Discussion Question #5: Which most powerfully influences gender differences in behavior-- biology or socialization? What are the implications (for individuals, institutions, and society) if biology is the most powerful? What are the implications if socialization is the most powerful?
Read attachment. Initial post should be no longer than 2-3 paragraphs and should show your understanding of the week’s readings and lectures. Follow APA format, citations and references.
4 Biology and Behavior
The data on how parents treat boys and girls reviewed in chapter 2 and
on how children learn about sex differences reviewed in chapter 3 raise
the question of whether all the sex differences we see are differences that
adults construct. What about built-in sex differences? In this chapter and
the next I summarize what we know about intrinsic differences between
males and females and consider the implications of those differences. One of the problems involved in discussions of sex differences is that
many people view biological influences as all-powerful and final. If a sex difference-such as mathematical ability-can be shown to have a bio-
logical component, it seems immutable and eternal. Our perception of
immutability stems, I think, from an inaccurate understanding of what it
means for a characteristic to have a biological basis. We interpret a bio-
logical sex difference as a difference that is a direct result of having a
certain set of sex chromosomes. Since we cannot change the set we have,
we are stuck with the differences that go along with being XX or XY.
This everyday interpretation of biology, however, is radically uninformed. Although chromosomal differences may be qualitative, none of the ensu-
ing differences are. Biology leaves us a lot of room to manoeuvre. Biology
is not necessarily destiny. In a way, it is odd that we should interpret sex differences as immutable,
when we do not accept biology as destiny in other aspects of human exis-
tence. For example, biology sets limits on the human life span, but we need not and do not for that reason accept a short average life span as
o~r fate. As a society, we put forth great efforts to understand the mecha- nisms of health and to cure disease and illness. We practice good hygiene,
we 11 westigate the roles of diet and exercise, we sterilize, we vaccinate, we
imx:ulate, we medi1:ate. Although we agree that no one can live forever,
wt· su1:1:essfully invest tremendous resources in trying to live longer. Our
su~·1:ess is evidence of the !ability and complexity of the processes that
underlie life and death and of humans' ability to intervene in those
pro1:esses. I propose that we adopt the same attitude toward biological sex differ-
en1:es. Biologi1:al sex differences arise through the actions of sex hor-
mones operating in our physical and social environments. Like the
pro1:esses relevant to life and death, those relevant to sex differences are
intrirnte and susceptible to change. Biology sets limits, but we need not
for that reason accept the differences we see as immutable. We have good
evidence from cultura
l,
situational, and temporal differences that the dif-ferences arc not immutable. In this 1:hapter and chapter 5, I have applied my own interpretation to
the resear1:h findings on physical, behavioral, and cognitive differences,
highlighting some findings the researchers themselves did not and deem-
phasizing others. Many different points of view on these data exist: while
some people deny the existence or importance of a biological influence,
others refer all differences to biology. My own position is that biological
influcn1:es exist and are important, but are only part of the story.
Sex Hormones
When we talk about "biological" sex differences, we are talking about
the influence of the hormones responsible for differentiation of the sexes.
The sex chromosomes themselves (XX for females and XY for males) do
not have an automatic and rigid set of consequences. They do not act like
on-off switches. Nor are they, even, the immediate agents of sex differen-
tiation. They exercise their influence through the differing hormonal de-
velopments they set into motion. Those developments in turn have their ' '
effects within one or another context and are inherently variable. The
same thing is true of genetic effects in general (see discussion in Neisser
et al. 1996). So, although we cannot change our chromosomes, they are, in a sense, irrelevant.
Biology and Behavior 69
Mammals with XX chromosomes develop ovaries that secrete one set
of hormones; mammals with XY chromosomes develop testes that secrete
the same set of hormones but in different amounts. The action of those
hormones underlies further physical sex differentiation. The three main
types of sex hormones-androgens, estrogens, and progestins-occur in
both sexes; the sexes differ, sometimes dramatically, in the amount of each hormone they produce. For example, testosterone levels in college-
age males and females are very different. The male-to-female ratio has
been reported to be as high as ten to one, with no overlap between the two groups (Udry & Talbert 1988). Average testosterone-concentration
levels in saliva have been reported to be about three times as high in males
as in females, again with no overlap between the two groups (Gouchie & Kimura 1991 ). Within each sex, however, there is a considerable range of
secretions of each hormone.
Within-Sex Variability It is in the realm of behavior that the variable effects of sex hormones within each sex are clearest. The effects vary depending on the social-
psychological context they occur in, for nonhuman animals as well as humans (Buchanan et al. 1992; Collaer & Hines 1995). Hormonal ef- fects, in other words, are context-dependent. Even in rats, the effects of
sex hormones differ, depending on the sort of handling the animals re- ceive, the type and amount of stimulation provided by their environment,
and the kind of maternal care they receive. (See Collaer & Hines 1995,
1
for a summary of these and other effects of gonadal steroids.) Although we commonly speak of the environment modifying or moder-
ating or influencing hormonal effects on human behavior, I find that ter-
minology misleading. It suggests a primacy and a univocality for the action of sex hormones that do not exist. It would be similarly inappro- priate to label environmental effects on human behavior as primary, and speak of hormones as modifying or moderating those environmental
effects. Rather, hormonal and environmental effects act together-they coact-to jointly influence people's and animals' traits and behaviors
(Moore 1985). Within the realm of behavior, there is no such thing as a
pure hormonal effect, because there is no such thing as a zero or neutral
70 Chapter 4
environment. Equally, there is no such thing as a pure environmental ef
fect, because there is no such thing as zero or neutral hormones. To under
stand any behavior, it is necessary to understand the contribution of both
hormones and the environment. We can get a small idea of the complexity of the interactions between
sex hormones and the environment in adult humans by looking at how
time of year and level of circulating testosterone affects men's ability to
rotate objects mentally (Kimura & Hampson 1994). In the fall, North American males have higher levels of testosterone than they have in the
spring; they also have lower scores on spatial rotation tests than they do in
the spring. As individuals, males who are below the average in circulating
testosterone also score higher on spatial rotation tests than those whose
levels are higher than average. Men's cycles occur on a daily basis as well.
When testosterone levels are higher, in the morning, men perform more
poorly on tests of spatial rotation. (See Kimura 1996 for a summary of
this and other experiments involving hormonal effects.)
Males' spatial rotation scores are linked to testosterone level in ways
that our gender schemas would not predict. Gender schemas represent
testosterone as contributing to masculinity and label certain kinds of
skills-like map reading and mental rotation-as masculine. But the data
show that, for mental rotation, having less testosterone is better than hav ing more-up to a point.
Being below the median is better than being above it, but for males
being very far below it is worse than being somewhat below. There ap
pears to be a curvilinear relationship between testosterone level and
scores on mental rotation tests. Aging males, who have very low testoster
one levels, have lower scores on tests of mental rotation than college-age
males whose scores are below the median (Kimura 1996). Women's tes tosterone levels are very low.
The data on men's variability show that our notions of people's abilities
are oversimplified. It is obvious that the component of spatial ability that
is due to knowledge of spatial relations cannot fluctuate over a twenty
four-hour period. Still, spatial ability undoubtedly involves more than
knowledge. It also requires registering the spatial properties of objects,
storing that information, and then mentally rearranging the objects to see
71 Biology and Behavior
what they would look like from different angles. Those processes appear
vulnerable to changes in testosterone level.2 Researchers do not know whether male-female differences in spatial abilities are due to underlying
differences in knowledge or to processing differences. There are at least some spatial tasks where males perform better than
females that seem unrelated to testosterone levels. For example, although male homosexuals and heterosexuals have comparable testosterone levels,
homosexuals generally perform the task of throwing a ball at a target less
well than heterosexuals do (Hall & Kimura 1995; Kimura 1996).3
Female scores on spatial-orientation tests fluctuate somewhat across
the menstrual cycle and show some individual variability. When estrogen
levels are very high, females score worse on some-but not all-tests of
spatial ability than they do when estrogen levels are moderate or low (Hampson 1990a, 1990b). For women, there is also some indication of a
curvilinear relation between estrogen and spatial skills, just as there is a
curvilinear relation between men's testosterone levels and spatial perfor
mance (Hampson 1990b). At the same time, female monthly fluctuations are considerably smaller
than male seasonal variations. The hormonal effects in women are rela tively small and do not show the other patterns we would expect to see
if estrogen level alone controlled behavior. For example, undergraduate
women who major in science score better on tests of spatial ability than do women who major in other fields. If estrogen level predicted spatial ability well, women in science would have estrogen levels different from
other women, but they do not (Hampson 1990b). Females' levels of testosterone, however, are related to scores on at least
some tests of spatial ability, even though those levels are much lower than males' levels. Females with above-average testosterone levels for females
score as well on spatial tests as males with below-average levels for males (Gouchie & Kimura 1991). That is so even though the testosterone levels of these women are only half as high as the male levels. A similarly low level in males would be associated with poor spatial performance. Per
haps because of differences in brain structure, a very low testosterone
level in males-a level too low to mediate good spatial performance-is
adequate for females to attain good spatial performance.
72 Ch<1pter 4
The resean:h on hormonal influences on behavior thus contradicts any
notion we might have that sex chromosomes determine our destiny, or
that hormones act independently of the environment in which they are
embedded. Hormones are relevant influences on our traits and behaviors,
but their effects are labile. Evaluation of the effects of hormones on the
specific physical and behavioral traits discussed in the remainder of the
chapter confirms this general conclusion.
Variation in Trait Expression
We can illustrate the relations between biology and trait expression with
an example unrelated to sex differences. Although some hearts are intrin-
sically less efficient pumps than others, diet and exercise can increase their
efficiency, just as a poor diet and lack of exercise can decrease it. The
environment influences how well a heart does its work. For most traits,
biology establishes a continuum of values rather than a specific value. In
the case of hearts, biology establishes the usual range within which they
operate, rather than a specific efficiency value that holds for all hearts.
In the same way, some sex differences change in value from one situa-
tion to another, while others are relatively invariant over an individual's
adult lifetime. Variability that occurs within an individual is variability
that can be altered.
Reproduction
Most people's everyday experience is that there are two, clearly different
sexes. Male and female genitalia differ markedly. (Ambiguous genitalia
exist but are rare.) A concomitant functional difference is as marked as
the anatomical difference: females are capable of birth and lactation and males are not.<
Unlike other sex differences that we pay attention to, anatomical differ-
ences hold to the same degree across all cultures, across all situations, and
across time. Cognitive and behavioral sex differences, on the other hand,
exist along a continuum, with a great deal of overlap between the sexes
and a great deal of variability within each sex. Only in the reproductive
area are the differences qualitative. All other differences between the
73 Biology and Behavior
sexes are quantitative and inherently variable, though the type and size of
the variability differs from trait to trait. In all the physical and behavioral
characteristics we consider here, the variability within each sex is greater
than the (average) difference between the sexes.
Height In all cultures, males are on average taller than females-a difference that
has persisted for thousands of years. Nonetheless, even with height, some
types of variability do exist. Within each sex, some people are taller than others. Some females are taller than others and are also taller than some
males. The difference between the extreme heights within each sex is
larger than the difference in the average height between the sexes. Height also varies from individual to individual as a function of nutri-
tion-which is environmentally determined-and development. Among
children born in the same place at the same time, those who are well
nourished grow taller than those who are poorly nourished. In old people,
height decreases because of bone loss. Changes in the social environment, however, do not affect adult height.
People are not, for example, taller at work than they are at home. Even
here, though, perceptions of height can be manipulated: people may wear
shoes that increase their apparent height, or they may, by slouching, re-
duce it. Compared to other characteristics we consider later in the chap-
ter, however, individual height has only limited variability.
Voice Pitch Other physical sex differences show much more mutability within an indi-
vidual. Consider voice pitch. Like height, it has a underlying biological
basis: on average, males have larger and more muscular larynxes, and
larger and more resonant throats. The result is that, within any culture,
males' voices tend to be deeper in pitch than females'. Unlike height, how-
ever, an individual's adult voice pitch is not fixed. We all possess a fairly
~ide range of possible pitches. Pitch is responsive, for example, to emo-
tional stress. If the expression of a trait can vary within an individual from one mo-
ment to the next, it follows that all other types of variability can also
l,1.
74 Chapter 4
exist. Again, voice pitch is a good example. The size of the sex difference
varies from culture to culture. In Italy, for instance, the difference is less
than it is in the United States (McConnell-Ginet 1983).
Unlike height, voice pitch could be, on average, almost the same in
males and females, because of the variability within an individual, and
because of the responsiveness of pitch to social and cultural factors. Fe
males could speak nearer the lower end of their range and males nearer
the upper end. (British Prime Minister Thatcher reportedly profited from
lessons in lowering her voice pitch so that she would appear more author
itative.) By the same token, the average difference in pitch between the
sexes would be extremely high if females spoke at the upper end of their
range and males spoke at the lower end of theirs. Thus, even for traits
with a clear underlying physiological connection, societies can "choose"
how extreme the behavioral differences will be. Traits whose expression
is inherently variable-like voice pitch-rather than more rigid-like
height-allow for the largest cultural and gender differences.
In essence, what matters in evaluating sex differences is not the existence
of a biological connection but the inherent variability in the expression
of a trait. Height and voice pitch both have a clear biological connection,
but adult height cannot be directly influenced by culture, whereas voice
pitch can. If a trait is inherently variable, like voice pitch, the expression of it is malleable. The variability makes it correspondingly difficult to as
sess the influence of biology. For some sex differences, biology limits the range of a behavior; further, the midpoint of the range for one sex may
be higher or lower than the midpoint for the other sex. The differences in
the midpoints may or may not have practical significance, depending on
how great the range is for each sex. To assess the sources of sex differ
ences we need to know both whether a given behavior has a direct biologi cal connection and, if it does, whether the proper analogy is to height or
to voice pitch. Yet we often do not know. For the traits that could be
relevant to professional achievement, voice pitch seems a better analogy,
because those traits are responsive to cultural, social, and psychological influences.
When I speak of influences, I am not suggesting that the environment
shapes or molds children or adults. Rather, people form nonconscious
75 Biology and Behavior
hypotheses based in part on the data they receive from the environment. Those hypotheses about sex differences-gender schemas-then guide
people's behavior. When a trait is malleable, gender schemas can affect
its expression.
Behavioral Traits and Professional Achievement
None of the traits discussed so far have any direct connection to people's
intellectual or professional abilities. Height and voice pitch have nothing to do with competence, even though we tend to think of competent people as tall and having a low voice. Other differences, however, have
more potential significance.
Activity Level Males are more active than females, a difference that appears to increase throughout childhood and into early adolescence (Eaton & Enns 1986).
6
Since few studies have examined activity differences after the age of fif- teen, we do not know whether the differences persist throughout the life span or begin to diminish at some point. Higher activity levels could indi- rectly cause more exploration and, even more indirectly, be related to
achievement. But that is speculation. The existence of infant sex differences in activity level has not gone
unquestioned. Relatively few studies have looked at infant sex differences; those that have found no differences in global activity level, although there is some suggestion that infant boys' movements are more vigorous than girls'. One careful study comparing girl and boy babies at two and a half months and at five months found almost no differences either in activity level or in vigor of movement ( Cossette, Malcuit, & Pomerleau 1991). The few differences observed were small and could have been due
to chance. Even fewer studies have looked at prenatal activity levels, but one or
two have found differences. If the existence of in utero sex differences were confirmed, that would suggest very strongly that activity differences
are indeed hormonally initiated, as male and female fetuses receive no differential social stimulation (Eaton & Enns 1986). The fact that sex differences in activity increase as development proceeds is compatible
76 Chapter 4
with two possibilities: (1) that high activity is encouraged in boys and
Jisrnuraged in girls; or (2) that activity levels mature and follow different
paths of biological development in boys and girls. One indication that differences in activity level are initiated by hor-
monal differences comes from data on children who experienced overly
high levels of androgens in utero. In a relatively rare genetic disorder (oc-
curring approximately once in every ten thousand births), an enzyme de-
ficiency leads to overproduction of androgens by the fetus's adrenal
glands. The disorder, commonly called congenital adrenal hyperplasia
(CAH), is usually visible in chromosomal females soon after birth, be-
cause the genitals are masculinized. As newborns, the girls typically re-
ceive genital surgery and hormonal treatment to counteract further effects
of the androgens. Both boys and girls with CAH have very high levels of
circulating androgens. Detection at birth is more difficult in boys, and
solid data on the effects on boys are not available.
Most studies have found that girls with CAH have higher activity levels
than normal girls, levels that are similar to those of normal boys (see dis-
cussion in Collaer & Hines 1995). Rough-and-tumble play, which is much more characteristic of boys than of girls, is also frequent in girls
with CAH. A study comparing three- to eight-year-old girls with and
without CAH found that the girls with CAH spent more time playing
with toys associated with males, such as vehicles and construction
toys, than the others did (Berenbaum & Hines 1992). A study of older children produced similar results (Berenbaum & Snyder 1995). The girls with CAH spent the same amount of time with masculine toys as boys
without CAH did. Toys typically associated with boys may lend them-
selves to high-activity play more easily than do toys typically associated with girls.
To summarize, activity level seems like a good candidate for a hormonally
influenced sex difference. Data both from boys and girls with normal pre-
natal hormonal development and from girls who experience excess an-
drogens in utero suggest hormonal involvement. But interactions with social-psychological factors cannot be ignored. We know that parents
treat boys and girls differently; parents of girls with CAH may be more
F
Biology and Behavior 77
tolerant of active play than the parents of girls with no history of atypical
hormonal production (see also chapter 5, n. 8).
There is no evidence that a high activity level or rough-and-tumble play
is either necessary or sufficient for later achievement. If either were im- portant, changes in the child's social environment could increase girls'
participation.
Aggression Another candidate for an intrinsic sex difference in behavior is hostile
physical aggression-defined as touching another person with the inten-
tion of inflicting harm, with or without the desire to obtain some goal.
There is a higher incidence of such hostility in boys, although the sex
differences appear to develop later-around the age of three-than dif- ferences in activity level (see review in Berk 1994).
Sex differences in aggression increase throughout adolescence, then di-
minish, apparently because society looks less and less favorably upon
physical aggression as children become adults. Overall, studies reliably
show males to be more physically aggressive than females, with little or no difference from one generation to the next (Eagly & Steffen 1986; Knight, Fabes, & Higgins 1996). Cultures vary in how often people ex- press hostile physical aggression, but there is usually a sex difference, with
males being more active and aggressive. Females of one culture may well be more physically active or aggressive than the males of another culture,
but they are likely to be less physically aggressive than males of their own culture. Even that difference, however, is not universal, as Mead demon- strated (1935).
Although males are usually more aggressive than females, the actual level of individual aggression is inherently highly variable. As with other
characteristics, the range of aggression within each sex is larger than the average difference between the sexes. Interestingly, boys seem less prone
to aggression if they have had experience looking after younger children (see Maccoby & Jacklin 1980; Tieger 1980). Hostile behavior can appar- ently be reduced by practicing nurturant behavior.
Most girls in most cultures either have actual child care experience, or the surrogate experience of playing with dolls. Unlike boys, girls may thus
hr Jiwrtnl from aggrrssion throughout their childhood. If more boys \\'t"rt' giwn thr opportunity to take care of young children, reduced ag-
grc~s1on might rrsult. Again, thrrr is no evidence that a tendency to physical aggression is
cithrr nrcrssary or sufficient for later achievement. I emphasize physical
aggrt·ss1011 hrcause that is where sex differences are most marked. Aggres-
siw thoughts anJ words, on the other hand, are more equally shared be-
twt·rn the sexes. StuJirs of white college-age males from the North or South of the
UnitrJ States Jemunstrate that a tendency to aggressive solutions is in-
tluenceJ by an individual's subculture. In a series of experiments, re-
st·archers arranged for the young men to be mildly insulted by a peer
(Cohen, Nisbett, Bowdle, & Schwarz 1996). After the insult, the students participated in various tasks designed to measure their aggression. North-
rrners and southerners responded differently in those follow-up tasks. In
one task, the students were asked to complete a story in which a male
studrnt's girlfriend complained to him that a male acquaintance had tried
to kiss her. Southerners who had been insulted completed the story more
aggressively than northerners who had been insulted, as well as more ag-
gressively than southerners or northerners who had not been insulted.
The researchers concluded that, for southern males, aggression and no-
tions of "honor" are interwoven: southerners may be easier to offend
than northerners and may also respond more aggressively to a perceived
offense (Cohen ct al. 1996). (Naturalistic data also indicate that white
southern males arc more likely to be violent in certain settings than their
northern counterparts (Cohen et al.)) For our purposes, the important
aspect of the experiment is its evidence that aggression is not just culturc- bound, but subculture-bound.
Aggression also varies from situation to situation (Eagly & Steffen 1986 ). A review of sixty-four experimental studies of aggression demon-
strates that, depending on the type of provocation to aggression, men may
be either more or less aggressive than women (Bettencourt & Miller 1996 ). Men, for example, react more aggressively than women do when
their intelligence is insulted. Women, however, react somewhat more ag- gressively than men to other types of insults.
79 Biology and Behavior
The wide variability in the expression of aggression in each sex, and
the concomitant variability in the presence and extent of sex differences
in aggression, indicate that cultures and subcultures play a major role in
determining people's reactions to frustration or provocation. As it does
with voice pitch, the variability suggests that there is great flexibility in the
amount of aggression males and females express. No particular amount is
natural or inevitable for either sex. Males may exhibit very little aggres-
sion or a great deal; females exhibit a similar range of behavior. Given that inherent variability, I interpret the pervasiveness of a sex
difference across cultures as an indication that the social arrangements of
many cultures have something in common, something that is conducive
to the development of similar gender schemas that set sex differences in
place and maintain them. Such an interpretation does not rule out a con-
tribution from sex hormones but, rather, suggests that the hormonal con-
tribution neither dominates nor effaces the environmental influences.
Putting together the data on activity-especially rough-and-tumble
play-and on aggression, we can see how they might be related. A high
activity level may be suppressed or encouraged, directed toward objects
or toward people, intended to help others or to hurt them (see discussion
in Parsons 1982). Hostile aggression is not a necessary outcome of high
activity. Girls with CAH are not more aggressive in their behavior than
other girls, even though their activity levels and liking for rough-and-
tumble play are greater. Although girls with CAH do score somewhat
higher on questionnaires asking about aggressive tendencies, they do not
act out those tendencies (see summary in Collaer & Hines 1995). Nonetheless, high activity level may be a predisposing factor to aggres-
sion. If combined with anger and lack of cultural prohibitions against the expression of anger, high activity can result in hostile aggression. Anger
may be a natural human emotion but there is no natural way to handle . ' it. As the North-South differences suggest, an individual's response is af-
fected by beliefs about what is appropriate. That hormones predispose but do not determine aggression is also ap-
parent from changes that occur during puberty, when boys' testosterone t levels increase greatly. Although the hormone increases in all boys, moS
_l___
180 Chapter 4
boys in most situations do not become more aggressive (Buchanan et al.
1992). Boys with higher levels of testosterone do seem to display more
physical aggression if they feel threatened or perceive a situation as unfair.
And boys who had behavior problems before puberty may have those
problems exacerbated by an increase in testosterone. Otherwise, however,
most boys do not show more aggression as a consequence of higher levels
of testosterone. To summarize, sex differences in aggression may be
linked to hormonal differences, but even physical aggression is strongly
influenced by social and psychological factors.
The Significance of Sex Differences
Physical and behavioral sex differences exist. Those differences include
reproductive role, height, voice pitch, activity level, and aggression. All
are affected to some degree by sex hormones, but in most cases we know
neither the extent nor exact nature of the influence.
To some degree, however, the presence of a hormonal influence on
behavior is irrelevant. Except for reproduction, sex differences are not
qualitative but average, quantitative differences. The variability in the ex
pression of a trait both within an individual and across individuals dem
onstrates the importance of the social environment. The variability tells
us that hormones and the social environment act together to produce be
havior. We need not change people's hormones to change their behavior; changing the social environment has clear effects.
The Origins of Sex Differences in Human Behavior
Evolved Dispositions Versus Social Roles
Alice H. Eagly
Wendy Wood
The origins of sex differences in human behavior can lie mainly in evolved dispositions that differ by sex or mainly in the differing placement of women and men in the social structure. The present article contrasts these 2 origin the- ories of sex differences and illustrates the explanatory power of each to account for the overall differences be- tween the mate selection preferences of men and women. Although this research area often has been interpreted as providing evidence for evolved dispositions, a reanalysis of D. M. Buss's (1989a) study of sex differences in the at- tributes valued in potential mates in 37 cultures yielded cross-cultural variation that supports the social structural account of sex differences in mate preferences.
s more research psychologists have become will- ing to acknowledge that some aspects of social behavior, personality, and abilities differ between
women and men (e.g., Eagly, 1995; Halpern, 1997), their attention has begun to focus on the causes of these differ- ences. Debates about causes center, at least in part, on determining what can be considered the basic or ultimate causes of sex differences. Theories of sex differences that address causes at this level are termed in this article origin theories (Archer, 1996). In such theories, causation flows from a basic cause to sex-differentiated behavior, and bi- ological, psychological, and social processes mediate the relation between the basic cause and behavior. In this article, we consider two types of origin theories: One of these implicates evolved psychological dispositions, and the other implicates social structure. Evolutionary psychol- ogy, as illustrated in the work of Buss (1995a), Kenrick and Keefe (1992), and Tooby and Cosmides (1992), thus rep- resents the first type of origin theory, and social psycho- logical theories that emphasize social structure represent the second type of origin theory (e.g., Eagly, 1987; Eagly, Wood, & Diekman, in press; Lorenzi-Cioldi, 1998; Ridge- way, 1991; West & Zimmerman, 1987; Wiley, 1995).
A
In the origin theory proposed by evolutionary psychol- ogists, the critical causal arrow points from evolutionary adaptations to psychological sex differences. Because women and men possess sex-specific evolved mechanisms, they differ psychologically and tend to occupy different social roles. In contrast, in the social structural origin theory, the critical causal arrow points from social structure to psychological sex differences. Because men and women
Northwestern University Texas A&M University
tend to occupy different social roles, they become psycho- logically different in ways that adjust them to these roles.
One important feature is shared by these two origin theories: Both offer a functional analysis of behavior that emphasizes adjustment to environmental conditions. How- ever, the two schools of thought differ radically in their analysis of the nature and timing of the adjustments that are most important to sex-differentiated behavior. Evolution- ary psychologists believe that females and males faced different pressures in primeval environments and that the sexes' differing reproductive status was the key feature of ancestral life that framed sex-typed adaptive problems. The resolutions of these problems produced sex-specific evolved mechanisms that humans carry with them as a species and that are held to be the root cause of sex- differentiated behavior. Although evolutionary psycholo- gists readily acknowledge the abstract principle that envi- ronmental conditions can influence the development and expression of evolved dispositions, they have given limited attention to variation of sex differences in response to individual, situational, and cultural conditions (e.g., Ar- cher, 1996; Buss, 1995b; Buss & Kenrick, 1998). For example, Buss (1998) emphasized "universal or near-uni-
Editor's note. Cheryl B. Travis served as action editor for this article.
Author's note. Alice H. Eagly, Department of Psychology, Northwest- ern University; Wendy Wood, Department of Psychology, Texas A&M University.
This article was completed while Alice H. Eagly was a Visiting Scholar at the Murray Research Center of Radcliffe College and was supported by a Sabbatical Award from the James McKeen Cattell Fund. The research received support from Grants SBR-9729449 and SBR- 9514537 from the National Science Foundation.
Thanks are extended to David Buss for making available for reanal- ysis data from his 37 cultures study (Buss, 1989b; Buss et al., 1990) and to Michael Bailey, April Bleske, Judith S. Bridges, Galen Bodenhausen, David Buss, Stephen M. Colarelli, Lee Cronk, Amanda Diekman, Steven W. Gangestad, Patricia Adair Gowaty, Judith Hall, Martie Haselton, Sarah Hrdy, Douglas T. Kenrick, Mary Kite, Richard Lippa, Dan McAdams, Anne McGuire, Felicia Pratto, Radmila Prislin, Dean Pruitt, Eshkol Rafaeli, Neal Roese, Alice Schlegel, and Jeffry Simpson for comments on a draft of the article. Also, Crystal Toures and Heather Franzese provided assistance in locating and entering relevant data.
Correspondence concerning this article should be addressed to Alice H. Eagly, Department of Psychology, 2029 Sheridan Road, Northwestern University, Evanston, IL 60208, or to Wendy Wood, Department of Psychology, Texas A&M University, College Station, TX 77843. Electronic mail may be sent to eagly@nwu.edu or wlw@psyc.tamu.edu.
June 1999 • American Psychologist
Copyright I 999 by the American Psychological Association, Inc. 0003-066X/99/$2.00Vol. 54. No. 6. 408-423
408
Alice H. Eaglyversal sex differences" (p. 421) in preferences for long- term mates.
Social structuralists maintain that the situations faced by women and men are quite variable across societies and his- torical periods as social organization changes in response to technological , ecological, and other transformations. From a social structural perspective, a society' s division of labor between the sexes is the engine of sex-differentiated behavior, because it summarizes the social constraints under which men and women carry out their lives. Sex differences are viewed as accommodations to the differing restrictions and opportunities that a society maintains for its men and women, and sex- differentiated behavior is held to be contingent on a range of individual, situational, and cultural conditions (see Deaux & Lafrance, 1998). Despite this emphasis on the social environ- ment, social structuralists typically acknowledge the impor- tance of some genetically mediated sex differences. Physical differences between the sexes, particularly men's greater size and strength and women's childbearing and lactation, are very important because they interact with shared cultural beliefs, social organization, and the demands of the economy to in fluence the role assignments that constitute the sexual division of labor within a society and produce psychological sex dif- ferences (Eagly, 1987; Wood & Eagly, 1999).
These thumbnail sketches of these two origin theories should make it clear that this debate about the origins of sex differences cannot be reduced to a simple nature-versus- nurture dichotomy. Both evolutionary psychology and so- cial structural theory are interactionist in the sense that they take both biological and environmental factors into ac- count, but they treat these factors quite differently. Evolu- tionary psychology views sex-specific evolved dispositions as psychological tendencies that were built in through genetically mediated adaptation to primeval conditions; the
theory treats contemporary environmental factors as cues that interact with adaptations to yield sex-typed responses . Social structural theory views sex-differentiated tendencies as built in through accommodation to the contemporaneous sexual division of labor; in this approach, physical differ- ences between the sexes serve as one influence on role assignment.
Another caution is that these theories do not merely reflect different levels of analysis. In some attempts to reconcile the two perspectives, writers have proposed that social structural theories identify proximal, contemporane- ous causes for the behavior of women and men, whereas evolutionary analyses invoke more distal causes that arose early in human history (e.g., Borkenau, 1992; Jackson, 1992; Schaller, 1997). Although the timing of the human adjustment to environmental conditions that is deemed critical is indeed different in the two theories, they propose causes that are similar in their position on the proximal versus distal continuum of causality. Both theories thus identify psychological causes (i.e., evolved dispositions, role expectations) that operate in the present and that exert their impact through more proximal processes (e.g. , emo- tions, perceptions). The social structural perspective is thus in stark contrast to evolutionary psychology models that attribute sex differences in contemporary society to sex- typed evolved mechanisms. The causes of sex differences in evolutionary psychology involve these mechanisms, which are intended to replace the social psychological mechanisms featured in theories that give a key role to social structure.
It also would be inappropriate to conclude that the social structural approach is incompatible with the general perspective of evolutionary theorizing. Social structural analyses suggest an evolved organism, but one in which evolutionary pressures yielded a variety of dispositions , such as the capacity for group living and for culture. These analyses do not imply that people's minds are blank slates, because humans possess facilities, such as for language, that develop in certain ways, given appropriate environ- ments. Moreover, our critique of theorizing in evolutionary psychology is not meant to apply to evolutionary principles in general. Evolutionary reasoning pertaining to humans is diverse (Smith, in press) and provides the basis, not only of evolutionary psychology, but also of models of the relation between biology and culture (Janicki & Krebs, 1998) and human behavioral ecology approaches that emphasize be- havioral variability in response to socioecological condi- tions (Cronk, 1991). The implications of these other evo- lutionary theories for psychological processes have yet to be fully developed and are not discussed in this article.
To illustrate the contrasting approaches of evolution- ary psychology and social structural theory, we first present and discuss each theory . Then we examine their predictions concerning the criteria men and women use in selecting mates. This domain of behavior has been central to evolu- tionary theorizing about human sex differences (e.g., Buss & Schmitt, 1993; Kenrick & Keefe, 1992), and the cross- cultural findings available in this area provide an opportu-
June 1999 • American Psychologist 409
Wendy Woodnity to examine empirically some of the predictions of evolutionary and social structural analyses.
Evolutionary Psychology as an Origin Theory of Sex Differences From the perspective of evolutionary psychology, human sex differences reflect adaptations to the pressures of the differing physical and social environments that impinged on females and males during primeval times (Buss, 1995a; Tooby & Cosmides, 1992). Evolutionary psychologists thus label the environment that produced a species' evolved tendencies as its environment of evolutionary adaptedness (EEA; Cosmides, Tooby, & Barkow, 1992; Symons, 1979, 1992; Tooby & Cosmides, 1990b). They loosely identify the Pleistocene era as the human EEA and generally as- sume that it was populated by hunter-gatherer groups. To the extent that males and females faced different adaptive problems as they evolved, the two sexes developed differ- ent str~tegies to ensure their survival and to maximize their reproductive success. The resolutions to these problems produced evolved psychological mechanisms that are spe- cific to each problem domain and that differ between women and men.
Although humans' evolved mechanisms developed in response to the types of problems consistently encountered by their ancestors and thus are presumed to be universal attributes of humans, environmental input affects how these mechanisms develop in individuals and how they are expressed in behavior (e.g., Buss & Kenrick, 1998). Be- cause culture influences developmental experiences and patterns current situational input, culture is in principle important to the expression of adaptive mechanisms (Tooby & Cosmides, 1992). However, evolutionary psy- chologists have devoted relatively little attention to the
interaction between such broader attributes of the social and cultural environment and the evolved mechanisms that may underlie sex differences. The contextual factors that have interested them generally relate directly to these hy- pothesized mechanisms. For example, Buss and Schmitt (1993) maintained that the characteristics that people seek in mates depend, not only on their sex, but also on whether they are engaging in short-term or long-term mating. Be- cause of a relative neglect of broader social context, evo- lutionary psychologists have generated little understanding of how variation in sex-differentiated behavior arises from developmental factors and features of social structure and culture (for an exception, see Draper & Harpending, 1982).
The aspect of evolutionary theory that has been ap- plied most extensively to sex differences is the theory of sexual selection initially proposed by Darwin (1871) and further developed by Trivers (1972). In the evolutionary psychologists' rendition of these views, sex-typed features of human behavior evolved through male competition and female choice of mates. Because women constituted the sex that devoted greater effort to parental investment, they were a limited reproductive resource for men, who were the less investing sex. Women were restricted in the number of children they could propagate during their life span because of their investment through gestating, bearing, and nursing their children; men did not have these restrictions. Men therefore competed for access to women, and women chose their mates from among the available men. As the more investing sex, women were selected for their wisdom in choosing mates who could provide resources to support their parenting efforts. Women's preferences for such men, in turn, produced sexual selection pressures on men to satisfy these criteria.
Proponents of sexual selection theory argue that sex differences in parental investment favored different strate- gies for reproductive success for men and women and consequently established different adaptive mechanisms governing mating behavior (Buss, 1996; Kenrick, Trost, & Sheets, 1996). It was to men's advantage in terms of fitness outcomes to "devote a larger proportion of their total mat- ing effort to short-term mating than do women" (Buss & Schmitt, 1993, p. 205)-that is, to be relatively promiscu- ous. Women, in contrast, benefited from devoting a smaller proportion of their effort to short-term mating and a larger proportion to long-term mating. Also, because of women's concealed fertilization, men were unable to determine eas- ily which children could proffer the fitness gains that fol- low from genetic relatedness. Men ostensibly adapted to this problem of paternity uncertainty by exerting sexual control over women and developing sexual jealousy and a motive to control women's sexuality (Daly & Wilson, 1998).
According to evolutionary psychologists (e.g., Buss, 1995b; Buss & Kenrick, 1998), sex differences in numer- ous psychological dispositions arose from differing fitness- related goals of women and men that followed from their contrasting sexual strategies. Because men competed with other men for sexual access to women, men's evolved dispositions favor violence, competition, and risk taking.
June 1999 • American Psychologist 410
Women in tum developed a proclivity to nurture and a preference for long-term mates who could support a family. As a result, men strived to acquire more resources than other men in order to attract women, and women developed preferences for successful, ambitious men who could pro- vide resources.
Critical to some of evolutionary psychologists' claims about sex differences is the assumption that an- cestral humans living in the EEA had a hunter-gatherer socioeconomic system (e.g., Buss, 1995b; Cosmides et al., 1992; DeKay & Buss, 1992). The idea of a division of labor in which men hunted while women gathered suggests sex-differentiated pressures linked to survival and reproduction. Such an ancestral division of labor might have favored men who were psychologically spe- cialized for hunting and women who were specialized for gathering. For example, cognitive abilities could have been affected, with men acquiring the superior spatial skills that followed from ancestral hunting, and women acquiring the superior spatial location memory that followed from ancestral gathering (e.g., Geary, 1995; Silverman & Phillips, 1998).
Various mediating processes are implied in evolution- ary psychology models of behavioral sex differences. The first and most important involves some means of retaining effective adaptations in human design and perpetuating them over time. Thus, sex-differentiated psychological mechanisms and developmental programs, like other adap- tations, are "genetic, hereditary, or inherited in the sense that ... their structured design has its characteristic form because of the information in our DNA" (Tooby & Cos- mides, 1990a, p. 37; see also Buss, Haselton, Shackelford, Bleske, & Wakefield, 1998; Crawford, 1998). Some evo- lutionary accounts also emphasize that genetic factors trig- ger biochemical processes that mediate psychological sex differences, especially by means of sex differences in hor- mone production (e.g., Daly & Wilson, 1983; Geary, 1995, 1996). In addition, sex-typed evolved mechanisms are translated into behavioral sex differences by various cog- nitive and affective processes. Establishing these links re- quires theoretical understanding and empirical documenta- tion of the range of processes by which the genetic factors implicated in innate dispositions might affect human be- havior (e.g., Collear & Hines, 1995).
Buss and Kenrick (1998) described evolutionary psy- chology's approach to understanding sex differences as a "metatheory" and summarized it as follows: "Men and women differ in domains where they faced different adap- tive problems over human evolutionary history" (p. 994). These theorists thus derive sex differences from heritable adaptations built into the human species. Because these differences are assumed to follow from evolutionary adap- tations, they are predicted to occur as central tendencies of male versus female behavior. Human behavior would thus be characterized by a deep structure of sex-differentiated dispositions, producing similar, albeit not identical, behav- ioral sex differences in all human societies.
Critique of the Evolutionary Origin Theory
A number of questions can be raised about evolutionary psychology's account of the origins of sex differences. One consideration is that evolutionary analyses have generally identified adaptations by relying on "informal arguments as to whether a presumed function is served with sufficient precision, economy, efficiency, etc. to rule out pure chance as an adequate explanation" (Williams, 1966, p. 10). Ex- planations that reflect this approach consist of an analysis of the functional relations served by a particular psycho- logical mechanism, along with the construction of a con- vincing story about how the adaptation might have made an efficient contribution to genetic survival or to some other goal contributing to reproduction in the EEA. These expla- nations serve as hypotheses that require additional valida- tion and thus can be useful for initiating scientific research.
In developing these analyses of the possible functions of behaviors, evolutionary scientists face special challenges in distinguishing adaptations from other possible products of evolution-for example, features that were random or that had utility for one function but were subsequently coopted to fulfill a new function (see Buss et al., 1998; Gould, 1991; Williams, 1966). Moreover, the products of evolution must be distinguished from the products of cul- tural change. Behaviors that provide effective solutions to problems of reproduction and survival can arise from in- ventive trial-and-error among individuals who are geneti- cally indistinguishable from other members of their living groups; such beneficial behaviors are then imitated and transmitted culturally.
An understanding of humans' primeval environment might help validate evolutionary hypotheses because adap- tations evolved as solutions to past environmental chal- lenges. Various bodies of science have some relevance, including observational studies of other primates, the fossil record, and ethnographic studies. However, models of hu- man nature constructed from the behavior of nonhuman primates do not yield a uniform picture that reflects key features of sex differences in modem human societies (see Fedigan, 1986; Strier, 1994; Travis & Yeager, 1991). Sim- ilarly ambiguous concerning sex differences are the models of early human social conditions that paleontologists and paleoanthropologists have developed from fossil evidence. Anthropologists continue to debate fundamental points- for example, whether hunting of dangerous prey might have emerged during the period that is usually identified as the human EEA (e.g., Potts, 1984; Rose & Marshall, 1996). As a consequence, assumptions that certain traits were adaptive and consequently are under genetic control cannot be firmly supported from analyzing attributes of the EEA. Moreover, early human societies likely took a wide variety of forms during the period when the species was evolving toward its modem anatomical form (Foley, 1996). Vari- ability in social organization is consistent with observations of more contemporary hunter-gatherer societies, which show great diversity in their social organization (Kelly, 1995). For example, studies of power relations between the sexes across diverse cultures show variability in the extent
June 1999 • American Psychologist 411
to which men control women's sexuality (Whyte, 1978), although evolutionary psychologists have assumed that this control is a defining feature of male-female relations. Therefore, because the EEA likely encompassed a variety of conditions, tracing humans' evolution requires under- standing of the timing, social organization, and ecological circumstances of multiple periods of adaptation (Foley, 1996). The ambiguity and complexity of the relevant sci- entific findings leave room for evolutionary psychologists to inadvertently transport relatively modem social condi- tions to humans' remote past by inappropriately assuming that the distinctive characteristics of contemporary rela- tions between the sexes were also typical of the EEA.
Given the difficulty of knowing the functions of be- haviors and the attributes of the EEA, other types of sci- entific evidence become especially important to validating the claims of evolutionary psychologists. The most con- vincing evidence that a behavioral pattern reflects an ad- aptation would be that individuals who possessed the ad- aptation enjoyed a higher rate of survival and reproduction than individuals who did not possess it. However, such evidence is difficult, if not impossible, to produce. Because humans' evolved mechanisms emerged in relation to past selection pressures, present reproductive advantage does not necessarily reflect past advantage, and evolutionary psychologists have warned against relying on measures of current reproductive success to validate hypothesized ad- aptations (Buss, 1995a; Tooby & Cosmides, 1992). In the absence of evidence pertaining to reproductive success, scientists might document the genetic inheritance of pos- tulated mechanisms and the processes by which genetic factors result in sex differences in behavior. However, for the psychological dispositions considered in this article, such evidence has not been produced. Instead, the scientific case for these sex-differentiated evolved dispositions rests on tests of evolutionary psychologists' predictions concern- ing the behavior of men and women in contemporary societies (e.g., Buss & Schmitt, 1993; Kenrick & Keefe, 1992). We evaluate some of these predictions in this article.
Social Structural Theory as an Origin Theory of Sex Differences A respected tradition in the social sciences locates the origins of sex differences, not in evolved psychological dispositions that are built into the human psyche, but in the contrasting social positions of women and men. In contem- porary American society, as in many world societies, women have less power and status than men and control fewer resources. This feature of social structure is often labeled gender hierarchy, or in feminist writing it may be called patriarchy. In addition, as the division of labor is realized in the United States and many other nations, women perform more domestic work than men and spend fewer hours in paid employment (Shelton, 1992). Although most women in the United States are employed in the paid workforce, they have lower wages than men, are concen- trated in different occupations, and are thinly represented at the highest levels of organizational hierarchies (Jacobs, 1989; Reskin & Padavic, 1994; Tomaskovic-Devey, 1995).
From a social structural perspective, the underlying cause of sex-differentiated behavior is this concentration of men and women in differing roles.
The determinants of the distribution of men and women into social roles are many and include the biolog- ical endowment of women and men. The sex-differentiated physical attributes that influence role occupancy include men's greater size and strength, which gives them priority in jobs demanding certain types of strenuous activity, es- pecially activities involving upper body strength. These physical attributes of men are less important in societies in which few occupational roles require these attributes, such as postindustrial societies. Also important in relation to role distributions are women's childbearing and in many soci- eties their activity of suckling infants for long periods of time; these obligations give them priority in roles involving the care of very young children and cause conflict with roles requiring extended absence from home and uninter- rupted activity. These reproductive activities of women are less important in societies with low birthrates, less reliance on lactation for feeding infants, and greater reliance on nonmatemal care of young children.
In general, physical sex differences, in interaction with social and ecological conditions, influence the roles held by men and women because certain activities are more efficiently accomplished by one sex. The benefits of this greater efficiency can be realized when women and men are allied in cooperative relationships and establish a division of labor. The particular character of the activities that each sex performs then determines its placement in the social structure (see Wood & Eagly, 1999). As historians and anthropologists have argued (e.g., Ehrenberg, 1989; Harris, 1993; Lerner, 1986; Sanday, 1981), men typically special- ized in activities ( e.g., warfare, herding) that yielded greater status, wealth, and power, especially as societies became more complex. Thus, when sex differences in status emerged, they tended to favor men.
The differing distributions of men and women into social roles form the basis for a social structural metatheory of sex differences, just as evolutionary theory provides a metatheory. The major portion of this social structural theory follows from the typical features of the roles of men and women. Thus, the first metatheoretical principle de- rives from the greater power and status that tends to be associated with male-dominated roles and can be succinctly stated as follows: Men's accommodation to roles with greater power and status produces more dominant behav- ior, and women's accommodation to roles with lesser power and status produces more subordinate behavior (Ridgeway & Diekema, 1992). Dominant behavior is con- trolling, assertive, relatively directive and autocratic, and may involve sexual control. Subordinate behavior is more compliant to social influence, less overtly aggressive, more cooperative and conciliatory, and may involve a lack of sexual autonomy.
The second metatheoretical principle follows from the differing balance of activities associated with the typical roles of each sex. Women and men seek to accommodate sex-typical roles by acquiring the specific skills and re-
June 1999 • American Psychologist 412
sources linked to successful role performance and by adapt ing their social behavior to role requirements. A variety of sex-specific skills and beliefs arise from the typical family and economic roles of men and women, which in many societies can be described as resource provider and home maker. Women and men seek to accommodate to these roles by acquiring role-related skills, for example, women learning domestic skills such as cooking and men learning skills that are marketable in the paid economy. The psy chological attributes and social behaviors associated with these roles have been characterized in terms of the distinc tion between communal and agentic characteristics (Bakan, 1966; Eagly, 1987). Thus, women's accommodation to the domestic role and to female-dominated occupations favors a pattern of interpersonally facilitative and friendly behav iors that can be termed communal. In particular, the as signment of the majority of child rearing to women encour ages nurturant behaviors that facilitate care for children and other individuals. The importance of close relationships to women's nurturing role favors the acquisition of superior interpersonal skills and the ability to communicate nonver bally. In contrast, men's accommodation to the employ ment role, especially to male-dominated occupations, fa vors a pattern of assertive and independent behaviors that can be termed agentic (Eagly & Steffen, 1984). This argu ment is not to deny that paid occupations show wide variation in the extent to which they favor more masculine or feminine qualities. In support of the idea that sex differentiated behaviors are shaped by paid occupations are demonstrations that to the extent that occupations are male dominated, they are thought to require agentic personal qualities. In contrast, to the extent that occupations are female dominated, they are thought to require communal personal qualities (Cejka & Eagly, 1999; Glick, 1991).
In social structural theories, differential role occu pancy affects behavior through a variety of mediating pro cesses. In social role theory (Eagly, 1987; Eagly et al., in press), an important mediating process is the formation of gender roles by which people of each sex are expected to have characteristics that equip them for the tasks that they typically carry out. These expectations encompass the pre ferred or desirable attributes of men and women as well as their typical attributes. Gender roles are emergents from the productive work of the sexes; the characteristics that are required to perform sex-typical tasks become stereotypic of women or men. To the extent that women more than men occupy roles that demand communal behaviors, domestic behaviors, or subordinate behaviors for successful role performance, such tendencies become stereotypic of women and are incorporated into a female gender role. To the extent that men more than women occupy roles that demand agentic behaviors, resource acquisition behaviors, or dominant behaviors for successful role performance, such tendencies become stereotypic of men and are incor porated into a male gender role. Gender roles facilitate the activities typically carried out by people of each sex. For example, the expectation that women be other-oriented and compassionate facilitates their nurturing activities within
the family as well as their work in many female-dominated occupations (e.g., teacher, nurse, social worker).
People communicate gender-stereotypic expectations in social interaction and can directly induce the targets of these expectations to engage in behavior that confirms them (e.g., Skrypnek & Snyder, 1982; Wood & Karten, 1986). Such effects of gender roles are congruent with theory and research on the behavioral confirmation of stereotypes and other expectancies (see Olson, Roese, & Zanna, 1996). Gender-stereotypic expectations can also affect behavior by becoming internalized as part of individuals' self-con cepts and personalities (Feingold, 1994). Under such cir cumstances, gender roles affect behavior through self-reg ulatory processes (Wood, Christensen, Hebl, & Rothgerber, 1997). The individual psychology that underlies these pro cesses is assumed to be the maximization of utilities. Peo ple perceive these utilities from the rewards and costs that emerge in social interaction, which takes place within the constraints of organizational and societal arrangements.
Gender roles coexist with specific roles based on factors such as family relationships and occupation. These specific social roles contribute directly to sex-differentiated behavior when women and men are differently distributed into them-for example, women into the homemaker role and men into the provider role. In contrast, when men and women occupy the same specific social role, sex differ ences would tend to erode because specific roles are con straining (e.g., Eagly & Johnson, 1990). However, gender roles ordinarily continue to have some impact on beha vior, even in the presence of specific roles (see Gutek & Morasch, 1983; Moscowitz, Suh, & Desaulniers, 1994; Ridgeway, 1997). Moreover, experimental evidence (e.g., Hembroff, 1982) suggests that people combine or average the expectations associated with specific roles and more diffuse roles such as gender roles in a manner that weights each set of expectations according to its relevance to the task at hand.
The social structural perspective provides a broad theoretical outline within which many social scientific the ories of sex-differentiated behavior can be placed. These theories focus on different aspects of the processes by which societies produce sex-differentiated behavior, and many theories have spawned detailed predictions and a substantial body of empirical research (see Beall & Stern berg, 1993; Canary & Dindia, 1998; England & Browne, 1992). For example, developmental psychologists have studied socialization in the family, school, and peer group. Social psychologists have examined the impact of gendered self-schemas, men's greater status, sex-differentiated ex pectations about behavior, and gendered patterns of social interaction. Sociologists have implicated organizational factors such as discriminatory employment practices, soci etal factors such as men's greater ownership of capital, and cultural factors such as the ideologies that legitimize gen der inequality. Social scientists have thus provided an array of interrelated theories, each of which illuminates certain aspects of the processes by which sex-differentiated behav ior is produced.
June 1999 • American Psychologist 413
In summary, in social structural accounts, women and men are differently distributed into social roles, and these differing role assignments can be broadly described in terms of a sexual division of labor and a gender hierarchy. This division of labor and the patriarchal hierarchy that sometimes accompanies it provide the engine of sex- differentiated behavior because they trigger social and psy- chological processes by which men and women seek some- what different experiences to maximize their outcomes within the constraints that societies establish for people of their sex. Sex differences in behavior thus reflect contem- poraneous social conditions.
Response to Critiques of the Social Structural Origin Theory A number of criticisms have been leveled against the social structural theory of sex differences and more specifically against social role theory (see Archer, 1996; Buss, 1996). At least some evolutionary psychologists have expressed skepticism that culture and social structure could have any independent causal role in relation to behavior. Instead, culture and social structure are seen as reflecting the un- derlying logic of evolved dispositions, and consequently they do not constitute the causal force underlying behav- ioral sex differences (Buss, 1995a; Tooby & Cosmides, 1992). However, from our perspective, culture and social structure can influence behavior. Culture consists of knowl- edge, beliefs, and evaluations shared among members of a society and reflects, not only the biological endowment of humans, but also the constraints of their social and physical environments. Social structure reflects culture and consists of "persisting and bounded patterns of behavior and inter- action among people or positions" (House, 1995, p. 390). Gender roles and other social roles are simultaneously aspects of culture, because they represent shared knowl- edge, and of social structure, because they represent bounded patterns of interaction.
Another criticism is that in social structural theories, individuals are treated as mere passive receptacles of the roles they are assigned (Buss, 1996). Although social sci- entists often do refer to role assignment, this term does not imply that people are typically assigned to roles arbitrarily, as if they were passive actors in the social system. On the contrary, social and organizational psychologists have demonstrated that the assumption of roles is a complex and dynamic process (e.g., Kerckhoff, 1995; Pfeffer, 1998). In deciding whether to attempt to assume particular roles at all, individuals take their own attributes, skills, and per- sonal preferences into account, although in some cultural contexts some roles are imposed on people regardless of their own preferences (e.g., the practice of early betrothal of girls). In general, social systems are arranged to shape people's self-concepts, skills, beliefs, and values so that the majority of people actively seek out experiences that help them to become appropriate occupants of existing social roles by meeting the expectations of these roles.
Evolutionary psychologists also claim that socio- cultural theorists view gender roles as "essentially arbi- trary" (Buss, 1996, p. 19) or as arising by "historical
accident" (Archer, 1996, p. 915). On the contrary, as we have explained, the content of gender roles is not arbitrary but is embedded in social structure and culture. Roles must thus facilitate the endeavors of a society, if its members are to prosper and survive. Therefore, different types of role systems become effective under differing circumstances. For example, in industrial economies, many roles are or- ganized by a market pricing system that takes into account factors such as ownership of property and contribution to production (see Fiske, 1992). The analytical frameworks for understanding how systems of social roles change over time have been developed by scholars in other disciplines (e.g., Diamond, 1997; Toynbee, 1934-1961). Yet, under- standing the principles by which women and men distribute themselves into a society's roles is part of the agenda of social psychologists as well as other social scientists.
A related criticism is that from a social structural perspective, "differences between cultures are random with respect to evolutionary hypotheses and therefore that, for example, sex differences should occur as frequently in one direction as the other" (Tooby & Cosmides, 1989, p. 37). However, our theoretical perspective is not consistent with random variation in sex differences across societies. In- stead, societal variation in the roles of men and women depends on multiple factors, including men's greater size and physical strength, women's reproductive activities, and the activities required by a society's economy and social organization, which in tum reflect technological develop- ments and the current ecology. Because these factors are not randomly distributed, certain types of social arrange- ments are more common than others, and sex differences appropriate to the common arrangements should be more frequent than reversals of these differences.
The social structural approach has also been criticized for treating the minds of women and men as identical except by virtue of the constraints that follow from exter- nally assigned roles (Buss, 1996 ). We acknowledge that the social structural perspective does imply that differences in the minds of women and men arise primarily from experi- ence and socialization, which reflect the physical attributes of women and men and the characteristics of the social and physical environment. This assumption that humans' psy- chological attributes are minimally constrained by geneti- cally encoded sex differences is consistent with the diver- sity of behaviors and skills exhibited by men and women across societies and within societies. Yet, our perspective is fully compatible with the idea that people possess evolved facilities, such as for language, that develop in predictable ways in appropriate environments.
Sex Differences in Mate Selection Criteria Predicted From Evolutionary Psychology and Social Structural Theory One reasonable area for comparing the predictive power of the evolutionary and the social structural origin theories of sex differences is human mating behavior, especially the criteria that people use for selecting mates. Evolutionary
June 1999 • American Psychologist 414
predictions have been articulated especially clearly for mating activities, and these behaviors can also be used to test a social structural perspective. Furthermore, empirical findings concerning mate selection preferences have been well-established for many years in the literature on the sociology of the family (e.g., Coombs & Kenkel, 1966). Powers's (1971) summary of 30 years of research con- cluded that at least in the United States, women generally prefer mates with good earning potential, whereas men prefer mates who are physically attractive and possess good domestic skills. Furthermore, women typically prefer a mate who is older than them, whereas men prefer a mate who is younger. Feingold's (1990, 1991, 1992a) meta- analyses of studies drawn from various research paradigms established that the sex differences in valuing potential mates' earning potential and physical attractiveness are robust, despite sex similarity on most criteria for selecting mates. Subsequent research based on a national probability sample of single adults provided further confirmation of the sex differences in age preferences as well as in valuing earning potential and physical attractiveness (Sprecher, Sullivan, & Hatfield, 1994).
Evolutionary psychologists have adopted mate prefer- ences as signature findings of their analysis. Women's valuing of mates' resources and men's valuing of mates' youth and physical attractiveness are thought to arise from the different parental investment of the sexes that was outlined in Trivers's (1972) sexual selection theory. It is commonly argued that women, as the more investing sex, seek mates with attributes that can support their parenting efforts. However, human mate selection does not follow a strict version of Trivers' s males-compete-and-females- choose model, because among humans, selection is a prod- uct of the behavior of both sexes, a process Darwin ( 1871) called "dual selection." In Buss's (1989a) account, male choice derives from women's time-limited reproductive capacity and the tendency for men to seek mates with attributes that suggest such capacity. In Kenrick and Keefe's (1992) account, men and women are both selective about potential mates and both invest heavily in offspring but with different kinds of resources. In particular, "males invest relatively more indirect resources (food, money, protection, and security), and females invest relatively more direct physiological resources ( contributing their own bodily nutrients to the fetus and nursing child)" (Kenrick & Keefe, 1992, p. 78). As a result, women prefer mates who can provide indirect resources, and men prefer healthy mates with reproductive potential. 1
In contrast, from a social structural perspective, the psychology of mate selection reflects people's effort to maximize their utilities with respect to mating choices in an environment in which these utilities are constrained by societal gender roles as well by as the more specific ex- pectations associated with marital roles. Consistent with these ideas, Becker's (1976) economic analysis of mating decisions characterized marriage as occurring between utility-maximizing men and women who can reach an equilibrium with a variety of types of exchanges, including, for example, an exchange between men's wages and worn-
en's household production and other attributes such as education and beauty. This cost-benefit analysis of mating appears even on occasion in the writings of evolutionary scientists. For example, Tattersall (1998) maintained that behavioral regularities, such as sex differences in mate selection criteria, are as likely to be due to rational eco- nomic decisions as to inherited predispositions, and Hrdy ( 1997) wrote that "a woman's preference for a wealthy man can be explained by the simple reality that ... males mo- nopolize ownership of productive resources" (p. 29).
The outcomes that are perceived to follow from mat- ing decisions depend on marital and family arrangements. To the extent that women and men occupy marital and family roles that entail different responsibilities and obli- gations, they should select mates according to criteria that reflect these divergent responsibilities and obligations. Consider, for example, the family system based on a male provider and a female domestic worker. This system be- came especially pronounced in industrial economies and is still prevalent in many world societies. To the extent that societies have this division of labor, women maximize their outcomes by seeking a mate who is likely to be successful in the economic, wage-earning role. In tum, men maximize their outcomes by seeking a mate who is likely to be successful in the domestic role.
The sex differences in the preferred age of mates also can be understood as part of the general tendency of men and women to seek partners likely to provide a good fit to their society's sexual division of labor and marital roles. Specifically, the marital system based on a male breadwin- ner and a female homemaker favors the age gap in mar- riage. Marriageable women who are younger than their potential mates tend to have lesser wages, social status, and education and knowledge than women who are the same age as potential mates. With the combination of a younger, less experienced woman and an older, more experienced man, it would be easier to establish the power differential favoring men that is normative for marital roles defined by a male breadwinner and a female domestic worker (Lips, 1991; Steil, 1997). Moreover, compared with somewhat older women, young women lack independent resources and therefore are more likely to perceive that their utilities are maximized in the domestic worker role. In complemen- tary fashion, older men are more likely to have acquired the economic resources that make them good candidates for the provider role. The older man and younger woman thus fit more easily than same-age partners into the culturally ex- pected pattern of breadwinner and homemaker.
1 Darwin (1871) expressed skepticism about the applicability of the processes of sexual selection to modern human societies. He argued that sexual selection was more powerful among early humans, who were guided by instinctive passions, than among contemporary members of society, who show greater foresight and reason in mating behavior. In fact, Darwin maintained that "civilized men are largely attracted by the mental charms of women, by their wealth, and especially by their social position" (Darwin, 1871, p. 178).
June 1999 • American Psychologist 415
Cross-Cultural Evidence for Sex Differences in Mate Preferences
Evolutionary psychologists' predictions that women select for resources and older age and men for attractiveness and younger age have been examined cross-culturally. Buss's (1989a; Buss et al., 1990) impressive study in 37 cultures of the characteristics that people desire in mates suggested that consistent with evolutionary psychology, these sex differences in mate preferences emerged cross-culturally. Similarly, Kenrick and Keefe (1992) examined the pre- ferred ages of mates in five countries and across various time periods in the 20th century and concluded that all provided evidence of sex differences in these preferences. Specifically, for dating and marriage, women preferred older men and men preferred younger women, although men's preferences were moderated by their age, with teen- age boys preferring girls of similar age.2
On the basis of these investigations, evolutionary ac- counts have emphasized the cross-cultural commonality in women's preference for resources and older age and men's preference for attractiveness and younger age. According to Buss (1989a) and Tooby and Cosmides (1989), unifor- mity across diverse cultures and social circumstances sug- gests powerful sex-differentiated evolved mechanisms that reflect an innate, universal human nature. Kenrick and Keefe (1992) also argued that "invariance across cultures is evidence that supports a species-specific, rather than a culture-specific, explanation" (p. 76).
Despite evidence for cross-cultural commonality in sex differences in mate selection criteria, these investiga- tions also yielded evidence for cultural variation. For ex- ample, Kenrick and Keefe (1992) found that the preference for younger wives was evident among Philippine men of all ages, but only among older men (i.e., age 30 or over) in the United States. However, the simple existence of uniformity or variability does not provide a definitive test of either the evolutionary or the social structural origin theory. Al- though evolutionary psychologists emphasize uniformity and social structural theorists emphasize variability, both perspectives have some power to explain both of these cross-cultural patterns. To account for uniformity, social structuralists can point to similarities in the sexual division of labor in the studied societies and can argue that these similarities produce these relatively invariant sex differ- ences. As Buss (1989a) noted, his 37 cultures, which were drawn from 33 nations, were biased toward urbanized cash-economy cultures, with 54% from Europe and North America. Furthermore, respondents selected from each so- ciety tended to be young, comparatively well-educated, and of relatively high socioeconomic status. To the extent that these societies similarly defined the roles of women and men and that the respondents were similarly placed in these societies' social structures, commonality in the sex differ- ences that follow from social structure should characterize these societies.
To account for cross-cultural variability, both evolu- tionary and social structural origin theories recognize that developmental processes and social factors that are unique
to each society direct behavior in ways that can yield variability in sex differences across cultures. Beyond this insight that some evidence of cross-cultural variability would not surprise theorists in either camp, the particular pattern of cross-cultural variation provides an informative test of the mechanisms underlying sex differences. Specif- ically, the social structural argument that a society's sexual division of labor and associated gender hierarchy are re- sponsible for sex differences in social behavior yields pre- dictions concerning cross-cultural variability in mate preferences.
In the nations included in Buss et al.' s (1990) cross- cultural sample, whose economies ranged from agrarian to postindustrial, some cultures were still strongly marked by this division of labor between the provider and domestic worker, whereas other cultures had departed from it. In advanced economies like the United States, women have entered the paid labor force and spend a smaller proportion of their time in domestic labor (Haas, 1995; Shelton, 1992). Although the tendency for men to increase their hours of domestic work is much more modest, the lives of men and women become more similar with greater gender equality. Therefore, people of both sexes should lessen their empha- sis on choosing mates whose value is defined by their fit to the division between domestic work and wage labor. Even in postindustrial economies such as the United States, however, the sex-typed division of labor remains in mod- ified form, with men devoting longer hours than women to wage labor and women devoting longer hours to domestic work (e.g., Ferree, 1991; Presser, 1994; Shelton, 1992). Therefore, the social structural prediction is that the sex differences in mate selection criteria that follow from the male-female division of labor should be substantially weakened in societies characterized by greater gender equality, albeit they should still be present to the extent that complete equality has not been achieved.3
Reanalysis of Buss et al.'s (1990) 37 Cultures Data
To evaluate whether the division of labor within a society could explain the mate preferences of men and women, we reanalyzed Buss et al.'s (1990) 37 cultures data. Our efforts focused on men's tendencies to select wives for domestic skill and younger age and women's tendencies to select husbands for earning capacity and older age. To test the hypothesis that a higher level of gender equality lessens these sex differences, we represented societies' gender equality in terms of archival data available from the United Nations (United Nations Development Programme, 1995).
Buss et al. (1990) derived the data on criteria for
2 Although Kenrick and Keefe (1992) showed that teenage boys prefer girls of similar age, this tendency is most likely a product of the lower age limits that exist for culturally and maturationally appropriate marital partners (Broude, I 992).
3 Prior efforts to test social structural hypotheses within Buss et al.' s (1990) 37 cultures data produced mixed or nonsignificant findings (Buss, 1989a; Glenn, 1989).
June 1999 • American Psychologist 416
selecting mates from questionnaire measures of preferences for a wide range of characteristics that might be desired in a mate: (a) One instrument obtained rankings of a set of 13 characteristics according to "their desirability in someone you might marry" (p. 11); (b) the other instrument obtained ratings on a 4-point scale of each of 18 characteristics on "how important or desirable it would be in choosing a mate" (p. 11). Buss et al. represented each culture by the male and female respondents' mean ranking of each of the 13 mate selection criteria and by their mean rating of each of the 18 criteria. A separate question inquired about pref erences for a spouse's age. The data that we reanalyzed consisted of mean preferences for each culture.
Our reanalysis confirmed Buss et al.' s ( 1990) con clusion that women placed more value than men on a mate's wage-earning ability. Furthermore, consistent with the greater domestic responsibility of women than men in most cultures, men valued good cook and house- keeper more than women did, a sex difference that has received little attention from evolutionary psychologists. When the sex differences in the mean preference ratings were averaged across the cultures, this difference was of comparable magnitude to those obtained on the at tributes most strongly emphasized by evolutionary psy chologists. Specifically, in both the rating and ranking data, the criteria of good earning capacity, good house- keeper and cook, and physically attractive produced the largest sex differences. The appropriateness of focusing on the criteria pertaining to earning ability and domestic skill within Buss et al.' s data was also supported by the good agreement across the ranking and rating data sets for sex differences in the valuation of the qualities of financial prospect, r(33) = .76, p < .001, and domestic skill, r(33) = .68, p < .001, whereas the agreement in the valuation of physical attractiveness was poorer, r(33) = .34, p < .05. In addition, as Buss et al. reported, the sex difference in the preferred age of mates was fully intact in the 37 cultures data. 4
Additional evidence for the social structural predic tions emerged when we evaluated the pattern of sex dif ferences in preferences across societies. Consistent with the division of labor principle, a substantial relation emerged between the sex difference in valuing a spouse's domestic skills and the sex difference in valuing a spouse's capacity to provide a good income. Specifically, on the basis of the ranking measure, the sex differences in the good earning capacity criterion and the good housekeeper criterion were correlated across the cultures, r(33) = .67, p < .001. On the basis of the rating measure, the sex differences in the financial prospect criterion and the housekeeper-cook cri terion were also correlated, r(35) = .38, p < .05. These positive correlations indicate that to the extent that women more than men reported seeking a mate who is a good breadwinner, men more than women reported seeking a mate who is a good homemaker. In addition, the sex difference in the preferred age of one's spouse bore a positive relation to the sex difference in preference for a good earner, r(3 3) = .34, p < .05 for the ranking data, and r(35) = .32, p < .06 for the rating data. Similarly, the sex
difference in preferred age bore a positive relation to the sex difference in preference for a good housekeeper and cook, r(33) = .58,p < .001 for the ranking data, and r(35) = .60, p < .001 for the rating data. These relationships show that to the extent that the sex difference in the preferred age of spouses was large, women more than men preferred mates who were good providers and men more than women preferred mates who were good domestic workers. The division of labor provides the logic of all of these relationships: Women who serve in the domestic role are the complement of men who serve as breadwinners, and the combination of older husbands and younger wives facilitates this form of marriage.
Analysis of gender equality. To test our hy pothesis that sex differences in mate preferences erode to the extent that women and men are similarly placed in the social structure, we sought cross-national indicators of gen der equality. Among the many such indicators compiled by United Nations researchers, the most direct indicator of gender equality is the aggregate Gender Empowerment Measure, which represents the extent to which women participate equally with men in economic, political, and decision-making roles (United Nations Development Pro gramme, 1995). This index increases as (a) women's per centage share of administrative and managerial jobs and professional and technical jobs increases, (b) women's percentage share of parliamentary seats rises, and (c) wom en's proportional share of earned income approaches parity with men's.
The Gender-Related Development Index is another useful indicator of societal-level gender equality provided by United Nations researchers. It increases with a society's basic capabilities to provide health (i.e., greater life expect ancy), educational attainment and literacy, and wealth, but imposes a penalty for gender inequality in these capabilities (United Nations Development Programme, 1995). Whereas this measure reflects equality in basic access to health care, education and knowledge, and income, the Gender Em powerment Measure is a purer indicator of equal partici pation in economic and political life.
In the set of 37 cultures, the Gender Empowerment Measure and the Gender-Related Development Index were correlated, r(33) = .74, p < .OOI, and both of these indexes were moderately correlated with general indexes of human development and economic development. One lim itation of the indexes of gender equality is that they are based on data from the early 1990s. Because Buss et al.' s (1990) data were collected in the mid-1980s, these indexes are from a slightly later time period, but the relative posi-
4 We did not also focus on the criterion of ambition and industri- ousness because it produced a substantially smaller sex difference in the 37 cultures data than the criteria of good earning capacity, good house- keeper and cook, and physically attractive. From the social structural perspective, industriousness is important for performance of domestic work as well as wage labor, and therefore both men and women should seek this quality in mates under the traditional division of labor between homemakers and providers.
June 1999 • American Psychologist 417
Table 1 Correlations of Mean Rankings and Ratings of Mate Selection Criteria With United Nations Indexes of Gender Equality for Buss et al.' s ( 1990) 3 7 Cultures Sample
Ranked criteria Roted criteria
Gender Gender-Related Gender Gender-Related Empowerment Development Empowerment Development
Measure Index Measure Index Mote selection criterion and rater (n = 33) (n = 34) (n = 35) (n = 36)
Good earning capacity (financial prospect) Sex difference -.43* __ 33t -_29t -.23 Women -.29 -.18 -.49** -.42** Men .24 .27 -.40* -.36*
Good housekeeper (and cook) Sex difference -.62*** -.54** -.61*** -.54** Women .04 -.01 .11 -.07 Men -.46** -.42* -.60*** -.61***
Physically attractive (good looks) Sex difference .13 -.12 .20 .18 Women Men
.14
.20 _34t .28
-.45** __ 33t
-.25 -.14
Note. The criteria were described slightly differently in the ranking and the roting tasks: The ranking term is given first, with the roting term following in parentheses. Higher values on the gender equality indexes indicate greater equality. For the preferences of women or men, higher values of the mean rankings and ratings of mote selection criteria indicate greater desirability in a mate; therefore, a positive correlation indicates on increase in the desirability of a criterion as gender equality increased, and a negative correlation indicates o decrease. Sex differences in these preferences were calculated as female minus male means for good earning capacity and male minus female means for good housekeeper and physically attractive. A positive correlation thus indicates an increase in the sex difference as gender equality increased, and a negative correlation indicates o decrease in the sex difference. t p < .10. * p < .05. •• p < .01. *** p < .001.
tions of the cultures should remain approximately the 5same.
To examine the relation between societal gender equality and mate preferences, we calculated the correla tions of these indexes with the sex differences in valuing a mate as a breadwinner and as a domestic worker-the two criteria most relevant to the traditional division of labor. These correlations for the ranking and the rating data, which appear in Table 1, are generally supportive of the social structural predictions. As the Gender Empowerment Measure increased in value, the tendency decreased for women to place greater emphasis than men on a potential spouse's earning capacity, although the correlation with the rated criterion was relatively weak. Also, as the Gender Empowerment Measure increased, the tendency decreased for men to place greater emphasis than women on a poten tial spouse's domestic skills. As expected in terms of the Gender-Related Development Index's less direct represen tation of the similarity of the roles of women and men, its correlations with these sex differences were somewhat weaker.
The preference data for each sex reported in Table I provide insight into these sex-difference findings. For good housekeeper and cook, the correlations for both the rating data and the ranking data indicated that as gender equality increased, men decreased their interest in choosing mates for their skill as domestic workers, and women showed no change in this preference. In contrast,
418
for good earning capacity, as gender equality increased, women decreased their emphasis on mates' earning po tential in the rating data (although nonsignificantly in the ranking data). However, men's preferences for good earning capacity are more difficult to interpret because their relations to gender equality were inconsistent across the ranking and rating measures. Inconsistencies between the two measures may reflect that rankings are judgments of the relative importance of the criteria in relation to the others in the list, whereas ratings are judgments of the absolute importance of the different criteria.
As shown in Table 2, examination of preferences for a spouse's age showed that as gender equality increased, women expressed less preference for older men, men ex pressed less preference for younger women, and conse quently the sex difference in the preferred age of mates became smaller. These relations suggest that sex differ-
5 Another compromise consisted of representing differing sub samples from the same broader culture (e.g., mainland United States and Hawaiian United States) with the same values of the United Nations indexes. For the Gender Empowerment Measure and the Gender-Related Development Index, data for all represented nations were published in 1995, with the exception of data for two nations published in 1996 and one in 1997 (United Nations Development Programme, 1995, 1996, 1997). For two cultures, ranking data for mate selection preferences were not available.
June 1999 • American PsychologistTable 2 Correlations of Mean Preferred Age Difference Between Self and Spouse With United Nations Indexes of Gender Equality for Buss et al.' s { 1990) 37 Cultures Sample
Gender Empowerment Gender-Related Measure Development Index
Rater (n = 35) (n = 36)
Sex difference -.73*** -.70*** Women -.64*** -.57*** Men .70*** .70***
Note. Higher values on the gender equality indexes indicate greater equality. Positive ages indicate preference for an older spouse, and negative ages indicate preference for o younger spouse. Therefore, for the preferences of women, a negative correlation indicates a decrease in the tendency to prefer an older spouse as gender equality increased, whereas for the preferences of men, a positive correlation indicates a decrease in the tendency to prefer a younger spouse. Because the sex difference in preferred age was calculated as female minus male mean preferred spouse age in relation to self, a negative correlation indicates a decrease in the sex difference in preferred age as gender equality increased. *** p < .001.
ences in age preferences reflect a sex-differentiated division of labor. 6
Interpretation of the magnitudes of the correlations reported in Tables 1 and 2 should take several consider- ations into account. One feature limiting the strength of these relationships is the assessment of the mate selection preferences with one-item questionnaire measures. Also, the indexes of gender equality imperfectly represented the critical conceptual variable, the extremity of the division of labor between male providers and female homemakers. In addition, the sampling of respondents was not uniformly implemented across the 37 cultures, nor would these sam- ples have corresponded to those that contributed to the indexes of gender equality. Finally, there may be a time lag between the social and economic changes reflected in these indexes and shifts in the individual preferences that con- stitute the 37 cultures data. For these several reasons, it is plausible to conclude that the correlations we report under- estimate the true magnitude of the predicted relationships.
Preference for physical attractiveness. As also shown in Table 1, correlations between the sex differ- ence in valuing potential mates' physical attractiveness and the United Nations indexes of gender equality were low and nonsignificant. These findings are not surprising, be- cause this mate selection criterion does not mirror the division between wage labor and domestic labor in the manner that earning potential, domestic skill, and age do. Nevertheless, under some circumstances, physical attrac- tiveness may be part of what people exchange for partners' earning capacity and other attributes.
Assuming that attractiveness is sometimes exchanged for other gains, the social structural perspective offers possibilities for understanding its value. Research on the physical attractiveness stereotype has shown that attractive-
ness in both sexes conveys several kinds of meaning- especially social competence, including social skills, socia- bility, and popularity (Eagly, Ashmore, Makhijani, & Longo, 1991; Feingold, 1992b). Therefore, men's greater valuing of attractiveness might follow from the greater importance of this competence in women's family and occupational roles, including women's paid occupations in postindustrial societies (Cejka & Eagly, 1999; Lippa, 1998), and the consequent inclusion of this competence in the female gender role. If women's roles demand greater interpersonal competence in societies with greater and lesser gender equality, the tendency for men to place greater value on mates' attractiveness would not covary with indexes that assess equality.
Another possibility is that the value of attractiveness stems from its perceived association with the ability to provide sexual pleasure. This idea receives support from research showing that attractiveness conveys information about sexual warmth (Feingold, 1992b). If so, men might seek sexiness in a mate in all societies, in addition to attributes such as domestic skill, whose importance varies with the society's level of gender equality. Given that the female gender role often includes sexual restraint and lack of sexual autonomy, women may place less emphasis on sexiness in mates than men do.
It is less certain that physical attractiveness conveys information about women's fertility, as should be the case if men's preference for attractiveness in mates developed because attractiveness was a cue to fertility (Buss, 1989a; Jones, I 995; Singh, 1993). It seems reasonable that per- ceptions of attractiveness and potential fertility would co- vary even in contemporary data, but these relations have proven to be inconsistent (e.g., Cunningham, 1986; Tassi- nary & Hansen, 1998). Moreover, Singh's (1993) research on judgments of female figures that varied in weight and waist-to-hip ratio suggested three somewhat independent groupings of attributes: health, attractiveness, and sexiness; capacity and desire for children; and youth.
Although little is known about the relation between women's attractiveness and their actual fecundity, Kalick, Zebrowitz, Langlois, and Johnson (1998) found that facial attractiveness in early adulthood was unrelated to number of children produced or to health across the life span. Although the few participants in their sample who did not marry were less attractive than those who did marry, once the nonmarried were excluded, physical attractiveness was unrelated to the number of children produced by male or female participants. Kali ck et al. ( 1998) concluded that "any relation between attractiveness and fecundity was due to mate-selection chances rather than biological fertility"
6 The United Nations indexes of economic development and fertility showed relationships to mate preferences that were similar to those displayed in Tables I and 2. The magnitude of these relationships was in general nonsignificantly smaller than those involving the Gender Empow- erment Measure. These relationships were expected, given that this mea- sure increased with economic development (real gross domestic product per capita), r(33) = .71, p < .001, and decreased with fertility, r(33) = -.61, p < .001.
June 1999 • American Psychologist 419
(p. 10). Of course, as we noted in our critique of evolu- tionary psychology in this article, proponents of the theory do not predict that hypothesized evolved dispositions, such as men's preference for physically attractive partners, would necessarily be related to current reproductive suc- cess. Evolutionary psychologists argue instead that actual fertility in modem societies may bear little relation to the factors indicative of reproductive success in the EEA.
In summary, several aspects of the findings from Buss et al.' s ( 1990) 37 cultures study are compatible with the social structural origin theory of sex differences. The idea that the extremity of the division between male providers and female homemakers is a major determinant of the criteria that people seek in mates fits with the observed covariation between men placing more emphasis than women on younger age and domestic skill and women placing more emphasis than men on older age and earning potential. The lessening of these sex differences with in- creasing gender equality, as represented by the United Nations indexes, is consistent with our claim that these sex differences are by-products of a social and family structure in which the man acts as a provider and the woman acts as a homemaker. More ambiguous are the sex differences in valuing mates' physical attractiveness. Without evidence that men's greater valuing of attractiveness follows from one or more specific mechanisms, the simple absence of a relation between gender equality and sex differences in valuing attractiveness in our reanalysis does not advance the claims of evolutionary psychology or the social struc- tural theory. Convincing evidence for either interpretation has yet to be generated. However, with respect to the other sex differences emphasized by evolutionary psychologists, their cross-cultural patterning suggests that they arise from a particular economic and social system.
As
Within-Society Effects of Social Position evidence that presumably counters the social structural
interpretation of sex differences in mate selection criteria, evolutionary psychologists (e.g., Buss & Schmitt, 1993) have sometimes cited studies that examined the relation within a given culture between individuals' mate prefer- ences and their economic resources (e.g., Buss, 1989b; Kenrick & Keefe, 1992; Townsend, 1989). In one of the most extensive of these studies, Wiederman and Allgeier ( 1992) assessed mate preferences and anticipated income of undergraduate students from a midwestem university and of a convenience sample of Ohio residents. Mate preference ratings from both samples yielded the typical sex differences in ratings of good looks and good financial prospect. The central finding was that women's anticipated income and their valuing of mates as a good financial prospect were positively related in the college sample, r(635) = .17, p < .001, and unrelated in the community sample, r( 165) = .04, ns. That women who expected to earn higher incomes still valued financial resources in their mates was taken as evidence in favor of the evolutionary theory of mate preferences.
On the basis of such data, any conclusions about the validity of the evolutionary or the social structural origin
theory are unwarranted because such studies confound women's income with their socioeconomic status. Women who themselves have higher incomes would tend to come from higher socioeconomic groups and would anticipate selecting mates from their own stratum of society. In the United States, both sexes' homogamous mating on the basis of education, occupation, and economic resources is a well-established phenomenon (e.g., Kalmijn, 1991, 1994; Mare, 1991). Therefore, women's socioeconomic status typically should be positively related to expectations con- cerning mates' financial prospects.
An additional consideration is that, because societal gender roles coexist with specific roles, achieving a high- paying job does not completely neutralize the impact of broader gender role expectations. Therefore, consistent with these broader norms, even women with higher-than- average income commonly regard themselves as secondary wage earners in their marriages (Ferree, 1991) and often prefer to leave the labor force entirely or to become em- ployed part-time while raising a family (Herzog, Bachman, & Johnston, 1983; Tittle, 1981). Despite earning a substan- tial income, most women likely anticipate being fully or partially dependent on their husband's income during a portion of their life span. Consequently, within-society analyses of mate preferences that seek to draw conclusions about the effects of women's own economic resources must control for the influences of expectations based on social class and education as well as actual and anticipated marital roles.
Conclusion Considered at the level of a general metatheory of sex differences, social structural theories provide alternative explanations of the great majority of the general predictions about sex-differentiated social behavior that have been featured in evolutionary psychology. Because the central tendencies of sex differences (see Eagly, 1995; Halpern, 1997; Hyde, 1996) are readily encompassed by both of these perspectives, neither the evolutionary metatheory nor the social structural metatheory is convincingly substanti- ated by a mere noting of the differences established in the research literature. It is far too easy to make up sensible stories about how these differences might be products of sex-differentiated evolved tendencies or the differing placement of women and men in the social structure. This overlap in general main-effect predictions calls for more refined testing of the two theoretical perspectives, and each perspective is associated with numerous more detailed pre- dictions and empirical tests.
Certainly there are many possibilities for distinguish- ing between the two approaches with appropriate research designs (see Jackson, 1992). Evolutionary psychologists have been especially resourceful in obtaining cross-cultural data intended to support their claims of invariance across cultures in sex-differentiated behavior. To be maximally informative about social structural factors, cross-cultural research should be systematically designed to represent cultures with differing forms of social organization and levels of gender equality. In addition; a variety of other
June 1999 • American Psychologist 420
research methods, including experiments and field studies, can yield tests of predictions that emerge from evolutionary and social structural perspectives.
Although this article contrasts social structural expla- nations of sex differences with those based on evolutionary psychology, social structural analyses may be generally compatible with some evolutionary perspectives, as we noted in the introductory section of this article. Our argu- ment that sex differences in behavior emerge primarily from physical sex differences in conjunction with influ- ences of the economy, social structure, ecology, and cul- tural beliefs is potentially reconcilable with theories of coevolution by genetic and cultural processes (Janicki & Krebs, 1998). Our position is also sympathetic to the in- terest that some evolutionary biologists and behavioral ecologists have shown in the maintenance of behavioral patterns from generation to generation through nongenetic, cultural processes (e.g., Sork, 1997). However, despite our acknowledgement of the importance of some evolved ge- netic influences on the behavior of women and men, an implicit assumption of our approach is that social change emerges, not from individuals' tendencies to maximize their inclusive fitness, but instead from their efforts to maximize their personal benefits and minimize their per- sonal costs in their social and ecological settings.
One test of the evolutionary psychology and social structural origin theories of sex differences lies in the future-that is, in the emerging postindustrial societies in which the division between men's wage labor and women's domestic labor is breaking down. Notable is the increase in women's paid employment, education, and access to many formerly male-dominated occupations. Accompanying these changes is a marked attitudinal shift toward greater endorsement of equal opportunity for women in the work- place and role-sharing in the home (e.g., Simon & Landis, 1989; Spence & Hahn, 1997; Twenge, 1997). Nonetheless, occupational sex segregation is still prevalent with women concentrated in occupations that are thought to require feminine qualities and with men in occupations thought to require masculine qualities (Cejka & Eagly, 1999; Glick, 1991). Given that occupational distributions currently take this form and that the homemaker-provider division of labor remains weakly in place, social structuralists would not predict that sex differences in behavior should have already disappeared. Instead, to the extent that the tradi- tional sexual division between wage labor and domestic labor disappears and women and men become similarly distributed into paid occupations, men and women should converge in their psychological attributes.
REFERENCES
Archer, J. (I 996). Sex differences in social behavior: Are the social role and evolutionary explanations compatible? American Psychologist, 51, 909-917.
Bakan, D. (1966). The duality of human existence: An essay on psychol ogy and religion. Chicago: Rand McNally.
Beall, A. E., & Sternberg, R. J. (Eds.). (1993). The psychology of gender. New York: Guilford Press.
Becker, G. S. (1976). The economic approach to human behavior. Chi- cago: University of Chicago Press.
Borkenau, P. (1992). Age preferences: The crucial studies have yet to be done. Behavioral and Brain Sciences, 15, 93-94.
Broude, G. J. (1992). The May-September algorithm meets the 20th century actuarial table. Behavioral and Brain Sciences, 15, 94-95.
Buss, D. M. (1989a). Sex differences in human mate preferences: Evolu- tionary hypotheses tested in 37 cultures. Behavioral and Brain Sci ences, 12, 1-14.
Buss, D. M. (1989b). Toward an evolutionary psychology of human mating. Behavioral and Brain Sciences, 12, 39-49.
Buss, D. M. (1995a). Evolutionary psychology: A new paradigm for psychological science. Psychological Inquiry, 6, 1-30.
Buss, D. M. (1995b). Psychological sex differences: Origins through sexual selection. American Psychologist, 50, 164-168.
Buss, D. M. (1996). The evolutionary psychology of human social strat- egies. In E.T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 3-38). New York: Guilford Press.
Buss, D. M. (l 998). The psychology of human mate selection: Exploring the complexity of the strategic repertoire. ln C. Crawford & D. L. Krebs (Eds.), Handbook of evolutionary psychology: Ideas, issues, and appli cations (pp. 405-429). Mahwah, NJ: Erlbaum.
Buss, D. M. et al. (I 990). International preferences in selecting mates: A study of 37 cultures. Journal of Cross-Cultural Psychology, 21, 5-47.
Buss, D. M., Haselton, M. G., Shackelford, T. K., Bleske, A. L., & Wakefield, J.C. (1998). Adaptations, exaptations, and spandrels. Amer ican Psychologist, 53, 533-548.
Buss, D. M., & Kenrick, D. T. (1998). Evolutionary social psychology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. 2, pp. 982-1026). Boston: McGraw-Hill.
Buss, D. M., & Schmitt, D. P. (1993). Sexual strategies theory: An evolutionary perspective on human mating. Psychological Review, 100, 204-232.
Canary, D. J., & Dindia, K. (Eds.). (1998). Sex differences and similarities in communication: Critical essays and empirical investigations of sex and gender in interaction. Mahwah, NJ: Erlbaum.
Cejka, M. A., & Eagly, A. H. (1999). Gender-stereotypic images of occupations correspond to the sex segregation of employment. Person ality and Social Psychology Bulletin, 25, 413-423.
Collear, M. L., & Hines, M. (I 995). Human behavioral sex differences: A role for gonadal hormones during early development? Psychological Bulletin, ll8, 55-107.
Coombs, R. H., & Kenkel, W. F. (1966). Sex differences in dating aspiration and satisfaction with computer-selected partners. Journal of Marriage and the Family, 28, 62-66.
Cosmides, L., Tooby, J., & Barkow, J. H. (1992). Introduction: Evolu- tionary psychology and conceptual integration. In J. H. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind: Evolutionary psy chology and the generation of culture (pp. 3-15). New York: Oxford University Press.
Crawford, C. (I 998). The theory of evolution in the study of human behavior: An introduction and overview. In C. Crawford & D. L. Krebs (Eds.), Handbook of evolutionary psychology: Ideas, issues, and appli cations (pp. 3-41). Mahwah, NJ: Erlbaum.
Cronk, L. (1991). Human behavioral ecology. Annual Review of Anthro pology, 20, 25-53.
Cunningham, M. R. (1986). Measuring the physical in physical attrac- tiveness: Quasi-experiments on the sociobiology of female facial beauty. Journal of Personality and Social Psychology, 50, 925-935.
Daly, M., & Wilson, M. (1983). Sex, evolution, and behavior (2nd ed.). Boston: Grant Press.
Daly, M., & Wilson, M. (1998). The evolutionary social psychology of family violence. In C. Crawford & D. L. Krebs (Eds.), Handbook of evolutionary psychology: Ideas, issues, and applications (pp. 431- 456). Mahwah, NJ: Erlbaum.
Darwin, C. (1871). The descent of man and selection in relation to sex. London: Murray.
Deaux, K., & Lafrance, M. (1998). Gender. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. I, pp. 788-827). Boston: McGraw-Hill.
DeKay, W. T., & Buss, D. M. (1992). Human nature, individual differ- ences, and the importance of context: Perspectives from evolutionary psychology. Current Directions in Psychological Science, l, 184-189.
June 1999 • American Psychologist 421
Diamond, J. (1997). Guns, germs, and steel: The fates of human societies. New York: Norton.
Draper, P., & Harpending, H. (1982). Father absence and reproductive strategy: An evolutionary perspective. Journal of Anthropological Re search, 38, 255-273.
Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation. Hillsdale, NJ: Erlbaum.
Eagly, A.H. (1995). The science and politics of comparing women and men. American Psychologist, 50, 145-158.
Eagly, A.H., Ashmore, R. D., Makhijani, M. G., & Longo, L. C. (1991). What is beautiful is good, but ... : A meta-analytic review of research on the physical attractiveness stereotype. Psychological Bulletin, 1JO, 109-128.
Eagly, A.H., & Johnson, B. T. (1990). Gender and leadership style: A meta-analysis. Psychological Bulletin, 108, 233-256.
Eagly, A.H., & Steffen, V. J. (1984). Gender stereotypes stem from the distribution of women and men into social roles. Journal of Personality and Social Psychology, 46, 735-754.
Eagly, A. H., Wood, W., & Diekman, A. (in press). Social role theory of sex differences and similarities: A current appraisal. In T. Eckes & H. M. Trautner (Eds.), The developmental social psychology of gender. Mahwah, NJ: Erlbaum.
Ehrenberg, M. (1989). Women in prehistory. London: British Museum Publications.
England, P., & Browne, I. (1992). Internalization and constraint in wom- en's subordination. In B. Agger (Ed.), Current perspectives in social theory (Vol. 12, pp. 97-123). Greenwich, CT: JAI Press.
Fedigan, L. M. (1986). The changing role of women in models of human evolution. Annual Review of Anthropology, 15, 25-66.
Feingold, A. (I 990). Gender differences in effects of physical attractive- ness on romantic attraction: A comparison across five research para- digms. Journal of Personality and Social Psychology, 59, 981-993.
Feingold, A. (1991 ). Sex differences in the effects of similarity and physical attractiveness on opposite-sex attraction. Basic and Applied Social Psychology, 12, 357-367.
Feingold, A. ( 1992a). Gender differences in mate selection preferences: A test of the parental investment model. Psychological Bulletin, 112, 125-139.
Feingold, A. (1992b). Good-looking people are not what we think. Psy chological Bulletin, 1JJ, 304-341.
Feingold, A. (1994). Gender differences in personality: A meta-analysis. Psychological Bulletin, 116, 429-456.
Ferree, M. M. (1991). The gender division of labor in two-earner mar- riages: Dimensions of variability and change. Journal of Family Issues, 12, 158-180.
Fiske, A. P. (1992). The four elementary forms of sociality: Framework for a unified theory of social relations. Psychological Review, 99, 689-723.
Foley, R. (1996). The adaptive legacy of human evolution: A search for the environment of evolutionary adaptedness. Evolutionary Anthropol ogy, 4, 194-203.
Geary, D. C. (1995). Sexual selection and sex differences in spatial cognition. Learning and Individual Differences, 7, 289-301.
Geary, D. C. (1996). Sexual selection and sex differences in mathematical abilities. Behavioral and Brain Sciences, 19, 229-284.
Glenn, N. D. (I 989). Intersocietal variation in the mate preferences of males and females. Behavioral and Brain Sciences, 12, 21-23.
Glick, P. (1991). Trait-based and sex-based discrimination in occupational prestige, occupational salary, and hiring. Sex Roles, 25, 351-378.
Gould, S. J. (1991). Exaptation: A crucial tool for an evolutionary psy- chology. Journal of Social Issues, 47, 43-65.
Gutek, B. A., & Morasch, B. (1983). Sex-ratios, sex-role spillover, and sexual harassment of women at work. Journal of Social Issues, 38(4), 55-74.
Haas, L. L. (1995). Household division of labor in industrial societies. In B. B. Ingoldsby & S. Smith (Eds.), Families in multicultural perspec tive: Perspectives on marriage and the family (pp. 268-296). New York: Guilford Press.
Halpern, D. F. (1997). Sex differences in intelligence: Implications for education. American Psychologist, 52, 1091-1102.
Harris, M. (1993). The evolution of human gender hierarchies: A trial
formulation. In B. D. Miller (Ed.), Sex and gender hierarchies (pp. 57-79). New York: Cambridge University Press.
Hembroff, L. A. (1982). Resolving status inconsistency: An expectation states theory and test. Social Forces, 61, 183-205.
Herzog, A. R., Bachman, J. G., & Johnston, L. D. (1983). Paid work, child care, and housework: A national survey of high school seniors' pref- erences for sharing responsibilities between husband and wife. Sex Roles, 9, 109-135.
House, J. S. (1995). Social structure, relationships, and the individual. In K. S. Cook, G. A. Fine, & J. S. House (Eds.), Sociological perspectives on social psychology (pp. 387-395). Boston: Allyn & Bacon.
Hrdy, S. B. (1997). Raising Darwin's consciousness: Female sexuality and the prehominid origins of patriarchy. Human Nature, 8, 1-49.
Hyde, J. S. (1996). Where are the gender differences? Where are the gender similarities? In D. M. Buss & N. M. Malamuth (Eds.), Sex, power, conflict: Evolutionary and feminist perspectives. New York: Oxford University Press.
Jackson, L. A. (1992). Physical appearance and gender: Sociobiological and sociocultural perspectives. Albany: State University of New York Press.
Jacobs, J. A. (1989). Revolving doors: Sex segregation and women's careers. Stanford, CA: Stanford University Press.
Janicki, M. G., & Krebs, D. L. (1998). Evolutionary approaches to culture. In C. Crawford & D. L. Krebs (Eds.), Handbook of evolutionary psychology: Ideas, issues, and applications (pp. 163-207). Mahwah, NJ: Erlbaum.
Jones, D. ( 1995). Sexual selection, physical attractiveness, and facial neoteny: Cross-cultural evidence and implications. Current Anthropol ogy, 36, 723-748.
Kalick, S. M., Zebrowitz, L.A., Langlois, J. H., & Johnson, R. M. (1998). Does human facial attractiveness honestly advertise health? Longitudi- nal data on an evolutionary question. Psychological Science, 9, 8-13.
Kalmijn, M. (1991). Status homogamy in the United States. American Journal of Sociology, 97, 496-523.
Kalmijn, M. (1994). Assortative mating by culture and economic occu- pational status. American Journal of Sociology, JOO, 422-452.
Kelly, R. L. (1995). The foraging spectrum: Diversity in hunter-gatherer lifeways. Washington, DC: Smithsonian Institution Press.
Kenrick, D. T., & Keefe, R. C. (1992). Age preferences in mates reflect sex differences in human reproductive strategies. Behavioral and Brain Sciences, 15, 75-91.
Kenrick, D. T., Trost, M. R., & Sheets, V. L. (1996). Power, harassment, and trophy mates: The feminist advantages of an evolutionary perspec- tive. In D. M. Buss & N. M. Malamuth (Eds.), Sex, power, and conflict: Evolutionary andfeminist perspectives (pp. 29-53). New York: Oxford University Press.
Kerckhoff, A. C. (1995). Social stratification and mobility processes: Interaction between individuals and social structures. In K. S. Cook, G. A. Fine, & J. S. House (Eds.), Sociological perspectives on social psychology (pp. 476-496). Boston: Allyn & Bacon.
Lerner, G. (1986). The creation of patriarchy. New York: Oxford Uni- versity Press.
Lippa, R. (1998). Gender-related individual differences and the structure of vocational interests: The importance of the "people-things" dimen- sion. Journal of Personality and Social Psychology, 74, 996-1009.
Lips, H. M. (1991). Women, men, and power. Mountain View, CA: Mayfield.
Lorenzi-Cioldi, F. (1998). Group status and perceptions of homogeneity. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 9, pp. 31-75). Chichester, England: Wiley.
Mare, R. D. (1991). Five decades of educational assortative mating. American Sociological Review, 56, 15-32.
Moscowitz, D. W., Suh, E. J., & Desaulniers, J. (1994). Situational influences on gender differences in agency and communion. Journal of Personality and Social Psychology, 66, 753-761.
Olson, J. M., Roese, N. J., & Zanna, M. P. (1996). Expectancies. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 211-238). New York: Guilford.
Pfeffer, J. (1998). Understanding organizations: Concepts and controver- sies. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. 2, pp. 733-777). Boston: McGraw- Hill.
June 1999 • American Psychologist 422
Potts, R. (1984). Home bases and early hominids. American Scientist, 72, 338-347.
Powers, E. A. (1971). Thirty years of research on ideal mate characteris- tics: What do we know? International Journal of Sociology of the Family, 1, 207-215.
Presser, H.B. (1994). Employment schedules among dual-earner spouses and the division of household labor by gender. American Sociological Review, 59, 348-364.
Reskin, B. F., & Padavic, I. (1994), Women and men at work. Thousand Oaks, CA: Pine Forge Press.
Ridgeway, C. L. (1991). The social construction of status value: Gender and other nominal characteristics. Social Forces, 70, 367-386.
Ridgeway, C. L. (1997). Interaction and the conservation of gender inequality: Considering employment. American Sociological Review, 62, 218-235.
Ridgeway, C. L., & Diekema, D. (1992). Are gender differences status differences? In C. L. Ridgeway (Ed.), Gender, interaction, and inequal- ity (pp. 157-180). New York: Springer-Verlag.
Rose, L., & Marshall, F. (1996). Meat eating, hominid sociality, and home bases revisited. Current Anthropology, 37, 307-338.
Sanday, P.R. (1981). Female power and male dominance: On the origins of sexual inequality. New York: Cambridge University Press.
Schaller, M. (1997). Beyond "competing," beyond "compatible." Ameri- can Psychologist, 52, 1379-1380.
Shelton, B. A. (1992). Women, men and time: Gender differences in paid work, housework, and leisure. New York: Greenwood Press.
Silverman, I., & Phillips, K. (1998). The evolutionary psychology of spatial sex differences. In C. Crawford & D. L. Krebs (Eds.), Handbook of evolutionary psychology: Ideas, issues, and applications (pp. 595- 612). Mahwah, NJ: Erlbaum.
Simon, R. J., & Landis, J.M. (1989). The polls-A report: Women's and men's attitudes about a woman's place and role. Public Opinion Quar- terly, 53, 265-276.
Singh, D. (1993). Adaptive significance of female physical attractiveness: Role of waist-to-hip ratio. Journal of Personality and Social Psychol- ogy. 65, 293-307.
Skrypnek, B. J., & Snyder, M. (1982). On the self-perpetuating nature of stereotypes about women and men. Journal of Experimental Social Psychology, 18, 277-291.
Smith, E. A. (in press). Three styles in the evolutionary study of human behavior. In L. Cronk, W. Irons, & N. Chagnon (Eds.), Adaptation and human behavior: An anthropological perspective. Hawthorne, NY: Aldine de Gruyter.
Sork, V. L. (1997). Quantitative genetics, feminism, and evolutionary theories of gender differences. In P. A. Gowaty (Ed.), Feminism and evolutionary biology: Boundaries, intersections, and frontiers (pp. 86- 115). New York: Chapman & Hall.
Spence, J. T., & Hahn, E. D. (1997). The Attitudes Toward Women Scale and attitude change in college students. Psychology of Women Quar- terly, 21, 17-34.
Sprecher, S., Sullivan, Q., & Hatfield, E. (1994). Mate selection prefer- ences: Gender differences examined in a national sample. Journal of Personality and Social Psychology, 66, 1074-1080.
Steil, J.M. (1997). Marital equality: Its relationship to the well-being of husbands and wives. Thousand Oaks, CA: Sage.
Strier, K. B. (1994). Myth of the typical primate. Yearbook of Physical Anthropology, 37, 233-271.
Symons, D. ( 1979). The evolution of human sexuality. New York: Oxford University Press.
Symons, D. (1992). On the use and misuse of Darwinism in the study of human behavior. In J. H. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind: Evolutionary psychology and the generation of culture (pp. 137-159). New York: Oxford University Press.
Tassinary, L. G., & Hansen, K. A. (1998). A critical test of the waist-to- hip-ratio hypothesis of female physical attractiveness. Psychological Science, 9, 150-155.
Tattersall, I. (1998). Becoming human: Evolution and human uniqueness. New York: Harcourt Brace.
Tittle, C. K. (1981). Careers and family: Sex roles and adolescent life plans. Beverly Hills, CA: Sage.
Tomaskovic-Devey, D. (1995). Sex composition and gendered earnings inequality: A comparison of job and occupational models. In J. A. Jacobs (Ed.), Gender inequality at work (pp. 23-56). Thousand Oaks, CA: Sage.
Tooby, J., & Cosmides, L. (1989). The innate versus the manifest: How universal does universal have to be? Behavioral and Brain Sciences, 12, 36-37.
Tooby, J., & Cosmides, L. (1990a). On the universality of human nature and the uniqueness of the individual: The role of genetics and adapta- tion. Journal of Personality, 58, 17-67.
Tooby, J., & Cosmides, L. (1990b). The past explains the present: Emo- tional adaptations and the structure of ancestral environments. Ethology and Sociobiology, 11, 375-424.
Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In J. H. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind: Evolutionary psychology and the generation of culture (pp. 19-136). New York: Oxford University Press.
Townsend, J.M. (1989). Mate selection criteria: A pilot study. Ethology and Sociobiology, JO, 241-253.
Toynbee, A. J. (1934-1961). A study of history (Vols. 1-12). New York: Oxford University Press.
Travis, C. B., & Yeager, C. P. (1991). Sexual selection, parental invest- ment, and sexism. Journal of Social Issues, 47(3), 117-129.
Trivers, R. (1972). Parental investment and sexual selection. In B. Camp- bell (Ed.), Sexual selection and the descent of man: 1871-1971 (pp. 136-179). Chicago: Aldine.
Twenge, J. M. (1997). Attitudes toward women, 1970-1995: A meta- analysis. Psychology of Women Quarterly, 21, 35-51.
United Nations Development Programme. (1995). Human development report 1995. New York: Oxford University Press.
United Nations Development Programme. (1996). Human development report 1996. New York: Oxford University Press.
United Nations Development Programme. (1997). Human development report 1997. New York: Oxford University Press.
West, C., & Zimmerman, D. H. (1987). Doing gender. Gender & Society, 1, 125-151.
Whyte, M. K. (1978). The status of women in preindustrial societies. Princeton, NJ: Princeton University Press.
Wiederman, M. W., & Allgeier, E. R. (1992). Gender differences in mate selection criteria: Sociobiological or socioeconomic explanation? Ethology and Sociobiology, 13, 115-124.
Wiley, M. G. (1995). Sex category and gender in social psychology. In K. S. Cook, G. A. Fine, & J. S. House (Eds.), Sociological perspectives on social psychology (pp. 362-386). Boston: Allyn & Bacon.
Williams, G. C. (1966). Adaptation and natural selection: A critique of some current evolutionary thought. Princeton, NJ: Princeton University Press.
Wood, W., Christensen, P. N., Heb!, M. R., & Rothgerber, H. (1997). Conformity to sex-typed norms, affect, and the self-concept. Journal of Personality and Social Psychology, 73, 523-535.
Wood, W., & Eagly, A.H. (1999). Social structure and the origins of sex differences in social behavior. Manuscript in preparation.
Wood, W., & Karten, S. J. (1986). Sex differences in interaction style as a product of perceived sex differences in competence. Journal of Personality and Social Psychology, 50, 341-347.
June 1999 • American Psychologist 423
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
