FOUR
Social Primates
In his country,
spotted crabs
born in their mother's death
grow up with crocodiles
that devour their young.
Why is he here now?
And why does he
take those women,
a jangle of gold bangles
as they make love,
only to leave them?
Competition and Cooperation
CHARLES DARWIN IS WIDELY THOUGHT to have bequeathed to us a vision of human society that, for all its conventional pieties about the importance of cooperation, is pitilessly competitive to its very core. “A war of all against all” is a phrase commonly quoted from Thomas Hobbes but often considered an appropriate description of Darwin's own social vision, particularly by those who know The Origin of Species but not Darwin's later works.1 In the closing paragraph of The Origin, Darwin wrote: “Thus, from the war of nature, from famine and death, the most exalted object which we are capable of conceiving, namely, the production of the higher animals, directly follows.”2
Despite the superficial similarity of the two concepts, Darwin's vision of the war of nature was utterly different from that of Hobbes. It was a war that, far from pitting all against all, needed and rewarded teamwork. The observant Darwin could hardly have failed to notice what any reader of this book can easily verify hundreds of times a day: human society surrounds us with people on whose cooperation we depend. From the moment you get up in the morning, you depend on other people to help make it through the day, whether it's the suppliers of your toothpaste and your morning coffee, the person who fixed the water leak you never knew about in the middle of the night so you could have your morning shower undisturbed, the driver of the bus you take to work, the journalist whose story made you giggle in the train, or the colleagues in your office, shop, or factory. In a true war of all against all, you would hope that all of these people had faltered, made mistakes, had in some way failed to do their jobs properly: that's what it means for someone to be your competitor. In fact, as a moment's reflection attests, we depend so much on others to do their jobs well that we hardly notice the smoothness of the social fabric they have made for us; the occasional incompetent stands out all the more annoyingly because most of the time the process works so well. It's a tribute to the effectiveness and omnipresence of cooperation that we can so easily forget that it happens at all.
Still, if human society is as cooperative as I have just made it sound, why does it all feel so stressful? A phrase like “the smoothness of the social fabric” hardly does justice to the tension almost everyone in a modern society feels as they juggle the demands of friends, family, and colleagues. It barely captures the pressure of other people's expectations, the anxiety as to whether we shall be able to deliver what other people want from us, whether we shall ever be considered up to the mark. You might be forgiven for thinking that human society is like one of those “good news, bad news” jokes they tell about doctors: “Congratulations, Homo sapiens, the good news is that you've evolved to be an extraordinarily cooperative species. The bad news is that cooperation is just as awful an experience as competition.”
Certainly cooperation is far from the cuddly ideal of the hippie generation: at the very least it involves setting higher standards for our behavior toward others than our natural inclinations recommend and eternal vigilance about whether our own and others' behavior really meets those standards. But a deeper explanation for all that stress is that cooperation rarely happens at the level of society as a whole: it takes place among groups within that society, groups whose membership can be highly fluid. Some of those groups are in competition with each other, and individuals may compete with one another to be accepted in the most influential and successful groups. It makes a big difference to your well-being which groups you can join, and their acceptance or rejection of you is a conspicuous and stress-inducing event. None of this will be surprising to anyone who has a clear memory of their school playground. None of it would be surprising to a primatologist, since we human beings are a species of social primates, and social primates are nature's experts at building (and breaking) coalitions. And none of it would have surprised Charles Darwin, who was acutely aware of the primate origins of our nature, even though he had drawn a veil over these in The Origin of Species. In fact Darwin had the remarkable insight that natural selection might positively encourage altruistic qualities because they could help coalitions to be successful in competition against rival coalitions. Here is a passage from The Descent of Man, Darwin's first serious discussion of how natural selection shaped human societies, published more than a decade after The Origin of Species:
When two tribes of primeval man, living in the same country, came into competition, if (other circumstances being equal) the one tribe included a great number of courageous, sympathetic and faithful members, who were always ready to warn each other of danger, to aid and defend each other, this tribe would succeed better and conquer the other. Selfish and contentious people will not cohere and without coherence nothing can be effected. A tribe rich in the above qualities would spread and be victorious over other tribes…thus the social and moral qualities would tend slowly to advance and be diffused throughout the world.3
It's clear that the qualities that Darwin thought human beings needed to survive in the “war of nature” were very different from those that later generations would associate with the term Darwinian. Our primate nature is the key to understanding this, because life as a social primate is all about managing cooperation in a world of shifting coalitions. This is very far from Hobbes's “war of all against all,” but it's also a long way from being a serene parade of agreeableness. Certainly, the determinants of fitness include our ability to cooperate with others in our group, the social intelligence to work out who else can be trusted to cooperate with us, and the ability to persuade other members of our group to work with us and not against us.4 But competition is present at many levels too. There's competition between individuals over basic access to economic resources: someone may be collaborating with you on an important project and still be trying to get a bigger share of the benefits than you get. Competition also takes place in other dimensions, notably between coalitions of individuals, and between individuals over the prize of access to powerful coalitions. The tension between those forms of competition is at the heart of primate social life. At any moment you are seeking to cooperate with some of the other members of your group, but at the same time you are intensely anxious about which groups you may be allowed to join (the groups are obviously more complex in human societies). Your fitness is going to depend not just on what you do but on what you can induce the other members of your group to do with you and for you, and on which groups you can persuade to count you as one of their own.
This is a social environment perfectly suited to inducing stress. In dominance hierarchies in most primate societies, low status is associated with characteristics that are usually good predictors of stress, such as a lack of autonomy and social control and a high degree of unpredictability in outcomes. Studies of human societies show similar results. There's a popular belief that stress goes with high-status jobs, but the truth is very different. A longitudinal study of British civil servants by Michael Marmot and his colleagues shows that stress-related illnesses such as cardiac disease show a strong negative correlation with rank.5 The lower down the hierarchy you are, the more likely you are to have somebody pushing you around. This is a very stressful situation, and it is associated over time with significantly greater risk of stress-related illnesses.
It was believed until recently that in nonhuman primate groups, exactly the same thing was true, but when it was first possible to measure cortisol levels (a physiological indicator of stress) among primate groups in the wild, some surprising results emerged. An interesting paper by Muller and Wrangham showed that cortisol levels were positively correlated with dominance among wild chimpanzees.6 The authors suggest that this is because dominant chimpanzees spend a great deal of energy maintaining their dominance—more, certainly, than senior civil servants do. It's a tough and uncertain life for chimpanzees at the top, and their cortisol levels are elevated because of the continual uncertainty about how long they're going to stay on top and the continual efforts they have to make to ensure they stay there. Frans de Waal's book Chimpanzee Politics describes such a life in gripping detail.7 Nor are the findings confined to chimpanzees: high levels of stress hormones have been reported for alpha male baboons.8
So primate existence is pretty stressful either way. You get stress if you win and stress if you lose. The explanation seems to be precisely that small differences in behavior may result in very big differences in outcomes: two individuals who are initially alike may have dramatically different life trajectories if one is accepted as a member of powerful coalitions while the other is not. (This may be part of the appeal of social structures such as caste systems.) It's not surprising, therefore, that even though we're surrounded by potential collaborators, we live in continual fear that our potential collaborators will decide to collaborate not with us but with somebody else.
Primate Negotiations
In our own species, unusually among social primates, a lot of the teamwork is undertaken by men and women together. There's a lot to gain from that cooperation, but the gains are often unequally shared. Why should that be? The answer lies in powerful economic forces that have shaped the relative bargaining power of men and women over the course of our evolution. One of the most robust findings in economics is that it pays to be indispensable. The only plumber available to fix your leak on a Sunday evening may be able to charge you an outrageous sum for the privilege. But if many plumbers happen to be in town for a convention, you can expect to get a better deal: competition drives down prices. What counts is not the indispensability of the service in general (fixing the leak) but of the particular plumber who does the job. Similarly, if you need a new set of keys to get into your house, the keys may be indispensable to you, but the particular locksmith who supplies them to you will not be. It is thanks to competition among locksmiths that you don't end up having to remortgage your house every time you lose your keys.
In general, then, the more fiercely you have to compete for the favors of the people with whom you hope to cooperate, and the fewer rivals they face for the privilege of collaborating with you, the less likely you are to be able to bargain for a large share of the benefits of that cooperation. We can therefore chart the way in which the gains from teamwork have been distributed between men and women by looking at how the intensity of competition among each sex for the favors of the other has changed during our evolution. In a phrase, men used to be the dispensable sex. The fact of sexual reproduction meant that the service they provided (sperm supply) was indispensable—but, like the locksmiths, the individual suppliers were not. However, over the course of the development of the human species, they managed to become progressively more indispensable, and they exacted a price from women as a result. That price is not inevitable: many features of modern society have made men more dispensable again. But we shall not understand the terms of the gender bargain until we understand the way competition shapes it, in other species as well as in our own.
Competition isn't just a matter of how much rivalry exists between males for the favors of females, as compared to the rivalry that exists between females for the favors of males: other factors make a difference, including how much the parties are committed to each other over the longer term. But that rivalry is a good place to start. Male dance flies have found a way to compete for the favors of females by offering food, and this has made the females compete in turn for access to this scarce resource. Human males have managed to do something rather similar: in a nutshell, they learned to contribute substantial parental care, and human females began to compete with each other to find and hold on to the most reliable providers of that care.
To see how that has happened, let's begin by noting what we share with other social primates, in particular with our closest cousins, the great apes: orangutans, gorillas, chimpanzees, and bonobos. Great apes differ from most other primate species in that their societies are patrilocal. This means that adolescent females leave the communities where they were born to join other communities. Females are wanderers, and in their wanderings they exercise some degree of choice over their mates. From that point on, there are significant differences in the different species' living arrangements. Orangutans remain famously solitary, whereas gorillas live in harems controlled by a single dominant male. He is much larger than the females as a result of the winner-takes-all rivalry among males, a rivalry whose reward is control of the entire harem. However, chimpanzees and bonobos, the closest to us of all the great apes, have a very different arrangement. Females of these two species live in small groups and mate promiscuously with many males. As you'd expect, because they don't have to compete for control of a harem, males of these species are relatively smaller than male gorillas. To be more accurate, they're smaller in almost all dimensions except one. They have massively larger testicles—nearly four times as large as the gorilla's, although their body weight is little more than a quarter of the gorilla's. As Roger Short, Alexander Harcourt and others have shown, large testicle size is closely associated with mating systems in which females mate with multiple males.9 This is for the simple reason that if males have no realistic hope of monopolizing sexual access to females, it is in the reproductive interest of each male to ensure that his sperm is as abundant as possible in order to increase the probability that it fertilizes the females in preference to that of rival males (this process is called “sperm competition”).10 There are important differences between chimpanzees and bonobos: in particular, bonobo females spend more time in stable groups than do chimp females, and their cooperation gives them more power over males.11 But the multiple matings in which each female engages are a common and striking feature of their social life.
Diverse as these mating arrangements may seem, they all involve females' exercising substantial choice. As the primatologist Craig Stanford has put it, “Among the four great apes, males try to control females…but females are difficult to control because they are following their own reproductive agenda.”12 Female gorillas give up some of that control of their reproductive agenda once they join the dominant male's harem, but chimp and bonobo females retain it both before and after they move out from the troop of their birth to join another troop, with bonobos in particular reinforcing that control by elaborate cooperation among the females in their troop. What makes it possible for female great apes to follow their own reproductive agenda is a relatively simple fact: once a female has been fertilized, she no longer needs the males, and certainly she has no particular need of the male that happens to have fathered her offspring.
This picture—of great ape females handling with aplomb a lot of hassle from males and maintaining their capacity for autonomy and choice throughout the process—sounds very different from the lot of most human females since the dawn of recorded history. Indeed, since written records began, and until very recently in most parts of the world, women were under the control of their fathers until that control passed to their husbands on their marriage, usually without their having any say in the matter.13 How and why did that arrangement develop? Why have human females been so comprehensively constrained by comparison with their great-ape cousins? Without written records it is difficult to be sure of the answer, but social behavior leaves records in other ways.
The Gender Bargain among Hunter-Gatherers
One thing we can be reasonably sure of is that the subordination of women must have happened very recently in evolutionary time. The evidence comes partly from the observation of hunter-gatherer societies in modern times, in which women have played a central part in foraging and therefore have exercised a substantial degree of autonomy. Agricultural societies have often been able to impose heavy constraints on women, either by ensuring that they do not work outside the home or by making them do routine work under close supervision in the fields. But foraging societies cannot make use of the labor of women who cannot move, cannot take decisions, and cannot think independently.
Still, the foraging societies that have been observed by modern scholars are few and highly untypical of the societies in which Homo sapiens evolved. While it seems a reasonable guess that human females can never have been as subordinate to males in hunter-gatherer communities as they became in agricultural societies, it would be good to have corroborating evidence. Two other kinds of evidence can be found directly in the human body, because one of the marvels of natural selection is that past behavior leaves its mark on current anatomy.
The first kind of evidence comes from our brains and from their cognitive capacities as psychologists have been able to measure them. As I discuss more fully in chapter 5, the average cognitive capacities of men and women differ in a number of respects. Men score significantly higher on average than women in some tests of spatial reasoning, such as tests of the ability to imagine whether the object in one picture is a spatial rotation of the object in another picture. Other cognitive tests, such as exercises in verbal reasoning or tests matching emotional states to faces in photographs, yield significantly higher scores on average for women than for men.14 There are also some well-documented differences in preferences between men and women, such as the often-replicated finding that women are on average more risk-averse than men.15 It's possible to debate whether such results are affected by the way in which the tasks are presented to the subjects, and there is also room for reasonable disagreement as to whether the dimensions in which women are stronger are more or less important than those in which men are stronger. But these qualifications are irrelevant in this context. Cognitive differences on average between men and women are very important for my argument here, but that is because of how small they are, not how large they are.
Because claims about gender differences in intelligence, like claims about racial differences in intelligence, have caused such acrimonious political debates in recent decades, many people who are averse to being considered racist or sexist have avoided trying to find out about them or even to think very much about them at all. This means that most people have never noticed something about gender and racial differences that was evident to Charles Darwin and that ought to become evident once again to us today. According to Darwin's view of sexual selection, the likelihood of finding gender differences is higher than the likelihood of finding racial differences. More exactly, this view implies that we should not expect to find important inherited differences in cognitive abilities between different human populations (including between races). Thus it's neither very surprising nor very informative when it proves difficult to find hard evidence for such inherited differences (raw differences reflecting environmental factors are relatively easy to find).16 Sexual selection, by contrast, should lead us to expect much bigger differences between men and women than in fact there are. So our failure to find larger differences tells us something very interesting about the conditions under which our species evolved.
Darwin spent many pages of The Descent of Man pointing out that sexual selection works in the first place on relatively superficial characteristics, such as skin color and type of facial hair. This meant that over quite a short period (by the standards of evolutionary time), two groups within a population (say the darker-skinned and the lighter-skinned) could become reproductively isolated even though they lived in the same territory. They would still be capable of interbreeding very easily; they just wouldn't usually want to. Their other traits might begin to diverge through the process known as genetic drift, but that would take much longer. And it would be less likely to happen in the case of traits (such as the various components of intelligence) that were subject to continuous selective pressure from the environment in both populations. Darwin was convinced, as we saw in chapter 3, that under the skin, human beings were very much alike. Indeed, his book The Expression of the Emotions in Man and Animals, published soon after The Descent of Man, underlined this point with evidence that facial gestures expressed emotions in a remarkably similar way across many different races and cultures.17
Modern DNA evidence, showing that human beings are descended from a common maternal ancestor who lived around 140,000 years ago and a common paternal ancestor who lived around half as long ago, reinforces that view.18 True, there are some important genetic changes that can become established in human populations in shorter periods than that, and the migration of populations out of Africa and across the globe would have produced some differences in selective pressure for such characteristics as height, skin color, and disease resistance. But it's most unlikely that the selective pressures on talent and ability would have produced big divergences between human populations on a comparatively short timescale.19 This wouldn't happen through sexual selection: since talent and ability are both less visible at first glance and more universally appealing when they can be discerned, their sexual attractiveness should not vary much from one human population to another. And it wouldn't happen through natural selection under environmental pressure either: the importance of talent and ability for general survival would have been very similar in Africa and in other regions of the world. So when modern researchers have difficulty finding hard evidence for large inherited differences in cognitive talent between populations, nobody should be very surprised.20 There have not been sufficiently important differences in selective pressures between populations since those populations became reproductively separated.
That situation contrasts strikingly with selective pressures affecting gender differences. Given the logic of sexual selection, we ought to be very surprised at how hard it is to find significant differences in overall cognitive abilities between men and women. There's no reason for the characteristics women have needed to appeal to men to be anything like the characteristics men have needed to appeal to women. In many other species, males and females are utterly different (think of all those gorgeously colored male birds). They differ in appearance, in behavior, and even in size (the male gorilla is twice the weight of the female). It's true that sex differences are less pronounced among mammals than among birds, for instance, but important differences remain. Men are 15-20 percent larger on average than women, and their upper body strength is substantially greater. That's because muscle tissue is expensive to grow and maintain, and natural selection has sensibly grown more of it among the males who needed it more in the conditions of our hunter-gatherer existence. Brain tissue is far more expensive than muscle to grow and maintain, yet natural selection has nevertheless given equally sophisticated brains to men and to women.
It's just possible that the mutations never arose that would have enabled the processes of embryonic and child development to produce less sophisticated brains in one than in the other (though there are qualitative differences between men's and women's brains that certainly have a genetic basis).21 But there's another compelling and much more plausible explanation: men and women developed equally sophisticated brains because both faced equally sophisticated cognitive challenges throughout almost all of human evolution. On this view, the subordinate and dependent condition of women that has characterized relatively recent centuries cannot have obtained for most of the time since we diverged from the chimpanzees and bonobos.
A second piece of anatomical evidence suggests that women in most hunter-gatherer societies were a good deal more autonomous than they came to be later, after the adoption of agriculture. This is the fact that the testicles of human males are of intermediate size between those of gorillas and those of chimps. This strongly suggests that human females were mating fairly often with more than one male during a single estrus cycle, albeit not nearly as often as among chimpanzees. They did so often enough, in fact, for it to be adaptive for males to engage in sperm competition, as chimpanzee males do. And this conclusion in turn makes it likely that women exercised a significant degree of choice over their mates.
In addition to testicle size, there is also evidence from penis size, where human anatomy is an even more striking exception to that of the other apes. Men have penises nearly twice as large as chimps and four to five times as large as gorillas do. (They also produce more seminal volume per ejaculate than either chimps or gorillas.) We don't know whether the explanation is sperm competition—the longer penis offers an advantage by depositing the sperm nearer to the uterus—or the greater stimulation that a large penis can provide and for which women might have had a direct preference. But it seems safe to assume that women were likely enough to be mating with more than one man that there was some adaptive value to the difference. That assumption in turn suggests that women were unlikely to have been entirely under the control of any one man. As we've seen, competition on one side of the relationship increases the bargaining power of the person on the other side.
There are many things we don't know about human mating during the six million years or so that separate us from our common ancestor with chimps and bonobos. We don't know, for instance, whether that common ancestor had a sex life that was more like ours, like that of chimps, or like that of modern gorillas: testicles don't fossilize. So we don't know whether the autonomy of human females was declining in the later hunter-gatherer period or whether it remained relatively stable, only to decline precipitously once agriculture arrived. In many other respects, hunter-gatherer societies changed radically during prehistory.22 The reduction of size asymmetry between men and women since the earliest species of hominins suggests that force was becoming less important than persuasion in relations between men and women from the earliest times. But at the same time, cooperation among men was becoming more sophisticated, and this development might have been used to women's disadvantage, so a lot of uncertainty remains about what those relations were like. Nor do we know much about the context in which multiple mating by females took place: was it open and acknowledged or furtive and secret?
Naturally, the lack of data hasn't discouraged speculation. The recent book Sex at Dawn by Christopher Ryan and Cacilda Jetha has claimed that “our hominid ancestors have spent almost all of the past few million years or so in small, intimate bands in which adults had several sexual relationships at any given time. This approach to sexuality probably persisted until the rise of agriculture and private property no more than ten thousand years ago. In addition to voluminous scientific evidence, many explorers, missionaries and anthropologists support this view, having penned accounts rich with tales of orgiastic rituals, unflinching mate sharing, and an open sexuality unencumbered by guilt or shame.”23
The evidence these authors cite for the existence of multiple female partners includes the fairly convincing anatomical evidence we have already noted, along with various kinds of evidence from behavior, including women's greater capacity than men for delayed orgasm as well as for multiple orgasms, and the common tendency for female orgasm to be quite noisy, which suggests that natural selection was not selecting strongly for discretion. But the evidence that this was “an open sexuality unencumbered by guilt or shame” is more difficult to interpret than they acknowledge. The anthropological evidence they cite is both conflicting and controversial. For instance, Margaret Mead's famous work on Samoa, claiming evidence of uncomplicated multiple sexuality among Samoan adolescents, has not survived the scrutiny of later ethnographers, and it seems that anthropologists (of all persuasions) have found it difficult to avoid reading into their evidence many of their own fears, hopes, and fantasies.24 Even solid evidence that such open sexuality has sometimes existed would not be the same as evidence that this was the general condition of mankind before agriculture. More important, though, Ryan and Jetha appear to consider that any social practices that persisted for a long period in prehistory would have to be ones that were generally accepted and free of conflict. For instance, they object to a claim by Matt Ridley that monogamy evolved very early among our ancestors, so that “long pair-bonds shackled each ape-man to its mate for much of its reproductive life.” Ryan and Jetha comment in response: “Four million years is an awful lot of monogamy. Shouldn't these ‘shackles' be more comfortable by now?”25
As we've seen, this common tendency to see physical or behavioral traits that have evolved over a very long time as necessarily “optimal” involves a serious misunderstanding of how evolution works. The interests of different individuals can and often do conflict (and even the interests of different genes within an individual can conflict). Predators and prey can evolve traits that are self-defeating; male and female scorpions and bedbugs can have mating strategies of gruesome unpleasantness, as we saw in chapter 1. Men and women can signal to each other in spectacularly wasteful ways, as we saw in chapter 2. Human sexuality could well have remained for four million years or more in a stable equilibrium in which the “shackles,” as Ridley calls them, continued to be very uncomfortable for both men and women. Whether prehistoric sex was like this, or whether it was a more relaxed and open affair, is something about which the evidence is simply too scant and conflicting to allow us to be sure. The duplicity and inconsistency that so often characterize our sexual behavior could be bugs in a system that once worked more easily and openly, or they could be talents favored by natural selection for coping with irreducible conflicts of interest.
The anthropologist Sarah Blaffer Hrdy has suggested a more subtle view than either of these, namely that polyandrous liaisons are likely to have been widespread in hunter-gatherer societies while also being a significant source of tension and conflict—conflict that certain cultural practices could nevertheless help to channel: “Given what a powerful emotion sexual jealousy is, polyandrous liaisons are a risky strategy, dangerous for all concerned. But widely held beliefs about ‘partible paternity' help ease some of this tension. In these cultures, semen from every man a woman has sex with in the months before her infant is born supposedly contributes to the growth of her fetus, resulting in chimeralike composite young sired by multiple men. Each possible father is subsequently expected to offer gifts of food to the pregnant woman and to help provide for the resulting child.”26
Not surprisingly, significant uncertainty about the paternity of a woman's children can have conflicting effects on the incentives of the fathers to contribute to the children's upkeep, as well as cause tension between the different men and between each of them and the mother.27 Each additional possible father brings into the tent another person who might contribute to child rearing (modern Western societies still have a vestige of this idea in the institution of godparenting). At the same time each addition makes the tent more crowded, potentially diluting the incentives of those already inside to make a contribution. Just where the trade-off has been made by different species, different societies, and different individuals within them is a question on which the evidence is scant and often hard to interpret.
I mention in chapter 1 that many socially monogamous birds are not sexually monogamous, and this naturally prompts the question of whether the extent of multiple mating makes a difference to how well the system of social monogamy works. Researchers who have compared levels of multiple female mating across socially monogamous bird species have observed a significant tendency for greater monogamy to be associated with a system known as “cooperative breeding,” in which the female receives help in rearing her chicks not only from her partner but also from other birds (either the partner's siblings or the offspring from previous broods).28 But there are many exceptions to this tendency, including relatively monogamous seabirds like the puffin that do not breed cooperatively and birds like the fairy wren that have some of the highest levels of multiple mating ever recorded but do breed cooperatively. And the comparison excludes openly polyandrous species such as the dunnock. In fact, it makes comparisons only among the species that practice social monogamy. So the study looks only at the effect of a female's multiple mating on those males whose incentives to contribute to raising offspring are likely to be diminished (namely the female's mate and his siblings) and excludes those whose incentives are increased (namely the other potential fathers). It may tell us that cooperation among potential fathers is reduced in socially monogamous species, but it can't tell us about the comparison between socially monogamous and polygamous ones. This absence of information did not stop the authors from claiming that “the evolution of cooperative behavior is favored by low levels of promiscuity” (they should have added “among socially monogamous species”). This example just shows how hard it is to draw the right conclusions even from a rigorous and impressive scientific study.
There remains a lot of uncertainty both about the degree of multiple sexuality during prehistory and about how openly women controlled their own sexual lives. For the moment, what matters is that our female ancestors almost certainly enjoyed more bargaining power in relation to men than they were ever to do once agriculture arrived. The fact that men were competing for their favors is an important part of the evidence for that claim.
The Changing Needs of Human Babies
Even under conditions of hunting and gathering, human females were more dependent on men than chimpanzee or bonobo females are, and this dependence almost certainly weakened their bargaining power long before agriculture was even imaginable. The reason for this is simply the needs of human infants. Early humans colonized a new and very risky evolutionary niche. They took a bet on large brains (or rather natural selection took that bet for them) and paid a significant behavioral price for doing so: the need to nurture their offspring for an exceptionally long period.
What were the large brains for? In a word, cooperation, of a scale and complexity that no primate had yet attempted. The more sophisticated the relations of cooperation and reciprocity within a group of individuals, and the larger the group of individuals concerned, the greater the cognitive challenge of keeping track of mutual obligations. Among primates, species with larger brains (relative to their body size) tend to live in larger groups.29 This suggests that larger brains are a necessary cost of living in larger groups (one that implies an increasing selective pressure in favor of larger brains as group size increases). It also suggests that the benefits that accrue to primates with larger brains include those that can be realized only by exploiting the cooperation possibilities in larger groups. Today human societies are based on elaborate networks of cooperation that span the entire globe, but our hunter-gatherer ancestors had no vision of such a consequence, and of course natural selection could have had no foresight of it either. Cooperation even within hunter-gatherer bands expanded the range of natural environments that could be exploited and foods that could be consumed, and these benefits of cooperation were large enough to offset the very important costs.
Large brains are expensive: they need a lot of protein to build and a lot of energy to run. Human young (and their pregnant or nursing mothers) need more meat and more calories than do the offspring of young chimps to supply their growing brains. Foraging for such a diet requires more ambitious social arrangements: greater cooperation among hunters and a greater willingness for gatherers to move large distances when necessary in pursuit of food. Careful studies of forager communities by modern social anthropologists have shown that their young members typically do not produce enough food to feed themselves until they are in their late teens. Although adolescents are strong and nimble, both hunting and gathering need skill and experience, and adolescents are usually big eaters. The time of life when foragers are producing the biggest food surplus is not in their early twenties, when they are strongest, but in their early forties.30 Children and adolescents depend not just on prime-age adults but also on older people, especially grandmothers, who contribute surprisingly large amounts to their overall nutritional needs.31
In addition, as Richard Wrangham and his coauthors have argued, the development of cooking brought about a radical shift in human social arrangements. Cooking enabled the extraction of much higher nutritional value from the foodstuffs human beings could find, but the time it required made these foodstuffs more vulnerable to theft. Human males were coopted as protectors by females in exchange for the superior food—and also for the promise of a greater stake in the family unit. So eating cooked food both required and permitted more cooperative social arrangements than any primate society had previously seen.32
The second cost is that large brains require large skulls to house them. At the same time that natural selection was favoring infants with larger skulls, it was also selecting among women for those who could stand upright and move quickly on two legs. Given the basic model of the human body on which natural selection had to work, bipedalism set a limit to the size of the pelvis: a woman with a narrow torso and wide legs would have waddled much too slowly away from her predators. Although hominin females grew in size over the course of our evolution (more than males did, in fact), the metabolic costs of a larger overall body size prevented them from growing proportionally to their babies' skulls. The skull of a baby at the end of pregnancy can only just fit through the space in the pelvis. The only way for natural selection to make human babies with larger skulls was to end gestation prematurely and allow them to be born in a state of dependence that no other animal (except marsupials) could possibly manage. So to the costs of feeding large-brained babies must be added the costs of protecting them, and protecting them requires sophisticated social arrangements. Our human ancestors took a very large gamble on those large-brained babies, a gamble that depended on their being able to realize cooperative benefits that would outweigh the costs. It was a gamble that nearly failed. Most branches of the hominin line died out, including the Neanderthals, who had slightly larger brains than ours. Only our branch survived into the modern era, and it too very nearly didn't make it. Humans needed to adapt to new habitats many times, until eventually those large brains gave them the flexibility and inventiveness to change the habitats themselves.
This direction of evolution might appear to have been good for women, because men were now contributing resources to their offspring—notably food and protection—t hat they had not contributed when our ancestors were more like bonobos and chimpanzees. In some absolute sense it probably was good for women. But it certainly changed women's bargaining power relative to that of men, and changed it for the worse. For if men were contributing more, their contributions also became more essential, both because the contributions were more necessary (protein was important for those growing infant brains), and because fewer men among those available might be counted on to provide it, given the longer-term nature of the commitment involved. Being essential enabled men to exact a price for their contributions. In a word, the resources controlled by men had become more scarce.
Scarcity and the Gender Bargain
An unequal exchange doesn't have to be the result of coercion: it can be the outcome of bargaining between individuals who have different degrees of control over the resources at stake. In most hunter-gatherer societies, men hunt for meat, and women gather roots, fruit, berries and so forth. There has been controversy over whether men and women contribute equally to subsistence in such societies, but in some sense that controversy is irrelevant, because there's no particular reason to expect the terms of the exchange to have been very favorable to women. Human beings need high-protein diets (especially for infants with growing brains). Men may or may not be effective at gathering calories (anthropologists disagree on this point),33 but they're the specialists at hunting large quantities of protein. Women can't usually hunt, not because of any lack of ability but because hunting would be too dangerous for their infants, or simply impractical in the circumstances. As a result, the men have a resource the women need, but the reverse is not true: men can get calories on their own if they have to. The men's resources are therefore more scarce and consequently command a higher price. There's nothing moral about this: it's simply the outcome of the bargain.
Consider a simple example. The point of the example isn't to describe literally how a hunter-gatherer society would have operated, any more than the point of a map is to look like the landscape it helps you to navigate. It's rather to see how men could have gained an advantage just by the fact that they could produce something that women needed, even if they hadn't also had other advantages, such as the power to coerce women physically. Suppose in a group there are ten men and ten women. Each of the ten men can gather one kilogram of starch (tubers, say) in a day. Let's suppose the women are more productive than the men: each one can gather two kilograms of starch in a day. However, the group of ten men can also hunt, and once every ten days or so, they bring back an animal that yields one hundred kilograms of meat, equivalent to one kilogram of meat per man per day. The men can choose whether to gather starch or to hunt meat. As it happens, both meat and starch are essential for life: each person has to eat, on average, a minimum of a quarter kilogram of meat and a half kilogram of starch every day. But let's suppose that meat is strongly preferred: if meat and starch cost the same, everyone would prefer to eat three times as much meat as starch. Let's assume that their relative preferences for meat and starch don't change as their overall intake increases (that won't be strictly realistic, since preferences for meat are more characteristic of better-fed people, but it makes the example simpler to understand).
The men decide to see how much starch they can get in exchange for their meat if they bargain with the women. They decide to negotiate a rate at which each person can exchange what they have against what they want. Before the bargaining, each man has one kilogram of meat, and each woman has two kilograms of starch. You might think the obvious rate to exchange would be one for two: half a kilogram of meat for every kilogram of starch, so that each person would give up half of what they started with, and everyone would end up eating half a kilogram of meat and one kilogram of starch. But remembering how much people prefer meat, at that rate of exchange, everyone would rather give up half their remaining starch in order to get half as much meat again. So everyone will want to get more meat for less starch, until it occurs to someone to offer a better rate, say three kilograms of starch for every kilogram of meat.
In fact, under some reasonable assumptions about how the need for meat increases as you have less and less of it, the rate at which meat exchanges for starch might rapidly go up as high as six to one! Given that everyone prefers meat enough to want to eat three times as much of it if the prices were equal, it takes a price ratio of six to one to induce them to eat only half as much. At that rate, after the bargaining the men will end up with three-quarters of a kilogram of meat and one and a half kilograms of starch, while the women will end up with one-quarter of a kilogram of meat and just half a kilogram of starch. Only at these values will the high price of meat finally restrain their desire to eat by enough to make them ration out the small amount of meat available. But the result is startling. The men have three times as much as the women, even though the women are twice as productive as the men at the one task at which their performance can be compared, which is the gathering of starch. And the men only really work one day in ten.
So why does it work out that way? It's because, in this scenario, meat is essential to life, it is highly desired, and men have all of it. Of course, starch is also essential to life, but it is less highly desired, and women don't have a monopoly on the starch (men can get it too, if they want). So meat is scarce in a way that starch is not, and for this scarcity men can extract a price. No physical compulsion is needed, just a hard bargain. If instead people could just substitute in their diet a kilogram of starch for every kilogram of meat, then meat and starch would exchange at an equal rate—nobody would pay more. Then women (who start with two kilograms) would end up eating twice as much as men. It's the fact that they don't have access to a substitute for the men's meat that puts them at a disadvantage, not greater merit on the part of the men.
Of course, real life is (and always has been) different from this example in many ways. Some of these differences make the outcomes less unequal: for instance, men and women are not always driving purely selfish bargains with each other but often care about each other's welfare. Nor are they always bargaining for food for their own consumption: they typically obtain food to share with children and other family members, and often with group members outside their immediate family as well.34 Another major difference is that, as Sarah Blaffer Hrdy has repeatedly emphasized, in most societies mothers receive help in looking after their infants from a whole network of adults other than the father (or presumed father). Siblings, grandparents, and unrelated adults can all play a part, and together they reduce the extent to which mothers are dependent on any particular man and even on the men in the group as a whole—which is particularly fortunate given the notorious unreliability of that dependence: “Does this mean that fathers are not important? No. However, it does mean that a mother giving birth to slow-maturing, costly young does so without being able to count on help from the father. The impact on child well-being of variable paternal commitment depends on local conditions and on who else is around, able, and willing to help.”35
However, there are also some ways in which real life differs from this example that tend to reinforce rather than diminish the inequality between the sexes. Most important is the fact that men are not just negotiating individually but often collude. In particular, they have colluded in the coercion of women and have often ostracized or physically assaulted those women who try to escape the coercion. This collusion has been a natural by-product of the increasingly sophisticated collaboration undertaken by males in all societies for a range of purposes, including food production and warfare. (Gathering can be cooperative but doesn't have to be, or not to the same extent.) Those activities have reduced the autonomy of females, as we can see even among the other great apes. For instance, female bonobos can assert themselves more successfully against males than can female chimpanzees. This difference appears to be due to differences in the distribution of food in their foraging range, as a result of which male bonobos have fewer reasons to cooperate than male chimpanzees do. Hunting in groups, for instance, is more common among male chimpanzees (females are at a disadvantage in hunting because of their infants).
Cooperation in one domain, such as hunting, can extend easily to another domain, such as cooperating in the sexual harassment of females or in their confinement under the watchful eye of their mates. Females face a disadvantage in cooperation because, in all great ape species, the adolescent females leave the group to mate and so are unlikely to find themselves in a group with other siblings (there's some evidence that this was true of most prehistoric human groups as well, though the ethnographic evidence for surviving groups that have been studied indicates a considerable amount of flexibility in residence patterns).36 This is a disadvantage that bonobos have managed to overcome in part by their exuberant use of sexual play both to reinforce ties between females and to defuse the aggression of males. In Homo sapiens (a less sexually inventive species), though it's true that harassment and confinement of women were less common among hunter-gatherers than they later became under agriculture, that would certainly not have been for want of trying on the part of the men.
There are many factors other than bargaining, then, that have affected the balance of power between men and women. Nevertheless, our example shows that to the extent that they do bargain, the result can be strikingly asymmetric because of the luck that gives one party the control of a particularly scarce resource that the other needs. An important reason why human males have exerted so much more control over human females than chimp or bonobo males exert over females of their own species is simply that human males have had things to offer that human females needed much more than the females of those other species do, namely food and protection. And they needed these because natural selection took Homo sapiens into an evolutionary niche dependent on big brains and the capacity for social cooperation. That looks like bad news for women everywhere, but it's actually good news for modern women. For, as I show later, the balance of scarcity, with all that it implies for dependence, can be changed—up to a point.
Goodbye to the Hunter-Gatherer Life
The ecological conditions under which most hunter-gatherers lived imposed a fairly strict constraint on the extent to which men were able to harass and confine women. Women needed a fair amount of autonomy in order to forage. Shares of food consumed were also probably much less unequal than our example suggests they could have been under purely selfish bargaining. There is evidence, for instance, that in many societies, males have competed to be seen as good providers, even if they have sometimes chosen ways of demonstrating their prowess that privilege display over nutritional efficiency (hunting large game, for instance, may be a less efficient way of finding protein than going after small game).37 The inequality that mattered, though, for the degree of autonomy enjoyed by women, was not in how much men and women consumed, but in how much each of them controlled.
Relations between men in forager societies were almost certainly fairly egalitarian, because hunting requires trust rather than compulsion, and attempts to coerce others mean that they will simply leave. Christopher Boehm's book Hierarchy in the Forest has argued that this was not because Homo sapiens lost the competitiveness and the status consciousness that characterizes ape societies but because individuals who abused their power and status would provoke a counterreaction by coalitions of the weaker members of their group.38 In a hunter-gatherer society, the cooperation of these weaker members was essential even to the strong. Whatever the roots of this comparative egalitarianism among men, the implications for women were profound. They enjoyed some freedom to choose their men, and the choice was not (as it would become later, if they had any choice in the matter at all) between becoming the only wife of a very poor man and the junior wife of a very rich one.
It's important all the same not to romanticize the hunter-gatherer life. It was highly violent by today's standards: the best current estimates place the proportion of deaths from violence at around 14 percent, roughly ten times the rate for the world as a whole in the twenty-first century.39 It was also unhealthy: skeletons of hunter-gatherers from North and South America reveal that more than half of them suffered from abscesses that must have caused agony, and the state of their teeth is painful even to imagine.40 On both these counts, though, farming was to make things at first a great deal worse.41 And the conditions of women were to deteriorate by even more than those of men.42 Men were to use their talent for cooperation—in particular, their talent for devising and enforcing complex systems of social rules—to restrain and confine women much more than had been possible during the long ages of hunter-gatherer life.
It is one of strangest paradoxes in the natural world that females, endowed with scarce biological resources (their eggs), should have become so powerless in the face of the males who started out as impoverished, controlling only their cheap and abundant sperm. If, as I've suggested, it was men's good luck in controlling the scarce meat needed by women that enhanced their bargaining power, how can it also be true that women have been weakened, not strengthened, by possession of the scarce gametes? The answer lies in the difference between assets that you control securely and assets that control you. It's the difference between having a million dollars kept in a safe place and having a gold filling worth a million dollars in your teeth. The first you can use at your leisure and free of any fear of coercion. But to turn the second into anything valuable you have to find a cooperative dentist, and in the meantime you need to be careful about where you go walking at night. Men's resources, when they have had them, have been ones they controlled; women's resources have often controlled them, because they were physically vested in women's bodies. In short, the value of scarce resources depends not just on how scarce they are and on how much people want them, but also on the ease with which they can escape them and the degree of security they enjoy in their use of them. And that in turn depends on the framework of rights and conventions that govern their use. For our hunter-gatherer ancestors, those rights were entirely informal and largely consensual. Agricultural societies were to redefine them in ways that were not usually to women's advantage.
It is also a paradox that the developments that led men to restrict women's autonomy so severely grew out of the most remarkable experiment in social cooperation that the world had ever seen. The adoption of agriculture transformed not just the technology of food production but humans' whole way of life. It obliged them to settle so as to guard their fields; it brought them together into villages, towns, and cities; it obliged them to defend themselves; and it created a surplus that made it possible for farmers to pay for the activities of soldiers, scribes, priests, and kings. Jared Diamond has called it the greatest catastrophe in the history of humanity; it was also, without doubt, the foundation of all modern civilization.43 Charles Darwin was in no doubt that human social cooperation had emerged because of, not in spite of, natural selection. And for him, sexual reproduction was central to that cooperation, indeed emblematic of it. To the man who famously set out a balance sheet of the advantages and disadvantages of marrying, and who depended for his researches on a network of collaborators who supplied him with observations of natural and social phenomena from across the globe, the study of sexual reproduction was a moving reminder that nothing worthwhile in life can be accomplished entirely alone.