Why is Sex Fun?: the evolution of human sexuality

CHAPTER 3. WHY DON'T MEN BREAST-FEED THEIR BABIES?: The Non-Evolution of Male Lactation

Today, we men are expected to share in the care of our children. We have no excuse not to, because we are perfectly capable of doing for our kids virtually anything that our wives can do. And so, when my twin sons were born in 1987,I duly learned to change diapers, clean up vomit, and perform the other tasks that come with parenthood.

The one task that I felt excused from was nursing my infants. It was visibly a tiring task for my wife. Friends kidded me that I should get hormone injections and share the burden. Yet cruel biological facts seemingly confront those who would bring sexual equality into this last bastion of female privilege or male cop-out. It appears obvious that males lack the anatomical equipment, the priming experience of pregnancy, and the hormones necessary for lactation. Until 1994, not a single one of the world's 4,300 mammal species was suspected of male lactation under normal conditions. The nonexistence of male lactation may thus seem to be a solved problem requiring no further discussion, and it may seem doubly irrelevant to a book about how the unique aspects of human sexuality evolved. After all, the problem's solution seems to depend on facts of physiology rather than on evolutionary reasoning, andexclusively female lactation is apparently a universal mammalian phenomenon not at all unique to humans.

In reality, the subject of male lactation follows perfectly from our discussion of the battle of the sexes. It illustrates the failure of strictly physiological explanations and the importance of evolutionary reasoning for understanding human sexuality. Yes, it's true that no male mammal has ever become pregnant, and that the great majority of male mammals normally don't lactate. But one has to go further and ask why mammals evolved genes specifying that only females, not males, would develop the necessary anatomical equipment, the priming experience of pregnancy, and the necessary hormones. Both male and female pigeons secrete crop “milk” to nurse their squab; why not men as well as women? Among seahorses it's the male rather than the female that becomes pregnant; why is that not also true for humans?

As for the supposed necessity of pregnancy as a primary experience for lactation, many female mammals, including many (most?) women, can produce milk without first being primed by pregnancy. Many male mammals, including some men, undergo breast development and lactate when given the appropriate hormones. Under certain conditions, a considerable fraction of men experience breast development and milk production even without having been treated hormonally. Cases of spontaneous lactation have long been known in male domestic goats, and the first case of male lactation in a wild mammal species has been reported recently.

Thus, lactation lies within the physiological potential of men. As we shall see, lactation would make more evolutionary sense for modern men than for males of most other mammal species. But the fact remains that it's not part of our normal repertoire, nor is it known to fall within the normal repertoire of other mammal species except for that single case reported recently. Since natural selection evidently could have made men lactate, why didn't it? That turns out to be a major question that cannot be answered simply by pointing to the deficiencies of male equipment. Male lactation beautifully illustrates all the main themes in the evolution of sexuality: evolutionary conflicts between males and females, the importance of confidence in paternity or maternity, differences in reproductive investment between the sexes, and a species' commitment to its biological inheritance.

As the first step in exploring these themes, I have to overcome your resistance to even thinking about male lactation, a product of our unquestioned assumption that it's physiologically impossible. The genetic differences between males and females, including those that normally reserve lactation for females, turn out to be slight and labile. This chapter will convince you of the feasibility of male lactation and will then explore why that theoretical possibility normally languishes unrealized.

Our sex is ultimately laid down by our genes, which in humans are bundled together in each body cell in twenty-three pairs of microscopic packages called chromosomes. One member of each of our twenty-three pairs was acquired from our mother, and the other member from our father. The twenty-three human chromosome pairs can be numbered and distinguished from each other by consistent differences in appearance. In chromosome pairs 1 through 22, the two members of each pair appear identical when viewed through a microscope. Only in the case of chromosome pair 23, the so-called sex chromosomes, do the two representatives differ, and even that's true only in men, who have a big chromosome (termed an X chromosome) paired with a small one (a Y chromosome). Women instead have two paired X chromosomes.

What do the sex chromosomes do? Many X chromosome genes specify traits unrelated to sex, such as the ability to distinguish red and green colors. However, the Y chromosome contains genes specifying the development of testes. In the fifth week after fertilization human embryos of either sex develop a “bipotential” gonad that can become either a testis or an ovary. If a Y chromosome is present, that bet-hedging gonad begins to commit itself in the seventh week to becoming a testis, but if there's no Y chromosome, the go-nnd waits until the thirteenth week to develop as an ovary.

That may seem surprising: one might have expected the second X chromosome of girls to make ovaries, and the Y chromosome of boys to make testes. In fact, though, people abnormally endowed with one Y and two X chromosomes turn out most like males, whereas people endowed with three or just one X chromosome turn out most like females. Thus, the natural tendency of our bet-hedging primordial gonad is to develop as an ovary if nothing intervenes; something extra, a Y chromosome, is required to change it into a testis.

It's tempting to restate this simple fact in emotionally loaded terms. As the endocrinologist Alfred Jost put it, “Becoming a male is a prolonged, uneasy, and risky venture; it is a kind of struggle against inherent trends towards femaleness.” Chauvinists might go further and hail becoming a man as heroic, and becoming a woman as the easy fallback position. Conversely, one might regard womanhood as the natural state of humanity, with men just a pathological aberration that regrettably must be tolerated as the price for making more women. I prefer merely to acknowledge that a Y chromosome switches gonad development from the ovarian path to the testicular path, and to draw no metaphysical conclusions.

But there's more to a man than testes alone. A penis and prostate gland are among the many other obvious necessities of manhood, just as women need more than ovaries (for instance, it helps to have a vagina). It turns out that the embryo is endowed with other bipotential structures besides the primordial gonad. Unlike the primordial gonad, though, these other bipolar structures have a potential that is not directly specified by the Y chromosome. Instead, secretions produced by the testes themselves are what channel these other structures toward developing into male organs, while lack of testicular secretions channels them toward making female organs.

For example, already in the eighth week of gestation the testes begin producing the steroid hormone testosterone, some of which gets converted into the closely related steroid dihydrotestosterone. These steroids (known as an-drogens) convert some all-purpose embryonic structures into the glans penis, penis shaft, and scrotum; the same structures would otherwise develop into the clitoris, labia minora, and labia majora. Embryos also start out bet-hedging with two sets of ducts, known as the Mullerian ducts and Wolffian ducts. In the absence of testes, the Wolffian ducts atrophy, while the Mullerian ducts grow into a female fetus's uterus, fallopian tubes, and interior vagina. With testes present, the opposite happens: androgens stimulate the Wolffian ducts to grow into a male fetus's seminal vesicles, vas deferens, and epididymis. At the same time, a testicular protein called Mullerian inhibiting hormone does what its name implies: it prevents the Mullerian ducts from developing into the internal female organs.

Since a Y chromosome specifies testes, and since the presence or absence of the testes' secretions specifies the remaining male or female structures, it might seem as if there's no way that a developing human could end up with ambiguous sexual anatomy. Instead, you might think that a Y chromosome should guarantee 100 percent male organs, and that lack of a Y chromosome should guarantee 100 percent female organs.

In fact, a long series of biochemical steps is required to produce all those other structures besides ovaries or testes. Each step involves the synthesis of one molecular ingredient, termed an enzyme, specified by one gene. Any enzyme can be defective or absent if its underlying gene is altered by a mutation. Thus, an enzyme defect may result in a male pseudohermaphrodite, defined as someone possessing some female structures as well as testes. In a male pseudoher-maphrodite with an enzyme defect, there is normal development of the male structures dependent on enzymes that act at the steps of the metabolic pathway before the defective enzyme. However, male structures dependent on the defective enzyme itself or on subsequent biochemical steps fail to develop and are replaced either by their female equivalent or by nothing at all. For example, one type of pseudohermaphrodite looks like a normal woman. Indeed, “she” conforms to the male ideal of female pulchritude even more closely than does the average real woman, because “her” breasts are well developed and “her” legs are long and graceful. Hence cases have turned up repeatedly of beautiful women fashion models not realizing that they are actually men with a single mutant gene until genetically tested as adults.

Since this type of pseudohermaphrodite looks like a normal girl baby at birth and undergoes externally normal development and puberty, the problem isn't even likely to be recognized until the adolescent “girl” consults a doctor over failure to begin menstruating. At that point, the doctor discovers a simple reason for that failure: the patient has no uterus, fallopian tubes, or upper vagina. Instead, the vagina ends blindly after two inches. Further examination reveals testes that secrete normal testosterone, are programmed by a normal Y chromosome, and are abnormal only for being buried in the groin or labia. In other words, the beautiful model is an otherwise normal male who happens to have a genetically determined biochemical block in his ability to respond to testosterone.

That block turns out to be in the cell receptor that would normally bind testosterone and dihydrotestos-terone, thereby enabling those androgens to trigger the further developmental steps of the normal male. Since the Y chromosome is normal, the testes themselves form normally and produce normal Mullerian inhibiting hormone, which acts as in any man to forestall development of the uterus and fallopian tubes. However, development of the usual male machinery to respond to testosterone is interrupted. Hence development of the remaining bipotential embryonic sex organs follows the female channel by default: female rather than male external genitalia, and atrophy of the Wolffian ducts and hence of potential male internal genitalia. In fact, since the testes and adrenal glands secrete small amounts of estrogen that would normally be overridden by androgen receptors, the complete lack of those receptors in functional form (they are present in small numbers in normal women) makes the male pseudohermaphrodite appear externally superfeminine.

Thus, the overall genetic difference between men and women is modest, despite the big consequences of that modest difference. A small number of genes on chromosome 23, acting in concert with genes on other chromosomes, ultimately determine all differences between men and women. The differences, of course, include not just those in the reproductive organs themselves but also all other postadolescent sex-linked differences, such as the differences in beards, body hair, pitch of voice, and breast development.

The actual effects of testosterone and its chemical derivatives vary with age, organ, and species. Animal species differ greatly in how the sexes differ, and not only in mammary gland development. Even among higher anthropoids— humans and our closest relatives, the apes-there are familiar differences in sexual distinctiveness. We know from zoos and photos that adult male and female gorillas differ obviously at a long distance by the male's much greater size (his weight is double the female's), different shape of head, and silver-haired back. Men also differ, though much less obviously, from women in being slightly heavier (by 20 percent on the average), more muscular, and bearded. Even the degree of that difference varies among human populations: for example, the difference is less marked among Southeast Asians and Native Americans, since men of those populations have on the average much less body hair and beard development than in Europe and Southwest Asia. But males and females of some gibbon species look so similar that you couldn't distinguish them unless they permitted you to examine their genitals.

In particular, both sexes of placental mammals have mammary glands. While the glands are less well developed and nonfunctional in males of most mammal species, that degree of male underdevelopment varies among species. At the one extreme, in male mice and rats, the mammary tissue never forms ducts or a nipple and remains invisible from the outside. At the opposite extreme, in dogs and primates (including humans) the gland does form ducts and a nipple in both males and females and scarcely differs between the sexes before puberty.

During adolescence the visible differences between the mammalian sexes increase under the influence of a mix of hormones from the gonads, adrenal glands, and pituitary gland. Hormones released in pregnant and lactating females produce a further mammary growth spurt and start milk production, which is then reflexly stimulated by nursing. In humans, milk production is especially under the control of the hormone prolactin, while the responsible hormones in cows includes somatotropin, alias “growth hormone” (the hormone behind the current debate over proposed hormonal stimulation of milk cows).

It should be emphasized that male/female differences in hormones aren't absolute but a matter of degree: one sex may have higher concentrations and more receptors for a particular hormone. In particular, becoming pregnant is not the only way to acquire the hormones necessary for breast growth and milk production. For instance, normally circulating hormones stimulate a milk production, termed witch's milk, in newborns of several mammal species. Direct injection of the hormones estrogen or progesterone (normally released during pregnancy) triggers breast growth and milk production in virgin female cows and goats-and also in steers, male goats, and male guinea pigs. The hor-monally treated virgin cows produced on the average as much milk as their half-sisters that were nursing calves to which they had given birth. Granted, hormonally treated steers produced much less milk than virgin cows; you shouldn't count on steer's milk in the supermarkets by next Christmas. But that's not surprising since the steers have previously limited their options: they haven't developed an udder to accommodate all the mammary gland tissue that hormonally treated virgin cows can accommodate.

There are numerous conditions under which injected or topically applied hormones have produced inappropriate breast development and milk secretion in humans, both in men and in nonpregnant or non-nursing women. Men and women cancer patients being treated with estrogen proceeded to secrete milk when injected with prolactin; among such patients was a sixty-four-year-old man who continued to produce milk for seven years after hormonal treatment was discontinued. (This observation was made in the 1940s, long before the regulation of medical research by human subjects protection committees, which now forbid such experiments). Inappropriate lactation has been observed in people taking tranquilizers that influence the hypothalamus (which controls the pituitary gland, the source of prolactin); it also has been observed in people recovering from surgery that stimulated nerves related to the suckling reflex, as well as in some women on prolonged courses of estrogen and progesterone birth-control pills. My favorite case is the chauvinist husband who kept complaining about his wife's “miserable little breasts,” until he was shocked to find his own breasts growing. It turned out that his wife had been lavishly applying estrogen cream to her breasts to stimulate the growth craved by her husband, and the cream had been rubbing off on him.

At this point, you may be starting to wonder whether all these examples are irrelevant to the possibility of normal male lactation, since they involve medical interventions such as hormone injections or surgery. But inappropriate lactation can occur without high-tech medical procedures: mere repeated mechanical stimulation of the nipples suffices to trigger milk secretion in virgin females of several mammal species, including humans. Mechanical stimulation is a natural way of releasing hormones by means of nerve reflexes connecting the nipples to hormone-releasing glands via the central nervous system. For instance, a sexually mature but virgin female marsupial can regularly be stimulated to lactate just by fostering another mother's young onto her teats. The “milking” of virgin female goats similarly triggers them to lactate. That principle might be transferable to men, since manual stimulation of the nipples causes a prolactin surge in men as well as in nonlac-tating women. Lactation is a not infrequent result of nipple self-stimulation in teenage boys.

My favorite human example of this phenomenon comes from a letter to the widely syndicated newspaper column “Dear Abby.” An unmarried woman about to adopt a newborn infant longed to nurse the infant and asked Abby whether taking hormones would help her to do so. Abby's reply was: Preposterous, you'll only make yourself sprout hair! Several indignant readers then wrote in to describe cases of women in similar situations who succeeded in nursing an infant by repeatedly placing it at the breast.

Recent experience of physicians and nurse lactation specialists now suggest that most adoptive mothers can begin producing some milk within three or four weeks. The recommended preparation for prospective adoptive mothers is to use a breast pump every few hours to simulate sucking, beginning about a month before the expected delivery of the birth mother. Long before the advent of modern breast pumps, the same result was achieved by repeatedly putting a puppy or a human infant to the breast. Such preparation was practiced especially in traditional societies when a pregnant woman was sickly and her own mother wanted to be ready to step in and nurse the infant in case the daughter proved unable to do so. The reported examples include grandmothers up to the age of seventy-one, as well as Ruth's mother-in-law Naomi in the Old Testament. (If you don't believe it, open a Bible and turn to the Book of Ruth, chapter 4, verse 16.)

Breast development occurs commonly, and spontaneous lactation occasionally, in men recovering from starvation. Thousands of cases were recorded in prisoners of war released from concentration camps after World War II; one observer noted five hundred cases in survivors of one Japanese POW camp alone. The likely explanation is that starvation inhibits not only the glands that produce hormones but also the liver, which destroys those hormones. The glands recover much faster than the liver when normal nutrition is resumed, so that hormone levels soar unchecked. Again, turn to the Bible to discover how Old Testament patriarchs anticipated modern physiologists: Job (chapter 21, verse 24) remarked of a well-fed man that “His breasts are full of milk.”

It has been known for a long time that many otherwise perfectly normal male goats, with normal testes and proven ability to inseminate females, surprise their owners by spontaneously growing udders and secreting milk. Billy-goat milk is similar in composition to she-goat milk but has even higher fat and protein content. Spontaneous lactation has also been observed in a captive monkey, the stump-tailed macaque of Southeast Asia.

In 1994, spontaneous male lactation was at last reported in males of a wild animal species, the Dyak fruit bat of Malaysia and adjacent islands. Eleven adult males captured alive proved to have functional mammary glands that yielded milk when manually expressed. Some of the males' mammary glands were distended with milk, suggesting that they had not been suckled and as a result milk had accumulated. However, others may have been suckled because they had less distended (but still functional) glands, as in lactating females. Among three samples of Dyak fruit bats caught at different places and seasons, two included lactating males, lactating females, and pregnant females, but adults of both sexes in the third sample were reproductively inactive. This suggests that male lactation in these bats may develop along with female lactation as part of the natural reproductive cycle. Microscopic examination of the testes revealed apparently normal sperm development in the lactating males.

Thus, while usually mothers lactate and fathers don't, males of at least some mammal species have much of the necessary anatomical equipment, physiological potential, and hormone receptors. Males treated either with the hormones themselves, or with other agents likely to release hormones, may undergo breast development and some lactation. There are several reports of apparently normal adult men nursing babies; one such man whose milk was analyzed secreted milk sugar, protein, and electrolytes at levels similar to those of mother's milk. All these facts suggest that it would have been easy for male lactation to evolve; perhaps it would have required just a few mutations causing increased release or decreased breakdown of hormones.

Evidently, evolution just didn't design men to utilize that physiological potential under normal conditions. In computing terminology, at least some males have the hardware; we merely haven't been programmed by natural selection to use it. Why not?

To understand why, we need to switch from physiological reasoning, which we have been using throughout this chapter, back to the evolutionary reasoning that we were using in chapter 2. In particular, recall how the evolutionary battle of the sexes has resulted in parental care being provided by the mother alone in about 90 percent of all mammal species. For those species, in which offspring will survive with zero paternal care, it's obvious that the question of male lactation never arises. Not only do males of those species have no need to lactate; they also don't have to bring food, defend a family territory, defend or teach their offspring, or do anything else for their offspring. The male's crass genetic interests are best served by chasing other females to impregnate. A noble male carrying a mutation to nurse his offspring (or to care for them in any other way) would quickly be outbred by selfish normal males that forewent lactation and thereby became able to sire more offspring.

Only for those 10 percent of mammal species in which male parental care is necessary does the question of male lactation even deserve consideration. Those minority species include lions, wolves, gibbons, marmosets-and humans. But even in those species requiring male parenting, lactation isn't necessarily the most valuable form that the father's contribution can take. What a big lion really must do is to drive off hyenas and other big lions bent on killing his cubs. He should be out patrolling his territory, not sitting home nursing the cubs (which the smaller lioness is perfectly capable of doing) while his cubs' enemies are sneaking up. The wolf father may make his most useful contribution by leaving the den to hunt, bringing back meat to the wolf mother, and letting her turn the meat into milk. The gibbon father may contribute best by looking out for pythons and eagles that might grab his offspring, and by vigilantly expelling other gibbons from the fruit trees in which his spouse and offspring are feeding, while marmoset fathers spend much time carrying their twin offspring.

All these excuses for male nonlactation still leave open the possibility that some other mammal species could exist in which male lactation might be advantageous to the male and his offspring. The Dyak fruit bat may turn out to be such a species. But even if there are mammal species for which male lactation would be advantageous, its realization runs up against problems posed by the phenomenon termed evolutionary commitment.

The idea behind evolutionary commitment can be understood by analogy to devices manufactured by humans. A manufacturer of trucks can easily modify one basic truck model for different but related purposes, such as transporting furniture, horses, or frozen food. Those different purposes can be fulfilled by making a few minor variations on the same basic design of the truck's cargo compartment, with little or no change in the motor, brakes, axles, and other major components. Similarly, an airplane manufacturer can with minor modifications use the same model of airplane to carry ordinary passengers, skydivers, or freight. But it is not feasible to convert a truck into an airplane or vice versa, because a truck is committed to truckhood in too many respects: heavy body, diesel motor, braking system, axles, and so on. To build an airplane, one would not start with a truck and modify it; one would instead start all over again.

Animals, in contrast, are not designed from scratch to provide an optimal solution for a desired lifestyle. Instead, they evolve from existing animal populations. Evolutionary changes in lifestyle come about incrementally through the accumulation of small changes in an evolutionary design adapted to a different but related lifestyle. An animal with many adaptations to one specialized lifestyle may not be able to evolve the many adaptations required for a different lifestyle, or may do so only after a very long time. For instance, a female mammal that gives birth to live young cannot evolve into a birdlike egg layer merely by extruding her embryo to the outside within a day of fertilization; she would have to have evolved birdlike mechanisms for synthesizing yolk, eggshell, and other avian commitments to egg laying.

Recall that, of the two main classes of warm-blooded vertebrates, birds and mammals, male parental care is the rule among birds and the exception among mammals. That difference results from birds' and mammals' long evolutionary histories of developing different solutions to tho problem of what to do with an egg that has just been fertilized internally. Each of those solutions has required a whole set of adaptations, which differ between birds and mammals and to which all modern birds and mammals are now heavily committed.

The bird's solution is to have the female rapidly extrude the fertilized embryo, packaged with yolk inside a hard shell, in an extremely undeveloped and utterly helpless state that is impossible for anyone except an embryologist to recognize as a bird. From the moment of fertilization to the moment of extrusion, the embryo's development inside the mother lasts only a day or a few days. That brief internal development is followed by a much longer period of development outside the mother's body: up to 80 days of incubation before the egg hatches, and up to 240 days of feeding and caring for the hatched chick until it can fly.

Once the egg has been laid, there is nothing further in the chick's development that uniquely requires its mother's help. The father can sit on the egg and keep it warm just as well as the mother can. After hatching out, chicks of most bird species eat the same food as their parents, and the father can collect and bring that food to the nest as well as the mother can.

In most bird species the care of the nest, egg, and chick requires both parents. In those bird species in which the efforts of one parent suffice, that parent is more often the mother than the father, for the reasons discussed in chapter 2: the female's greater obligate internal investment in the fertilized embryo, the greater opportunities foreclosed for the male by parental care, and the male's low confidence in paternity as a result of internal fertilization. But in all bird species the female's obligate internal investment is much less than that in any mammal species, because the developing young bird is “born” (laid) in such an early stage of development compared to even the least developed newborn mammal. The ratio of development time outside the mother-a time of duties that in theory can be shared by the mother and the father-to development time inside the mother is much higher for birds than for mammals. No mother bird's “pregnancy”-egg formation time— approaches the nine months of human pregnancy or even the twelve days of the briefest mammalian pregnancy.

Hence female birds are not as easily bluffed as female mammals into caring for the offspring while the father deserts to philander. That has consequences for the evolutionary programming not only of birds' instinctive behaviors but also of their anatomy and physiology. In pigeons, which feed their young by secreting “milk” from their crops, both the father and the mother have evolved to secrete milk. Biparental care is the rule in birds, and while in those bird species that practice uniparental care the mother is usually the sole caretaker, in some bird species it is the father, a development unprecedented among mammals. Care by the father alone characterizes not only those bird species characterized by sex-role-reversal polyandry but also some other birds, including ostriches, emus, and tina-mous.

The bird solution to the problems posed by internal fertilization and subsequent embryonic development involves specialized anatomy and physiology. Female but not male birds possess an oviduct of which one portion secretes albumin (the egg white protein), another portion makes the inner and outer shell membranes, and still another makes the eggshell itself. All of those hormonally regulated structures and their metabolic machinery represent evolutionary commitment. Birds must have been evolving along this pathway for a long time, because egg laying was already widespread in ancestral reptiles, from which birds may have inherited much of their egg-making machinery. Creatures that are recognizably birds and no longer reptiles, such as the famous Archaeopteryx, appear in the fossil record by 150 million years ago. While the reproductive biology of Archaeopteryx is unknown, a dinosaur fossil from about 80 million years ago has been found entombed on a nest and eggs, suggesting that birds inherited nesting behavior as well as egg laying from their reptilian ancestors.

Modern bird species vary greatly in their ecology and lifestyle, from aerial fliers to terrestrial runners and marine divers, from tiny hummingbirds to giant extinct elephant birds, and from penguins nesting in the Antarctic winter to toucans breeding in tropical rainforests. Despite that variation in lifestyle, all existing birds have remained committed to internal fertilization, egg laying, incubation, and other distinctive features of avian reproductive biology, with only minor variations among species. (The principal exceptions are the brush turkeys of Australia and the Pacific islands: they incubate their eggs with external heat sources, such as fermentative, volcanic, or solar heat, rather than with body heat.) If one were designing a bird from scratch, perhaps one could come up with a better but entirely different reproductive strategy, such as that of bats, which fly like birds but reproduce by pregnancy, live birth, and lactation. Whatever the virtues of that bat solution, it would require too many major changes for birds, which remain committed to their own solution.

Mammals have their own long history of evolutionary commitment to their solution to the same problem of what to do with an internally fertilized egg. The mammalian solution begins with pregnancy, an obligate period of embryonic development within the mother that lasts much longer than in any mother bird. Pregnancy's duration ranges from a minimum of twelve days in bandicoots to twenty-two months in elephants. That big initial commitment by a female mammal makes it impossible for her to bluff her way out of further commitment and has led to the evolution of female lactation. Like birds, mammals have evidently been committed to their distinctive solution for a long time. Lactation does not leave fossil traces, but it is shared among the three living groups of mammals (monotremes, marsupials, and placentals), which had already differentiated from each other by 135 million years ago. Hence lactation presumably arose in some mammal-like reptilian ancestor (so-called therapsid reptiles) even earlier.

Like birds, mammals are committed to much specialized reproductive anatomy and physiology of their own. Some of those specializations differ greatly between the three mammalian groups, such as placental development resulting in a relatively mature newborn in placental mammals, earlier birth and relatively longer postnatal development in marsupials, and egg-laying in monotremes. These specializations have probably been in place for at least 135 million years.

Compared to those differences between the three mammalian groups, or compared to the differences between all mammals and birds, variation within each of the three groups of mammals is minor. No mammal has re-evolved external fertilization or discarded lactation. No marsupial or placental mammal has re-evolved egg laying. Species differences in lactation are mere quantitative differences: more of this, less of that. For instance, the milk of Arctic seals is concentrated in nutrients, high in fat, and almost devoid of sugar, while human milk is more dilute in nutrients, sugary, and low in fat. Weaning from milk to solid food extends over a period of up to four years in traditional human hunter-gatherer societies. At the other extreme, guinea pigs and jackrabbits are capable of nibbling solid food within a few days of birth and dispensing with milk soon thereafter. Guinea pigs and jackrabbits may be evolving in the direction of bird species with precocial young, such as chickens and shorebirds, whose hatchlings already have open eyes, can run, and can find their own food but cannot yet fly or fully regulate their own body temperature. Perhaps, if life on Earth survives the current onslaught by humans, the evolutionary descendants of guinea pigs and jackrabbits will discard their inherited evolutionary commitment to lactation-in a few more tens of millions of years.

Thus, other reproductive strategies might work for a mammal, and it would seem to require few mutations to transform a newborn guinea pig or jackrabbit into a newborn mammal that requires no milk at all. But that has not happened: mammals have remained evolutionarily committed to their characteristic reproductive strategy. Similarly, even though we have seen that male lactation is physiologically possible, and although it also would seem to require few mutations, female mammals have nevertheless had an enormous evolutionary head start on males in perfecting their shared physiological potential for lactation. Females, but not males, have been undergoing natural selection for milk production for tens of millions of years. In all the species I cited to demonstrate that male lactation is physiologically possible-humans, cows, goats, dogs, guinea pigs, and Dyak fruit bats-lactating males still produce much less milk than do females.

Still, the tantalizing recent discoveries about Dyak fruit bats make one wonder whether out there today, undiscovered, might be some mammal species whose males and females share the burden of lactation-or one that might evolve such sharing in the future. The life history of the Dyak fruit bat remains virtually unknown, so we cannot say what conditions favored in it the beginnings of normal male lactation, nor how much milk (if any) the male bats actually supply to their offspring. Nevertheless, we can easily predict on theoretical grounds the conditions that would favor the evolution of normal male lactation. Those conditions include: a litter of infants that constitute a big burden to nourish; monogamous male-female pairs; high confidence of males in their paternity; and hormonal preparation of fathers, while their mate is still pregnant, for eventual lactation.

The mammal species that some of these conditions already best describe is-the human species. Medical technology is making others of these conditions increasingly applicable to us. With modern fertility drugs and high-tech methods of fertilization, births of twins and triplets are becoming more frequent. Nursing human twins is such an energy drain that the daily energy budget of a mother of twins approaches that of a soldier in boot camp. Despite all our jokes about infidelity, genetic testing shows the great majority of American and European babies tested to have been actually sired by the mother's husband. Genetic testing of fetuses is becoming increasingly common and can already permit a man to be virtually 100 percent sure that he really sired the fetus within his pregnant wife.

Among animals, external fertilization favors, and internal fertilization mitigates against, the evolution of male parental investment. That fact has discouraged male parental investment by other mammal species but now uniquely favors it in humans, because in-vitro external fertilization techniques have become a reality for humans within the past two decades. Of course, the vast majority of the world's babies are still conceived internally by natural methods. But the increasing number of older women and men who wish to conceive but have difficulty doing so, and the reported modern decline in human fertility (if it is real), combine to ensure that more and more human babies will be products of external fertilization, like most fish and frogs.

All these features make the human species a leading candidate for male lactation. While that candidacy may take millions of years to perfect through natural selection, we have it in our power to short-circuit that evolutionary process by technology. Some combination of manual nipple stimulation and hormone injections may soon develop the latent potential of the expectant father-his confidence in paternity buttressed by DNA testing-to make milk, without the need to await genetic changes. The potential advantages of male lactation are numerous. It would promote a type of emotional bonding of father to child now available only to women. Many men, in fact, are jealous of the special bond arising from breast-feeding, whose traditional restriction to mothers makes men feel excluded. Today, many or most mothers in first-world societies have already become unavailable for breast-feeding, whether because of jobs, illness, or lactational failure. Yet not only parents but also babies derive many benefits from breast-feeding. Breast-fed babies acquire stronger immune defenses and are less susceptible to numerous diseases, including diarrhea, ear infections, early-age-onset diabetes, influenza, necrotizing enterocolitis, and SIDS (Sudden Infant Death Syndrome). Male lactation could provide those benefits to babies if the mother is unavailable for any reason.

It must be acknowledged, however, that the obstacles to male lactation are not only physiological ones, which can evidently be overcome, but also psychological ones. Men have traditionally regarded breast-feeding as a woman's job, and the first men to breast-feed their infants will undoubtedly be ridiculed by many other men. Nevertheless, human reproduction already involves increasing use of other procedures that would have seemed ridiculous until a few decades ago: procedures such as external fertilization without intercourse, fertilization of women over the age of fifty, gestation of one woman's fetus inside another woman's womb, and survival of prematurely delivered one-kilogram fetuses by high-tech incubator methods. We now know that our evolutionary commitment to female lactation is physiologically labile; it may prove psychologically labile as well. Perhaps our greatest distinction as a species is our capacity, unique among animals, to make counter-evolutionary choices. Most of us choose to renounce murder, rape, and genocide, despite their advantages as a means for transmitting our genes, and despite their widespread occurrence among other animal species and earlier human societies. Will male lactation become another such counter-evolutionary choice?