The Rise and Fall of the Third Chimpanzee

Chapter two traced our evolutionary history through the appearance of humans with fully modern anatomy and behavioural capabilities, but that chapter does not prepare us to go straight on to consider in more detail the development of human cultural hallmarks, such as language and art. That is because Chapter Two took up only the evidence of bones and tools. Yes, our evolution of large brains and upright posture was prerequisite to language and art, but that was not enough by itself. Human bones alone do not guarantee humanity. Instead, our rise to humanity also required drastic changes in our life-cycle, which will be the subject of Part Two of this book. For any species one can describe what biologists term its 'life-cycle'. That means traits such as the number of offspring produced per litter or birth; the interval between births; the parental care (if any) that offspring receive from the mother or father; social relations between adult individuals; how a male and female select each other to mate with; frequency of sexual relations; and longevity. We take the forms of these traits as they exist in humans for granted, as the norm, but our life-cycle is actually bizarre by animal standards. All the traits that I have just mentioned vary greatly between species, and we are extreme in most respects. To mention only some obvious examples, most animals produce litters much larger than one baby at a time, most animal fathers provide no parental care, and few other animal species live even a small fraction of three-score years and ten.

Of these exceptional features of ours, some are shared by apes, suggesting that we merely retained traits already acquired by our ape-like ancestors. For instance, apes too usually give birth to one baby at a time, have births spaced several years apart, and live for several decades. None of these things is true of the other animals most familiar (but less closely related) to us, such as cats, dogs, songbirds, and goldfish.

In others of these respects, we are greatly different even from apes. Here are some obvious differences whose functions are well understood. Human babies continue to have all food brought to them by their parents even after weaning, whereas weaned apes gather their own food. Most human fathers as well as mothers, but only chimpanzee mothers, are closely involved in caring for their young. Like seagulls but unlike apes or most other mammals, we live in dense breeding colonies of nominally monogamous couples, some of whom also pursue extramarital sex. All these traits are as essential as large brain-cases for the survival and education of human offspring. That is because our elaborate, tool-dependent methods of obtaining food make weaned human infants incompetent to feed themselves. They first require a long period of food-provisioning, training, and protection—an investment much more taxing than that facing the ape mother. Hence human fathers who want their offspring to survive to maturity have generally assisted their mate with more than just sperm, the sole parental input of an orangutan father. Our life-cycle also differs from that of wild apes in more subtle respects whose functioning is nevertheless still discernible. Many of us live longer than most wild apes: even hunter-gatherer tribes include some elderly individuals who are enormously important as repositories of experience. Men's testes are much larger than those of gorillas but smaller than those of chimps, for reasons that will become apparent in Chapter Three. We regard human female menopause as inevitable, and Chapter Seven will show why it makes good sense for humans, but it is almost unprecedented among other animals. The closest mammalian parallel is among some tiny mouselike marsupials in Australia, and it is their males, not their females, that undergo menopause. Our longevity, testis size, and menopause were thus also prerequisites to our humanity. Still other features of our life-cycle differ far more drastically from those of apes than do our testes, yet the functions of those remaining novel features of ours remain hotly debated. We are unusual in having sex mainly in private and for fun, rather than mainly in public and only when the female is able to conceive. Ape females advertise the time when they are ovulating; human females conceal it even to themselves. While anatomists understand why men's testes are the size that they are, an explanation for men's relatively enormous penis still escapes us. Whatever their explanation, all these-features, too, are part of what defines humanity. Certainly, it is hard to picture how fathers and mothers could cooperate harmoniously in rearing their children if human females resembled some primate females in having their genitalia turn bright red at the time of ovulation, becoming sexually receptive only at that time, flaunting their red badge of receptivity, and proceeding to have sex in public with any male in the vicinity. Human society and child-rearing rest therefore not only on the skeletal changes mentioned in

Chapter Two, but also on these remarkable new features of our life-cycle. Unlike the case with our skeletal changes, however, we cannot follow through our evolutionary history the timing of each of these life-cycle changes, because they leave no direct fossil imprint. As a result, they receive only brief attention in paleontology texts despite their importance. Archaeologists have recently discovered a Neanderthal hyoid bone, one of the key pieces of our speech-producing equipment, but as yet no trace of a Neanderthal penis. We do not know whether Homo erectus was already on the road to evolving a preference for having sex in private, in addition to having evolved his and her well-documented large brain. Our sole clues about the dating of these life-cycle changes are that something about longevity can be inferred from skeletons, and that size differences between fossil men and women may be indirect reflections of their mating system (more of that in Chapter Three). We cannot even prove through fossils, as we can for our large brain size, that we rather than living apes are the ones whose life-cycles diverged most from the ancestral condition. Instead, we have to be content with merely inferring that conclusion from the fact that our life-cycles are exceptional compared not just to living apes but also to other primates, suggesting that we were the ones who did more changing.

Darwin established in the mid-Nineteenth Century that the anatomy of animals has evolved through natural selection. Within this century, biochemists have similarly traced how the chemical make-up of animals has evolved through natural selection. But so has the behaviour of animals, including reproductive biology and sexual habits in particular. Life-cycle traits have some genetic basis, as we shall see below, and vary quantitatively among individuals of the same species. For instance, some women are genetically predisposed to give birth to twins, while genes for long lifespan run in some families more than in others. Life-cycle traits affect our success in passing on our genes, through affecting our success in wooing mates, conceiving and rearing babies, and surviving as adults. Just as natural selection tends to adapt an animal's anatomy to its ecological niche and vice versa, so natural selection also tends to mould animals' life-cycles. Those individuals leaving the most numerous surviving offspring promote their genes for life-cycle traits as well as for bones and chemical make-up.

A difficulty with this reasoning is that it seems as if some of our traits, such as menopause and aging, would reduce (rather than enhance) our output of offspring and should not have resulted from natural selection. It often proves profitable to try to understand these paradoxes through the concept of'trade-offs'. In the animal world there is nothing that is free or pure good. Everything involves costs as well as benefits, by using space, time, or energy that could have been devoted to something else. You might otherwise have thought that women who never underwent menopause would leave more descendants than women who do. But consideration of the hidden costs of foregoing menopause (Chapter Seven) will help us understand why evolution did not design these strategies into us. The same considerations illuminate such painful questions as why we grow old and die (Chapter Seven), and whether we are better off (even in a narrow evolutionary sense) in being faithful to our spouse or in pursuing extramarital affairs (Chapter Four).

I have been assuming in this discussion that our distinctively human life-cycle traits have some genetic basis. The comments that I made in Chapter One about the function of genes in general apply here as well. Just as our height and most of our observable traits are not influenced by only a single gene, there surely is not a single gene specifying menopause, testis size, or monogamy. In fact, we know little about the genetic bases of human life-cycle traits, though selective breeding experiments in mice and sheep have illuminated the genetic control of their testis size. Enormous cultural influences obviously operate on our motivation for providing child care or seeking extramarital sex, and there is no reason to believe that genes contribute significantly to differences among individual people in these traits. However, genetic differences between humans and the other two chimpanzee species probably do contribute to the consistent differences in many life-cycle traits between all human populations and all chimpanzee populations. There is no human society, regardless of its cultural practices, whose men have chimpanzee-sized testes and whose women forego menopause. Among those 1.6 % of our genes that differ between us and chimps and that have any function, a significant fraction is likely to be involved in specifying traits of our life-cycle.

The story of our uniquely human life-cycle occupies the five chapters of Part Two. Chapter Three begins by taking up the distinctive features of human social organization and of sexual anatomy, physiology, and behaviour. As already mentioned, features that make us strange among animals include our societies of nominally monogamous couples, our genital anatomy, and our constant and generally private pursuit of sex. Our sex lives are reflected not only in our genitalia but also in the relative sizes of men's and women's bodies (much more equal than are the bodies of male and female gorillas or orangutans). We shall see how some of these familiar and distinctive features have known functions, while others continue to defy understanding. No honest discussion of the human life-cycle could get away with noting that we are nominally monogamous and just leaving it at that. Pursuit of extramarital sex is obviously greatly influenced by each individual's particular upbringing and by the norms of the society in which the individual lives. Despite all that cultural influence, we are left with having to explain the facts that both the institution of marriage and the occurrence of extramarital sex have been reported from all human societies; but that extramarital sex is unknown in gibbons, although they do practise 'marriage' (that is, lasting male/female pairing to rear offspring); and that the question of extramarital sex is meaningless for chimpanzees because they do not practise 'marriage'. Hence an adequate discussion of our uniquely human life-cycle must account for our combination of marriage with extramarital sex. As Chapter Four will show, animal precedents exist to help us make evolutionary sense of our combination: men and women tend to differ in their attitudes towards extramarital sex much as geese and ganders do.

Chapter Five turns to another distinctive human life-cycle trait: how we select our sex partners, marital or otherwise. That problem scarcely arises for baboon troops, in which there is little selection: any male tries to mate with each female as she comes into heat. While common chimpanzees practise some selection of their sex partners, they are still much less selective and much more promiscuously baboon-like than are humans. Mate selection is a decision of major consequence in the human life-cycle, because married couples share parental responsibilities as well as sexual involvement. Precisely because care of human children demands such heavy and prolonged parental investment, we have to select our co-investor much more carefully than does a baboon. Nevertheless, Chapter Five will show that we can find animal precedents for our procedure in choosing sex partners, by going beyond primates to rats and birds. Our mate selection criteria, explored in Chapter Five, are relevant to human racial variation, as will be discussed in Chapter Six. Humans native to different parts of the globe vary conspicuously in external appearance, as do gorillas, orangutans, and most other animal species occupying a sufficiently extensive geographic range. Our visible geographic variability has often been taken as a pretext for exercising a human hallmark to be discussed in Chapter Sixteen: genocidal killings. Some of the geographic variation in our appearance surely reflects natural selection moulding us to local climate, just as weasels in areas with winter snow develop white fur in winter for better camouflage and survival. But I shall argue in Chapter Six, as Darwin maintained, that our visible geographic variability arose mainly through sexual selection, as a result of those mate-choice procedures of ours discussed in Chapter Five.

Chapter Seven brings the discussion of our life-cycle to an end, by asking why our lives have to come to an end. Aging is another feature of our life-cycle so familiar that we take it for granted: of course we shall all grow old and eventually die. So will all individuals of all animal species, but different species age at very different rates. Among animals we are relatively long-lived and became even more so around the time that Cro-Magnons replaced Neanderthals. Our longevity has been important for our humanity, by permitting effective transmission of learned skills between generations. But even humans grow old. Why is aging inevitable, given our extensive capacity for biological self-repair?

Here, more than in any other chapter, the importance of thinking in terms of evolutionary tradeoffs becomes clear. If measured by the ability to leave increased numbers of offspring, it just would not pay us to make the increased investment in self-repair mechanisms required to live longer. We shall see that the trade-off concept also illuminates the puzzle of menopause: a shutdown of child-bearing, paradoxically programmed by natural selection so that women can leave more surviving children.

THREE THE EVOLUTION OF HUMAN SEXUALITY

Human sexuality seems normal to us but is bizarre by the stmdards of other animals. Our bizarre sex lives were as crucial to cur rise to human status as were our large brains.

No week passes without publication of yet another book about sex. Our desire to read about sex is surpassed only by our desire to practise it. You might suppose that the basic facts of human sexuality must be familiar to lay people and understood by scientists. Just test your own grasp of sex by trying to answer these five easy questions:

Among the various ape species and man, which has by tar the biggest penis, and what for?

Why should men be bigger than women?

How can men get away with having much smaller testes than chimpanzees?

Why do humans copulate in private, while all other social animals do it in public?

Why don't women resemble almost all other female mammals in having easily recognized days of fertility, with sexual receptivity confined to those days?

If your answer to the first question was 'the gorilla', put on a dunce's cap; the correct answer is man. If you gave any intelligent answers to the next four questions, publish them; scientists are still debating rival theories.

These five questions illustrate how hard it is to explain the most obvious facts of our sexual anatomy and physiology. Part of the problem is our hang-ups about sex: scientists did not even begin to study the subject seriously until recently, and they still have [rouble being objective.

Another difficulty is that scientists cannot do controlled experiments on the sexual practices of us humans, as they can on our cholesterol intake or tooth-brushing habits. Finally, sex organs do not exist in isolation: they are adapted to their owners' social habits and life-cycle, which are in turn adapted to food-gathering habits. In our own case that means, among other things, that evolution of human sex organs has been intertwined with that of human tool use, large brains, and child-rearing practices. Thus, our progress from being just another species of big mammal to being uniquely human depended on the remodelling not only of our pelvises and skulls, but also of our sexuality. Given knowledge of how an animal feeds, a biologist can often predict that animal's mating system and genital anatomy. If we want to understand how human sexuality came to be the way it is, we have to begin by understanding the evolution of our diet and our society. From the vegetarian diet of our ape ancestors, we diverged within the last several million years to become social carnivores as well as vegetarians. Yet our teeth and claws remained those of apes, not of tigers. Our hunting prowess depended instead on large brains: by using tools and operating in coordinated groups, our ancestors were able to hunt successfully despite their deficient anatomical equipment, and they regularly shared food with each other. Our ability to gather roots and berries also came to depend on tools and thus to require large brains.

As a result, human children took years to acquire the information and the practice needed to be an efficient hunter-gatherer, just as they still take years to learn how to be a farmer or computer programmer today. During those many years after weaning, our children are still too dumb and helpless to acquire their own food; they depend entirely on their parents to bring food to them. These habits are so natural to us that we forget that baby apes gather food as soon as they are weaned.

The reasons why human infants are totally incompetent at food-gathering are actually two-fold—mechanical and mental. Firstly, making and wielding the tools used to obtain food requires fine finger coordination that children take years to develop. Just as my three-year-old sons still cannot tie their own shoelaces, three-year-old hunter-gatherer children cannot sharpen a stone axe, weave a net, or build a dugout canoe. Secondly, we depend on much more brainpower than do other animals in acquiring food, because we have a much more varied diet and more varied and complicated food-gathering techniques. For instance, New Guineans with whom I work typically have separate names for about a thousand different species of plants and animals living in the vicinity. For each of those species they know something about its distribution and life history, how to recognize it, whether it is edible or otherwise useful, and how best to capture or harvest it. All this information takes years to acquire.

Weaned human infants cannot support themselves because they lack these mechanical and mental skills. They need adults to teach them, and they also need adults to feed them for the decade or two that they are being taught. As is true of so many other human hallmarks, these problems of ours have animal precedents. In lions and many other species, the young must be trained to hunt by their parents. Chimpanzees too have a varied diet, employ varied foraging techniques, and assist their young in obtaining food, while common (but not pygmy) chimps make some use of tools. Our distinction is not absolute but one of degree: for us the necessary skills and hence the parental burden are far greater than for lions or chimpanzees.

The resulting parental burden makes care by the father as well as the mother important for a child's survival. Orangutan fathers provide their offspring with nothing beyond their initial donation of semen; gorilla, chimpanzee, and gibbon fathers go beyond that to offer protection; but hunter—gatherer human fathers provide some food and much teaching as well. Hence human food-gathering habits required a social system in which a male retained his relationship with a female after fertilizing her, in order to assist in rearing the resulting child. Otherwise, the child would be less likely to survive, and the father less likely to pass on his genes. The orangutan system, in which the father departs after copulation, would not work for us. The chimpanzee system, in which several adult males are likely to copulate with the same oestrus female, also would not work for us. The result of that system is that a chimpanzee father has no idea which infants in the troop he has sired. For the chimp father that is no loss, as his exertions on behalf of troop infants are modest. The human father, however, who will contribute significantly to the care of what he thinks is his child, had better have some confidence in his paternity—for example, through having been the exclusive sexual partner of the child's mother. Otherwise, his child-care contribution may help pass on some other man's genes.

Confidence in paternity would be no problem if humans, like gibbons, were scattered over the landscape as separate couples, so that each female would only rarely encounter a male other than her consort. But there are compelling reasons why almost all human populations have consisted of groups of adults, despite the paranoia about paternity that this causes. Among the reasons: much human hunting and gathering involves cooperative group efforts among men, women, or both; much of our wild food occurs in scattered but concentrated patches, able to sustain many people; and groups offer better protection against predators and aggressors, especially against other humans.

In short, the social system we evolved to accommodate our un-apelike food habits seems utterly normal to us, but is bizarre by ape standards and is virtually unique among mammals. Adult orangutans are solitary; adult gibbons live as separate monogamous male/female pairs; gorillas live in polygamous harems, each consisting of several adult females and usually one dominant adult male; common chimpanzees live in fairly promiscuous communities consisting of scattered females plus a group of males; and pygmy chimpanzees form even more promiscuous communities of both sexes. Our societies, like our food habits, resemble those of lions and wolves: we live in bands containing many adult males and many adult females. Furthermore, we diverge from even lions and wolves in how those societies are organized: our males and females are paired off with each other. In contrast, any male lion within a lion pride can and regularly does mate with any of the pride's lionesses, making paternity unidentifiable. Our peculiar societies instead have their closest parallels in colonies of seabirds, like gulls and penguins, which also consist of male/female pairs.

At least officially, human pairing is more or less monogamous in most modern political states, but is 'mildly polygynous' among most surviving hunter-gatherer bands, which are better models for how mankind lived over the last million years. (This description omits consideration of extramarital sex, through which we become effectively more polygamous and whose scientifically fascinating aspects I shall discuss in Chapter Four.) By 'mildly polygynous', I mean that most hunter-gatherer men can support only a single family, but a few powerful men have several wives. Polygyny on the scale of elephant seals, among which powerful males have dozens of wives, is impossible for hunter-gatherer men, because they differ from elephant seals in having to provide child care. The big harems for which some human potentates are famous didn't become possible until the rise of agriculture and centralized government let a few princes tax everyone else in order to feed the royal harem's babies.

Now let's see how this social organization shapes the bodies of men and women. Take first the fact that adult men are slightly bigger than similarly aged women (about eight per cent taller and twenty per cent heavier, on the average). A zoologist from outer space would take one look at my 5-foot 8-inch wife next to me (5 foot 10 inches), and would instantly guess that we belonged to a mildly polygynous species. How, you may ask, can one possibly guess mating practices from relative body size?

It turns out that, among polygynous mammals, average harem size increases with the ratio of the male's body size to the female's body size. That is, the biggest harems are typical of species in which males are much larger than females. For example, males and females are the same size in gibbons, which are monogamous; male gorillas, with a typical harem of three to six females, weigh nearly double the weight of each female; but the average harem is forty-eight wives for the southern elephant seal, whose 3-ton male dwarfs his 700-pound wives. The explanation is that, in a monogamous species, every male can win a female, but in a very polygynous species most males languish without any mate, because a few dominant males have succeeded in rounding up all the females into their harems. Hence, the bigger the harem, the fiercer is the competition among males and the more important it is for a male to be big, since the bigger male usually wins the fights. We humans, with our slightly bigger males and slight polygyny, fit this pattern. (However, at some point in human evolution, male intelligence and personality came to count for Wore than size: male basketball players and sumo wrestlers don't tend to have more wives than male jockeys or coxswains.) Because competition for mates is fiercer in polygynous than in monogamous species, the polygynous species also tend to have more marked differences between males and females in other respects besides body size. These differences are the secondary sexual characteristics that play a role in attracting mates. For instance, males and females of the monogamous gibbons look identical at a distance, while male gorillas (befitting their polygyny) are easily recognized by their crested heads and silver-haired backs. Here too, our anatomy reflects our mild polygyny. The external differences between men and women are not nearly as marked as sex-related differences in gorillas or orangutans, but the zoologist from outer space could probably still distinguish men and women by the body and facial hair of men, men's unusually large penis, and the large breasts of women even before first pregnancy (in this we are unique among primates).

Proceeding now to the genitalia themselves, the combined weight of the testes in the average man is about \Vi ounces. This may boost the macho man's ego when he reflects on the slightly lower testis weight in a 450-pound male gorilla. But wait—our testes are dwarfed by the 4-ounce testes of a 100-pound male chimpanzee. Why is the gorilla so economical, and the chimp so well-endowed, compared to us?

The Theory of Testis Size is one of the triumphs of modern physical anthropology. By weighing the testes of thirty-three primate species, British scientists identified two trends: species that copulate more often need bigger testes; and promiscuous species in which several males routinely copulate in quick sequence with one female need especially big testes (because the male that injects the most semen has the best chance of being the one to fertilize the egg). When fertilization is a competitive lottery, large testes enable a male to enter more sperm-tickets in the lottery.

Here is how these considerations account for the differences in testis size among the great apes and humans. A female gorilla does not resume sexual activity until three or four years after giving birth, and she is receptive for only a couple of days a month until she becomes pregnant again. So even the successful male gorilla with a harem of several females experiences sex as a rare treat—if he is lucky, a few times a year. His relatively tiny testes are quite adequate for those modest demands. The sex life of a male orangutan may be somewhat more demanding, but not much. However, each male chimp in a promiscuous troop of many females lives in sexual nirvana, with nearly daily opportunities to copulate for a common chimp and several daily copulations for the average pygmy chimp. That, plus his need to outdo other male chimps in semen output if he is to fertilize the promiscuous female, explains his need for gigantic testes. We humans make do with medium-sized testes because the average man copulates more often than gorillas or orangutans but less often than chimps. In addition, the typical woman in a typical menstrual cycle does not force several men into sperm competition to fertilize her.

Thus, primate testis design well illustrates the principles of trade-offs and evolutionary cost/benefit analyses explained on page 52. Each species has testes big enough to do their job, but not unnecessarily larger ones. Bigger testes would just entail more costs without proportional benefits, by diverting space and energy from other tissues and increasing the risk of testicular cancer.

From this triumph of scientific explanation we descend to a glaring failure: the inability of twentieth-century science to formulate an adequate Theory of Penis Length. The length of the erect penis averages 1 Vi inches in a gorilla, 1 Vz inches in an orangutan, 3 inches in a chimp, and 5 inches in a man. Visual conspicuousness varies in the same sequence: a gorilla's penis is inconspicuous even when erect because of its black colour, while the chimp's pink erect penis stands out against the bare white skin behind it. The flaccid penis is not even visible in apes. Why does the human male need his relatively enormous, attention-getting penis, which is larger than that of any other primate? Since the male ape successfully propagates his kind with much less, does not the human penis represent largely wasted protoplasm that would be more valuable if devoted, say, to cerebral cortex or improved fingers?

Biologist friends to whom I pose this conundrum usually think of distinctive features of human coitus where they suppose a long penis might somehow be useful: our frequent use of the face-to-face position, our acrobatic variety of coital positions, and the supposedly long duration of our coital bouts. None of these explanations survives close scrutiny. The face-to-face position is also a preferred one for orangutans and pygmy chimps, and is used occasionally by gorillas. Orangutans vary face-to-face copulation with dorso-ventral and sideways positions, and do it while hanging from branches of trees—surely that demands more penile acrobatics than our comfortable boudoir exercises. Our mean duration of coitus (about four minutes for Americans) is much longer than for gorillas (one minute), pygmy chimps (fifteen seconds), or common chimps (seven seconds), but shorter than for orangutans (fifteen minutes) and lightning-fast compared to the twelve-hour-long copulations of marsupial mice. (Are you listening, ghosts of Errol Flynn and Don Juan?)

Since it thus seems unlikely that special features of human coitus demand a large penis, a popular alternative theory is that the human penis has also become an organ of display, like a peacock's tail or a lion's mane. This theory is reasonable but begs the question, what type of display, and to whom?

Proud male anthropologists unhesitatingly answer, an attractive display, to women, but this represents mere wishful thinking. Many women say that they are turned on by a man's voice, legs, and shoulders more than by the sight of his penis. A telling point is that the women's magazine Viva initially published photos of nude men but dropped them after surveys showed lack of female interest. When Vivas nude men disappeared, the number of female readers increased, and the number of male readers decreased. Evidently, the male readers were the ones buying Viva for its nude photos. While we can agree that the human penis is an organ of display, the display is intended not for women but for fellow men. icn.

Other facts confirm the role of a large penis as a threat or status display towards other men. Recall all the phallic art created by men for men, and the widespread obsession of men with their penis size. Evolution of the human penis was effectively limited by the length of the female vagina: a man's penis would damage a woman if it were significantly larger. However, I can guess what the penis would look like if this practical constraint were removed and if men could design it themselves. It would resemble the penis sheaths (phallocarps) used as male attire in some areas of New Guinea where I do field work. Phallocarps vary in length (up to 2 feet), diameter (up to 4 inches), shape (curved or straight), angle made with the wearer's body, colour (yellow or red), and decoration (such as a tuft of fur at the end). Each man has a wardrobe of several sizes and shapes from which to choose each day, depending on his mood that morning. Embarrassed male anthropologists interpret the phallocarp as something used for modesty or concealment, to which my wife had a succinct answer on seeing a phallocarp: 'The most immodest display of modesty I've ever seen!

Thus, astonishing as it seems, important functions of the human penis remain obscure. Here is a rich field for research.

Passing now from anatomy to physiology, we are immediately confronted by our sexual activity pattern, which must be considered freakish by the standards of other mammal species. Most mammals are sexually inactive most of the time. They copulate only when the female is in oestrus—that is, when she is ovulating and capable of being fertilized. Female mammals apparently 'know' when they are ovulating, for they solicit copulation then by presenting their genitals towards males. Lest a male miss the point, many female primates go further; the area around the vagina, plus in some species the buttocks and breasts, swells up and turns red, pink, or blue. This visual advertisement of female availability affects male monkeys in the same way that the sight of a seductively dressed woman affects male humans. In the presence of females with brightly swollen genitals, male monkeys stare much more often at the female's genitals, develop higher testosterone levels, attempt to copulate more often, and penetrate more quickly and after fewer pelvic thrusts than in the presence of females not displaying their wares. Human sexual cycles are quite different. The human female maintains her sexual receptivity more or less constantly, instead of having it sharply confined to a short oestrus phase. Indeed, despite numerous studies aimed at settling whether a woman's receptivity varies at all through her cycle, there is still no agreement about the answer—nor about the cycle phase when receptivity is maximal if it does vary.

So well concealed is human ovulation that we did not have accurate scientific information on its timing until around 1930. Before that, many physicians thought that women could conceive at any point in their cycle, or even that conception was most likely at the time of menstruation. In contrast to the male monkey who has only to scan his surroundings for brightly swollen lady monkeys, the unfortunate human male has not the faintest idea which ladies around him are ovulating and capable of being fertilized. A woman herself may learn to recognize sensations associated with ovulation, but it is often tricky, even with the help of thermometers and ratings of vaginal mucus quality. Furthermore, today's would-be mother, who tries in such ways to sense ovulation in order to achieve (or avoid) fertilization, is responding by cold-blooded calculation to hard-won, modern book knowledge. She has no other choice; she lacks the innate, hot-blooded sense of sexual receptivity that drives other female mammals. Our concealed ovulation, constant receptivity, and brief fertile period in each menstrual cycle ensure that most copulations by humans are at the wrong time for conception. To make things worse, menstrual cycle length varies more between women, or from cycle to cycle in a given woman, than for other female mammals. As a result, even young newlyweds who omit contraception and make love at maximum frequency have only a twenty-eight per cent probability of conception in each menstrual cycle. Animal breeders would be in despair if a prize cow had such low fertility, but in fact they can schedule a single artificial insemination so that the cow has a seventy-five per cent chance of being fertilized!

Whatever the main biological function of human copulation, it is not conception, which is just an occasional by-product. In these days of growing human overpopulation, one of the most ironic tragedies is the Catholic Church's claim that human copulation has conception as its natural purpose, and that the rhythm method is the only proper means of birth control. The rhythm method would be terrific for gorillas and most other mammal species, but not for us. In no species besides humans has the purpose of copulation become so unrelated to conception, or the rhythm method so unsuited for contraception.

For animals, copulation is a dangerous luxury. While occupied in acto flagrante, an animal is burning up valuable calories, neglecting opportunities to gather food, vulnerable to predators eager to eat it, and vulnerable to rivals eager to usurp its territory. Copulation is something to be accomplished in the minimum time required to do the job of fertilization. In contrast, human sex, as a device to achieve fertilization, would have to be rated a huge waste of time and energy, an evolutionary failure. Had we retained a proper oestrus cycle like other mammals, the wasted time could have been diverted by our hunter-gatherer ancestors to butchering more mastodons. By this results-oriented view of sex, any hunter-gatherer band whose females advertised their oestrus period could thereby have fed more babies and out-competed neighbouring bands.

Thus, the most hotly debated problem in the evolution of human reproduction is to explain why we nevertheless ended up with concealed ovulation, and what good all our mistimed copulations do us. For scientists, it is no answer just to say that sex is fun. Sure, it's fun, but evolution made it that way. If we were not getting big benefits from our mistimed copulations, mutant humans who had evolved not to enjoy sex would have taken over the world.

Related to this paradox of concealed ovulation is the paradox of concealed copulation. All other group-living animals have sex in public, whether they are promiscuous or monogamous. Paired seagulls mate in the middle of the colony; an ovulating female chimpanzee may mate consecutively with five males in each other's presence. Why are we unique in our strong preference for copulating in private?

Biologists are currently arguing over at least six different theories to explain the- origin of concealed ovulation and concealed copulation in humans. Interestingly, the debate proves to be a

Rohrschach test for the gender and outlook of the scientists involved. Here are the theories and their proponents:

1. Theory preferred by many traditional male anthropologists.

According to this view, concealed ovulation and copulation evolved in order to enhance cooperation and reduce aggression among male hunters. How could cavemen bring off the precise teamwork needed to spear a mammoth, if they had been fighting that morning for the public favours of a cavewoman in oestrus? The implicit message of this theory is that women's physiology is important chiefly for its effect on bonds between men, the real movers of society. However, one can broaden this theory to make it less blatantly sexist. Visible oestrus and sex would disrupt human society by affecting female/female and male/female as well as male/male bonds.

To illustrate this broadened version of the prevalent theory, consider the following scene from an imaginary soap opera, showing what life would be like for us modern hunter-gatherers if we did not have concealed ovulation and private copulation. Our soap opera stars Bob and Carol and Ted and Alice and Ralph and Jane. Bob, Alice, Ralph, and Jane work together in an office where the men hunt contracts and the women gather accounts payable. Ralph is married to Jane. Bob's wife is Carol, and Alice's husband is Ted. Carol and Ted work elsewhere.

One morning, Alice and Jane both discover on awakening that they have turned bright red in order to advertise impending ovulation and sexual receptivity. Alice and Ted make love at home before they go off in their separate directions to work. Jane and Ralph go together to work, where they copulate occasionally on the office sofa in the presence of their co-workers.

Bob cannot help lusting for Alice and Jane when he sees them bright red and sees Jane and Ralph copulating. He is unable to concentrate on his work. He repeatedly propositions Jane and Alice.

Ralph drives Bob away from Jane.

Alice is faithful to Ted and rejects Bob, but the hassle also interferes with her work.

All day, Carol in her office elsewhere is seething with jealousy at the thought of Alice and Jane, because Carol knows that Alice and Jane are bright red and attractive to Bob, while she (Carol) is not.

As a result, the office succeeds in bagging few contracts and accounts. In the meantime, other offices, where ovulation is concealed and where copulation is private, prosper. Eventually, Bob's,

Alice's, Ralph's, and Jane's office goes extinct. The only offices that survive are those with concealed ovulation and copulation.

This parable suggests that the traditional theory, by which concealed ovulation and copulation evolved to promote cooperation within human societies, is plausible. Unfortunately, there are other, equally plausible theories that I will now explain more briefly.

2. Theory preferred by many other traditional male anthropologists.

Concealed ovulation and copulation cement the bonds between a particular man and woman, thereby laying the foundations of the human family. A woman remains sexually attractive and receptive so that she can satisfy a man sexually all the time, bind him to her, and reward him for his help in rearing her baby. The sexist message: women evolved to make men happy. Left unexplained by this theory is the question of why pairs of gibbons, whose unflinching devotion to monogamy should make them role models for the Moral Majority, remain constantly together despite having sex only every few years.

3. Theory of a more modern male anthropologist (Donald Symons).

Symons noted that a male chimpanzee who kills a small animal is more likely to share the meat with an oestrus female than with a non-oestrus female. This suggested to Symons that human females might have evolved a constant state of oestrus, in order to ensure a frequent meat supply from male hunters by rewarding them with sex. As an alternative theory, Symons noted that women in most hunter—gatherer societies have little say in selection of a husband. The societies are male-dominated, and male clans just suit themselves by exchanging daughters in marriage. However, by being constantly attractive, even a woman wed to an inferior male could privately seduce a superior male and secure his genes for her children. Symons' theories, while still male-orientated, at least represent a step forward in that he views women as cleverly pursuing their own goals.

4. Theory produced jointly by a male biologist and a female biologist (Richard Alexander and Katherine Noonan).

If a man could recognize signs of ovulation, he could use that knowledge to fertilize his wife by copulating with her only while she is ovulating. He could then safely neglect her the rest of the time and go off and philander, secure in the knowledge that the wife he left behind was unreceptive, if not already fertilized. Hence women evolved concealed ovulation to force men into a permanent marriage bond, by exploiting male paranoia about fatherhood. Not knowing the time of ovulation, a man must copulate often with his wife to have a chance of fertilizing her, and that leaves him less time to develop dalliances with other women. The wife benefits, but so does the husband. He gains confidence in his paternity of his children, and he need not worry that his wife will suddenly attract many competing men by turning bright red on a particular day. At last, we have a theory seemingly grounded in sexual equality.

5. Theory of a female sociobiologist (Sarah Hrdy).

Hrdy was impressed by the frequency with which many primates—including not only monkeys but also baboons, gorillas, and common chimps—kill infants not their own. The bereaved mother is thereby induced to come into oestrus again and often mates with the murderer, thus increasing his output of progeny. (Such violence has been common in human history: male conquerors kill the vanquished men and children but spare the women.) As a counter-measure, Hrdy reasoned, women evolved concealed ovulation in order to manipulate men by confusing the issue of paternity. A woman who distributed her favours widely would thereby enlist many men to help feed (or at least not to kill) her infant, since many men could suppose themselves to be the infant's father. Whether this theory is right or wrong, we must applaud Hrdy's overturning of conventional masculine sexism and transferring sexual power to women.

6. Theory of another female sociobiologist (Nancy Burley).

The average 7-pound newborn human weighs double a newborn gorilla, but the 200-pound gorilla mother dwarfs the average human mother. Because the newborn human is so much larger in relation to its mother than are newborn apes, birth is exceptionally painful and dangerous in humans. Until the advent of modern medicine, women often died in childbirth, whereas I have never heard of such a fate befalling a female gorilla or chimpanzee. Once humans had evolved enough intelligence to associate conception with copulation, oestrous women could have chosen to avoid copulating at the time of ovulation, and could have thereby spared themselves the pain and peril of childbirth, but such women would have left fewer descendants than women who could not detect their ovulation. Thus, where male anthropologists saw concealed ovulation as something evolved by women for men (Theories 1 and 2), Nancy Burley sees it as a trick that women evolved to deceive themselves.

Which of these six theories for the evolution of concealed ovulation is correct? Not only are biologists uncertain; it is only in recent years that the question has begun to receive serious attention. This dilemma exemplifies a pervasive problem in establishing causation in evolutionary biology, as well as in history, psychology, and many other fields where one cannot manipulate variables to perform controlled experiments. Such experiments would afford the most convincing way to demonstrate cause or function. If we could remodel one tribe of people so that all women advertised their day of ovulation, we could then see whether cooperation within or between couples broke down, or whether the Women used their new knowledge to avoid becoming pregnant. In the absence of such experiments, we can never be certain what human society Would really be like today without concealed ovulation. If it is hard to determine the function of things happening today under our eyes, how much harder must it be to determine functions in the vanished past! We know that human bones and tools were different hundreds of thousands of years ago, when concealed ovulation may have been evolving. Probably human sexuality, including the function of concealed ovulation, may also have been different then, in ways now hard for us to picture. Interpretation of our past runs the constant risk of degenerating into mere 'paleopoetry' stories that we spin today, stimulated by a few bits of fossil bone, and expressing like Rohrschach tests our own personal prejudices, but devoid of any claim to validity about the past.

Nevertheless, having mentioned six plausible theories, I cannot just walk away from the problem without attempting some synthesis. Here again, we come up against another pervasive problem in establishing causation. It is rare for complex phenomena such as concealed ovulation to be influenced by only a single factor. It would be as silly to seek a single cause of concealed ovulation as to claim that there was a single root cause of the First World War. Instead, there were many independent factors in the period 1900–1914 pushing towards war, others pushing towards peace. War finally broke out when the net weight of factors tipped towards war. Yet that does not excuse going to the opposite extreme of 'explaining' complex phenomena by an unweighted laundry list encompassing every conceivable factor.

As a first step towards pruning down our laundry list of six theories, let's realize that, whatever factors caused our distinctive sexual habits to evolve in the distant past, they would not be persisting today if there were not some factors still sustaining them. But the factors responsible for their initial appearance need not have been the same as the ones now operative. In particular, although the factors behind Theories 3, 5, and 6 may have been major ones long ago, they do not seem to be so now. Only a minority of modern women uses sex either to obtain food or other resources from a number of men, or to confound paternity and induce many men simultaneously to support a woman's child. Postulates of their former role are paleopoetry, albeit plausible paleopoetry. Let's just content ourselves with trying to understand why concealed ovulation and frequent private copulation might make sense now. At least, our guesses can be guided by introspection about ourselves plus observations of others.

The factors behind Theories 1, 2, and 4 seem to me still operative today, and to be facets of the same paradox, the most distinctive feature of human social organization. That paradox is that a man and woman desirous for their child (and genes) to survive must cooperate with each other for a long time to rear their child, and must also cooperate economically with many other couples living close by. It is obvious that regular sexual relations between a man and woman intensifies their connection, compared to their connections with other women and men whom they see daily but with whom they are not involved sexually. Concealed ovulation and constant receptivity advance this 'new' function of sex (new by the standards of most mammals) as a social cement, not just a device for fertilization. This function is not, as in the traditional male chauvinist version of Theories 1 and 2, a sop thrown by a cold, calculating woman to a sex-starved man, but instead an inducement for both sexes. Not only have all signs of female ovulation vanished, but the act of sex itself takes place privately, to emphasize the distinction between sexual and non-sexual partners within the same close group. As for the objection that gibbons remain monogamously involved without the reward of constant sex, that is easy to explain: each gibbon couple has minimal social—and no economic—involvement with other gibbon couples.

Human testis size also seems to me an outcome of that same basic paradox of human social organization. While our testes are larger than a gorilla's, because we have frequent sex for fun, they are still smaller than a chimpanzee's, because we are more monogamous. The oversized human penis may have evolved as an arbitrary sexual display symbol, as arbitrary as a lion's mane or a woman's enlarged breasts. Why were lionesses not the ones to develop enlarged breasts, lions an oversized penis, and men a mane? If they had, those permuted signals could have functioned equally well. That it did not come out that way may bejust an accident of evolution, a result of each species' and sex's relative ease of evolving those various structures. But there is still something basic missing from our discussion so far. I have talked about an idealized form of human sexuality: monogamous couples (plus a few polygynous households), husbands confident in their paternity of their wives' children, and husbands helping their wives with child-rearing rather than neglecting the kids in order to philander. As justification for discussing this fictitious ideal, I maintain that actual human practice is much closer to this ideal than to baboon or chimpanzee practice. But the ideal is still fictitious. Any social system with rules of conduct is open to the risk of individuals cheating when they find the advantages of cheating to outweigh the burden of sanctions. The question is thus a quantitative one. Does cheating become so regular that the whole system collapses, or does cheating occur but not so often as to destroy the system, or is cheating vanishingly rare? As translated for human sexuality, that question becomes one of whether ninety, thirty or one per cent of human babies are fathered extramaritally. That question ^and its consequences will be the subject of the next chapter.

FOUR THE SCIENCE OF ADULTERY

Cold-blooded analysis of adultery views life as an evolutionary contest whose winners are those individuals leaving the largest number of surviving offspring. This view helps one understand why humans reinvented adultery after the other two chimps had bypassed it.

People have many reasons to lie when asked whether they have committed adultery. Consequently, it is notoriously difficult to get accurate scientific information about this important subject. One of the few existing sets of hard facts emerged as a totally unexpected by-product of a medical study, performed nearly half-a-century ago for a different reason. That study's findings have never been revealed until now.

I recently learned those facts from the distinguished medical scientist who ran the study. (Since he does not wish to be identified in this connection, I shall refer to him as Dr X.) In the late 1940s Dr X was studying the genetics of human blood groups, which are molecules that we acquire only by inheritance. Each of us has dozens of blood-group substances on our red blood cells, and we inherit each substance either from our mother or from our father. The study's research plan was straightforward: go to the obstetrics ward of a highly respectable US hospital; collect blood samples from 1,000 newborn babies and their mothers and fathers; identify the blood groups in all the samples; and then use standard genetic reasoning to deduce the inheritance patterns.

To Dr X's shock, the blood groups revealed nearly ten per cent of those babies to be the fruits of adultery! Proof of the babies' illegitimate origin was that they had one or more blood groups lacking in both alleged parents. There could be no question of mistaken maternity—the blood samples were drawn from an infant and its mother soon after the infant emerged from the mother. A blood group present in a baby but absent in its undoubted mother could only have come from its father. Absence of the blood group from the mother's husband as well showed conclusively that the baby had been sired by some other man, extramaritally. The true incidence of extramarital sex must have been considerably higher than ten per cent, since many other blood-group substances now used in paternity tests were not yet known in the 1940s, and since most bouts of intercourse do not result in conception. At the time that Dr X made his discovery, research on American sexual habits was virtually taboo. He decided to maintain a prudent silence, never published his findings, and it was only with difficulty that I got his permission to mention his results without betraying his name. However, his results have more recently been confirmed by several similar genetic studies whose results did get published. Those studies variously showed between about five and thirty per cent of American and British babies to have been adulterously conceived. Again, the proportion of the tested couples of whom at least one practised adultery must have been higher, for the same two reasons as in Dr X's study.

We can now answer the question posed at the end of the last chapter: whether extramarital sex is for humans a rare aberration, a frequent exception to a 'normal' pattern of marital sex, or so frequent as to make a sham of marriage. The middle alternative proves to be the correct one. Most fathers really are raising their own children, and human marriage is not a sham. We are not just promiscuous chimpanzees pretending to be otherwise. Yet it is also clear that extramarital sex is an integral, albeit unofficial, part of the human mating system. Adultery has also been observed in many other animal species whose societies resemble ours in being based on male and female co-parents with a lasting bond. Since such lasting bonds do not characterize common chimpanzee or pygmy chimpanzee society, it is meaningless to talk of adultery in chimps. We must have reinvented it after our chimp-like ancestors had rendered it obsolete. Therefore, we cannot discuss human sexuality, and its role in our rise to humanity, without carefully considering the science of adultery. Most of our information about adultery's incidence has come from researchers asking people about their sex lives, rather than from blood-grouping their babies. Since the 1940s, the myth that marital infidelity is rare in the US has been publicly exploded by a long succession of surveys, beginning with the Kinsey report. Nevertheless, even though this is the supposedly liberated 1990s, we are still profoundly ambivalent about adultery. It is thought of as exciting; no television soap opera could attract many viewers without it. It has few rivals as a basis of humour. Yet, as Freud pointed out, we often use humour to deal with things that are intensely painful. Thus, throughout history, adultery has also had few rivals as a cause of murder and human misery. In writing about this subject, it is impossible to remain completely serious, but it is also impossible not to be revolted at the sadistic institutions by which societies have attempted to deal with extramarital sex.

What makes a married person decide to seek or avoid adultery? Scientists have theories to explain many other things, so it should not be surprising that there is also a theory of extramarital sex (abbreviated to EMS, and not to be confused with premarital sex or PMS, in turn not to be confused with premenstrual syndrome, also PMS). With many species of animals the problem of EMS never arises, because they do not opt for marriage in the first place. For instance, a female Barbary macaque in heat copulates promiscuously with every adult male in her troop and averages one copulation per seventeen minutes. However, some mammals and most bird species do opt for 'marriage'. That is, a male and a female form a lasting pair-bond to devote care or protection to their joint offspring. Once there is marriage, there is also the possibility of what socio-biologists euphemistically term 'the pursuit of a mixed reproductive strategy' (abbreviated to MRS). In plain English, that means being married while simultaneously seeking extramarital sex. Married animals vary enormously in the degree to which they mix their reproductive strategies. There appears to be no recorded instance of EMS in the little apes called gibbons, while snow geese indulge regularly. Human societies similarly vary, but I suspect that none comes close to the faithful gibbons. To explain all this variation, sociobiologists have found it useful to apply the reasoning of game theory. That is, life is considered an evolutionary contest whose winners are those individuals leaving the largest number of surviving offspring. Contest rules are set by the ecology and reproductive biology of the particular species. The problem is then to figure out which strategy is most likely to win the contest: rigid fidelity, pure promiscuity, or a mixed strategy. But I must make one thing clear right at the outset. While this sociobiological approach certainly proves useful for understanding adultery in animals, its relevance for human adultery is an explosive issue and one to which I shall return. The first thing one realizes is that the best game strategy differs between males and females of the same species. This is because of two profound differences between the reproductive biology of males and females, in the minimum necessary reproductive effort, and in the risk of being cuckolded. Let's consider these differences, which are painfully familiar to humans.

For men, the minimum effort needed to sire an offspring is the act of copulation, a brief expenditure of time and energy. The man who sires a baby by one woman one day is biologically capable of siring a baby by another woman the same day. For women, however, the minimum effort consists of copulation plus pregnancy plus (throughout most of human history) several years spent nursing—a huge commitment of time and energy. Thus, a man potentially can sire far more offspring than can a woman. A nineteenth-century visitor who spent a week at the court of the Nizam of Hyderabad, a polygamous Indian potentate, reported that four of the Nizam's wives gave birth within eight days, and that nine more births were anticipated for the following week. The record lifetime number of offspring for a man is 888, sired by Emperor Moulay Ismail the Bloodthirsty of Morocco, while the corresponding record for a woman is only sixty-nine (a nineteenth-century Moscow woman specializing in triplets). Few women have topped twenty children, whereas some men easily do so in polygynous societies. As a result of this biological difference, a man stands to gain much more from EMS or polygamy than does a woman—if one's sole criterion is the number of offspring born. (To female readers about to stop reading in outrage, or to male readers about to cheer, I warn you now—keep reading, there is much more to the question of EMS.) For human EMS the statistical evidence is naturally hard to come by, but for human polygamy it is readily available. In the sole polyandrous society for which I could find data, the Tre-ba of Tibet, women with two husbands average fewer children, not more children, than women with one husband. In contrast, nineteenth-century American Mormon men realized big benefits from polygyny: men with one wife averaged only seven children, but men with two wives averaged sixteen children, and those with three wives averaged twenty. Polygynous Mormon men as a group averaged 2.4 wives and fifteen children, while polygynous Mormon church leaders in particular averaged five wives and twenty-five children. Similarly, among the polygynous Temne people of Sierra Leone, a man's average number of children increases from 1.7 to seven as his number of wives increases from one to five. The other sexual asymmetry relevant to the best game strategy involves confidence that one really is the biological parent of one's putative offspring. A cuckolded animal, deceived into rearing offspring not its own, has thereby lost the evolutionary game while advancing the victory of another player, the real parent. Barring a switch of babies in the hospital nursery, women cannot be cuckolded; they see their baby emerge from their bodies. Nor can there be cuckoldry of males in animal species practising external fertilization (that is, fertilization of eggs outside the female's body). For instance, some male fish watch a female shed eggs, then immediately deposit sperm on the eggs and scoop them up to care for them, secure in their paternity. However, men and other Wale animals practising internal fertilization—fertilization of eggs inside the female's body—can readily be cuckolded. All that the putative father knows for sure is that his sperm went into the mother, and eventually an offspring came out. Only observation of the female throughout her whole fertile period can absolutely exclude the possibility that some other male's sperm also entered and did the actual fertilizing.

An extreme solution to this simple asymmetry is the one formerly adopted by southern India's

Nayar society. Among the Nayar, women freely took many lovers simultaneously or in sequence, and husbands accordingly had no confidence in paternity. To make the best of a bad situation, a Nayar man did not live with his wife or care for his supposed children, but he instead lived with his sisters and cared for his sisters' children. At least, those nieces and nephews were sure to share one-quarter of his genes.

Bearing in mind these two basic facts of sexual asymmetry, we can now examine what is the best game strategy, and when EMS pays. Let's examine three game plans of increasing complexity:

Game Plan 1.

A man should always seek EMS, because he has so little to lose and so much to gain.

Consider the hunter-gatherer conditions prevailing throughout most of human evolution, under which a woman could at best rear about four children in the course of her life. Through one dalliance, her otherwise faithful husband could increase his lifetime reproductive output from four to five: an enormous increase of twenty-five per cent, for only a few minutes' work.

What is wrong with this dazzlingly naive reasoning?

Game Plan 2.

A moment's reflection should expose a basic flaw of Game Plan 1; it considers only the potential benefits of EMS to a man and ignores his potential costs. Obvious costs would include the risk of detection and injury or murder by the husband of the woman sought as

EMS partner; the risk that one's own wife will desert; the risk of being cuckolded by one's wife while one is off seeking EMS; and the risk that one's legitimate children will suffer through one's neglect of them. Thus, according to Game Plan 2 the would-be Casanova, like a sophisticated investor, should seek to maximize his gains while minimizing his losses. What reasoning could be more impeccably judicious?

Game Plan 3.

The man silly enough to be satisfied with Game Plan 2 has obviously never approached a lady with an offer of EMS or PMS. Worse yet, the silly man has never even thought about the statistics of human heterosexual intercourse, which dictate that, for every bout of

EMS by a man, there must be one bout of EMS (or at least PMS) by a woman.

Game Plans 1 and 2 share the flaw that they ignore considerations of the woman's strategy, without which any male strategy is doomed to failure. Hence Game Plan 3 must combine a male strategy and a female strategy. But, since one husband suffices to realize a woman's maximum reproductive potential, what could possibly attract a woman to EMS or PMS? This question puzzles the current generation of theoretical sociobiologists with a purely intellectual interest in EMS, just as it has taxed the ingenuity of would-be male adulterers throughout human history.

To proceed further with our theoretical exploration of Game Plan 3, we need rigorous empirical data on EMS. As surveys of people's sexual habits are notoriously unreliable, let's first turn to some recently published studies of birds that nest as mated pairs in large colonies. These, rather than our closest relatives the apes, are the animals whose mating system most closely resembles our own. Compared to us, birds have the disadvantage that one cannot ask them about their motives for EMS, but this is no great loss, as our answers are often lies anyway. The great virtue of colonial birds for EMS research is that one can band the birds in a colony, then sit nearby for hundreds of hours and determine exactly who does what with whom. I am unaware of equivalent information for a large human population. Important recent observations of adultery among birds were made on five species of herons, gulls, and geese. All five nest in dense colonies composed of nominally monogamous male/female pairs. One parent alone is incapable of rearing a chick, as an unguarded nest is likely to be destroyed while the parent is off gathering food, nor is a male capable of feeding or guarding two families simultaneously. Consequently, among the ground rules of sexual strategy for these colonial birds are the following: polygamy is forbidden; copulation with or by an unmated female is pointless, unless she soon acquires a mate to care for the resulting offspring; but surreptitious fertilization by one male of another male's mate is a viable strategy.

The first study involved great blue herons and great egrets at Hog Island, Texas. In these species the male builds a nest and stays there to court visiting females. Eventually a male and female accept each other ^and copulate about twenty times. The female then lays eggs and goes off ^to spend most of the daylight hours feeding, while the male remains to guard the nest and eggs. During the first day or two after pairing, the male often resumes courting any passing female as soon as his mate leaves to feed, but EMS does not result. Instead, the male's halfway-unfaithful behaviour seems to constitute 'divorce insurance' that reserves a back-up mate for him in case his own mate deserts (she does desert him in up to twenty per cent of the pairings reported). The passing 'back-up' females pursue the courtship out of ignorance. They are seeking a mate and have no way of knowing that the male is already mated, until his spouse returns (which she does at frequent intervals) and drives them off. Eventually, the male gains complete confidence that he will not be deserted, and he ceases to court any passing females.

In the second study, of little blue herons in Mississippi, behaviour that might have originated as divorce insurance took a more serious turn. Sixty-two cases of EMS were documented, mostly between a female on her nest and a male from the neighbouring nest while the female's mate was busy finding food. Most females initially resisted but then ceased resisting, and some females engaged in more EMS than marital sex. To reduce his own risk of being cuckolded, the adulterous male did his feeding as quickly as possible, returned often to his own nest to guard his mate, and travelled no further than neighbouring nests to seek EMS. EMS was usually timed to occur when the chosen female had not yet completed egg-laying and could still be fertilized. However, adulterous copulations were quicker than marital copulations (eight versus twelve seconds), hence possibly less effective at fertilizing, and nearly half of all nests involved in EMS were subsequently abandoned.

Among herring gulls in Lake Michigan, thirty-five per cent of mated males were observed to engage in EMS. This percentage is nearly the same as the value of thirty-two per cent reported for young American husbands in a study published by Playboy Press in 1974, but there is a big difference between gulls and humans in female behaviour. Whereas Playboy Press reported EMS for twenty-four per cent of young American wives, every mated female gull virtuously rejected adulterous male advances and never solicited the neighbouring male in her own mate's absence. Instead, all cases of male EMS involved unmated female gulls practising PMS. To decrease his own risk of being cuckolded, the male spent more time chasing intruders away from his nest when his mate was fertile than when she was not fertile. As for how the male induced his mate to remain faithful during the time that he was off seeking EMS, his secret—like that of some married men similarly pursuing a mixed reproductive strategy—consisted of feeding her diligently and copulating often whenever she was receptive.

Our final set of rigorous data involves snow geese breeding in Manitoba. Just as I explained in the case of little blue herons, EMS in snow geese mainly involves a male approaching an initially resisting female on a neighbouring nest in the absence of her mate. The mate's absence is usually due to the fact that he himself is off seeking EMS. It may seem as if the male thereby loses as much as he gains, but a male goose is not so dumb. As long as the female is still laying eggs, her mate remains to guard her. (A nesting female is propositioned fifty times less often in her mate's presence than in his absence.) Only after the female has finished laying does her mate go off on EMS quests, with his paternity assured at home. Such bird studies illustrate the value of a scientific approach to adultery. They have revealed a series of sophisticated strategies by which adulterous male birds try to have it both ways, so as to obtain confidence of paternity at home while sowing their seed abroad. The strategies include wooing unmated females for 'divorce insurance', as long as one feels unsure of one's wife's fidelity; guarding one's fertile spouse; feeding her copiously and copulating with her often, to induce her to remain faithful in one's absence; and coveting one's neighbour's spouse at a time when she is fertile and one's own spouse is no longer fertile. However, not even these applications of the scientific method in all its power sufficed to clarify what, if anything, female birds gain from EMS. One possible answer is that female herons weighing desertion of their mates may use EMS to shop around for a new mate. Another is that some unmated female gulls in colonies with a deficit of males may get fertilized by PMS, and then try to rear the chicks with the help of another, similar female.

The chief limitation of these colonial bird studies is that the females often seem to be unwilling participants in EMS. For further understanding of a more active female role, we have no choice but to turn to human studies, riddled as they are with problems of cultural variation, observer bias, and dubiously reliable survey responses.

Surveys comparing men with women in various cultures scattered around the world typically purport to find the following differences: men are more interested in EMS than are women; men are more interested than women in seeking a variety of sexual partners for the sake of variety 'itself; women's motives for EMS are more likely to be marital dissatisfaction and/or a desire for a lasting new relationship; and men are less selective in taking on a casual female sexual partner than vice versa. For example, among the New Guinea highlanders with whom I work, the Men say they seek EMS because sex with their own wife (or even wives, m the case of polygynous men) inevitably becomes boring, while the women who seek EMS do so mainly because their husband cannot satisfy them sexually (for example, because of old age). In the questionnaires that several hundred young Americans filled out for a computer dating service, women expressed stronger partner preferences than men did in almost every respect: intelligence, status, dancing ability, religion, race, etc. The only category in which men were more selective than women was physical attractiveness. After a date the men and women then filled out 'debriefing' questionnaires, with the result that two-and-a-half times as many men as women expressed a strong romantic attraction to their computer-selected partner. Thus, the women were choosier, the men more undiscriminating, in their reactions to partners.

Obviously, we are on shaky ground if we expect an honest answer when we ask people their attitudes about EMS. However, people also express their attitudes in laws and in their behaviour. In particular, some widespread hypocritical and sadistic features of human societies stem from two fundamental difficulties that men face in seeking EMS. Firstly, a man who pursues an MRS is trying to have it both ways: he wishes to obtain sex with other men's wives, while denying sex with his own wife (or wives) to other men. Some men therefore inevitably gain at the expense of other men. Secondly, as we have discussed, there is a realistic biological basis for men's widespread paranoia about being cuckolded.

Adultery laws provide a clear example of how men have dealt with these dilemmas. Until recently, essentially all such laws—Hebraic, Egyptian, Roman, Aztec, Moslem, African, Chinese, Japanese, and others—were asymmetrical. They existed to secure a married man's confidence in his paternity of his children, and for no other purpose. Consequently these laws define adultery by the marital status of the participating woman; that of the participating man is irrelevant. EMS by a married woman is considered an offence against her husband, who is commonly entitled to damages, often including violent revenge or else divorce with refund of the bride price. EMS by a married man is not considered an offence against his wife. Instead, if his partner in adultery is married, the offence is against her husband; if she is unmarried, the offence is against her father or brothers (because her value as a prospective bride is reduced).

No criminal law against male infidelity even existed until a French law of 1810, and that law only forbade a married man to keep a concubine in his conjugal house against his wife's wishes. Viewed from the perspective of human history, the absence or near-symmetry of modern Western adultery laws is a novelty that only appeared in the last 150 years. Even today, prosecutors, judges, and juries in the US and England often reduce a homicide charge to manslaughter of the lowest degree, or else acquit altogether, when a husband kills an adulterous wife or her lover caught in the act.

Perhaps the most elaborate system to uphold confidence of paternity was that maintained by Chinese emperors of the T'ang Dynasty. For each of the emperor's hundreds of wives and concubines, a team of court ladies kept records on dates of menstruation, so that the emperor could copulate with that wife on a date likely to result in fertilization. Dates of copulation were also recorded, and as an auxiliary form of record-keeping, were commemorated by an indelible tattoo on the woman's arm and by a silver ring on her left leg. It goes without saying that equal thoroughness was applied to excluding men other than the emperor from the harem. Men of other cultures have resorted to less complicated but even more repulsive means of ensuring paternity. These measures limit sexual access to wives, or else to daughters or sisters who would command a high bride price if delivered as proven virgin goods. Relatively mild measures include close chaperoning or virtual imprisonment of women. Similar purposes are served by the code of 'honour and shame' widespread in Mediterranean countries. (Translation: EMS for me but not for you; only the latter is a shame to my honour.) Stronger measures include the barbaric mutilations euphemistically and misleadingly termed 'female circumcision'. These consist of removal of the clitoris or most of the external female genitalia to reduce female interest in sex, marital or extramarital. Men bent on total certainty invented infibulation, suturing a woman's labia majora nearly shut, so as to make intercourse impossible. An infibulated wife can be de-infibulated for childbirth or for re-insemination after each child is weaned, and can be re-infibulated when the husband takes a long trip. Female circumcision and infibulation are still practised in twenty-three countries today, from Africa through Saudi Arabia to Indonesia. When adultery laws, imperial records, and coercive restraint still fail to ensure paternity, murder is available as a last resort. The role of sexual jealousy as one of the commonest causes of homicide emerges from studies in many American cities and in many other countries. Usually, the murderer is a husband while the victim is his adulterous wife or her lover, or else the lover kills the husband. The table on the following page gives some actual numbers for murders committed in Detroit in 1972. Until the formation of centralized political states provided soldiers with loftier motives, sexual jealousy also loomed large in human history as a cause of war. It was the seduction (abduction, rape) by Paris of Mefielaus's wife Helen that provoked the Trojan War. In the modern New Guinean highlands, only disputes over ownership of pigs rival disputes over sex in triggering war.

Asymmetric adultery laws, tattooing of wives after insemination, virtual imprisonment of women, genital mutilation of women—these behavioural habits are unique to the human species, defining humanity as much as does invention of the alphabet. More exactly, they are new means to the old evolutionary goal of males promoting their genes. Some of our other means to this goal are ancient ones shared with many animals, including jealous murder, infanticide, rape, inter-group warfare, and adultery itself. Human male infibulators stitch the vagina closed; some male animals achieve the same result by cementing a female's vagina after copulating with her.

Sociobiologists have had considerable success at understanding the marked differences among animal species in the details of these practices. As a result of recent research, it is no longer controversial to conclude that natural selection caused animals to evolve behavioural patterns, as well as anatomical structures, that tend to maximize the number of their descendants. Few scientists doubt that natural selection moulded human anatomy. However, no theory has caused such bitter divisions among my fellow biologists today as the claim that natural selection likewise moulded our social behaviour. Most of the human behaviour discussed in this chapter is considered barbaric by modern Western society. Some biologists are outraged not only by the behaviours themselves, but also by sociobiological explanations for the evolution of the behaviours. To 'explain' a behaviour seems uncomfortably close to defending it.

Like nuclear physics and all other knowledge, sociobiology is available for abuse. People have never lacked pretexts to justify the abuse or killing of other people, but ever since Darwin formulated his theory of evolution, evolutionary reasoning has also been abused as such a pretext. Sociobiological discussions of human sexuality can be seen as seeking to justify men's abuse of women, analogous to the biological justifications advanced for whites' treatment of blacks or Nazis' treatment of Jews. In the critiques that some biologists have directed at sociobiology, two fears recur: that a demonstrated evolutionary basis for a barbaric behaviour would seem to justify it; and that a demonstrated genetic basis for the behaviour would imply the futility of attempts at change.

In my view, neither fear is warranted. As for the first, one can seek to understand how something arose, regardless of whether one considers that something admirable or abominable. Most books analysing the motives of murderers are not written in an effort to justify murder, but instead to understand its causes as a way of preventing it. As for the second fear, we are not mere slaves to our evolved characteristics, not even to our genetically acquired ones. Modern civilization is fairly successful at thwarting ancient behaviours like infanticide, and one of the main objectives of modern medicine is to thwart the effects of our harmful genes and microbes, despite our having come to understand why it is natural for those genes and microbes to tend to kill us. The case against infibulation does not collapse even if the practice can be shown to be genetically advantageous to male infibulators. Instead, we condemn it because we hold the mutilation of one person by another to be ethically loathsome.

While sociobiology is therefore useful for understanding the evolutionary context of human social behaviour, this approach still should not be pushed too far. The goal of all human activity cannot be reduced to the leaving of descendants. Once human culture was firmly in place, it acquired new goals. Many people debate today whether to have children, and many decide that they prefer to devote their time and energy to other activities. We shall reach a simiiar perspective in later chapters for other attributes as uniquely human as our sexuality, including our art and our abuse of drugs. For these activities too, one can identify animal precursors and discern original roles in promoting survival and gene transmission, but these activities also proceeded to take on a life of their own. Hence I claim only that evolutionary reasoning is valuable for understanding the origin of such human practices, and not that it is necessarily the only way to understand their current forms.

In short, we evolved, like other animals, to win the reproduction game. That contest has a single aim, to leave as many descendants as possible. Much of the legacy of that game strategy is still with us. But we have also chosen to pursue ethical goals, which can conflict with the goals and methods of the sexual contest. Having that choice among goals represents one of our most radical departures from other animals.

FIVE HOW WE PICK OUR MATES AND SEX PARTNERS

Most humans are choosier about their sex partners than are the (other two) chimpanzees. By what criteria do we select our spouse or bedmate, and how does each of us develop our individual standard of beauty?

One evening, while I was camping with some New Guinea men of the Fore tribe, the conversation turned to women and sex, and my Fore friends proceeded to explain to me their tastes: The most beautiful women are Fore women. They have gorgeous black skin, thick, dark frizzy hair, full lips, broad noses, small eyes, a nice smell, and perfectly shaped breasts and nipples. Women of other New Guinea tribes are less attractive, and white women are unspeakably hideous. Just compare your white women with our women to see why—white skin like a sick albino's, straight hair like strings, sometimes even hair coloured yellow like dead grass or red like a poisonous snail, thin lips and narrow noses like axe blades, big eyes like a cow's, a repulsive smell when they sweat, and breasts and nipples of the wrong shape. When you get ready to buy a wife, find a Fore if you want someone beautiful.

Among the reasons I did not follow that advice was that I happen to find those 'unspeakably hideous' women attractive. But then I was conditioned by my own society's ideals, just as my Fore friends were by theirs. Darwin commented that every people he knew about—Chinese, Hottentots, black Africans, Fijians, and others—measure beauty by their own appearance. Are there really no universal rules of human beauty and sex appeal? If not, do we inherit our particular taste in marriage partner through our genes, or do we learn it by looking at other members of our society? How, really, do we pick our sex partners and spouse?

It may be surprising to realize that this problem is one that arose anew during the evolution of the human species—or at least became much more important for us than for the other two chimpanzees. As we saw in Chapter Three, our familiar human mating-system, based ideally on couples maintaining on-going involvement, is a human innovation. Pygmy chimps are the opposite of sexually selective; females mate in sequence with many males, and there is much sexual activity between females and between males as well. Common chimps are not so completely promiscuous—a male and female may sometimes go off and 'consort' with each other for a few days—but they still rank as promiscuous by human standards. However, humans are much more selective sexually, since rearing a human child is difficult (at least for hunter-gatherers) without a father's help, and since sex becomes part of the cement that differentiates co-parents from other men and women frequently encountered. Choosing a mate or sex partner is not so much a human invention as a reinvention of something practised by many other (nominally) monogamous animals with lasting pair-bonds, and lost by our chimpanzee-like ancestors. Those choosy animals include many bird species, plus our distant ape relatives, the gibbons. We saw in Chapter Four that this ideal depiction of a human society based on monogamous couples coexists with a good deal of extramarital sex. That activity also involves selection of sex partners, with adulterous women tending to be more selective than adulterous men. Thus, selection of spouses and sex partners is another important piece of what defines humanity. It is as basic to our rise from chimpanzee status as is the remodelled pelvis discussed in detail in physical anthropology texts. We shall see in the next chapter that our sexual choosiness may be central to the origin of the most conspicuous visible variability in modern humans. That is, much of what we think of as human racial variation may have arisen as a by-product of the beauty standards by which we choose our sex partners.

In addition to this theoretical interest, the question of how we select our mates and sex partners is of much personal interest. It preoccupies most of us for much of our lives. Those of us who are still unattached spend daily hours dreaming about whom we will consort with or marry. The question becomes more intriguing when we compare what turns on different people within the same culture. Think of the men or women that you find sexually attractive. If you are a man, for instance, do you prefer women who are blonde or brunette, flat-chested or buxom, and with big or small eyes? If you are a woman, do you like men who are bearded or smooth-shaven, tall or short, and smiling or scowling? Probably you do not go for just anyone, only certain types attract you. Everyone can name friends who got divorced, then chose a second spouse who was the spitting image of the first. A colleague of mine went through a long series of plain, slim, brown-haired, round-faced girlfriends, until he finally found one he got along with and married her. Whatever your own preference, you will have noticed that some of your friends have completely different tastes.

The particular ideal that each of us pursues is an example of what are called 'search images'. (A search image is a mental picture against which we compare objects and people around us in order to be able to recognize something quickly, like a Perrier bottle amidst all the other bottled waters on the supermarket shelf, or one's child at a playground with other kids.) How do we develop our private search image for a mate? Do we seek someone familiar and similar to us, or are we more turned on by someone exotic? Would most European men really marry a Polynesian woman if given the chance? Do we seek someone complementary to us so as to fulfil our needs? For instance, there undoubtedly are some dependent men who marry a mothering woman, but how typical are such pairings? Psychologists have tackled this question by examining many married couples, measuring everything conceivable about their physical appearance and other characteristics, and then trying to make sense out of who married whom. A simple numerical way of describing the result is by means of a statistical index called the correlation coefficient. If you line up 100 husbands in order of their ranking for some characteristic (say, their height), and if you also line up their 100 wives with respect to the same characteristic, the correlation coefficient describes whether a man tends to be at the same position in the husbands' line-up as his wife is in the line-up of wives. A correlation coefficient of plus one would mean perfect correspondence: the tallest man marries the tallest woman, the thirty-seventh tallest man marries the thirty-seventh tallest woman, and so on. A correlation coefficient of minus one would mean perfect matching by opposites: the tallest man marries the shortest woman, the thirty-seventh tallest man marries the thirty-seventh shortest woman, and so on. Finally, a correlation coefficient of zero would mean that husbands and wives assort completely randomly by height: a tall man is as likely to marry a short woman as a tall woman. These examples are for height, but correlation coefficients can also be calculated for anything else, such as income and IQ.

If you measure enough things about enough couples, here is what you will find. Not surprisingly, the highest correlation coefficients—typically around +0.9—are for religion, ethnic background, race, socioeconomic status, age, and political views. That is, most husbands and wives prove to be of the same religion, ethnic background, and so on. Perhaps you also will not be surprised that the next highest correlation coefficients, usually around +0.4, are for measures of personality and intelligence, such as extroversion, neatness, and IQ. Slobs tend to marry slobs, though the chances of a slob marrying a compulsively neat person are not as low as the chances of a political reactionary marrying a left-winger.

What about matching of husbands and wives for physical characteristics? The answer is not one that would leap out at you immediately if you just looked at a few married couples. That is because we do not select our own mates for their bodies as carefully as we select the mates of our show dogs, racehorses, and beef cattle. But we select nevertheless. If you measure enough couples, the answer that finally emerges is unexpectedly simple. On the average, spouses resemble each other slightly but significantly in almost every physical feature examined. That is true of all the obvious traits you would first think of when asked to design your ideal beloved—his or her height, weight, hair colour, eye colour, and skin colour—but it is also true of an astonishing variety of other traits that you probably would not have mentioned in your description of the perfect sex partner. Those other traits include ones as diverse as breadth of nose, length of ear lobe or middle finger, circumference of wrist, distance between eyes, and lung volume! Experimenters have made this finding for people as diverse as Poles in Poland, Americans in Michigan, and Africans in Chad. If you do not believe it, try noting eye colours (or measuring ear lobes) the next time you are at a dinner party with many couples, and then get your pocket calculator to give you the correlation coefficient.

Coefficients for physical traits are on the average +0.2- not so high as for personality traits (+0.4) or religion (+0.9), but still significantly higher than zero. For a few physical traits the correlation is even higher than 0.2-for instance, an astonishing 0.61 for length of middle finger. At least unconsciously people care more about their spouse's middle finger length than about his or her hair colour and intelligence!

Thus, like tends to marry like. Among the obvious explanations that contribute to these results is propinquity: we tend to live in neighbourhoods defined by socioeconomic status, religion, and ethnic background. For instance, in large American cities one can point to the rich neighbourhoods and the poor neighbourhoods, and also to the Jewish section, Chinese section, Italian section, black section, and so on. We meet people of the same religion when we go to church, and we tend to meet people of similar socioeconomic status or political views in many of our daily activities. Since we thus have far more opportunities to meet people like us than unlike us in these respects, of course we are more likely to marry someone of our religion, socioeconomic status, and so on. But we don't live in neighbourhoods grouped by length of ear lobe, so there must be some other reason why spouses tend to be matched in that respect as well.

Another obvious reason why like tends to marry like is that marriage is not just a choice; it is a negotiation. We do not go out searching until we find a person with the right eye colour and length of middle finger, then announce to that person, 'You are marrying me'. For most of us, marriage results from a proposal rather than a unilateral announcement, and the proposal is the culmination of some sort of negotiation. The more similar a man and woman are in political views, religion, and personality, the smoother will be the negotiation. Hence the match in personality traits is on the average closer for married couples than for dating couples, closer for happily than unhappily married couples, and closer for couples who stay married than for those who get divorced. But this still does not explain spousal resemblance in ear lobe length, which is only rarely cited as a factor in divorce. The remaining factor deciding whom you will marry, besides propinquity and smoothness of negotiation, is surely sexual attraction based on physical appearance. That in itself is no surprise. Most of us are aware of our preferences in obvious visible features like height, build, and hair colour. What is initially surprising is the importance of so many other physical traits that we usually do not consciously notice, such as ear lobes, middle fingers, and interocular distances. Nevertheless, all those other traits contribute unconsciously to the snap decisions we make when we are introduced to someone and a voice inside tells us, 'She's my type! Here is an example. When my wife and I were introduced to each other, I instantly found Marie attractive and vice versa. In retrospect, I can understand why: we are both brown-eyed, similar in height and build and hair colour, and so on. But, on the other hand, I also had a sense that there was something about Marie that did not quite match my ideal, even though I could not figure out what exactly it was. Not until Marie and I first went to a ballet together did I solve the puzzle. I lent Marie my opera glasses, and when she passed them back to me, I found that she had pushed the eye-pieces so close together that I could not see through them until I had spread them apart again. I then realized that Marie has more close-set eyes than I do, and that most women whom I had pursued before had wide-set eyes like my own. Thanks to Marie's ear lobes and other merits, I have been able to make peace with my and her mismatched interocular distances. Nevertheless, the episode with the opera glasses made me appreciate for the first time that I have always found wide-set eyes a turn-on, even though

I had not been explicitly aware of it.

So, we tend to marry someone who looks like us. But—wait a minute. The men who look most similar to a woman are the men who share half of her genes—her father or brother!

Similarly, the best-matched mate for a man would be his mother or sister! Yet most of us obey the incest taboo and certainly do not marry our parent or sibling of the opposite sex.

Instead, I am saying that people tend to marry a person who looks like the parent or sibling of the opposite sex. Our actual behaviour is summed up by a popular song of the 1920s.

The reason we tend to resemble our mates is that many of us are looking for someone who reminds us of our parent or sibling of the opposite sex, who in turn resembles us. As children, we already begin to develop our search image of a future sex partner, and that image is heavily influenced by the people of the opposite sex whom we see most often.

For most of us that is our mother (or father) and sister (or brother), plus close childhood friends.

At this point, you are probably turning to your spouse or Significant Other, pulling out your tape measure, and discovering a gross mismatch between your and his (or her) ear lobes. Or perhaps you have pulled out a photo of your mother or sister, and you detect not the faintest resemblance when you hold it beside your spouse. You may be about to throw away this book as patent nonsense. But if your wife is not a dead ringer for your mother, don't stop reading, and conversely don't get worried that you should see a psychiatrist about your pathological search image. After all, remember:

1. Studies consistently show that factors like religion and personally influence our choice of spouse much more strongly than physical appearance. All I am making is the obvious point that physical traits have some influence. In fact, I would predict much higher correlation coefficients for physical traits between casual sex partners than between spouses. That is because we can select casual sex partners solely on the basis of physical attraction, without regard to religion or political views. This prediction awaits testing.

2. Remember also that your search image could have been influenced by any of the people of the opposite sex that you regularly saw around you as you were growing up. That includes playmates and siblings as well as parents. Perhaps your spouse resembles the little girl next door, rather than Mother.

3. Finally, remember that lots of independent physical traits enter into our search image, so most of us end up with a mild average resemblance to our spouses in many traits, rather than with a very close resemblance in a few traits. This idea is known as the 'buxom redhead theory'. If a man's mother and sister were both buxom redheads, he might grow up to consider buxom redheads very exciting, but redheads are relatively rare, and buxom redheads still rarer. Furthermore, the man's preference even in a casual sex partner is likely to depend on some other physical traits as well, and his preference in a wife will certainly depend on her views about children, politics, and money. Hence, in a group of sons of buxom redheads, a few lucky ones will find a girl like Mother in those two respects, some will have to settle for buxom non-redheads, others for non-buxom redheads, and most for run-of-the-mill non-buxom brunettes.

You may also be objecting at this point that my argument applies only to societies where spouses pick each other. As friends from India and China are quick to remind me, that is a peculiar custom of the twentieth-century US and Europe. It was not true of the US and Europe in the past, and it is still not true of most of the world today, where marriages are instead arranged by the families involved. The bride and groom often are not even introduced until the wedding day. How could my argument possibly apply to such marriages?

Of course it couldn't, if one is talking just about legal marriages. But my argument would still apply to the choice of extramarital sex partners, who may father a non-trivial fraction of children, just as blood-group studies proved for American and British children (Chapter Four). In fact, I would expect that if extramarital fathering is frequent even in societies where a woman already exercises her sexual preferences in choosing a husband, it may be even more frequent in societies with arranged marriages, where a woman's choice can only be expressed extramaritally. It is not just the case, then, that Fore men prefer Fore women over Californian women, and vice versa: our search images are much more specific. However, these insights still leave questions unanswered. Did I inherit or learn my search image for someone like Mother? If I were offered the choice of sex with my sister or a strange woman, I would certainly reject the offer of my sister and probably my first cousin, but would I prefer my second cousin over a strange woman (because the cousin probably resembles me more)? There are some crucial experiments that would settle these questions—for instance, keeping a man in a large cage with his female first, second, third, fourth, and fifth cousins, counting how many times he had sex with each, and repeating the experiment with many men (or women) and their cousins. Alas, such experiments are hard to do with humans, but they have been done for several animal species, with instructive results. I shall give just three examples, the cousin-loving quail, and the perfumed mice and rats. (We cannot use our closest relatives the chimpanzees for these examples, since they are so unselective.)

Consider first the case of Japanese quail, which are either brown or white. Quail normally grow up with their biological parents and siblings. However, it is also possible to 'cross-foster' quail by switching eggs between quail mothers and their nests before the eggs hatch. In that way, a baby quail may be reared by foster-parents and grow up with 'pseudo-siblings'—that is, littermates among whom the baby hatched but to whom the baby is not genetically related. The preferences of male quail have been tested by putting a male in a cage with two females and observing with which female the male spent more time or copulated. It turns out that males preferred whichever colour of female they grew up with. Furthermore, when a brown-loving male was given a choice between brown females that he had never seen before (although some were his relatives from whom he had been separated before hatching), he preferred his first cousin to his third cousin or an unrelated female, but he also preferred his first cousin to his sister. Evidently, male quail as they grow up learn the appearance of their sisters (or mother) with whom they are reared, then seek a mate that is very similar but not too similar. In fancy technical language, biologists term this the Principle of Optimal Intermediate Similarity. Like other things in life, inbreeding seems to be good in moderation—a little inbreeding, but not too much. For instance, among unrelated brown females a male prefers an unfamiliar one over a familiar one with whom he grew up (a pseudo-sister', who pushes the male's not-too-much-incest button). Mice and rats similarly learn in childhood what to look for in a mate, but they choose by smell more than by appearance. When infant female mice were reared by parents sprayed repeatedly with Parma Violet perfume, the females on reaching adulthood sought out Parma-Violet-scented males in preference to unscented males. ('I want a boy, just like the boy, that smells like dear old Dad'.) In another experiment, infant male rats were reared by mother rats whose nipples and vagina were sprayed with lemon odour, then the male on reaching adulthood was put in a cage with a lemon-smelling or unscented female rat. Each such encounter was videotaped and played back to note the times of key events. It turned out that males with scented mothers mounted and ejaculated more quickly when placed with a scented female than with an unscented one, while the reverse was true for males with unscented mothers. For example, sons of scented mother rats were so excited by a scented sex partner that they ejaculated in only eleven-and-a-half minutes, while they took over seventeen minutes to ejaculate with an unscented female. But sons of unscented mother rats took over seventeen minutes with the scented partner and only twelve minutes with the unscented partner. Obviously, the males had learned to be sexually excited by their mother's smell (or lack of smell); they did not inherit the knowledge. What do these experiments on quail, mice, and rats show? The message is clear. Animals of those species learn to recognize their parents and siblings as they grow up, then are programmed to seek out an individual fairly similar to the parent or sibling of the opposite sex—but not Mother or Sister herself. They may inherit some search image of what constitutes a rat, but they evidently learn their search image of who in particular is a beautiful, eligible rat. We can immediately appreciate what experiments are needed to get unequivocal proof of this theory for humans. We should take an average happy family, spray Father every day with Parma Violet, spray Mother's nipples daily with lemon oil while she is nursing, and then wait twenty years to seejvhom the sons and daughters marry. Alas, we would be frustrated by the many obstacles to establishing Scientific Truth for humans. But some observations and accidental experiments still let us tip-toe towards the truth.

Take the incest taboo. Scientists debate whether the taboo itself in humans is instinctive or learned. However, this chapter is concerned with a separate question: given that we somehow acquire an incest taboo, do we learn to whom to apply it, or do we inherit that information in our genes? Normally we grow up with our closest relatives (parents and siblings), so our subsequent avoidance of them as sex partners could equally well be genetic or learned, but adoptive brothers and sisters also tend to avoid incest, suggesting learned avoidance.

This conclusion is strengthened by an interesting set of observations made in Israeli kibbutzim—the collective settlements whose members house, school, and care for all their children together as a large group. Thus, kibbutz children live from birth until young adulthood in intimate association with each other, like a gigantic family of brothers and sisters. If propinquity were the main factor influencing whom we marry, most kibbutz children should marry within the kibbutz. In fact, a study of 2,769 marriages contracted by kibbutz-reared children turned up only thirteen between children from the same kibbutz. All the other children married outside the kibbutz on reaching maturity.

Even those thirteen cases turned out to be the exceptions that proved the rule: all involved couples of whom one had moved into that kibbutz only after the age of six! Among children reared in the same peer group since birth, there were not only no marriages, but also no adolescent or adult heterosexual activity at all. This is astonishing restraint on the part of nearly 3,000 young men and women who enjoyed daily opportunities for sexual involvement with each other, and who had far fewer opportunities for involvement with outsiders. It illustrates dramatically that the period between birth and the age of six is a critical time for formation of our sexual preferences. We learn, however unconsciously, that our intimate associates from that period are ineligible as sex partners when we become mature.

We also appear to learn the part of our search image that tells us whom to seek, not just the part that tells us whom to avoid. For instance, a friend of mine who is 100 % Chinese herself happened to grow up in a community in which every other family was white. Eventually she moved as an adult to an area with many Chinese men, and for some time she dated both Chinese and white men, but came to realize that it was the whites who attracted her. She has been married twice, both times to white men. Her own experiences led her to ask her Chinese women friends about their backgrounds. It turned out that most of her friends reared in white enclaves also ended up marrying white men, while those reared in Chinese neighbourhoods married Chinese men—although all had plenty of men of both types from whom to choose during their young adult years.

Hence those who surround us as we grow up, though ineligible themselves as eventual mates, nevertheless shape our standards of beauty and search image.

I Think to yourself: what sort of men or women do you find physically attractive, and where did you develop that taste? I would guess that most people, like myself, can trace their preference to the appearance of parents, siblings or childhood friends. So do not be discouraged by all those old generalizations about sex appeal—'Gentlemen prefer blondes, 'Men seldom make passes at girls who wear glasses, etc. Each such 'rule' applies only to some of us, and there are plenty of men out there whose mothers were myopic brunettes. Fortunately for my wife and me—both of us brulettes raring glasses, born of brunette glass-wearing parents—beauty is in the eye of the beholder.

SIX SEXUAL SELECTION, AND THE ORIGIN OF HUMAN RACES

People from different parts of the world can be distinguished at a glance by so-called racial characteristics. But those same traits - ones such as the colour of our skin and hair and eyes, or the shapes of our breasts and genitals—play a big role in how we select our mates and sex partners. Thus, our outward appearances and our beauty standards have evolved in tandem to different local end points.

'White man! Lookim this-feller line three-feller man. This-feller number-one he belong Buka Island, na 'nother-feller number-two he belong Makira Island, na this-feller number-three he belong Sikaiana Island. Yu no savvy? Yu no enough lookim straight? I think, eye-belong-yu he bugger-up finish? . No, damn it, my eyes-belong-me were not ruined beyond repair. It was my first visit to the Solomon Islands in the Southwest Pacific, and I told my scornful guide through the medium of pidgin English that I saw perfectly well the differences between those three men in a row over there. The first one had jet-black skin and frizzy hair, the second had niuch lighter skin and frizzy hair, and the third had straighter hair and more slanty eyes. The only thing the matter with me was that I had no experience of what people from each particular Solomon island looked uke. By the end of my first trip through the Solomons, I too could match People to their islands by their skin and hair and eyes. In those variable features, the Solomons are a microcosm of humanity.

Simply by looking at a person, even laymen can often tell what part of the world that person comes from, and trained anthropologists may be able to 'place' him or her in the right part of the right country. For example, given one person each from Sweden, Nigeria, and Japan, none of us would have any trouble deciding at a glance which person was from which country. The most visibly variable features in clothed people are of course skin colour, the colour and form of the eyes and hair, body shape, and (in men) the amount of facial hair. If the people to be identified were undressed, we might also notice differences in amount of body hair, the size and shape and colour of a woman's breasts and nipples, the form of her labia and buttocks, and the size and angle of a man's penis. All those variable features contribute to what we know as human racial variation. Those geographic differences among humans have long fascinated travellers, anthropologists, bigots, and politicians, as well as the rest of us. Since scientists have solved so many arcane questions about obscure unimportant species, surely you might expect them to have answered one of the most obvious questions about ourselves: 'Why do people from different areas look different? Our understanding of how humans came to differ from other animals would remain incomplete if we did not also consider how, in the process, human populations acquired their most visible differences from each other. Nevertheless, the subject of human races is so explosive that Darwin excised all discussion of it from his famous 1859 book On the Origin of Species. Even today, few scientists dare to study racial origins, lest they be branded racists simply for being interested in the problem. There is another reason why we do not understand the significance of human racial variation: it proves to be an unexpectedly difficult problem. Twelve years after Darwin wrote his book attributing the origin of species to natural selection, he wrote another book 898 pages long, attributing the origin of human races to our sexual preferences which I described in the last chapter, and entirely rejecting a role of natural selection. Despite that verbal overkill, many readers were unconvinced. To this day, Darwin's theory of sexual selection (as he called it) remains controversial. Instead, modern biologists generally invoke natural selection to explain the visible differences among human races—especially the differences in skin colour, whose relation to sun exposure seems obvious. However, biologists cannot even agree on why natural selection led to dark skin in the tropics. I shall explain why I believe natural selection to have played only a secondary role in our racial origins, and why Darwin's preference for sexual selection seems to me correct. I therefore consider visible human racial variation to be largely a byproduct of the remodelled human life-cycle that forms the subject of Part Two of this book.

Firstly, to place matters in perspective, let's realize that racial variation is not at all confined to humans. Most animal and plant species with sufficiently wide distributions, including all higher ape species except the geographically localized pgymy chimp, also vary geographically. So marked is variation in some bird species, such as North America's white-crowned sparrow and Eurasia's yellow wagtail, that experienced birdwatchers can identify an individual bird's approximate birthplace by its plumage pattern. Variation in apes encompasses many of the same characteristics that vary geographically in humans. For example, among the three recognized races of gorillas, western lowland gorillas have the smallest bodies and rather grey or brown hair, while mountain gorillas have the longest hair, and eastern lowland gorillas share black hair with mountain gorillas. Races of white-handed gibbons similarly vary in hair colour (variously black, brown, reddish, or grey), hair length, tooth size, protrusion of the jaws, and protrusion of the bony ridges over the eyes. All these traits that I have just mentioned as varying among gorilla or gibbon populations also differ among human populations. How does one decide whether recognizably distinct animal populations from different localities constitute different species, or belong instead to the same species and just constitute different races (also known as subspecies)? As explained in Chapter Two, the distinction is based on interbreeding under normal circumstances: members of the same species may interbreed normally if given the opportunity, while members of different species do not. (But closely related species that would not normally interbreed in the wild, like lions and tigers, may do so if a male of one is caged with a female of the other and given no other choice.) By this criterion, all living human populations belong to the same species, since some interbreeding has occurred whenever humans from different regions have come into contact—even people as dissimilar in appearance as African Bantus and Pygmies. With humans as with other species, populations may intergrade into each other, and it becomes arbitrary to decide which populations to group as races. By the same criterion of interbreeding, the large gibbons known as siamangs are a distinct species from the smaller gibbons, since both occur together in the wild without hybridizing. This is also the criterion for considering Neanderthals possibly as a species distinct from Homo sapiens, since hybrid skeletons have not been identified despite apparent Cro-Magnon/Neanderthal contact (see Chapter Two).

Racial variation has characterized humans for at least the past several thousand years, and possibly much longer. Already around 450 BC, the Greek historian Herodotus described the Pygmies of West Africa, the black-skinned Ethiopians, and a blue-eyed red-haired tribe in Russia. Ancient paintings, mummies from Egypt and Peru, and bodies of people preserved in European peat bogs confirm that people several thousand years ago differed in their hair and facial features much as they do today. Origins of modern races can be pushed back still further, to at least ten thousand years ago, since fossil skulls of that age from various parts of the world differ in many of the same respects that modern skulls from the same regions differ. More controversial are the studies of some anthropologists, contested by others, reporting continuity of racial skull characteristics for hundreds of thousands of years. If those studies are correct, then some of the human racial variation that we see today may predate the Great Leap Forward, and may have gone back to the times of Homo erectus. Now let's turn to the question of whether natural selection or sexual selection has made the larger contribution to those visible geographic differences of ours. Take first the arguments about natural selection, the selection of traits that enhance survival. No scientist denies today that natural selection does account for many of the differences between species, such as why lions have paws with claws while we have grasping fingers. No one denies either that natural selection explains some geographic variation ('racial variation') within some animal species. For instance, Arctic weasels that live in areas covered by winter snow change colour from brown in summer to white in winter, while more southerly weasels stay brown all year. That racial difference enhances survival, because white weasels against a brown background would be glaringly conspicuous to their prey if they were not camouflaged against snow.

By the sajne token, natural selection surely explains some geographic variation in humans. Many black Africans but no Swedes have the sickle-cell haemoglobin gene, because the gene protects against malaria, a tropical disease that would otherwise kill many Africans. Other localized human traits that surely evolved through natural selection include the big chests of Andean Indians (good for extracting oxygen from thin air at high altitudes), the compact shapes of Eskimos (good for conserving heat), the slender shapes of southern Sudanese (good for losing heat), and the slit-like eyes of northern Asians (good for protecting eyes against cold and against sun glare off the snow). All these examples are easy to understand. Can natural selection similarly explain the racial differences that we think of first, those in skin colour and eye colour and hair? If so, one might expect that the same trait (for instance, blue eyes) would reappear in different parts of the world with similar climates, and that scientists would agree on what the trait is good for.

Seemingly the simplest trait to understand is skin colour. Our skins run the spectrum from various shades of black, brown, copper, and yellowish to pink with or without freckles. The usual story to explain this variation by natural selection goes as follows. People from sunny Africa have blackish skins. So too (supposedly) do people from other sunny places, like southern India and New Guinea. Skins are said to get paler as one moves north or south from the equator, until one reaches northern Europe, with the palest skins of all. Obviously, dark skins evolved in those people who were exposed to much sunlight. That is just like the skins of whites tanning under the summer sun (or in tanning salons!), except that tanning is a reversible response to sun rather than a permanent genetic one. It is equally obvious what good a dark skin does in sunny areas: it protects against sunburn and skin cancer. Whites who spend lots of time outdoors in the sun tend to get skin cancer, and they get it on exposed parts of their body like their head and hands. Does that not all make sense?

Yes, but… it is really not so simple at all. To begin with, skin cancer and sunburn cause little debilitation and few deaths. As agents of natural selection, they have an utterly trivial impact compared to infectious diseases of childhood. Hence many other theories have been proposed to explain the supposed pole-to-equator gradient in skin colour.

One favourite competing theory notes that the sun's ultraviolet rays promote vitamin D formation in a layer of our skin beneath the main pigmented layer. Thus, people in sunny tropical areas might have evolved dark skin to protect them against the risk of kidney disease caused by too much vitamin D, while people in Scandinavia with its long dark winters evolved pale skins to protect them against the risk of rickets caused by too little vitamin D. Two other popular theories are that dark skins are to protect our internal organs against overheating by the tropical sun's infrared rays, or—just the opposite—dark skins help keep tropical people warm when the temperature drops. And if those four theories are not enough for you, consider four more: that dark skins provide camouflage in the jungle, or that pale skins are less sensitive to frostbite, °r that dark skins protect against beryllium poisoning in the tropics, or that pale skins cause deficiency of another vitamin (folic acid) in the tropics.

With at least eight theories in the running, we can hardly claim to understand why people from sunny climates have dark skins. That in rtself does not refute the idea that, somehow, natural selection caused the evolution of dark skins in sunny climates. After all, dark skins could have multiple advantages, which scientists may sort out some day. Instead, the heaviest objection to any theory based on natural selection is that the association between dark skins and sunny climates is a very imperfect one. Native peoples had very dark skins in some areas receiving relatively little sunlight, like Tasmania, while skin colour is only medium in sunny areas of tropical Southeast Asia. No American Indians have black skins, not even in the sunniest parts of the New World. When one takes cloud cover into account, the world's most dimly lit areas, receiving a daily average of under three-and-a-half hours of sunlight, include parts of equatorial West Africa, southern China, and Scandinavia, inhabited respectively by some of the world's blackest, yellowest, and palest peoples! Among the Solomon Islands, all of which share a similar climate, jet-black people and lighter people replace each other over short distances. Evidently, sunlight has not been the sole selective factor that moulded skin colour.

The first response of anthropologists to these objections is to raise a counter-objection, the time factor. This argument tries to explain away the cases of pale-skinned people in the tropics by claiming that those particular peoples migrated to the tropics too recently to have evolved black skins. For example, the ancestors of American Indians may have reached the New World only 11,000 years ago (Chapter Eighteen): perhaps that has not been long enough to evolve black skins in the tropical Americas. But if you are going to evoke the time factor to explain away objections to the climate theory of skin colour, then you also have to consider the time factor for peoples who supposedly support that theory. One of the prime supports of the climate theory is the pale skin of Scandinavians, living in the cold, dark, foggy North. Unfortunately, Scandinavians have been in Scandinavia for an even shorter time than American Indians have been in the Amazon. Until about 9,000 years ago, Scandinavia was covered by an ice-sheet and could hardly have supported any people, pale-skinned or dark-skinned. Modern Scandinavians reached Scandinavia only around 4,000 or 5,000 years ago, as a result of the expansion of farmers from the Near East (Chapter Ten) and of Indo-European speakers from southern Russia (Chapter Fifteen). Either Scandinavians acquired their pale skins long ago in some other area with a different climate, or else they acquired them in Scandinavia within half the time that Indians have spent in the Amazon without becoming dark-skinned.

The sole people in the world about whom we can be certain that they spent the last 10,000 years in the same location were the natives of Tasmania. Lying south of Australia, at the temperate latitude of Chicago or Vladivostok, Tasmania used to be connected to Australia until it was cut off by rising sea levels 10,000 years ago and became an island. Since modern Tasmanian natives did not have boats capable of going more than a few miles, we know that they were derived from colonists who walked out to Tasmania at the time of its connection to Australia, and who remained there continuously until they were exterminated by British colonists in the Nineteenth Century (Chapter Sixteen). If any people had enough time for natural selection to match their skin colour to their local temperate-zone climate, it was the Tasmanians. Yet they had blackish skins, supposedly adapted to the Equator.

If the case for natural selection of skin colour seems weak, that for hair colour and eye colour is virtually non-existent. There are no consistent correlations with climate, and not even any half-plausible theories for the supposed advantage lent by each colour type. Blonde hair is common in cold, wet, dimly lit Scandinavia and also among Aborigines of the hot, dry, sunny desert of central Australia. What do those two areas have in common, and how does being blonde help both Swedes and Aborigines to survive? Do freckles and red hair help Irishmen catch leprechauns? Blue eyes are common in Scandinavia and supposedly help their owners see farther in dim, misty light, but that speculation is unproven, and all my friends in the even dimmer, mistier mountains of New Guinea see just fine with their dark eyes.

The racial traits for which it seems most absurd to seek an explanation based on natural selection are our variable genitalia and secondary sex characteristics. Are hemispherical breasts an adaptation to summer rainfall and conical breasts an adaptation to winter fog, or vice versa? Do the protruding labia minora of Bushmen women protect them against pursuing lions, or reduce their water losses in the Kalahari Desert? You surely don't think that men with hairy chests can thereby keep warm while going shirtless in the Arctic, do you? If you do think so, then please explain why women do not share hairy chests with men, since women also have to keep warm. Facts such as these were what made Darwin despair of imputing human racial variation to his own concept of natural selection. He finally Qismissed the attempt with a succinct statement: 'Not one of the external differences between the races of man are of any direct or special service to him. When Darwin came up with a theory that he preferred, he termed it sexual selection' to contrast with natural selection, and he devoted an entire book to explaining it. The basic notion behind this theory is easily grasped. Darwin noted many animal features that had no obvious survival value but that did play an obvious role in securing mates, either by attracting an individual of the opposite sex or by intimidating a rival of the same sex. Familiar examples are the tails of male peacocks, the manes of male lions, and the bright red buttocks of female baboons in oestrus. If an individual male is especially successful at attracting females or intimidating rival males, that male will leave more descendants and will tend to pass on his genes and traits—as a result of sexual selection, not natural selection. The same argument applies to female traits as well.

For sexual selection to work, evolution must produce two changes simultaneously: one sex must evolve some trait, and the other sex must evolve in tandem a liking for that trait. Female baboons could hardly afford to flash red buttocks if the sight revolted male baboons to the point of their becoming impotent. As long as the female has it and the male likes it, sexual selection could lead to any arbitrary trait, just as long as it does not impair survival too much. In fact, many traits produced by sexual selection do seem quite arbitrary. A visitor from outer space who had yet to see humans could have no way of predicting that men rather than women would have beards, that the beards would be on the face rather than above the navel, and that women would not have red and blue buttocks.

That sexual selection really can work, at least in birds, was proved by an elegant experiment carried out by the Swedish biologist Make Andersson on the long-tailed widowbird of Africa. In this species the male's tail in the breeding season grows to 20 inches long, while the female's tail is only 3 inches. Some males are polygamous and acquire up to six mates, at the expense of other males who get none. Biologists had guessed that a long tail served as an arbitrary signal by which males attracted females to join their harem. Andersson's test was to cut off part of the tail from nine males until their tails were only 6 inches long. He then glued those cut segments to the tails of nine other males to give them 30-inch tails, and he waited to see where the females built their nests. It turned out that the males with the artificially lengthened tails attracted on the average over four times as many mates as the males with artificially shortened tails. Perhaps our first reaction to Andersson's experiment is: those dumb birds! Imagine a female selecting a particular male to father her offspring merely because his tail is longer than other males' tails! But before we get too smug, let's consider again what we learned in the last chapter about how we humans select our own mates. Are our criteria such good indicators of genetic worth? Do not some men and women set disproportionate value on the size or form of certain body parts, which are really nothing more than arbitrary signals for sexual selection? Why did we evolve to pay any attention at all to a beautiful face, which is useless to its owner in the struggle for survival?

In animals some of the traits that vary racially are ones produced by sexual selection. For instance, lions' manes vary in length and in colour. Males of the astrapia birds of paradise in New Guinea have fancy tails to display to females, but different populations evolved tails of different shapes and colours. From west to east, the tails are broad and purple, short and white-based, very long and white, long and purple, and broad and purple again. Similarly, snow geese occur in two colour phases, a blue phase commoner in the western Arctic and a white phase commoner in the eastern Arctic. Birds of each phase prefer a mate of the same phase. Could human breast shape and skin colour similarly be the outcome of sexual preferences that vary arbitrarily from area to area? After 898 pages Darwin convinced himself that the answer to this question was a resounding 'yes'. He noted that we pay inordinate attention to breasts, hair, eyes, and skin colour in selecting our mates and sex partners. He noted also that people in different parts of the world define beautiful breasts, hair, eyes, and skin by what is familiar to them. Thus, Fijians, Hottentots, and Swedes each grow up with their own learned, arbitrary beauty standards, which tend to maintain each population in conformity with those standards, since individuals deviating too far from the standards would find it harder to obtain a mate. Darwin died before his theory could be tested against rigorous studies of how people actually do select their mates. Such studies have proliferated in recent decades, and I summarized the results in Chapter Five. There I showed that people tend to marry individuals who resemble themselves in every conceivable character, including hair and eye and skin colour. To explain that seeming narcissism of ours, I reasoned that we develop our beauty standards by imprinting on the people we see around us in childhood—especially on our parents and siblings, the people of which we see the most. But our parents and siblings are also the people to whom we bear the strongest physical resemblance, since we share their genes. Thus, if you are a fair-skinned, blue-eyed blonde who grew up in a family of fair-skinned, blue-eyed blondes, that is the sort of person whom you will consider most beautiful and will seek as a mate. In the meantime, my dark-skinned, dark-haired New Guinea friends were growing up with other New Guineans and learning to regard fair-skinned, blue-eyed blondes as grotesquely revolting.

To test that imprinting theory of human mate choice rigorously, one would have to do experiments like shipping some Swedish babies to adoptive parents in New Guinea, or painting some Swedish parents permanently black. Then, after waiting twenty years for the babies to grow up, one could study whether they preferred Swedes or New Guineans as sex partners. Alas, once again, the Search for Truth about humans founders on practical problems, but such tests can be performed with full experimental rigour on animals. Take snow geese, for example, with their blue or white colour phases. Do white geese learn or inherit their preference in the wild for white geese over blue ones? Canadian biologists hatched gosling eggs in an incubator, then put the goslings into a nest of goose 'foster-parents'. When those goslings grew up, they chose a mate with the colour of the foster-parents. When goslings were reared in a large mixed flock of both blue and white birds, they showed no preference between blue and white prospective mates on reaching adulthood. Finally, when the biologists dyed some white parents pink, their offspring came to prefer pink-dyed geese. Thus, geese do not inherit but learn a colour preference, by imprinting on their parents (and on their siblings and playmates). How, then, do I think that people in different parts of the world evolved their differences? Our insides remained invisible to us and were moulded only by natural selection, with results such as that of tropical Africans but not Swedes evolving the anti-malarial defence of a sickle-cell haemoglobin gene. Many visible features of our outsides also got moulded by natural selection. But, just as in animals, sexual selection had a big effect in moulding the external traits by which we pick our mates.

For us humans those traits are especially the skin, eyes, hair, breasts, and genitals. In each part of the world those traits evolved in tandem with our imprinted aesthetic preferences to reach different, somewhat arbitrary results. Which particular human population ended up with any given eye or hair colour may have been partly an accident of what biologists term the 'founder effect'. That is to say, if a few individuals colonize an-empty land and their descendants then multiply to fill the land, the genes of those few founding individuals may still dominate the resulting population many generations later. Just as some birds of paradise ended up with yellow plumes and others with black plumes, so some human populations ended up with yellow hair and others with black hair, some with blue eyes and others with green eyes, some with orange nipples and others with brown nipples.

I do not mean thereby to claim that climate has nothing whatsoever to do with skin colour. I acknowledge that tropical peoples tend on the average to have darker skins than temperate-zone peoples, though there are many exceptions, and that this is probably due to natural selection, though we are unsure of the exact mechanism. Instead, I am saying that sexual selection has been strong enough to render the correlation between skin colour and sun exposure quite imperfect.

If you are still sceptical about how traits and aesthetic preferences can evolve together to different and arbitrary end points, just think about our changing fashion preferences. When I was a schoolboy in the early 1950s, women rated men with crew-cuts and clean-shaven faces a's handsome. Since then, we have seen a parade of men's fashions, including beards, long hair, earrings, purple-dyed hair, and the Mohawk hair style. A man daring to flaunt any of those fashions in the 1950s would have revolted the girls and enjoyed zero mating success. That is not because crew-cuts were better adapted to atmospheric conditions of Stalin's last years, while a purple Mohawk has higher survival value in our post-Chernobyl era. Instead, men's appearances and women's tastes changed in tandem, and the changes occurred far more rapidly than evolutionary changes in skin colour, since no gene mutations were required. Either women came to like crew-cuts because good men had them, or men adopted crew-cuts because good women liked them, or something of both happened. The same goes for women's appearances and men's tastes.

To a zoologist, the visible geographic variability that sexual selection produced in humans is impressive. I have argued in this chapter that much of our variability is a by-product of a distinctive feature of the human life-cycle, our choosiness with respect to our spouses and sex partners. I do not know of any other wild animal species in which eye colour of different populations can be green, blue, grey, brown, or black, while skin colour varies geographically from pale to black and hair is either red, yellow, brown, black, grey, or white. There may be no limits, except those imposed by evolutionary time, on the colours with which sexual selection can adorn us. If humanity survives another 20,000 years, I predict that there will be women with naturally green hair and red eyes—plus men who think such women are the sexiest.

SEVEN WHY DO WE GROW OLD AND DIE?

We constantly invest resources in the repair of our bodies, just as we do with our cars. Unfortunately for us and for all other animals, there is a limit to the resources that natural selection found it worthwhile to programme into our self-repair. As a result, we eventually grow old and die, but at least we age more slowly than our ape relatives. 'Mother, why did Grandpa have to die? Will you die some day? Will I die too? Why?

Death and aging constitute a mystery that we often ask about as children, deny in youth, and reluctantly come to accept as adults. I scarcely reflected on aging when I was a college student. Now that I am fifty-three years old, I find it decidedly more interesting. Life expectancy among US white adults is, presently about seventy-eight years for men, eighty-three for women. But few of us will survive to 100. Why is it so easy to live to eighty, so hard to live to 100, and almost impossible to live to 120? Why do humans with access to the best medical care, and animals kept in a cage with plenty of food and no predators, inevitably grow infirm and die? It is the most obvious fact of life, but there is nothing obvious about what causes it.

In the bare fact of our aging and dying, we resemble all other animals. In the detarh, however, we have improved considerably over the course of our evolutionary history. Not a single individual of any ape species has been recorded as achieving the current life expectancy of US whites, and only exceptional apes reach their fifties. Hence we age more slowly than do our closest relatives. Some of that slowdown may have developed recently, around the time of the Great Leap Forward, since quite a few Cro-Magnons lived into their sixties while few Neanderthals passed forty.

Slow aging is crucial to the human lifestyle because the latter depends on transmitted information. As language evolved, far more information became available to us to pass on than previously. Until the invention of writing, old people acted as the repositories of that transmitted information and experience, just as they continue to do in tribal societies today. Under hunter-gatherer conditions, the knowledge possessed by even one person over the age of seventy could spell the difference between survival and starvation or defeat for a whole clan. Thus, our long lifespan was important for our rise from animal to human status.

Obviously, our ability to survive to a ripe old age depended ultimately upon advances in culture and technology. It is easier to defend yourself against a lion if you are carrying a spear than just a hand-held stone, and easier yet with a high-powered rifle. However, advances in culture and technology alone would not have been enough, unless our bodies had also become redesigned to last longer. No caged ape in a zoo, enjoying all the benefits of modern human technology and veterinary care, reaches eighty. We shall see in this chapter that our biology became remoulded to the increased life expectancy that our cultural advances made possible. In particular, I would guess that Cro-Magnon tools were not the sole reason why Cro-Magnons lived on the average longer than Neanderthals. Instead, around the time of the Great Leap Forward our biology must have changed so that we aged more slowly. That may even have been the time when menopause, the concomitant of aging that paradoxically functions to let women live longer, evolved.

In short, cultural and biological change had to develop hand-in-hand to permit our long lives.

Along with the changes in our sexual anatomy, physiology, behaviour, and preferences discussed in Chapters Three to Six, retarded aging is the last of the life-cycle changes that made possible the third chimpanzee's rise.

The way in which scientists think about aging depends on whether they are interested in so-called proximate explanations or ultimate explanations. To appreciate this difference, consider the question, 'Why do skunks smell bad? A chemist or molecular biologist would answer,

'It's because skunks secrete chemical compounds with certain particular molecular structures. Due to the principles of quantum mechanics, those structures result in bad smells. Those particular chemicals would smell bad no matter what the biological function of their bad smell was.

But an evolutionary biologist would instead reason,

'It's because skunks would be easy victims for predators if they didn't defend themselves with bad smells. Natural selection made skunks evolve to secrete bad-smelling chemicals; those skunks with the worst smells survived to produce the most baby skunks. The molecular structure of those chemicals is a mere incidental detail; any other bad-smelling chemicals would suit skunks equally well.

The chemist has offered a proximate explanation: that is, the mechanism immediately responsible for the observation that was to be explained. The evolutionary biologist has instead offered an ultimate explanation: the function or chain of events that caused that mechanism to be present. The chemist and the evolutionary biologist would each dismiss the other's answer as not being 'the real explanation'.

Similarly, studies of aging are pursued independently by two groups of scientists who scarcely communicate with each other. One group seeks a proximate explanation, the other an ultimate explanation. Evolutionary biologists try to understand how natural selection could ever permit aging to occur, and they think that they have found an answer to this question. Physiologists inquire instead into the cellular mechanisms underlying aging, and admit that they do not yet have an answer. But I shall argue that aging cannot be understood unless we seek both explanations simultaneously. In particular, I expect that the evolutionary (ultimate) explanation will help us find the physiological (proximate) explanation of aging that has so far eluded scientists. Before I can pursue this reasoning, I must anticipate objections of my physiologist friends. They tend to believe that something about our physiology somehow makes aging inevitable, and that evolutionary considerations are irrelevant. For instance, one such theory attributes aging to the progressive difficulties that our immune system is said to face in distinguishing our own cells from foreign cells. Physiologists subscribing to this view make an implicit assumption that natural selection could not lead to an immune system without that fatal defect. Is this belief warranted? -

To evaluate this objection, let's consider biological repair mechanisms, because aging may be thought of simply as unrepaired damage or deterioration. Our first association with the word 'repair' is likely to be to those repairs that cause us the most frustration, car repairs. Our cars tend to grow old and die, but we spend money to postpone their inevitable fate. Similarly, we are unconsciously but constantly repairing ourselves too, at every level from that of molecules to that of tissues or whole organs. Our own self-repair mechanisms, like those we lavish on our cars, are of two sorts—damage control, and regular replacement.

An automotive example of damage control is that we replace a car's bumper only if it is bashed in; we do not routinely replace the bumper at every regular oil change. The most visible example of damage control applied to our bodies is wound healing, by which we repair damage to our skin. Many animals can achieve more spectacular results: lizards regenerate severed tails, starfish and crabs their limbs, sea cucumbers their intestines, and ribbon worms their poison stylets. At the invisible molecular level our genetic material, DNA, is repaired exclusively by damage control. We have enzymes that recognize and fix damaged sites in the DNA helix while ignoring intact DNA.

The other type of repair, regular replacement, is also familiar to every car-owner. We periodically change the oil, air filter, and ball-bearings to eliminate slight wear, without waiting for the car to break down first. In the biological world, teeth are similarly replaced on a pre-scheduled basis: humans go through two sets, elephants six sets, and sharks an indefinite number, during their lifetimes. Though we humans go through life with the same skeleton with which we were born, lobsters and other arthropods regularly replace their exoskeleton by moulting it and growing a new one. Still another highly visible example of scheduled repair is the continual growth of our hair: no matter how short we cut it, its growth will replace the cut portion.

Regular replacement also goes on at a microscopic or submicroscopic level. We constantly replace many of our cells about once every few days for the cells lining our intestine, once every two months for the cells lining the urinary bladder, and once every four months for our red blood cells. At the molecular level, our protein molecules are subject to continuous turnover at a rate characteristic of each particular protein; we thereby avoid the accumulation of damaged molecules. If you compare your beloved's appearance today with a photograph taken a month ago, he (or she) may look the same, but many of the individual molecules forming that beloved body are different. While all the king's horses and men couldn't put Humpty Dumpty together again, Nature is taking us apart and putting us back together every day. Thus, much of an animal's body can be repaired as needed, or is regularly replaced anyhow, but the details of how much is replaceable vary greatly with the part and with the species. There is nothing physiologically inevitable about the limited repair capabilities of us humans. Since starfish can regrow amputated limbs, why can't we? What prevents us from having six sequential sets of teeth like an elephant, rather than just baby teeth and adult teeth? With four more natural sets, we would not need fillings, crowns, and dentures as we got older. Why don't we protect ourselves against arthritis? — all we would need is to replace our joints periodically, as crabs do. Why don't we guard against heart disease by periodically replacing our hearts, as ribbon worms replace their poison stylets? One might suppose that natural selection would favour the man or woman who did not die of heart disease around the age of eighty but continued to live and produce babies at least until the age of 200. Why, for that matter, cannot we repair or replace everything in our bodies?

The answer surely has something to do with the expense of repair. Here again, the analogy of car repair is helpful. If the boasts of the Mercedes-Benz company are to be believed, their cars are so well built that, even should you do no maintenance whatsoever—not even lubrication or oil changes—your Mercedes will still run for years. At the end of that time, of course, it will fall apart from accumulated irreversible damage. So Mercedes-owners generally do choose to service their cars regularly. My Mercedes-owning friends tell me that Mercedes service is very expensive, hundreds of dollars every time they drive into the workshop. Nevertheless, they consider the expense worth it. A serviced Mercedes lasts much longer than an unserviced Mercedes, and it is much cheaper to service your old Mercedes regularly than to discard it and buy a new one every few years.

That is how Mercedes-owners reason in Germany and the US. But suppose you were living in Port Moresby, the capital of Papua New Guinea, automobile accident capital of the world, where any car is likely to be written off within a year no matter how you maintain it. Many car-owners in New Guinea do not go to the expense of maintaining their car; they use the saved money to help buy the inevitable next car.

By analogy, how much an animal 'should' invest in biological repair depends on the expense of the repairs, and on a comparison of the animal's expected lifespan with and without the repairs. But such 'should' questions belong to the realm of evolutionary biology, not physiology. Natural selection tends to maximize one's rate of producing offspring that survive to leave offspring of their own. Evolution can thus be regarded as a strategy game, in which the individual whose strategy leaves the most descendants wins. Hence the type of reasoning used in game theory is helpful in understanding how we came to be the way we are.

This problem of lifespan, and of investment in biological repair, is in turn one of an even broader class of evolutionary problems addressed by game theory: the mystery of what sets the maximum limit on any advantageous trait. There are lots of other biological traits, besides lifespan, that beg the question why natural selection has not made them longer or bigger or faster or made more of them. For instance, people who are big or smart or can run fast have obvious advantages over small, dumb, slow people—especially throughout most of human evolution, when we were still fending off lions and hyenas. Why did we not evolve to become on the average even bigger, smarter, and faster than we now are? The complication that makes these evolutionary design problems less simple than they might at first seem is this: natural selection acts on whole individuals, not on single parts of an individual. It is you, not your big brain or fast legs, that does or does not survive and leave offspring. Increasing one part of an animal's body may be beneficial in some obvious respect but harmful in other respects. For instance, that one larger part might not fit in well with other parts of the same animal, or it might drain off energy from other parts.

To evolutionary biologists, the magic word that expresses this complication is 'optimize'. Natural selection tends to mould each trait to the size, speed, or number that maximizes the survival and reproductive success of the whole animal, given the animal's basic design. Hence each trait in itself does not tend towards a maximal value. Instead, each trait converges on some optimal intermediate value, neither too big nor too small. The whole animal is thereby more successful than it would be if that trait were bigger or smaller.

Should this reasoning about animals seem abstract, think instead of our everyday machines. Essentially the same principles apply to engineering design, of machines by humans, as to evolutionary design, of animals by natural selection. For example, consider my pride and joy among my machines, my 1962 Volkswagen Beetle, the only car I have ever owned. (Car buffs will remember 1962 as the year that Volkswagen introduced the big rear window in the Beetle.) On a smooth, level road with an assisting tailwind, my VW can go at 65 mph. To BMW owners, that may sound distinctly submaximal. Why don't I junk my puny 4-cylinder, 40-horsepower engine, install instead the 12-cylinder, 296-horsepower engine from my neighbour's BMW 750IL, and roar off at 180 mph down the freeway?

Well, even I, dodo about cars that I am, know that that would not work. To begin with, that huge BMW engine would not fit into my VW's engine compartment, which would need enlarging. Then, the BMW engine is meant to go in front, but the VW engine compartment is m the back, so I would have to change the gearbox and transmission and other things. I would also have to change the shock absorbers and brakes, designed to smooth the ride and stop a car at 65 mph but not at 180 mph. By the time I had finished modifying my VW to take the BMW engine, there would not be much remaining from my original Beetle, and the Modifications would have cost me a big pile of money. I suspect that my puny 40-horsepower engine is optimal, in the sense that I could no increase my cruising speed without sacrificing other performance features of my car—as well as sacrificing other money-requiring features of my lifestyle.

While the marketplace eventually eliminates engineering monstrosities like a VW with a BMW engine, all of us can think of monstrosities that took quite a while to eliminate. To those of you who share my fascination with naval warfare, British battle-cruisers are a good example. Before and during the First World War, the British navy launched thirteen warships called battle-cruisers, designed to be as large and with as many big guns as battleships but much faster. By maximizing speed and firepower, the battle-cruisers immediately caught the public imagination and became a propaganda sensation. However, if you take a 28,000-ton battleship, keep the weight of the big guns nearly constant, and greatly increase the weight of the engines while still maintaining total weight around 28,000 tons, you have to skimp on the weight of some other parts. The battle-cruisers skimped especially on weight of armour, but also on weight of small guns, internal compartments, and anti-aircraft defence. The results of this suboptimal overall design were inevitable. In 1916 H.M.S. Indefatigable, Queen Mary, and Invincible all blew up almost as soon as they were hit by shells at the Battle ofjutland. H.M.S. Hoodblew up in 1941, a mere eight minutes after entering battle with the German battleship Bismarck. H.M. S. Repulse was sunk by Japanese bombers a few days after the Japanese attack on Pearl Harbor, thereby acquiring the dubious distinction of being the first large warship to be destroyed from the air while in combat at sea. Faced with this stark evidence that some spectacularly maximal parts do not make an optimal whole, the British navy let its programme of building battle-cruisers become extinct. In short, engineers cannot tinker with single parts in isolation from the rest of a machine, because each part costs money, space, and weight that might have gone into something else. Engineers instead have to ask what combination of parts will optimize a machine's effectiveness. By the same reasoning, evolution cannot tinker with single traits in isolation from the rest of an animal, because every structure, enzyme, or piece of DNA consumes energy and space that might have gone into something else. Instead, natural selection favoured that combination of traits that maximizes the animal's reproductive output. Thus, both engineers and evolutionary biologists have to evaluate the trade-offs involved in increasing anything; that is, its costs, as well as the benefits that it would bring. An obvious difficulty in applying this reasoning to our life-cycles is that they have many features seeming to reduce, not to maximize, our ability to produce offspring. Growing old and dying is just one example; other examples are human female menopause, bearing one baby at a time, producing babies only once every year or so at most, and not even starting to produce babies until the age of twelve to sixteen. Would not natural selection favour the woman who reached puberty at age five, completed gestation in three weeks, regularly bore quintuplets, never underwent menopause, put lots of biological energy into repair of her body, lived to 200, and thereby left hundreds of offspring?

But posing the question in that form pretends that evolution can change our bodies one piece at a time, and ignores the hidden costs. For example, a woman certainly could not reduce the length of pregnancy to three weeks without changing anything else about herself or her baby. Remember that we only have a finite amount of energy available to us. Even people doing hard exercise and eating rich food—lumberjacks, or marathon runners in training—cannot metabolize much more than about 5,000 calories per day. How should we allocate those calories between repairing ourselves and rearing babies, if our goal is to raise as many babies as possible? At the one extreme, if we put all our energy into babies and devoted no energy to biological repair, our bodies would age and disintegrate before we could rear our first baby. At the other extreme, if we lavished all our available energy on keeping our bodies in shape, we might live a long time but would have no energy left for the exhausting process of making and rearing babies. What natural selection must do is to adjust an animal's relative expenditures of energy on repair and on reproduction, so as to maximize its reproductive output, averaged over its lifetime. The answer to that problem varies among animal species, depending on factors such as their risk of accidental death, their reproductive biology, and the costs of various types of repair. This perspective can be employed to make testable predictions about how animals should differ in their repair mechanisms and rates of aging. In 1957 the evolutionary biologist George Williams cited some striking facts about aging that become comprehensible only from an evolutionary perspective. Let's consider several of Williams's examples and re-express them in the physiological language of biological repair, by taking slow aging as an indication of good repair mechanisms.

The first example concerns the age at which an animal first breeds and produces offspring. That age varies enormously among species: few humans are so precocious as to produce babies before the age of twelve years, while any self-respecting mouse a mere two months old can already make baby mice. Animals belonging to a species whose age of first breeding is late, like us, need to devote much energy to repair, in order to ensure that they survive to that reproductive age. Hence we expect investment in repair to increase with age at first reproduction.

For instance, correlated with our having a much later age of first reproduction than do mice, we humans age far more slowly than mice and are thus presumed to repair our bodies much more effectively. Even with plenty of food and the best medical care, a mouse is lucky to reach its second birthday, while we would be unlucky not to reach our seventy-second birthday. The evolutionary reason: a human who invested no more of his/her energy in repair than does a mouse would be dead long before reaching puberty. Hence it is more worthwhile to repair a human than a mouse.

What might that postulated extra energy expenditure of ours actually consist of? At first, our human repair capabilities seem unimpressive. We cannot regrow an amputated arm, and we do not regularly replace our skeleton, in the way that some short-lived invertebrates do. However, such spectacular but infrequent replacements of a whole structure probably are not the biggest items in an animal's repair budget. Instead, the biggest expense is all that invisible replacement of so many of your cells and molecules, day after day. Even if you spend all day every day just lying in bed, you need to eat about 1,640 calories per day if you are a man (1,430 for a woman) just to maintain your body. Much of that maintenance metabolism goes to our invisible scheduled replacement. And so I would guess that we cost more than a mouse in the respect of putting a bigger fraction of our energy into self-repair, and a smaller fraction into other purposes like keeping warm or caring for babies.

The second example I shall discuss involves the risk of irreparable injury. Some biological damage is potentially reparable, but there is also damage that is guaranteed to be fatal (for example, being eaten by a lion). If you are likely to be eaten by a lion tomorrow, there is no point paying a dentist to start expensive orthodontic work on your teeth today. You would do better to let your teeth rot and start having babies immediately. But if an animal's risk of death from irreparable accidents is low, then there is a potential payoff, in the form of increased lifespan, from putting energy into expensive repair mechanisms that retard aging. This is the reasoning by which Mercedes-owners decide to pay for lubrication of their cars in Germany and the US but not in New Guinea.

Biological analogies are that the risk of death from predators is lower for birds than for mammals (because birds can escape by flying), and lower for turtles than for most other reptiles (because turtles are protected by a shell). Thus, birds and turtles stand to gain a lot from expensive repair mechanisms, compared to flightless mammals and shell-less reptiles that will soon be eaten by predators anyway. Indeed, if one compares longevities of well-fed pets protected from predators, birds do live longer (that is, do age more slowly) than similarly sized mammals, and turtles live longer than similarly sized shell-less reptiles. The bird species best protected from predators are seabirds like petrels and albatrosses that nest on remote oceanic islands free of predators. Their leisurely life-cycles rival our own. Some albatrosses do not even breed until they are ten years old, and we still do not know how long they live: the birds themselves last longer than the metal rings that biologists began putting on their legs a few decades ago in order to age them. In the ten years that it takes an albatross to start breeding, a mouse population could have gone through sixty generations, most of which would already have succumbed to predators or old age. As our third example, let's compare males and females of the same species. We expect more potential payoff from repair mechanisms, and lower rates of aging, in that sex with the lower accidental mortality rate. For many or most species, males suffer greater accidental mortality than females, partly because males put themselves at greater risk by fighting and bold displays. This is certainly true of human males today and has probably been so throughout our history as a species—men are the sex most likely to die in wars against men of other groups, and in individual fights within a group. Also, in many species the males are bigger than the females, but studies of red deer and of New World blackbirds show that males are thereby more likely than females to die when food becomes scarce.

Correlated with this greater accidental death rate of men, men also age faster and have a higher non-accidental death rate than women. At present, women's life expectancy is about six years greater than that of men; some of this difference is because more men than women are smokers, but there is a sex-linked difference in life expectancy even among non-smokers. These differences suggest that evolution has programmed us so that women put more energy into self-repair, while men put more energy into fighting. Expressed another way, it just is not worth as much to repair a man as it is to repair a woman. I do not thereby mean to denigrate male fighting, which serves a useful evolutionary Purpose for a man: to gain wives and to secure resources for his children and his tribe, at the expense of other men and their children and tribe.

My remaining example of how some striking facts of aging become comprehensible only from an evolutionary perspective concerns the distinctively human phenomenon of survival past reproductive age, Specially past female menopause. Since transmitting one's genes to the next generation is what drives evolution, other animal species rarely survive past reproductive age. Instead, Nature programmes death to coincide with the end of fertility, because there is then no longer an evolutionary benefit to gain from keeping one's body in good repair. It is an exception in need of explanation to realize that women are programmed to live for decades after menopause, and that men are programmed to live to an age when most men are no longer busy siring babies. But the explanation becomes apparent on reflection. The intense phase of parental care is unusually protracted in the human species and lasts nearly two decades. Even those older people whose own children have reached adulthood are tremendously important to the survival of not just their children but of their whole tribe. Especially in the days before writing, they acted as the carriers of essential knowledge. Thus, Nature has programmed us with the capacity to keep the rest of our bodies in reasonable repair even at an age when the female reproductive system itself has fallen into disrepair.

Conversely, though, we have to wonder why natural selection programmed female menopause into us in the first place. It too, like aging, cannot be explained away as something physiologically inevitable. Most mammals, including human males plus chimps and gorillas of both sexes, merely experience a gradual decline and eventual cessation of fertility with age, rather than the abrupt shutdown of women's fertility. Why did that peculiar, seemingly counter-productive feature of ours evolve? Would not natural selection favour the woman who remained fertile until the bitter end?

Human female menopause probably resulted from two other distinctively human characteristics: the exceptional danger that childbirth poses to the mother, and the danger that a mother's death poses to her offspring. Recall from Chapter Three the enormous size of the human infant at birth relative to its mother: our big 7-pound babies emerging from 100-pound mothers, compared to little 4-pound gorilla babies emerging from a 200-pound gorilla mother. As a result, childbirth is dangerous to women. Especially before the advent of modern obstetrics, women often died in childbirth, whereas mother gorillas and chimps virtually never do.

Now recall also from Chapter Three the extreme dependence of human infants on their parents, especially on their mother. Because human infants develop so slowly and cannot even feed themselves after weaning (unlike young apes), the death of a hunter-gatherer mother would have been likely to be fatal to her offspring up to a later age in childhood than for any other primate. Hence a hunter-gatherer mother with several children was gambling the lives of those children at every subsequent childbirth. Since her investment in those prior children increased with their age, and since her own risk of death in childbirth also increased with her age, the odds of her gamble paying off got worse and worse as she got older. When you already have three children alive but still dependent on you, why risk those three for a fourth?

Those worsening odds probably led through natural selection to menopausal shutdown of human female fertility, in order to protect a mother's prior investment in children. Since childbirth carries no risk of death for fathers, men did not evolve menopause. Like aging, menopause illustrates how an evolutionary approach illuminates features of our life-cycle that would otherwise be counterintuitive. It is even possible that menopause evolved only within the past 40,000 years, when Cro-Magnons and other anatomically modern humans began frequently to survive to the age of sixty or more. Neanderthals and earlier humans usually died before the age of forty anyway, so that menopause would have brought their women no benefits if it were to occur at the same age as in modern Femina sapiens.

Thus, the longer lifespan of modern humans than of apes rests not only on cultural adaptations, such as tools to acquire food and deter predators. It also rests on the biological adaptations of menopause and increased investment in self-repair. Whether those biological adaptations developed especially at the time of the Great Leap Forward or earlier, they rank among the life history changes that permitted the rise of the third chimpanzee to humanity. The last conclusion that I wish to draw from an evolutionary approach to aging is that it undermines the approach which has long dominated the physiological study of aging. The gerontological literature is obsessed with a search for The Cause of Aging—preferably a single cause, certainly not more than a few major causes. Within my own lifetime as a biologist, hormonal changes, deterioration in the immune system, and neural degeneration have vied in popularity for the title of The Cause, without compelling support having been adduced to date for any of the candidates. But evolutionary reasoning suggests that this search will remain futile. There should not be just one, or even a few, dominant physiological mechanisms of aging. Instead, natural selection should act to match rates of aging in all physiological systems, with the result that ^aging involves innumerable simultaneous changes.

The basis of this prediction is as follows. There is no point doing expensive maintenance on one piece of the body if other pieces are deteriorating more rapidly. Conversely, natural selection should not permit a few systems to deteriorate long before all the others, as the cost of extra repairs on just those few systems would then buy a big increase in life expectancy and would be worth it. By analogy, Mercedes-owners should not install cheap ball-bearings when they are lavishing expense on all other parts of the car. Had they been so foolish, they could have doubled the lifetime of their costly car just by spending a few more dollars for better ball-bearings. But it would not pay either to go to the expense of installing diamond ball-bearings, when all the rest of the car would have rusted away before those ball-bearings wore out. Thus, the optimal strategy for Mercedes-owners, and for us, is to repair all parts of our cars or bodies at such rates that everything finally collapses all at once. It seems to me that this depressing prediction is borne out, and that we come closer to this evolutionary ideal than to the physiologists' long-sought Cause of Aging. Signs of aging can be found wherever one looks for them. Already I am conscious in myself of tooth wear, considerable decreases in muscle performance, and significant losses in hearing, vision, smell, and taste. For all these senses, the acuity of women is greater than that of men of equal age, whatever the age group compared. Ahead of me lies the familiar litany: weakening of the heart, hardening of the arteries, increasing brittleness of bones, decreases in kidney filtration rates, lower resistance of the immune system, and loss of memory. The list could be extended almost indefinitely. Evolution seems indeed to have arranged things so that all our systems deteriorate, and that we invest in repair only as much as we are worth.

From a practical standpoint, this conclusion is disappointing. If there had been one dominant cause of aging, curing that cause would have provided us with a fountain of youth. This thought, operating at a time when aging was thought to be largely a hormonal phenomenon, inspired some attempts at miraculous rejuvenation of old people by hormonal injections or implantation of young gonads. Such an attempt was the subject of Sir Arthur Conan Doyle's story, The Adventure of the Creeping Man, in which the aged Professor Presbury becomes infatuated with a young woman, desperately wants to rejuvenate himself, and instead is found creeping around like a monkey after midnight. The great Sherlock Holmes discovers the reason: the Professor has been seeking youth by injecting himself with the serum of langur monkeys.

I could have warned Professor Presbury that his myopic obsession with proximate causation would lead him astray. Had he thought of ultimate evolutionary causation, he would have realized that natural selection would never permit us to deteriorate through a single mechanism with one simple cure. Perhaps it is just as well. Sherlock Holmes worried greatly about what would happen if such an elixir of life were found: There is danger there—a very real danger to humanity. Consider Watson, that the material, the sensual, the worldly would all prolong their worthless lives… It would be the survival of the least fit. What sort of cesspool may not our poor world become? Holmes would be relieved to know that his worries now appear unlikely to materialize.