Causal Cascades and Causal Thickets
Gender, Nature, and Nurture: Looking to the Future
The developmental tracks portrayed in Fig. 7.1 constantly interact with one another, often in complex ways. They form causal thickets, hard-to-analyze tangles of influences that interact via many interlocking feedback loops. Consider the following examples. Genes held in common by parents and children influence how parents treat their children and also how children respond to parents (Tracks 1 and 2 interact). Biological predispositions in girls and boys foster sex segregation, and conversely, sex segregation amplifies biological predispositions in girls and boys (Tracks 1 and 3 interact). Parental socialization molds the ways children interact with their peers (Tracks 4 and 5 interact). Peer influences determine which TV shows children watch and the resulting gender messages children take from TV (Tracks 2 and 3 interact). Parent and teacher stereotypes influence the educational choices of boys and girls, which then influence their subsequent occupational choices (Tracks 2 and 4 interact). Work and educational settings influence individuals' gender stereotypes, stress levels, and even hormone levels (Tracks 1, 4, and 5 interact). The list of possible interactions goes on without end, with feedback loops swirling in all direction, all of them inextricably intertwined.
Gender as a Complex Causal Cascade
The interweaving developmental processes'—the causal thickets-portrayed in Fig. 7.1 suggest three major conclusions:
1. It is often hard to partition the overall causes of gender into two clear categories labeled nature and nurture.
2. On a practical level, changing any single causal factor in gender development may produce at best modest effects, if all the other factors that create and sustain gender remain in place. And predicting the effect of a change in any single factor is often difficult, because its effects may ripple through the total system in unexpected ways.
3. On a more theoretical level, the whole of gender development is often greater than the sum of its parts.
Stanford University psychologist Eleanor Maccoby (1998) provided a concrete example of the emergent complexity of gender development when she discussed the relationship between family gender socialization and childhood sex segregation. Maccoby proposed that early parental socialization—fathers' high levels of rough-and-tumble play with their sons and parents' high levels of verbal discussion with daughters—may have increasingly powerful consequences as children increasingly interact with same-sex peers.
Although Maccoby acknowledged that biological factors may influence boys' and girls' playstyles. innate readiness always requires environmental stimulation to show itself. Nature needs nurture and nurture needs nature, according to this point of view. Maccoby's proposed causal cascade can be summarized as follows. Parental socialization of boys and girls, in interaction with biological predispostions, leads boys and girls to interact in distinctive ways with their peers, and this in turn fosters sex segregation and the development of distinctively different boy cultures and girl cultures.
Here are some additional causal cascades that may contribute to the development of gender:
· Cascade 1: Both biological predispositions ana early social learning lead to sex-typed toy preferences in children. These toy preferences in turn lead to sex differences in child-parent and child-peer interactions and to the development of different motor skills and cognitive abilities in boys and girls. Ultimately, this cascade affects the classes children take in school and the occupations they choose as adults.
· Cascade 2: Genetic predispositions influence boys' and girls' play-styles, which influence children's preference for male or female playmates. Most boys prefer to play with other boys, but some prefer instead to piay with girls. Similarly, most girls prefer to play with other girls, but some prefer instead to piay with boys. Playing in largely same-sex versus opposite-sex groups influences individuals' attributions of arousal, their developing erotic reactions to peers, and ultimately, their adult sexual orientation. (See D. J. Bem, 1996, 2000, for a more complete description of this "exotic become erotic" theory of sexual orientation.)
· Cascade 3: Adults' beliefs about boys' and girls' math abilities affect children's self-concepts and feelings of self-competence regarding math, which then influence the classes boys and girls take, which ultimately influence later choices of college majors and adult careers.
· Cascade 4: Biological predispositions, doll play, and mass media influences lead girls to be more interested in babies than boys are. As a result, girls learn more about babies, spend more time with them, and develop the skills needed to care for babies and young children. Parents and neighbors foster this early bias by often assigning girls the task of babysitter and surrogate parent. After marriage and child-birth, both men and women agree that women are naturally more suited to caring for babies than men are, and in family life, women-even full-time working women—assume much more responsibility for child care than men do.
· Cascade 5: Parental treatment, sex-typed grooming, physical cues such as body shape and voice pitch, and constant social classification by gender lead toddlers to quickly learn the categories of male and female. Children readily apply these labels to themselves and to their peers, and they use these labels to organize gender-related behaviors that they observe in themselves and others. After achieving accurate gender labeling, children exaggerate the sex differences they perceive in others and they develop in-group feelings toward their own sex and out-group feelings toward the other sex. With the internalization of gender standards that occurs between ages 3 and 4 years, perceived differences between the sexes are transformed into moral imperatives. Then, children not only believe that boys and girls are different, but that they should be different.
Causal Cascades and the Nature—Nurture Debate
When applied to the nature-nurture debate, the notion of a causal cascade raises a centra! question. In a complex, interacting, dynamic, causal system, like that portrayed in Fig. 7.1, is it ever possible to partition the causes of any particular gender-related behavior exclusively into one of two simple and mutually exclusive categories: nature or nurture? The answer suggested by Fig. 7.1 is, probably not.
Why not? One reason is that causes are rarely pure in the sense that they have just biological or just environmental antecedents. For example, gene expression (a seemingly biological cause) is influenced by both DNA codes and environmental factors (cellular environments, uterine environment, external stressors). Social causes, such as parental treatment and peer influences, are influenced by both biological factors (e.g., parents' genes and peers' X and Y chromosomes) and social factors (e.g., gender roles and cultural traditions). An individual's choice of settings (e.g., a child's choice of male or female playmates) is genetically as well as socially influenced. Individuals' levels of sex hormones can be influenced by environments (by stress, by success, by the presence or absence of members of the opposite sex) as well as by their sex chromosomes.
Assigning causes to nature or to nurture depends, in part, on how far back you want to look in the causal chain. It depends on the particular developmental instant at which you take your causal snapshot; for example, the relative influence of nature and nurture on physical aggression probably differs, both quantitatively and qualitatively, for 3-year-olds and for adults.
In developmental terms, there is one way in which nature seems to have a head start on nurture; an individual's prenatal development, which is largely biologically driven, precedes his or her exposure to social environments. However, even in the case of prenatal development, environments (e.g., the uterine environment, the mother's social setting) can have significant impacts on the developing fetus. Still, these environmental inputs are likely to have their immediate effects on the fetus through biological mediators such as hormone levels, immunological factors, blood chemistry, physical traumas, or infectious agents.
Recent behavior genetic research suggests that the heritability of adult intelligence (estimated to be about 60% to 30%) is higher than the heritability of childhood intelligence (about 40% to 50%) (Jensen, 1998; McClearn et ai, 1997; McGue, Bouchard, lacono, Lykken, 1993). The greater genetic contribution to variations in adult intelligence may result from the fact that adults have greater freedom to choose their intellectual (or non-intellectual) environments than children do. After all, children must go to school, and some children are exposed, against their wills, to enrichment programs prescribed by their parents. But after leaving home, people are freer to do their own thing. Intelligent people tend to place themselves in settings that continue to develop their intellect; nonintelligent people do not.
The heritability of masculinity and femininity may similarly vary with age (as some new, unpublished behavior genetic studies are beginning to hint). A somewhat feminine boy may be pressured by peers and parents to behave in an acceptably masculine manner, but when he leaves home he may be freer to express his true self. The broader point is this: There may be no overall answer to behavior genetic questions such as, What is the heritability of masculinity-femininity? Instead, there may by multiple answers, which depend on age and on other factors as well.
Similarly, there may be no global answers to the following' nature-nurture questions:
· How much are sex differences in aggression due to socialization?
· To what degree are sex differences in visual-spatial ability influenced by hormonal variations?
· To what degree are individual differences in sexual orientation due to variations in social environments?
The answer to each question may vary, depending on other factors such as age, education levei, social milieu, and cultural background. This does not mean that nature-nurture questions are meaningless. Rather it means that we should expect a range of answers to such questions.
In discussing possible factors that influence the relative impact of nature and nurture, we should not ignore the obvious; one such factor may be gender itself. The causal cascades sketched in Fig. 7.1 may sometimes differ for males and females. The following findings are consistent with this hypothesis: Parents police gender more strongly in sons than in daughters (see Chapter 5). The process of childhood sex segregation is more extreme and intense in boys than it is in girls, and boys seem to police other boys' gender-related behavior more strongly than girls police girls (see Chapters 1 and 5). Prenatal testosterone in maternal blood is related to preschool girls' but not boys' sex-typed behaviors (see Chapter 4). Boys' sex-typed behaviors appear to be more impervious to adult influences than girls' sex-typed behaviors are (see Chapter 5). After achieving gender labeling, young girls show behavioral effects (e.g., reduced levels of aggression) that boys do not (see Chapter 5).
Sociologist Richard Udry (2000) proposed that girls, because of their lower testosterone levels, may be more responsive to gender socialization—whatever direction it takes—whereas boys may be more rigidly channeled by innate factors. In a similar vein, psychologist Roy Baumeister (2000) proposed that women's sexual behavior may be more variable, flexible, and responsive to social factors, whereas men's sexuality may be more fixed, rigid, and driven by innate factors. Emotion research has suggested that women's subjective emotions are more responsive to social feedback, whereas men's emotions are more "read out" from their current physiological states (Roberts & Pennebaker, 1995). Men and women appear to respond to stress differently. Women show more of a tend-and-befriend response, which leads to social interaction and comparison, whereas men show more of a fight-or-flight response, which leads more often to social isolation (Taylor et al., 2000). Taken together, these varied findings suggest that in a host of ways, women's gender-related behaviors may be more responsive to nurture (social environments, social comparisons, and social pressures), and that men's gender-related behaviors may be guided more by nature (genes, hormones, inner physiology).
Perhaps it's no accident, then, that female gender theorists have tended to emphasize the nurture of gender, the influence of socialization, gender-schemas, and social roles (Bem, 1981b; Deaux & Major, 1987; Eagly, 1987; Maccoby, 1998), whereas male gender theorists have tended to emphasize more the nature of gender, that is, evolutionary pressures, genes, hormones, and brain structures (Browne, 2002; Buss, 1999; Geary 1998; Ken rick, 1987). Like the rest of us, gender scientists form their intuitions, in part, based on their own life experiences, and the life experiences of female and male scientists differ, on average, just as do the experiences of women and men more generally
Will future researchers succeed in developing a unified held theory of gender that accounts for the development of gender in all people, at all times? Or will they need instead to develop subtheories of gender: theories for males and females; theories for toddlers, teenagers, and adults; theories for disadvantaged and middle-class people; theories for people from individualist and from collectivist cultures? To date, most gender theorists have striven to create all-purpose theories (e.g., social learning theories, gender schema theories, social role theories) that attempt to explain the development of gender in all people, using universal principles. However, the truth may turn out to be more complex than this. Rather than developing a universal theory of gender and honing in on a single answer to the nature-nurture question, researchers may instead need to be satisfied with multiple theories and multiple answers. They may come to learn that different causal cascades lead to gender in different ways, in different groups, at different stages of life.
Cascades, Fulcrums, and Social Interventions
The notion of a causal cascade raises an important practical question: If our goal Is to effect real-life change in gender-related behaviors (e.g., to encourage girls to study math and natural sciences more or to induce boys and men to be less aggressive), where should we intervene to produce the largest effects? Do the causal thickets portrayed in Fig. 7.1 give us guidance? Are there especially sensitive points in the causal web—what I'll call fulcrums—where modest interventions can lead to large effects? Or, on the other hand, is the thicket of factors leading to gender so overdetermined—with so many interlinked causes pushing in the same direction—that the system as a whole possesses an inertia that resists quick and easy fixes? Is childhood sex segregation the key to sex typing? Which is more important, parental rearing or peer pressures? Can parents change children's gender-related behaviors if teachers do not cooperate? What role does each thread play in the overall web of gender?
The term cascade implies a sequence of interlocking causal events, where small initial effects may combine, over time, to produce large ultimate effects. Psychologists Richard Martell, David Lane, and Cynthia Erarich (1996) demonstrated such a process in a study that investigated gender-related hiring bias. These researchers conducted a computer simulation that postulated a business hierarchy like those found in many corporations, with eight job levels and with fewer employees in top level jobs than in lower level jobs. They further assumed that both men and women randomly varied in their qualifications (e.g., in their test scores, their job experience) but that the two sexes were, on average, equally qualified. Company officials who decided on promotions were slightly biased (d = 0.2) in favor of men. In operational terms, this meant that the simulation boosted each male worker's job qualification score by a few points.
The organization started with equal numbers of men and women at each job level. New employees always started at the bottom, and higher level employees were selected from the most qualified people at the next lower level. Twenty employment cycles were simulated, in which 15% of the employees were lost to attrition at the start of each cycle. Although the simulation started with equal numbers of men and women at all job levels, by the 20th cycle, 53% of the lowest-level workers but only 35% of the highest-level workers were women. Assuming a somewhat higher level of gender bias (d = 0.45), the simulation generated an even more extreme result. After 20 promotion cycles, 58% of the lowest level workers but only 29% of highest level workers were women
This principle, that repeated small effects can produce large cumulative effects has been discussed earlier. For example, Chapter 1 described a meta-analysis showing that men, on average, end up with slightly better outcomes than women after face-to-face negotiations (Stuhlmacher & Walters, 1999). Although this difference is small (d = 0.09 to 0.25, depending on the kind of negotiation), it could lead to more sizable differences over the course of repeated negotiations, for example, repeated salary negotiations over the course of an entire career. Like compound interest, small advantages build over time.
Or consider a second example. If children have a slight preference to play with same-sex peers during early childhood, then as the choice tournament for playmates repeats, day after day, this slight bias accumulates; more and more of a child's friends will be same-sex friends. Later, when children form into groups, these groups become increasingly sex segregated, based in part on small biases in individual boy's and girl's preferences. The increasing sex segregation of boys' and girls' groups serves to amplify differences in boys' and girls' playstyles and thereby further strengthen preferences for same-sex playmates.
Or consider a third example. After marriage, a husband and wife both pursue their respective careers. When the possibility for promotion and increased work responsibilities arise, a bias exists in favor of pursuing the husband's promotion over the wife's, particularly if the promotion involves moving to a new city or working at a distance from home. Over time and with repeated promotions, husband-wife differences in career success compound.
Although the examples just listed tend to emphasize small environmental causes that snowball over time, the same can be true of biological causes. Indeed, because genes and hormones may produce fairly constant biases toward certain kinds of behavior (toward rough-and-tumble play, toward aggressiveness, toward verbal communication of feelings, toward playing with mechanical toys and objects), their effects, even if small, may steadily accumulate over the course of a lifetime. It may be hard to counter such biological biases with environmental interventions such as brief classroom programs or gender-neutral parenting because the cascading biological tendencies operate 24 hours a day, inside and outside of school, inside and outside of the home. Although social pressures may come and go, genes and hormones, in an important sense, are forever.
Although theorists often tend to portray nature and nurture as standing in opposition to one another; in fact, nature and nurture often reinforce one another. For example, biological factors (e.g., toward rough and-tumble play in boys) may foster childhood sex segregation, and simultaneously, social and cultural factors may also foster sex segregation. Male biological predispositions toward physical aggressiveness are often amplified by cultural learning. Sex differences in visual spatial abilities may be exaggerated by the play and school activities into which boys and girls are channeled. Clearly, biological and social factors that work in concert will be more potent than biological and social factors that oppose one another, and the mutually reinforcing effects of nature and nurture will accumulate more rapidly than will effects that do not superimpose.
It's important to note that a cascade is not simply a process in which repeated small causes yield large cumulative effects. It is also a process in which causal factors at one level trigger increasingly complex chains of causal events at subsequent levels. These multiplying consequences then become causes themselves, feeding back to influence, and alter their original causes (see Fig. 7.1). Such proliferating feedback loops of cause and effect are ubiquitous at the cellular level. For example, DNA is read by chemicals in the cell, which then construct new proteins based on DNA instructions. The synthesized proteins then feed back to influence ongoing chemical reactions in the cell and to turn on and turn off segments of DNA. Rather than viewing DNA as the chemical mastermind that directs all other processes in a cell, we might more accurately envision that everything causes everything else, in an unimaginably complex, self-regulating, Rube Goldberg machine.
To complicate matters even further, the causal factors that feed back to influence DNA expression do not exist Just at the cellular level. Feedback loops also cut across causal levels (like those portrayed in Fig. 7.1). For example, events in the adrenal glands can affect DNA expression in brain cells. Mothers' stress levels and immune reactions can feed back to influence the action of DNA in fetal cells. Even external and social environments—stress, nutrition, and the presence of a sexual partner-can influence DNA expression in one's cells (see Gottlieb, Wahlsten, & Lickliter, 1998).
Lest the notion of intertwined causal cascades seem hopelessly complex to the point of suggesting impenetrable causal thickets that are inaccessible to analysis, it is important to note, optimistically, that science has often made great progress by imposing artificial simplicity on very complex causal systems (e.g., atomic nuclei, living cells, marine ecologies, planetary climates, and spiral galaxies ). Scientists enforce simplicity on complexity, in part by developing theories that everyone knows to be oversimplifications. Such theories can nonetheless provide useful approximations to reality.
Although specific theories of gender often strive for simplicity, gender theories as a whole may sometimes seem to be a study in confusion and contradiction. It is certainly true that current theories embrace a broad diversity of viewpoints (see Chapter 3). However, there is strength in diversity. Contemporary nurture theories have moved beyond simple socialization accounts of gender to propose models that include the influence of social roles, sex segregation and peer influences, gender schemas and stereotypes, and current social settings. And nature theories have moved beyond the simplistic notions that "anatomy is destiny" or "heredity is destiny." They now probe gender in increasingly subtle ways, from the vantage points of evolutionary psychology, behavior genetics, molecular biology, and neuroendocrinology. Most contemporary biological theorists acknowledge that environments interact with biological factors at all levels of analysis.
Both nature and nurture now have seats at the theoretical table, and so the really hard work now begins, that is, to specify, in nitty-gritty detail, exactly how the many biological and social-environmental factors identified by recent theories weave together to create the complex tapestry known as gender.
Causal Cascades and the Two Faces of Gender
The term gender serves as a kind of shorthand for two different phenomena: (a) sex differences in behavior, and (b) individual differences in masculinity and femininity (see Chapters 1 and 2). This raises an obvious question: Are the causal cascades sketched out in Fig. 7.1 the same for these two sides of gender? More specifically, are the causal factors that generate sex differences in behavior the same as those that generate individual differences in masculinity and femininity?
Let's consider these questions in relation to a specific finding that masculine people are more likely to die than feminine people at any given age (Lippa, Martin, & Friedman, 2000). Is this finding relevant to the topic of sex differences? The answer is almost certainly, yes. Epidemiological studies consistently show that men die at a younger age than women do. In the United States, for example, the mean difference in life expectancy for men and women is 6 or 7 years. Thus, the finding that masculinity is linked to mortality is matched by the parallel finding that men, on average, die younger than women do.
But are the causal factors that lead to sex differences in mortality the same as the causal factors that lead masculinity to be linked to mortality within each sex? We do not yet know the answer to this question. However, I believe that the answer is likely to be, yes. Some of the common causal factors may be biological (e.g., men have higher testosterone levels than women do, and similarly, masculine individuals have higher testosterone levels than feminine individuals). Other common factors may be behavioral (men smoke more than women do, and similarly, masculine individuals smoke more than feminine individuals). And still other common factors may be environmental (on average, men work and play in more dangerous settings than women do, and similarly, masculine individuals work and play in more dangerous settings than feminine individuals do), if the factors that lead masculinity to be linked to mortality strongly overlap with the factors that lead to sex differences in mortality, then masculinity and maleness (and similarly, femininity and femaleness) will prove to have deep as well as surface similarities.
Research on antisocial behavior provides another good example of how careful studies can probe the causal factors that contribute both to sex differences in behavior and to variations within each sex (see Rowe, Vazsonyi, Flannery, 1995). In the Dunedin longitudinal study of more than 900 New Zealand young people, Moffit, Caspi, Rutter, and Silva (2001) assessed a variety of risk factors that predicted adolescents' antisocial behaviors. These risk factors included maternal problems (e.g., mother's psychiatric symptoms and criminality), general family risks (e.g., family conflict, frequent moves, lower socioeconomic status), children's cognitive and neurological risks (e.g., neurological abnormalities, low IQ), childhood behavior problems (e.g., hyperactivity, difficult temperament), and peer-reiated risks (e.g., peer rejection, peer delinquency).
In combination, these risk factors predicted both boys' and girls' antisocial behaviors quite strongly, and risk factors that predicted boy's antisocial behaviors also tended to predict girls' antisocial behaviors. Moreover, sex differences in antisocial behavior were largely explained by sex differences in specific risk factors, in particular, by boys' higher levels of cognitive and behavioral risks, hyperactivity, and peer rejection and hanging out with "bad sorts." It is worth noting, however, that there was one kind of risk factor that predicted variations in antisocial behavior within each sex but did not predict sex differences in antisocial behaviors, namely, general family-related problems. This makes sense because family problems (e.g., frequent moves, family conflict, inconsistent and harsh parental discipline, lower socioeconomic status, and the presence of only one parent in a family) affect equal numbers of boys and girls and so they should not lead to sex differences in antisocial behaviors.
Comparing the factors responsible for sex differences with those responsible for within-sex variations may be worthwhile when studying other sorts of behaviors as well. For example, consider the following questions:
· Are the causal factors that lead men and women to have different sexual orientations the same as the factors that lead to individual differences in sexual orientation within each sex?
· Are the causal factors that lead men, on average, to be more physically aggressive than women the same as factors that lead to individual differences in aggression within each sex?
· Are the causal factors that lead men and women, on average, to choose different kinds of occupations the same as factors that lead to individual differences in occupational choices within each sex?
The answer to each of these questions is not clear, but the very act of posing such questions encourages researchers to study and compare the causal cascades that contribute to the two sides of gender.
Some sex differences may actually result from differences between subgroups of men and women that are defined in terms of their masculinity and femininity levels. Once again, the masculinity and mortality study provides a concrete example. Lippa, Martin, and Friedman (2000) found that men in their study were more likely to die than women at any given age. Further analyses showed, however, that sex differences in mortality were strongest for masculine men and feminine women, but they were much smaller for feminine men and masculine women. Thus, what appeared to be a sex difference, may have in fact been largely a difference between just some men (those high on masculinity) and some women (those high on femininity).
Consider another example. Sex differences in homicide rates may be due particularly to differences between some men (hypermasculine young men) and women, but not between most men and women, thus, what appears to be a sex difference from one perspective may appear, from another perspective, to be largely a difference between a male subgroup and women. Or consider another example. Sex differences in sexual orientation (an individual's degree of sexual attraction to men or to women) may be stronger for some groups of men and women (masculine men vs. feminine women), and weaker for others (feminine men vs. masculine women). All of these examples suggest that research on sex differences should move beyond the simple question,"Do men differ from women?"—to consider the more subtle question, "Which men differ most strongly from which women?"
There is a final way in which the study of sex differences is linked to the study of masculinity and femininity: The very size of sex differences may depend, in part, on individual differences in masculinity and femininity. Recall that the most common measure of sex differences—the d statistic—depends both on mean differences between the sexes and on the amount of variation observed within each sex (see Chapter 1). The more variation there is within each sex—variation that is due, in part, to individual differences in masculinity and femininity—the smaller the d statistic.
This statistical point brings us back to the nature-nurture debate, for the relative magnitude of sex differences and within-sex individual differences may provide another way to probe the relative impact of nature and nurture. The following thought experiment will make this clearer, Imagine a society in which all boys are sent at an early age to military camps where they are trained to be stoic, competitive, and aggressive. In contrast, all girls remain at home, where they are sequestered, shrouded in confining robes, kept illiterate, and educated only to raise children and carry out domestic tasks. Such a gender-polarized society would likely produce very large differences between the two sexes and strong homogeneity within each sex.
In contrast, imagine a second society in which boys and girls are treated exactly alike from birth on. Boys and girls attend the same schools, study the same curricula, and play the same sports. Both boys and girls wear the same unisex clothes and all read the same nonsexist children's books. Parents give boys and girls the same toys to play with, and teachers treat boys and girls alike. Such a society would probably produce much smaller differences between the two sexes and it would permit much more variability within each sex.
Thus, to the extent a culture's gender socialization practices influence both sex differences in behavior and individual differences in masculinity and femininity, we should expect a negative relationship, across cultures, between the magnitude of sex differences and the magnitude of within-sex variations in gender-related behaviors. However, to the extent that biological factors are responsible for both sex differences and individual differences in masculinity and femininity, the magnitude of sex differences may often be unrelated, across cultures, to within-sex variations (think of sex differences and within-sex variations in height as an example).
In short, the relation between sex differences and variations in masculinity and femininity, across cultures, can provide researchers with another tool to use in studying the contribution of nature and nurture to gender. This example shows, once again, that although sex differences and variations in masculinity and femininity are conceptually distinct, they are also intimately intertwined, like so many other aspects of gender.