‘‘All this cell, brain, and organ development stuff is fascinating,’’ a frustrated parent might say to me. ‘‘But I still want to know why my little boy rushes around shooting imaginary laser guns, while my little girl prefers jump rope.’’ Many Loveweb participants raise similar challenges, citing studies showing that gender differences appear at an early age—surely, they believe, an argu­ment for inborn difference. How can I reconcile the observations of countless parents and the multitude of studies by sociologists and developmental psy­chologists with a systems approach to gender acquisition? Here I fit together already existing pieces of the puzzle.

‘‘Gender,’’ argue some sociologists, ‘‘is a situated accomplishment. . . not merely an individual attribute but something accomplished in interaction with others.’’46 Both children and adults learn through direct feedback from

others to ‘‘do gender.’’47 Classmates, parents, teachers, and even strangers on the street evaluate how a child dresses. A boy who wears pants conforms to social norms, while one who dons a skirt does not. And he hears about it right away! Gender, then, is never merely individual, but involves interactions between small groups of people. Gender involves institutional rules. If a gay man made up as a woman walks down the street, he soon learns that he has deviated from a gender norm. The same man in a gay bar will receive compli­ments as he partakes in a subculture that plays by a different set of guidelines. Furthermore, we ‘‘do gender’’ as part of ‘‘doing difference.’’ We establish identities that include race and class as well as gender, and we do gender differently depending upon our location in racial and class hierarchies.48

In America and Europe, boys and girls begin to behave differently during the preschool years. By middle school each group thinks the other has ‘‘coo­ties,’’ but during the years of hormonal hell, they return to each other for sex and socializing. As adults they live and work in overlapping but gender-divided institutions, and as old people they are separated once more, this time by the differential death rates for men and women. Developmental psychologists, sociologists, and systems theorists have some tantalizing findings about how children acquire gender, although obtaining similar information for the rest of the life cycle remains for future scholars.49

Traditionally, psychology has offered three approaches to understanding gender development: Freudian psychodynamics, social learning, and cogni­tive development. For Freud, the child’s own awareness of his or her genitals produces erotic fantasies, which in turn lead to identification with a suitable adult figure and the development of an appropriate gender role.50 Social learn­ing proponents focus on adult awareness of an infant’s genitals, which leads to differential reinforcement, the offer of gender-appropriate models, and thus the development of gender role and identity.51 Cognitive theory also starts with others’ awareness of a child’s genitals. This leads to labeling and thence to gender identity and finally to the acquisition of an appropriate gender role.52 Feminist social scientists have used each of these paradigms to produce infor­mation about the development of sexual difference. A primary goal in the past has been to produce better accounts of female development, since in their original forms all three theories primarily produced narratives about how boys became men. More recently, however, a number of feminist voices have begun to challenge the very structure of the field, calling for more complex accounts of difference and a return to the study of male-female similarities.53 Here I depend especially on the work of cognitive and social learning re­searchers. Regardless of the particular approach, the goal remains that of un­derstanding the development of the self: ‘‘behavior, experience, and identifi­cations, including sexual desire and object choice, [that] are relatively stable or fixed or that, at least, . . . [are] a basic or primary ‘core’ of identity.’’54

Gender and sexuality often appear to us as universal features of human existence. Need such apparent universality mean that human sexuality and gender are inborn and only superficially shaped by social experience? We can see that this is the wrong way to ask the question by looking at the develop­ment of another apparently universal human behavior: smiling.55 Newborns have a simple smile: the face relaxes while the sides of the mouth stretch out­ward and up. An identical ‘‘smile’’ has been seen in fetuses as young as twenty – six weeks of gestation. This suggests that, initially, a basic set of neural con­nections develops that enables a developing human to ‘‘smile’’ as a reflex, even in utero. In the newborn, smiling occurs spontaneously in rapid eye move­ment (REM) sleep states, but at first does not function as a mode of emo­tional expression.

By two weeks after birth, smiles begin to appear infrequently when the baby is awake, and more body parts are recruited into the event. The lips curl up farther, ‘‘cheek muscles contract, and the skin around the eyes wrinkles.’’ Three-month-old babies smile much more frequently when awake, and they do so in nonrandom bursts, in response to stimuli in the environment. By the time an infant is from half a year to two years old, smiling blends with a wide variety of other facial expressions—surprise, anger, excitement. Further­more, the facial expressions have become both more complex and individually varied. Accompanying the smile may be ‘‘nose wrinkles, jaw drops, blinks, blows, and brow raises that served to communicate affects from pleasure to mischief.’’56 Thus, over two years, smiling changes in shape (and all that shape implies in terms of muscle and nerve recruitment), timing, and connection to other expressive actions. A smile is not a smile is not a smile (to butcher Gertrude Stein a bit).

At the same time that the muscles and nerves that govern smiling develop and become more complex, so too do the functions and social contexts that elicit smiling. While at birth, drowsiness and a decrease in sensory input elic­its smiling, soon infants respond by smiling at familiar voices and sounds, and less regularly to touch. By six weeks, a baby smiles mostly while awake, in response to visual cues. By three to six months, a baby is more likely to smile at its mother than at inanimate objects, and by the end of the first year ‘‘smil­ing serves a variety of communicative functions, including the intent to flirt or do mischief.’’57 At first blush, smiling seems to be a simple reflexive re­sponse, but over time it changes in complex ways—in terms of the nerves and muscles involved, but also in terms of what social situations elicit smiling and how the child uses smiling as part of a complex system of communication— with other humans. Thus a physiological response becomes ‘‘socialized’’ not only in terms of intentional use but also in terms of the actual body parts (which nerves and muscles are used and what stimulates them) themselves.

Looking at the smile response as a developmental system enables us to exchange meaningless claims such as that ‘‘smiling is inborn and genetic’’ for carefully designed experimental studies ‘‘that systematically vary the condi­tions. . . which. . . may influence the form, timing and function of smiling’’ over different parts of the life cycle.58 The psychologist Alan Fogel and his colleagues have used their studies of the smile response to develop what they call a dynamic systems perspective on emotion.59 First they argue that emo­tions are relational rather than individual. Young infants, for example, smile in response to other people or objects. Second, they view emotions as self­organizing, stable systems. But stability does not imply permanence. Thus visual induction of the smiling response is stable in infants for three to four months, but is eventually replaced by a new stable system involving a variety of forms of physical interactions with its mother (or other caretaker).60

Little if any of the work on dynamic developmental systems has made its way into the study of human sexual development, but its applicability seems obvious. First, we need to stop looking for universal causes of sexual behavior and gender acquisition and instead learn more about (and from) individual difference. Second, we need to think harder about how to study sex and gen­der as part of a developmental system. Third, we need to become more imag­inative and specific about what we mean by the word environment. At the moment I think we are pretty clueless about the environmental components of human sexual development, but the idea provided by Fogel and others— that behaviors go through periods of instability (when they are more easily changed) and stability (when they seem fixed)—is helpful.

We do have some starting points. Since the mid-1980s, several groups of developmental psychologists have asked a set of interrelated questions about gender: What do children know about sex (the body parts), and when do they know it? Does such knowledge correlate with or affect gender-related behaviors such as differing patterns of play? A story outline has begun to emerge.61 Psychologists have introduced the concept of a schema or schematic processing, which enable children to use rudimentary knowledge to make choices about ‘‘appropriate’’ play, peers, and behaviors. According to this line of thought, children adopt particular sex roles as they integrate their own sense of self with their developing gender schema, a process—like the devel­oping smile—that takes several years. It is a reasonable (and testable) guess that during this time certain forms of bodily gender expression (such as ‘‘throwing like a girl’’) develop stability. But—also like the developing smile—stability need not suggest permanence, as observing top girl Little Leaguers would make quite clear.

Anybody who has observed a young child as he or she learns about the world has seen schema in operation. I remember, for example, when my tod­dler niece pointed at a clock with a schematic outline of an owl’s face. ‘‘Owl,’’ she proudly pronounced. I recall being amazed that she could recognize such a featureless representation when her storybooks all showed detailed drawings of these nocturnal birds. But she had internalized an owl schema, which en­abled her to recognize this bird on the basis of minimal information. Beverly Fagot and her colleagues studied gender schema in children ranging in age from і.75 to 3.25 years. They gave the kids a ‘‘gender task’’—to correctly classify pictures of adults and children as ‘‘mommy,’’ ‘‘daddy,’’ ‘‘boy,’’ or ‘‘girl.’’ The younger children (those averaging about two years old) could not pass the test—that is, they apparently had no working concept of gender. The older children, however, (those averaging about 2.5 years), correctly classified both adults and children. Furthermore, those children who had developed boy-girl labels behaved differently from those who had not. The older kids, for example, preferred same-sex play groups, and girls who passed the label­ing test were less aggressive.62

Fagot and Leinbach also observed і.5-year-old kids at home. At this age they could neither pass gender-labeling tests nor engage in sex-typed play. By the time the children were 2. 25 years old about half, called early labelers, could accurately label boys and girls. Two differences emerged between the early and late labelers. First, ‘‘parents of future early labelers gave more posi­tive and negative responses to sex-typed toy play’’ and, by 2.25 years, ‘‘early labelers showed more traditional sex-typed behavior than late labelers.’’63 By age 4, early and late labelers did not differ in preference for sex-stereotyped play. The early labelers maintained a greater awareness of sex stereotypes, however. Fagot and colleagues conclude that ‘‘the child’s construction of a gender schema reflects back the behavioral, cognitive, and affective dimen­sions of the familial environment.’’64

I used to ride my bike to grade school, ruminating as I traveled the subur­ban New York landscape. For a time one problem in particular held my atten­tion. I knew that boys had short hair, girls had long hair, and babies were born bald. How, I puzzled, did adults have the awesome power to declare immediately the sex of a newborn? I knew about genitalia, of course. I had an older brother, and we bathed together until I was four or five. Occasionally, also, I caught a glimpse of my father in the altogether. But I never connected such information to my puzzlement with birth announcements. Then, one day when I was about ten years old, biking home from school, the answer just popped into my head. ‘‘So that’s how they know,’’ I thought. As I look back now, through feminist theory-fogged scrim, I realize that as a child, gender had been clear on my horizon many years before sex became visible.65

I was not alone in my confusion, just a bit slow to resolve it. In America, at least, small children seem to base their initial, rudimentary gender schema on cultural markers of gender, not knowledge of genital differences. In one study, the psychologist Sandra Bem showed 3-, 4- and 3-year-olds photo­graphs of either a naked boy or a naked girl and then of the same child dressed either in girls’ or boys’ clothing. Children younger than three had a hard time labeling the naked children as a boy or a girl, but successfully used social clues—clothes and hairstyles—to classify the dressed ones.66 About 40 per­cent of the 3-, 4- and 3-year-old children accurately identified sex in all the photos once they had knowledge of genitalia. The rest, however, had not yet acquired a notion of sex constancy—that is, they used gender signals such as hairstyle and clothing to decide who was a boy and who was a girl. This also meant that some of these children believed that they could become the oppo­site sex by dressing as one. Their own gender identity was not yet fixed.

Children’s understanding of anatomical constancy didn’t seem to affect sex role preferences. Instead, early gender schema proved critical. ‘‘First, chil­dren learned to label the sexes, and only later did they show strong sex-typed toy and peer preferences and knowledge about sex differences in toys and clothing.’’ Even though children did not need a concept of sex stability to develop sex-stereotyped preferences, having such knowledge strengthened the level of such preferences. It may be that ‘‘children who can label the sexes but do not understand anatomical stability are not yet confident that they will always remain in one gender group.’’67 In keeping with the above findings, older children (aged 6 to 10 years) make more extreme stereotypic gender judgments than do younger ones. Not surprisingly, they first learn to associate characteristics relevant to their own sex and only later stabilize their expecta­tions of the other sex (see figure 9.3).68