fact that one individual would often walk away from the ritual, unimpressed by his or her would-be partner. The evidence for mutual choice was there, but most biologists neglected Darwin’s theory of sexual selection in monogamous species.
Birds offer many examples. If, among emus, only males had bright blue bare patches on their cheeks and necks, biologists would probably have called the patches sexual ornaments. But since females have them too, they are usually relegated to the status of species recognition markers. Likewise for the dramatic yellow eyebrow-tufts sprouting from both male and female rockhopper penguins. And the 11-foot wingspans of both male and female wandering albatrosses, which are displayed during mutual courtship by stretching the black tips of the white wings as far apart as possible for the inspection of the opposite sex. All, we are told, for mere species recognition. This viewpoint implies that the hours of mutual conversation during human courtship are likewise nothing more than a way for us to tell that the other individual is a human rather than a chimpanzee. Amotz Zahavi has mocked the species recognition idea as attributing a very high degree of stupidity and very poor mate choice to animals. I agree with his view. These same animals show good discrimination ability when it comes to food and predators, so why should they need such dramatic markers to tell whether a potential mate is of their own species? Fitness matching, a form of mutual mate choice based on fitness indicators, may be a more sensible explanation for most sexual ornaments that show very small sex differences.
The question remains of how our ancestors actually made their sexual choices. Perhaps during large tribal gatherings, they formed huge mixed-sex aggregations like sage grouse, where individuals could weigh up hundreds of prospects. This would have made mutual choice extremely easy. However, such Pleistocene singles bars were probably rare.
Much more likely, individuals would encounter a slow trickle of new sexual possibilities, one at a time. The search for a good sexual partner was sequential and opportunistic. Success would depend on one’s ability to manipulate which band one joins, and who joins one’s band. (A band is the small group of individuals with whom a hominid would forage and spend most nights; clans and tribes are larger sets.) New individuals might join an existing band. The band may encounter other bands at water sources. Individuals might leave their band, looking for new groups that offer more sexual opportunities.
Contact between bands may have been tense and brief, with the threat of violent confrontation balanced against the possible benefits of trade, gossip, and the exchange of sexual partners. Selection would have favored a capacity for very fast decisions about which individuals were attractive enough to pursue. These snap judgments could have been based on information like physical appearance, bodily ornamentation, apparent social status, and public display behavior (such as sports, music, and story-telling). Our ability to judge the physical attractiveness of a human face in a seventh of a second is a legacy of selection for such fast decision-making. Since males would usually have been more motivated to pursue sexual prospects, they would have been more active in this initial phase of searching through bands, looking for attractive potential mates, and trying to switch bands to court good possibilities.
Once mutually attracted individuals arranged to be in the same band, they could split off into temporary courting pairs. Their interaction would resemble the consortships formed by chimpanzee pairs who go off into the bush together for several days. During this most intense phase of courtship, hominids could get to know each other much better, bringing into play all of the psychological levels of courtship discussed in this book. Before language evolved, they would have groomed each other, played, canoodled, shared food, and done all the usual primate things to form social relationships. After language, they would have talked endlessly. During these consortships, the male would usually have been trying to copulate because he would have little to lose from a short-term sexual relationship. If he succeeded, he might grow bored and go away, or he might stay around.
Male and female mate choice waxed and waned in importance at different stages of courtship. Basically, males would scan for physically attractive females and pursue them, trying to establish consortships. This would be a major stage of male mate choice, subjecting females to intense sexual selection for immediate physical appeal. Once a male tried to approach a female to form a consortship, the first stage of female mate choice would be triggered. On the basis of his appearance and behavior, she would reject him (usually) or provisionally agree to continue interacting. This would impose sexual selection on males to create a positive impression during the first few minutes of interaction. After several hours or days of consorting, the female would decide whether to have sex. If she agreed, they would probably copulate frequently for several days or weeks. At that stage, male mate choice would once again reassert itself: will he stay with this female, or grow bored and abandon her in search of someone who would make a more interesting long-term partner? The female would be deciding the same thing: does he offer anything beyond a few orgasms and some good times?
How smart did our ancestors have to be to make all these complicated mate choices? A cognitive psychologist might try to construct mathematical models of how all the information about sexual cues gets integrated, and how all the individuals get compared. This makes the mate choice task look daunting. However, my research on simple rules for mate choice suggests that very good sexual choices can result from very fast, very simple decision rules.
Fitness indicators themselves make sexual choice simple. When a female long-tailed widowbird chooses a mate, she can get a pretty good estimate of his fitness simply by looking at the length and symmetry of his tail feathers. She does not need a complete DNA profile highlighting all his mutations—the tail is all she needs to see. The fitness indicators that our ancestors evolved also made sexual choice much easier. They could just pay attention to a few cues like height and facial appearance, and get a pretty good estimate of an individual’s fitness. Each trait that we consider sexually attractive already summarizes a huge amount of information about an individual’s genes, body, and mind.
We do not need to combine the information about these sexual traits in very complicated ways, either. It might seem difficult to compare two possible mates who differ in dozens of ways. It seems that the mathematically correct procedure would be to take each of their features, multiply it by its importance, add up all the results, and then compare the total score for each individual. But this is not necessary. Psychologist Gerd Gigerenzer and his colleagues have found that if you have to pick between two prospects based on a number of features, you can make extremely good decisions by doing something much, much simpler. You can rank the features you find most important, then compare the prospects on each feature until you find a feature where one prospect is clearly superior. For example, if you think intelligence and beauty are the most important two features in a sexual partner, you can just go down your list and compare each prospect. Is one significantly more intelligent than the other? If so, pick the bright one. If not, then is one significantly more physically attractive than the other? If so, pick the beautiful one. If not, choose randomly, because it doesn’t matter. Gigerenzer’s team has a lot of evidence that this very simple rule, which they call “Take the Best,” makes decisions almost as good as the most sophisticated mathematical decision rules in almost every situation. It has astonishing power as a decision rule, yet it is very simple. If our ancestors used a rule of thumb like Take the Best to choose mates, they could have made very good decisions without needing to process a great deal of information using very complicated rules.
Although sexual decision-making can itself be fast and efficient, it sometimes takes time to acquire the relevant information about a potential mate. If a woman is interested in assessing a man’s personality, intelligence, and experiences, it may take weeks of conversation before she has (unconsciously) gathered all the information she needs to fall in love. As we shall see in Chapter 10, conversations during courtship are how we learn the most about potential mates, and these conversations take time. Insofar as men may be satisfied with certain minimal standards of physical appearance before their sexual interest is aroused, their sexual decision-making may appear faster—but only because physical appearance can be judged much faster than character. When it comes to making long-term sexual commitments based on traits that are more than skin deep, men may take even longer than women.
Another challenge is to decide when to form a serious relationship while one is searching through a sequence of encounters and consortships. Economists and statisticians have developed mathematical models of optimal search that look appropriate. But here again a simple rule can do much better. The standard optimal search strategy is called the 37 percent rule. It is useful when you are looking for the best candidate for a position, and you encounter the candidates one at a time, and you have to offer the position on the spot to the first candidate you like, without going back to previously interviewed candidates. This is somewhat like looking for a long-term mate. The 37 percent rule says that you should estimate how many total candidates are likely to apply for the position, interview the first 37 percent of them, and remember the best out of that initial sample. Then, keep interviewing until you find a candidate who seems even better than that. Once you find that better candidate, stop searching and stick with that one. The trouble with this rule is that the time and energy costs of searching can grow very large if you have a large number of possible candidates. For single New Yorkers, it is infeasible to date 37 percent of Manhattan’s population before finding a spouse.
In our research on mate search strategies, colleague Peter Todd and I found that a rule we call “Try a Dozen” performs as well as the 37 percent rule under a wide range of conditions. Try a Dozen is simple: interview a dozen possible mates, remember the best of them, and then pick the very next prospect who is even more attractive. You do not have to estimate the total number of potential mates you will encounter in your reproductive lifetime; you only have to bet that you will meet at least fifty or so. Humans seem to follow something like the Try a Dozen rule: we get to know a number of opposite-sex friends during adolescence, fall in love at least once, remember that loved one very clearly, and tend to marry the next person who seems even more attractive. Each individual is “satisficing”–looking for someone who is pretty good and good enough, rather than the absolute best they could possibly find. But at the evolutionary level, these satisficing rules impose sexual selection that is almost as strong as the most complicated, perfectionist decision strategy.
In general, very simple rules of thumb can result in sexual choices that are almost as good as the best strategies developed through mathematical analysis. Our ancestors did not have to have sexual supercomputers in their heads in order to make very good sexual choices under Pleistocene conditions of great uncertainty, limited information, and potential deception. Sexual selection does not require a sophisticated set of sexual choice rules. What matters is how efficient the rules are at distinguishing between mates. If very simple rules can make fairly good sexual decisions, then, across many matings and many generations, those rules can impose very strong sexual selection.
When trying to attract a sexual partner, heritable fitness is not the only thing worth advertising. When males and females cooperate to rear offspring, they should care about more than each other’s good genes. They should seek mates in good health because they are more likely to survive as partners and parents. They should seek mates capable of efficient cooperation and coordination, so they make an effective team. Since health and future cooperation cannot be assessed directly, they must be estimated using indicators such as energy level and kindness. Those indicators can evolve according to the same principles as fitness indicators.
Usually, there is a lot of overlap between basic fitness and these other qualities. Condition-dependent indicators can advertise both heritable fitness and the aspects of bodily and mental condition that are important for shared parenting. An individual who is grossly incompetent at finding food may have bad genes, bad condition, and bad parenting potential.
In principle, sexual choice could sometimes put non-heritable qualities ahead of heritable fitness. If the environment is so demanding that a female simply cannot raise a child by herself, then she might favor an attentive, experienced father, even if he has a lower general fitness than a charming athletic genius who is hopelessly incompetent with babies. However, she might still prefer to have an affair with the genius and let the experienced father raise the resulting child. New DNA methods for establishing paternity have shown that this sort of eugenic cuckoldry is surprisingly common in birds previously thought to be monogamous, and in humans.
Until recently, evolutionary psychology emphasized the non-genetic benefits of mate choice. This emphasis may have come in part from sexual selection terminology favored by biologists in the 1980s. Food gifts, nests, territories, and fertility were termed the “direct” benefits of mate choice, and good genes were termed the “indirect” benefits; it sounds more secure to receive a direct than an indirect benefit. In particular, leading evolutionary psychologists such as Don Symons, David Buss, and Randy Thornhill focused on the material benefits that high-status men could offer women, and the fertility benefits that healthy young women could offer men. This has been a powerful research strategy for explaining many sex differences in human mating behavior.
However, many male human courtship behaviors that appear to give purely material benefits to females may have evolved mainly as fitness indicators. Males of many species give females food during courtship. Male scorpionflies give females the prey they have caught. Our male ancestors probably gave females a share of the meat from the hunt. Until recently, men in modern societies brought home almost all of the money necessary to sustain their families. Don’t females in all cases simply want a good meal instead of good genes? I think the analogy is deceptive. Male scorpionflies give females a significant proportion of all the calories the female will need to produce her next batch of eggs. Modern men used to give women all the money they needed to live in a market economy. But the meat provided by our male ancestors may have been only a minor contribution to the energy needs of a mother and her children. A pregnant hominid would have needed about four pounds of food a day for 280 days, about a thousand pounds in total. If a male hominid gives her ten pounds of meat during a month-long courtship, that’s fairly generous by modern hunter-gatherer standards, but it is less than 1 percent of the food she will need just during the pregnancy.
Of course, given a choice between a fitness indicator that offers zero material benefits (such as an impressive courtship dance) and one that happens to produce a material benefit (such as an impressive hunting success), evolution may favor females who appreciate the material benefit. From a fitness indicator viewpoint, the material benefits simply bias evolution to favor fitness indicators that happen to deliver practical benefits in addition to information about mutation load.
Likewise, male defense of good territories may have evolved as a fitness indicator as well as a material benefit. Generally, female animals forage where they want, exploiting the available food resources. Males follow the females around and try to mate with them. The strongest males often succeed in driving the weakest away from the prime food-patches where the females have already decided to forage. Since the females might as well prefer a stronger to a weaker male, they might as well mate with the male who happens to be defending their food-patch. To a human observer used to the idea of land ownership, it might look as if the strong male has “acquired ownership” of the territory, which he generously allows the females to use. Perhaps even in the male animal’s mind, he “owns” the territory. But to the females, they are just foraging wherever they want. The males may be running around and fighting each other, and large, muscular males may happen to last longer and stay closer to the females. The females have little incentive to go chasing after the smaller, weaker males that were driven away, so they may tend to mate with the stronger males. The females thus use the male’s ability to defend the territory from other males as a fitness indicator. Sometimes the strategies of sexual choice are so efficient that they hardly look like active sexual choice at all. As long as the females do not stumble across any male trait that is a better fitness indicator than resource-defense ability, it may look as if the male automatically wins “the right to mate” by “owning the territory.” But that would be missing the point. The females may be using the cue of resource-defense ability mainly to get good genes, not to get food.
In modern market economies people put a high value on wealth indicators during courtship. This can be rational, given the range of goods and services that money can buy, and the difference it can make to one’s quality of life. As Thorstein Veblen argued a century ago, modern culture is basically a system of conspicuous consumption in which people demonstrate their wealth by wasting it on luxuries. Wealth indicators follow the handicap principle just as fitness indicators do, but this makes it easy to mistake one for the other. David Buss has amassed a lot of evidence that human females across many cultures tend to prefer males who have high social status, good income, ambition, intelligence, and energy—contrary to the views of some cultural anthropologists, who assume that people vary capriciously in their sexual preferences across different cultures. He interpreted this as evidence that women evolved to prefer good providers who could support their families by acquiring and defending resources. I respect his data enormously, but disagree with his interpretation.
The traits women prefer are certainly correlated with male abilities to provide material benefits, but they are also correlated with heritable fitness. If the same traits can work both as fitness indicators and as wealth indicators, so much the better. The problem comes when we try to project wealth indicators back into a Pleistocene past when money did not exist, when status did not imply wealth, and when bands did not stay in one place long enough to defend piles of resources. Ancestral women may have preferred intelligent, energetic men for their ability to hunt more effectively and provide their children with more meat. But I would suggest it was much more important that intelligent men tended to produce intelligent, energetic children more likely to survive and reproduce, whether or not their father stayed around. In other words, I think that evolutionary psychology has put too much emphasis on male resources instead of male fitness in explaining women’s sexual preferences.
The most important quality that indicators advertise other than heritable fitness is age. Obviously, age is not directly heritable. A 40-year-old woman will give birth to a nine-month-old, just as a 20-year-old woman will. However, age has a dramatic effect on fertility, especially in women. Individuals before puberty are infertile. Female adolescents are significantly less fertile than 20-year-olds. Female fertility declines gradually during the thirties, and declines steeply after age 40. Women after menopause are infertile. This female fertility profile is a basic fact of life to which male mate choice systems have adapted. Youth is an important cue of fertility.
There may have been male hominids who preferred to start exciting relationships with wise, fulfilled, 60-year-old females. But if they did so exclusively, they would have left no offspring to inherit that preference. Any sexual choice mechanism that preferred infertile individuals to fertile individuals would have died out in one generation. Since male sperm production ability declines more slowly with age, female mate preferences need not have paid so much attention to a man’s age as a cue of his reproductive ability. This reasoning, as developed by Don Symons, David Buss, and other evolutionary psychologists, explains the universal, cross-cultural pattern that men care more about a partner’s age than women do, men generally preferring partners younger than themselves, and women generally preferring partners older than themselves.
However, male hominids may not have been quite so youth-obsessed as men from agricultural, pastoral, and modern civilizations. In most cultures with recorded history, men were under social, legal, economic, and religious pressures to stay monogamously married for life. The younger their bride, the more offspring they could produce. This put a huge premium on youth, and men competed to claim young women before another man could.
A woman’s youth may not have been quite so crucial in the Pleistocene, as long as the woman was still reasonably fertile. If our hominid ancestors had several medium-term relationships in sequence, males need not have been so picky about female age. If the relationship was likely to end after five years—as anthropologist Helen Fisher has argued that they usually did in prehistory—it would have mattered little whether she was 10 years or 30 years away from menopause.
During her reproductive years, a woman’s age does have a negative correlation with her fertility. But under challenging Pleistocene conditions, age would have had a positive correlation with heritable fitness because low-fitness individuals would have died younger. Any woman who managed to reach her mid-thirties and raise several children successfully, while staying physically and psychologically attractive, might have made a better genetic bet for a choosy male than an untested teenager of unproven fertility. Other male primates tend to shun adolescent females without offspring, and prefer older, high-ranking females with offspring who have already demonstrated their fertility, survival ability, social intelligence, and mothering skills.
There is strong evidence from evolutionary psychology that men in modern societies generally prefer the physical appearance of women around 20 years old to those who are older (or younger). But I have argued that this preference may have been amplified somewhat by the economic and religious pressures for monogamy since civilization arose, which makes finding a young bride crucial to a man’s reproductive success.
More importantly, there has been much less research on the age at which women’s minds are most attractive. Perhaps mature men tend to find young women beautiful but boring, and older women slightly less physically attractive but much more interesting. If so, we should not view the preference for youthful appearance as any less of a legitimate adaptation than the preference for a worldly mind. Data gathered by Doug Kenrick shows that older men generally prefer women closer to their own age—in their mid-thirties rather than their early twenties, for example—as long-term sexual partners. Presumably this is because women in their mid-thirties are typically more intriguing, multifaceted people who display the mental aspects of their fitness in richer ways that can be more reliably assessed. Evolutionary psychology has rightfully emphasized the strong male human interest in young female bodies, but I think its scope should be broadened to include the romantic interest aroused in both sexes by mature, worldly minds.
In any case, chronological age, like heritable fitness, could not be perceived directly during human evolution. To distinguish children from adults, our ancestors had to rely on cues of sexual maturity such as male musculature, beard growth, and voice pitch, and female breast and hip development. To distinguish young adults of peak fertility from other adults of declining fertility, they had to rely on age cues such as wrinkles, gray hair, sagging skin, slow gait, and memory loss.
Like fitness indicators, age indicators leave some room for deception. This may have some relation to our apparent “neoteny,” which means that we have, it has been argued, retained some of the physical and mental traits of juvenile apes into our adulthood. Our faces look more like the faces of very young chimpanzees than they do like those of adult chimpanzees. Our playful creativity resembles the behavior of young primates more than it does the stern, lazy brutality of adult apes. Stephen Jay Gould has argued that our neotenization was a key trend in human evolution, and he sees our behavioral flexibility as a side-effect of our general neoteny.
But neoteny can be viewed very differently. Our neotenous features may have evolved through sexual choice as somewhat deceptive cues of youth. If male hominids preferred younger, more fertile females to older, less fertile females, then there would have been sexual selection pressures on females to appear physically and behaviorally younger than they really were. They could do this by evolving younger-looking faces, and by being more playful, creative, spontaneous, and uninhibited throughout their adult life. The result would be neotenized female hominids. The same argument could apply to males, insofar as female choice favored signs of youthful energy. (It is not clear why our lineage evolved these neotenous youth-cues while other primates did not—one could invoke sexual selection’s unpredictability, though that is not a very satisfying explanation.) In my view, Gould’s neoteny theory identified a set of somewhat deceptive youthfulness indicators that must have evolved through some form of sexual or social selection. It is not a competing theory of human evolution, but a description of some physical and psychological trends that still require an evolutionary explanation.
Apparent preferences for youth are not as simple as they seem. It is often hard to distinguish indicators of youth from indicators of fitness. This is because fitness indicators usually work by being very dependent on condition, and condition is highest during the flower of youth. All things being equal, any mate choice mechanism that evolved to favor a condition-dependent indicator will tend to favor youth over age simply because youths will display the indicator in a healthier condition. However, the fact that women often prefer older men suggests that mate choice mechanisms can easily evolve to compensate for this youth-bias whenever it proves maladaptive.
Sexual selection was not the only kind of social selection during human evolution. For humans, as for most primates, all kinds of social relationships affect survival and reproduction. In forming and maintaining many of these relationships there are good reasons to advertise one’s fitness, just as one does to potential sexual partners. Friends of higher fitness may survive longer, offer more competencies, and give better advice. Allies of higher fitness may help one to win fights and wars. Trading partners of higher fitness may live longer, travel longer distances to acquire more valuable commodities, and have the social intelligence to keep their promises. None of these social relationships entails any merging of genes, so they are not subject to positive-feedback processes as powerful as runaway sexual selection. But they still offer plenty of scope for all kinds of socially selected indicators to evolve.
We can often use the same fitness indicators in non-sexual relationships as we do in sexual relationships. If vigorous dancing all night displays our physical fitness to potential mates, it equally displays our fitness to potential friends and allies. Whenever a fitness indicator evolved in our ancestors through sexual selection, it was probably generalized to other social relationships rather quickly. Conversely, any indicator that evolved in the context of friendships or tribal alliances could easily have been modified for courtship.
The overlapping use of fitness indicators in sexual and non-sexual relationships is why making friends so often feels like a variant of sexual courtship. There is the same desire to present oneself to best advantage, emphasizing skills, downplaying weaknesses, revealing past adventures, investing extra energy in the interaction. This does not mean that friendships always have a sexual undercurrent, or that friendship is maintained through some kind of sexual sublimation. It simply means that the same principles of self-advertisement work in both kinds of relationship. If friendships gave important survival and social advantages during human evolution, and if our ancestors were choosy about their friends, then many of our fitness indicators may have evolved for friendship as well as for sexual relationships.
An especially important non-sexual relationship is that between parents and offspring. Children often compete to display their fitness to their parents, older siblings, and older relatives. They may shout “Hey dad, look at this!,” and then try to do something that is challenging for a child of their age and abilities. At first glance it seems odd that they should bother. According to modern social norms, parents are supposed to love their children unconditionally, regardless of their fitness or abilities. But Pleistocene Africa did not always permit such unconditional support. Times were sometimes tough. Just as birds often have to choose which chick gets the worm and which starves, human parents may have had to choose how much support to invest in a particular child. Evolutionary psychologists Martin Daly and Margo Wilson have called this the problem of “discriminative parental solicitude.” Parents must sometimes discriminate about which child deserves their solicitude. Older children are often favored because they have already survived the risky phase of infancy. But parents may also be sensitive to a child’s fitness, which mean its prospects of successful survival and reproduction. Investment in a very low-fitness child means investment in an individual very unlikely to pass one’s genes on to grandchildren. For better or worse, evolution considers that an unwise investment, and favors a more discriminating attitude. In every culture, children with physical deformities and serious psychological disorders are at enormously greater risk of neglect, abuse, beating, and infanticide by parents.
Given parents who discriminate between children based on their apparent fitness, children have incentives to evolve fitness indicators. As when people initiate friendships, children can use many of the same strategies that work in courtship, without there being any hidden sexual motive to the display. This is where I believe Freud went wrong with his hypotheses about Oedipus and Electra complexes. He observed a set of fitness indicators that children directed at parents—energetic play, humorous storytelling, flirtatious conversation—and inferred a secret children’s desire to have sex with their parents. That inference seems evolutionarily incredible. Presumably our hominid ancestors evolved a set of sexual choice mechanisms for judging the fitness of potential mates. Perhaps children found it convenient to play upon some of the same mechanisms to advertise their fitness to their parents, to solicit more attention and care. This does not mean that children want incest—it means that they want parental support.
Homosexuality has not been mentioned so far in this book. My heterosexual emphasis comes not from homophobia, religious conviction, or moral conservatism. My subject is human evolution, and homosexual behavior is just not very important in evolution. Not a single ancestor of any living human was exclusively homosexual. Any hominid that was would not have produced any offspring, and would not have become anyone’s ancestor. There may have been many gay and lesbian hominids, but if they were exclusively homosexual, they are not our ancestors, and we are not their descendants. In any case, it is unlikely that there were many exclusively homosexual hominids. Any genetic propensity towards exclusive homosexuality would have been eliminated in just one generation of selection. No biologist has ever offered a credible theory explaining how exclusive homosexuality could evolve in a sexually reproducing species. Its existence in 1 or 2 percent of modern humans is a genuine evolutionary enigma that I cannot explain.
There is no such evolutionary problem with bisexuality, in which individuals enjoy sex with both sexes. Certainly bisexual behavior occurs in other species. Bonobos (previously known as “pygmy chimpanzees”) engage in a lot of sexual activity with same-sex individuals, including kissing, genital rubbing, and genital licking. This does not impair their heterosexual reproduction in the slightest. Evolution does not respect our hunger for simplistic political categories of sexual behavior, in which every individual can be put on a continuum of “sexual orientation.” Ordinary bonobos enjoy heterosexual behavior, and homosexual behavior, and they have lasted a million years as a species, about ten times longer than we have so far. There is nothing “unnatural” about homosexual behavior.
Moreover, many male humans with strong homosexual desires get married and produce offspring, as Oscar Wilde did. Many female humans with strong lesbian desires produce children too. Evolution has no moralistic motive to punish homosexual behavior. As long as homosexual behavior does not displace heterosexual behavior, it has little impact on evolution. Homosexual behavior—as an adjunct to heterosexual behavior—would be expected to evolve whenever its fitness benefits (making friends, appeasing threats, making peace after arguments) exceed its costs (energy, time, and the increased risk of sexually transmitted disease).
Our hominid ancestors might have been almost exclusively heterosexual, like chimpanzees, or very homoerotic like bonobos. We do not know. Even male chimpanzees hold each other’s penises for comfort when they are frightened. Perhaps, like bonobos, humans evolved some adaptations for homoerotic flirtation and same-sex sexual friendships. If the social benefits of homosexual relationships were strong enough, homosexual preferences could, in principle, have shaped human physical appearance and mental capacities. However, these preferences had no direct reproductive consequences, so they would have had much weaker evolutionary effects than heterosexual preferences. As a result, we have to focus on heterosexual behavior when considering the role of sexual choice in the mind’s evolution.
Sexual choice and courtship in human evolution was not just a matter of boy meets girl. We have seen that our ancestors were highly social primates living in groups with children, relatives, and friends. Sexual relationships began and ended within family and tribal contexts.
If mate choice favors good genes, it can be useful to meet a potential mate’s blood relatives, because they share some of the same genes. An individual’s kin give additional information about their heritable fitness. If an intelligent man has foolish brothers or a beautiful woman has ugly sisters, this may lower their attractiveness as potential parents of one’s children. Siblings share half of their genes, as do parents and offspring. The apparent fitness of a woman’s mother or daughter carries half as much information about the woman’s own genetic quality as her own fitness indicators. Given two sexual prospects who appear to display equal fitness, the one whose relatives appear healthier, brighter, more attractive, more fertile, and more successful probably has higher actual fitness. Since our ancestors tended to live in kin groups, there were plentiful opportunities for mate choice to take into account this sort of kin quality. Our mate choice systems would have evolved to exploit this gold mine of genetic information.
If sexual choice paid attention to the fitness of a potential mate’s relatives, then those relatives would have been under sexual selection to display high fitness. This would have been a much weaker pressure than ordinary sexual selection, but it could still have been significant in shaping our instincts for display. If parents could help their offspring attract better mates by appearing intelligent, healthy, and successful, then the copies of their own genes that are carried in their offspring would benefit. Likewise, if children could help their mothers appear more attractive by demonstrating that they carry good genes, then the copies of their genes in their mothers would be passed on to larger numbers of half-siblings. Any courtship effort that helps your relatives to find good mates helps your own genes to spread. (Of course, there may be conflicts of interest between relatives over these courtship displays, as when adolescents wish that their parents made more effort to act reasonably cool when their friends visit, or divorced parents wish that their adolescents would behave better towards potential step-parents.)
Mate choice that takes into account the qualities of a potential mate’s relatives would have favored hominids who spread their courtship effort out across their lifetimes. In childhood and old age their courtship would be vicarious, carried out on behalf of their relatives. In the prime of life it would be mostly for themselves, but also for their sexually active relatives. We should not expect to see fitness indicators used exclusively after puberty and before menopause, only that they are then directed at different targets.
Vicarious, collective courtship by relatives might explain why humans are so good at producing certain kinds of cooperative display. Evolutionary psychologists have usually assumed that human cooperation evolved for survival benefits. Cooperation can certainly help the cooperators survive better—if they are doing something that is actually useful together. But what about religious rituals, dances, and feasts that have high time and energy costs and no credible survival payoffs? Consider the huge Thanksgiving feasts that American families organize when a daughter first brings home a potential husband. The family members are not improving their collective survival chances; they are improving the daughter’s mating prospects by demonstrating their wealth, health, family size, and other aspects of familial fitness. The prodigious waste of uneaten turkey even follows the predictions of the handicap principle. Across many cultures, marriage rituals serve similar functions, wasting vast resources so that a kin group can display its fitness to a group of possible in-laws. American families also advertise their wealth and status by producing costly rituals when one of them reaches sexual maturity—as in bar mitzvahs, debutante balls, and “sweet sixteen” parties. Rich parents even advertise familial fitness by paying over a hundred thousand dollars for each child to attend a private university, whereas in Britain, they pay even more for preuniversity private schooling.
Modern human families compete to attract good mates for their young people. Perhaps Pleistocene kin groups and tribes did so too, inventing various rituals, myths, legends, totems, and dances to display their superiority over other groups competing in the same sexual market. To the extent that mating occurred across group boundaries, cooperative group activities may have evolved as collective courtship displays through sexual selection. This may explain the observation by anthropologists Chris Knight and Camilla Power that a great deal of human ritual behavior consists of collective displays by female relatives on behalf of their youngest female kin when they reach sexual maturity. Of course, once the mental capacities for collective fitness displays evolved through sexual selection, those capacities might prove useful for other functions as well, such as intimidating rival groups competing for the same territories and resources.
So much for collective courtship. As for collective sexual choice, each individual’s mate choice decisions probably took into account the views of their parents, siblings, offspring, and companions. Sometimes they may have immediately discounted this advice, realizing that their relatives’ interests did not coincide with theirs. But sometimes other individuals would have offered useful information about a potential mate. They may have interacted with the prospect in other contexts, or heard useful gossip. Older relatives may have offered words of wisdom from their past experience of sexual choices and sexual relationships. During the Pleistocene, when social conditions were less volatile than today, one generation’s experiences of courtship and parenting would have been much more relevant to the next generation. Before the evolution of language, relatives could have revealed their attitudes about a sexual prospect through the usual primate signals: threat displays and attacks, or friendly grooming and food sharing. After language evolved, the relative merits of sexual rivals must have become subjects of impassioned discussion. Parents may have been especially vocal about their views, because the sexual choices of their children were so important to the number and quality of grandchildren who would carry their genes. However, parental influence on sexual choice does not imply some sort of arranged marriage system in which sexual selection no longer operates. On the contrary, by integrating information from several individuals our ancestors could have made much more accurate estimates of each prospect’s strengths and weaknesses, driving sexual selection more strongly in particular evolutionary directions.
Biologists have not developed models of how sexual selection works when mate choice and courtship are socially distributed. I would guess that the runaway process would not work so strongly when sexual preferences and sexually selected traits are spread across different bodies. It would have a harder time establishing the genetic correlations between preferences and traits that drive runaway. However, there may be fewer such problems with sensory bias effects and preferences for fitness indicators. For example, the sensory and cognitive biases of friends and relatives could influence an individual’s sexual choices just as their own biases would. Ornaments and indicators could still evolve even if parents were choosing sexual partners for their children, and even if aunts were producing courtship displays on behalf of their nieces.
It should go without saying, but I’ll say it anyway: all of the significant evolution in our species occurred in populations with brown and black skins living in Africa. At the beginning of hominid evolution five million years ago, our ape-like ancestors had dark skin just like chimpanzees and gorillas. When modern Homo sapiens evolved a hundred thousand years ago, we still had dark skins. When brain sizes tripled, they tripled in Africans. When sexual choice shaped human nature, it shaped Africans. When language, music, and art evolved, they evolved in Africans. Lighter skins evolved in some European and Asian populations long after the human mind evolved its present capacities.
The skin color of our ancestors does not have much scientific importance. But it does have a political importance given the persistence of anti-black racism. I think that a powerful antidote to such racism is the realization that the human mind is a product of black African females favoring intelligence, kindness, creativity, and articulate language in black African males, and vice versa. Afrocentrism is an appropriate attitude to take when we are thinking about human evolution.