But we always find the patterning. We know of no culture that has said, articulately, that there is no difference between men and women except in the way they contribute to the creation of the next generation; that othewise in all respects they are simply human beings with varying gifts, no one of which can be exclusively assigned to either sex. We find no culture in which it has been thought that all identified traits—stupidity and brilliance, beauty and ugliness, friendliness and hostility, initiative and responsiveness, courage and patience and industry—are merely human traits. However differently the traits have been assigned, some to one sex, some to the other, and some to both, however arbitrary the assignment must be seen to be (for surely it cannot be true that women's heads are both absolutely weaker —for carrying loads—and absolutely stronger—for carrying loads—than men's), although the division has been arbitrary, it has always been there in every society of which we have any knowledge.

So in the twentieth century, as we try to re-assess our human resources, and by taking thought to add even a jot or a tittle to the stature of our fuller humanity, we are faced with a most bewildering and confusing array of apparently contradictory evidence about sex differences. We may well ask: Are they important? Do real differences exist, in addition to the obvious anatomical and physical ones—but just as biologically based—that may be masked by the learnings appropriate to any given society, but which will nevertheless be there? Will such differences run through all of men's and all of women's behaviour? Must we expect, for instance, that a brave girl may be very brave but will never have the same kind of courage as a brave boy, and that the man who works all day at a monotonous task may learn to produce far more than any woman in his society, but he will do it at a higher price to himself? Are such differences real, and muse we take them into account? Because men and women have always in all societies built a great superstructure of socially defined sex differences that obviously cannot be true for all humanity—or the people just over the mountain would not be able to do it all in the exactly opposite fashion—must some such superstructures be built? We have here two different questions: Are we dealing not with a must that we dare not flout because it is rooted so deep in our biological mammalian nature that to flout it means individual and social disease? Or with a must that, although not so deeply rooted, still is so very socially convenient and so well tried that it would be uneconomical to flout it—a must which says, for example, that it is easier to get children born and bred if we stylize the behaviour of the sexes very differently, teaching them to walk and dress and act in contrasting ways and to specialize in different kinds of work? But there is still the third possibility. Are not sex differences exceedingly valuable, one of the resources of our human nature that every society has used but no society has as yet begun to use to the full?
-------------------------------------------------------------

We speak in our current folk-language of the beast in man, of the thin veneer of civilization, and either statement simply means that we do not trust mankind to be continuously human.

For our humanity rests upon a series of learned behaviours, woven together into patterns that are infinitely fragile and never directly inherited. The ant we discover imbedded in a block of Baltic amber, which the geologist dates 20,000,000 years ago, may be trusted to reproduce its typical ant behaviour wherever it can survive. It is trustworthy for two reasons: first, because its complex behaviour, by which its society will be divided into minute castes that carry out predetermined tasks, is built into the very structure of its body, and second, because even if it should learn something new, it cannot teach it to the other ants. The repetitious pattern in which countless generations of a single species repeat a pattern more complicated than the dreams of a technocratic Utopia is protected by these two circumstances: behaviour imbedded in physical structure and an inability to communicate new learnings. But man does not even carry the simplest forms of his behaviour in such a way that a human child without other human beings to teach it can be relied upon to produce spontaneously a single cultural item. Long before his small fist is strong enough really to deliver a blow, the angry gestures of the human child bear the stamp not of his long mammalian past, but of the club-using or spear-throwing habits of his parents. The woman left alone to bear her child calls not upon some reliable instinctive pattern that will guide her through the complexities of cutting the umbilical cord and cleansing the new infant from the traces of birth, but fumbles helplessly among bits of folk-lore and old wives' tales that she has overheard. She may act from the memory of what she has seen animals do, but in her own living nature she finds no reliable cues.

We may cherish our noses or our lips, our relatively hairless bodies, our graceful arms and cunning hands, but when we shrink from some human deformity that makes a human being look more like an animal, when we shudder away from members of other races and identify them by the particular ways in which they may be felt to be more animal than we—for example, the thin lips and hairiness of the Caucasian, the bridgeless noses of some Mongolians, the pigmentation of the Negroid type—under the manifest fear of miscegenation lies the knowledge that all forms of cultural behaviour can be lost, that they are dearly purchased and dearly kept. Whenever men's fear is expressed in social terms—in great group rituals in which the sun is sought to shine again, in the Balinese New Year when all men are quiet for a day that life may flow on, among the Iroquois when once a year men lived out their dreams, confessed their sins, and plunged naked into icy rivers at the behest of a dream—the expression of the fear becomes also a way of assuaging it. These rituals are ways of restating that only together can men be human, that their humanity depends not on individual instinct, but on the traditional wisdom of their society. When men lose this sense that they can depend upon this wisdom, either because they are thrown among those whose behaviour is to them no guarantee of the continuity of civilization or because they can no longer use the symbols of their own society, they go mad, retreating slowly, often fighting a heart-breaking rear action as they relinquish bit by bit their cultural inheritance, reamed with such difficulty, never learned so that the next generation is safe.

This is a fear that may indeed be counted to men for wisdom and not chalked up under some attribution of peculiar irrationality. It is so deep that it may include the most minute and the most irrelevant actions. The smallest details of manners—which food is eaten, when it is eaten, with whom and on whatshaped plates—may become the conditions on which man feels his humanity is held. In caste societies, such as India or the Southeastern United States, where one's culturally defined humanity is inextricably bound up with membership in a caste group, to associate with the other caste in forbidden ways means the loss of one's humanity itself. The sense of membership in one's own sex is deeply imbued with such attitudes. So the Cossack woman in Sholokhov's novel, prying into the ways of the strange Turkish woman brought among the Cossacks, reports: "I saw them myself. She wears trousers.... When I saw them my blood ran cold." In cultures where tablemanners are the insignia of humanity, people may be unable to eat their food at the table with some one who eats different—especially if the table-manners are also class- or caste-marked so that the presence of one who eats differently automatically classifies one with that ocher. The strong men of western Europe feel unmanned when they meet people from eastern Europe, where men squat to urinate, and in modern Australia women feel strangely embarrassed when American women tell their husbands to fetch the cocktails. Each little courtesy to or restraint or deference from others is cherished for what it is, something laboriously learned and easy to lose. Against such a background we can look at the arrangements that surround the relationships between the sexes which have been essential to the preservation of human society. Beneath the thousand fleeting and irrelevant symbols —the lifted hat of the gentleman, the downcast eyes of the lady, the geranium on the window-sill of the German burgher, or the scrubbed white steps of the mill workers of the English Midlands—is there a core of common practice to which all societies everywhere have had to cling in order to keep the dearly bought, learned aspects of their humanity?

When we survey all known human societies, we find everywhere some form of the family, some set of permanent arrangements by which males assist females in caring for children while they are young. The distinctively human aspect of the enterprise lies not in the protection the male affords the females and the young—this we share with the primates. Nor does it lie in the lordly possessiveness of the male over females for whose favours he contends with other males—this too we share with the primates. Its distinctiveness lies instead in the nurturing behaviour of the male, who among human beings everywhere helps provide food for women and children. The sentimental figures of speech so common in the modern Western world in which the bees and the ants and the flowers are invoked to illustrate the more suspect aspects of human beings have obscured our recognition of how much of an invention this behaviour of of human males is. True, father-birds do feed their young, but men are a long way from birds on the evolutionary tree. Male fighting-fish do make bubble-nests and only capture the female long enough to squeeze her eggs out of her, and then, after driving her away, devote themselves, rather unsuccessfully, to retrieving the eggs that fall out of the bubble-nest, and—when they don't eat up the eggs or the young—some young survive. But these analogies from the world of birds and fish are far from man. Among our structurally closest analogues—the primates—the male does not feed the female. Heavy with young, making her way laboriously along, she fends for herself. He may fight to protect her or to possess her, but he does not nurture her.

Somewhere at the dawn of human history, some social invention was made under which males started nurturing females and their young. We have no reason to believe that the nurturing males had any knowledge of physical paternity, although it is quite possible that being fed was a reward meted out to the female who was not too fickle with her sexual favours. In every known human society, everywhere in the world, the young male learns that when he grows up, one of the things which he must do in order to be a full member of society is to provide food for some female and her young. Even in very simple societies, a few men may shy away from the responsibility, become tramps or ne'er-do-wells or misanthropists who live in the woods by themselves. In complex societies, a large number of men may escape the burden of feeding females and young by entering monasteries—and feeding each other—or by entering some profession that their society will classify as giving them a right to be fed, like the Army and the Navy, or the Buddhist orders of Burma. But in spite of such exceptions, every known human society rests firmly on the learned nurturing behaviour of men.
-------------------------------------------------------------

So the three themes, complementariness, reciprocity, and symmetry, weave in and out of the long learning process, interpenetrating and informing each other, until one side of the complementariness may be so emphasized as to become a form of symmetrical behaviour, and with difference in age to provide the only asymmetry—as among the Arapesh, where husbands are ideally much older than their wives— because their receptivity and responsiveness are so much stressed. Or the assertive, invasive side of the suckling relationship may become dominant for both mother and child, with both sexes becoming assertive and demanding. Through the body, the ways of the body are learned.
-------------------------------------------------------------

In any human group it is possible to arrange men and women on a scale in such a way that between a most masculine group and a most feminine group there will be others who seem to fall in the middle, to display fewer of the pronounced physical features that are more characteristic of one sex than of the other. This is so whether one deals entirely in secondary sex characters, such as arrangement of pubic hair, beard, layers of fat, and so on, or whether one deals with such primary sex characters as breasts, pelvic measurements, hip-torso proportions and so on. These differences are even more conspicuous when one considers such matters as skin sensitivity, depth of voice, modulation of movement. Also, one finds in most groups of any size that there are very few individuals who insist on playing the role of the opposite sex in occupation or dress or interpersonal sex activities. Whether full transvestitism will occur seems to be a question of cultural recognition of this possibility. Among many American Indian tribes the berdache, the man who dressed and lived as a woman, was a recognized social institution, counterpointed to the excessive emphasis upon bravery and hardiness for men. In other parts of the world, such as the South Pacific, although a large number of ritual reversals of sex on ceremonial occasions may occur, there are many tribes where there is no expectation that any single individual will make the complete shift. Peoples may provide sexreversal roles for both sexes-as among the Siberian aborigines, where sex reversal is associated with shamanism; they may permit it to men but deny it to women; or they may not provide any pattern at all.
-------------------------------------------------------------

The solution of the Oedipus situation will depend a great deal of the way in which a boy or a girl accepts primary sex membership. But it is not enough for a child to decide simply and fully that it belongs to its own sex, is anatomically a male or a female, with a given reproductive role in the world. For growing children are faced with another problem: "How male, how female, am I?" He hears men branded as feminine, women condemned as masculine, others extolled as real men, and as true women. He hears occupations labelled as more or less manly, for a man, or more or less likely to derogate her womanhood, for a woman. He hears types of responsiveness, fastidiousness, sensitivity, guts, stoicism, and endurance voted as belonging to one sex rather than the other. In his world he sees not a single model but many as he measures himself against them; so that he will judge himself, and feel proud and secure, worried and inferior and uncertain, or despairing and ready to give up the task altogether.
-------------------------------------------------------------

A society can provide elaborate roles that will attract many individuals who would never spontaneously seek them. Fear that boys will be feminine in behaviour may drive many boys into taking refuge in explicit femininity. Identification of a little less hairiness on the chin, or a slightly straighter bust line, as fitting one for membership in the opposite sex may create social deviance. If we are to interpret these experiences, which all children have, we look for some theory of what these differences mean.

We strip away all this superstructure when we have invoked the presence or absence, the recognition and toleration, of transvestite social institutions, or the explicit suppression of homosexual practice, but we still find differences that need explanation. After we have gathered together the insights from detailed case-histories in Western society that show how accidents of upbringing, faulty identifications with the wrong parent, or excessive fear of the parent of the opposite sex may drive both boys and girls into sexual inversion, still we are left with a basic problem. Set end to end, standing in a line, the men of any group will show a range in explicit masculinity of appearance as well as in masculinity of behaviour. The females of any group will show a comparable variety, even more, in fact, if we have X-ray pictures to add to their deceptive pelvic profiles, which do not reveal their feminine reproductive capacities accurately. Is this apparent range to be set down to differences in endocrine balance, set against our recognition that each sex depends for full functioning upon both male and female hormones and the interaction between these hormones and the other endocrines?? Has every individual a bisexual potential that may be physiologically evoked by hormone deficit or surplus, which may be psychologically evoked by abnormalities in the process of individual maturation, which may be sociologically invoked by rearing boys with women only, or segregating boys away from women entirely, or by prescribing and encouraging various forms of social inversion? When human beings—or rats— are conditioned by social circumstances to respond sexually to members of their own sex as adults and in preference to members of the opposite sex, is this conditioning playing on a real bisexual base in the personality, which varies greatly in its structure as between one member of a group and another?

At first blush, it seems exceedingly likely that we have to advance some such hypothesis. If one looks at a group of little boys, it would seem fairly obvious that it would be easier to condition those who now appear "girlish" to an inverted role, and that from a group of little girls, the "boyish" girl would be the easiest to train into identification with the opposite sex. And does not "easiest" here mean the greatest degree of physical bisexuality?? Yet the existing data make us pause. The most careful research has failed to tie up endocrine balance with actual homosexual behaviour. Those rare creatures who have both male and female primary sex organs present of course major anomalies and confusions, but so far they have thrown little light on the general problem. The extraordinary lack of correlation between physique that can be regarded as hypermasculine and hyperfeminine and successful reproductivity is marked in every group. The man who shows the most male characteristics may have no children, while some pallid, feminine-looking mouse of a man fathers a large brood. The woman with ample bosom and wide hips may be sterile, or if she bears children she may be incapable of suckling them. Yet we are still continually confronted with what looks like a correlation between the tendency towards sexual inversion of the men and women who deviate most towards the expected physique of the opposite sex. In the primitive tribe that does not recognize inversion, the boy who decides to make mats will look more like the female type for that tribe, the woman who goes out hunting will tend to look more like the male. Does this apparent physical correspondence mean nothing, is it sheerly an accident within a normal range of variation? If the tribe sets hairiness up as a desirable male characteristic, will the less hairy become confused about their sex role, while if the tribemen think that hairiness is simply a brutish characteristic, the very hairy may be almost sexually ostracized and the most hairless will not thereby be regarded as less male? This would be the extreme environmental answer, while the invocation of some very subtle, as yet unplumbed structural and functional variation in the biological basis of sex membership would be the extreme genetic answer.

I suggest another hypothesis that seems to me to fit better the behaviour of the seven South Seas peoples whom I have studied. A Balinese male is almost hairless—so hairless that he can pluck his whiskers out one by one with pincers. His breasts are considerably more developed than are a Westerner's. Almost any Balinese male placed in a series of western-European males would look "feminine." A Balinese female, on the other hand, has narrow hips and small high breasts, and almost any Balinese female placed among a group of westem-European women would look "boyish." Many of them might be suspected of being unable to suckle children, perhaps accused of having infantile uteruses. But should these facts be interpreted to mean that the Balinese is more bisexual, less sexually differentiated, than the western European, that the men are less masculine, the women less feminine, or simply that the Balinese type of masculinity and femininity is different? The extreme advocates of a varying bisexual balance would claim that in some races the men are less differentiated, are more feminine and so on, than in others, and might also apply the same argument to the women. But on the whole, it would be agreed that at least some of the respects in which a Balinese male would seem feminine are matters that do not really affect his masculinity at all: his height, girth, hairiness, and the like. So it might be fairly readily admitted that as between racial strains that vary as greatly as Balinese and northern Europeans, Andamanese pygmies and Nubian giants, not only would certain of the criteria for masculinity and femininity be inoperative, but also that actual cross-correspondence might occur, as all Andamanese males would fall within the height range for females in some much taller group.

But all human groups of which we have any knowledge show evidence of considerable variation in their biological inheritance. Even among the most inbred and isolated groups, very marked differences in physique and apparent temperament will be found, and despite the high degree of uniformity that characterizes the child-rearing practices of many primitive tribes, each adult will appear as more or less masculine, or more or less feminine, according to the standards of that particular tribe. There will be, furthermore, orders of variation that seem, at least on inspection —for we have no detailed records—to apply from one group to another. Although almost every Balinese would fall within the general configuration that might be classified technically as asthenic, yet the asthenic Balinese continues to contrast with the Balinese who is heavier in bony structure, or shorter and plumper. Within the limits set by the general type, these same differences occur, in both men and women. Not until we have far more delicate methods of measurement, which allow not only for individual constitution but for ancestral strains, will we have any way of knowing whether there is any genuine correspondence, on a behavioural level, among the slender, narrow-bodied of the Arapesh, Tchambuli, Swede, Eskimo, and Hottentot, or whether their behaviour, although possibly in some way constitutionally based, is still in no way referable to something they may be said to have in common. Until such measures are developed, and such studies made, one can only speculate on the basis of careful observation, with no better instru ment for comparison than the human eye. But use of this instrument on seven different peoples has suggested to me the hypothesis that within each human group we will find, probably in different proportions and possibly not always in all, representatives of the same constitutional types that we are beginning to distinguish in our own population. And I further suggest that the presence of these contrasting constitutional types is an important condition in children's estimate of the completeness of their sex membership.

If we recognized the presence of comparable ranges of constitutional types in each human society, any single continuum that we now construct from the most masculine to the least masculine can be seen to be misleading, especially to the eye of the growing child. We should instead define a series of continuums, distinguishing between the most masculine and the least sexually differentiated male within each of these several types. The slender little man without beard or muscle who begets a whole brood of children would not then seem such an anomaly, but could be regarded as the masculine version of a human type in which both sexes are slender, small, and relatively hairless. The tall girl whose breasts are scarcely discernible, but who is able to suckle her baby perfectly satisfactorily as her milk seems to spread in an almost even line across her chest, will be seen not as an imperfectly developed female—a diagnosis that is contradicted by the successful way in which she bears and suckles children, and her beautiful carriage in pregnancy—but as the female of a particular constitutional type in which women's breasts are much smaller and less accentuated. The big heman with hair on his chest, whose masculinity is so often claimed to be pallid and unconvincing, will be seen to be merely a less masculine version of a type in which enormous muscularity and hairiness are the mode. The woman whose low fertility contrasts so strangely with her billowing breasts and hips may be seen as only one of a type of woman with very highly emphasized breasts and hips—her low fertility only conspicuous because most of the women with whom she is co mpared have smaller bosoms and less full hips. The apparent contradiction between pelvic X-rays and external pelvic measurements might also be resolved if it were considered from this point of view.

And as with physical type, so with other aspects of personality. The fiery, initiating woman would be classified only with fiery, initiating men of her own type, and might be found to look like not a lion, but merely like a lioness in her proper setting. When the meek little Caspar Milque toast was placed side by side not with a prize-fighter, but with the meekest female version of himself, he might be seen to be much more masculine than she. The plump man with soft breasttissue, double chin, protruding buttocks, whom one has only to put in a bonnet to make him like a woman, when put beside the equally plump woman will be seen not to have such ambiguous outlines after all; his masculinity is still indubitable when contrasted with the female of his own kind instead of with the male of another kind. And the slender male and female dancers, hipless and breastless, will seem not a feminine male and a boyish female, but male and female of a special type. Just as one would not be able to identify the sex of a male rabbit by comparing its behaviour with that of a lion, a stag, or a peacock as well as by comparing rabbit buck with doe, lion with lioness, stag with doe, and peacock with peahen —so it may well be that if we could disabuse our minds of the habits of lumping all males together and all females, the pattern of male and female roles may make them roles that can be played. In every such tightly patterned picture there will be some who rebel, will commit suicide—if suicide is a culturally recognized way out—will become promiscuous or frigid or withdrawn or insane, or, if they are gifted, will become innovators of some variation in the pattern. But most of them will learn the pattern, alien though it be.

So in each of the societies I have studied it has been possible to distinguish those who deviated most sharply from the expected physique and behaviour, and who made different sorts of adjustment, dependent upon the relationship between own constitutional type and cultural. ideal The boy who grow up into a tall, proud, restive man whose very pride makes him sensitive and liable to confusion suffers a very different fate in Bali, Samoa, Arapesh and Manus. In Manus, he takes refuge in the vestiges of rank the Manus retain, takes more interest in ceremonial than in trading mixes the polemics of acceptable trading invectives with much deeper anger. In Samoa such a man is regarded as too violent to be trusted With. the headship of a family for many, many years; the village waits until his capacity for anger and intense feeling has been worn down by years of erosive soft resistance to his unseemly over-emphases. In Bali, such a man may take more initiative than his fellows only to be thrown back into sulkiness and confusion, unable to carry it through. Among the Maori of New Zealand, it is probable that he would have been the cultural ideal, his capacity for pride matched by the demand for pride, his violence by the demand for violence, and his capacity for fierce gentleness also given perfect expression, since the ideal woman was as proud and fiercely gentle as himself.

But in complex modern societies, there are no such clear expectations, no such perfectly paired expectancies, even for one class or occupational group or rural region. The stereotyped roles for men and for women do not necessarily correspond, and whatever type of man is the ideal, there is little likelihood that the corresponding female type will also be the ideal. Accidents of migration, of crossclass marriage, of frontier conditions, may take the clues for the female ideal from quite another type from which the male ideal is taken. The stereotype may itself be blurred and confused by several different expectations, and then split again, so that the ideal lover is not the ideal brother or husband. The pattern of inter relationships between the sexes, of reserve or intimacy, advance or retreat, initiative and response, may be a blend of several biologically congruent types of behaviour instead of clearly related to one, We need much more material on the extent to which this sort of constitutional types may actually be identified and studied before we can answer the next questions about the differential strength and stability and flexibility of cultures in which ideals are a blend, or a composite, or a single lyric theme, ideals that are so inclusive that every male and female finds a rather blurrily defined place within them, or so sharp and narrow that many males and females have to develop counterpointed patterns outside them.

A recognition of these possibilities would change a great deal of our presentday practices of rearing children. We would cease to describe the behaviour of the boy who showed an interest in occupations regarded as female, or a greater sensitivity than his fellows, as "on the female side, and could ask instead what kind of male he was going to be. We would take instead the primary fact of sex membership as a cross-constitutional classification, just as on a wider scale the fact of sex can be used to classify together male rabbits and male lions and male deer, but would never be permitted to obscure for us their essential rabbit, lion, and deer characteristics. Then the little girl who shows a greater need to take things apart than most