| 摘要 | Grateful acknowledgment is due to the John D. and Catherine T. MacArthur Foundation, which has supported the Research Network on Successful Midlife Development, on which I serve as a member. I am particularly grateful to my colleagues on the network for their helpful feedback on all my work over the past several years. Most relevant to this paper is grant support for my review of the literature in evolutionary psychology, reproductive biology, and adolescent development in connection with a book manuscript in process on the reproductive phase of women’s lives. I have drawn heavily from this review in the preparation of this paper.
Abstract
This paper applies a biopsychosocial perspective to adolescent sexual feelings and behavior. In particular, it explores sexual attraction, desire, and mate selection as evolutionary adaptations just as necessary to species survival as sweating to cope with summer heat or shivering in response to winter cold. Sex differences in mating strategies are described as part of this evolutionary adaptation, and recent research findings are summarized that demonstrate how these strategies explain contemporary sexual and reproductive behavior in Western societies today, as they do sexual behavior in the past or across diverse cultures. The implications of these findings are discussed for their significance to intervention efforts to postpone sexual initiation and avert births outside marriage. The author urges greater emphasis on the early pubertal years; increased attention to teaching adolescents more about their own sexual development; a less absolutist focus on sexual abstinence, which may be appropriate and feasible for 12-year-olds but not for 16-year-olds; better and more widespread sex education at earlier ages and throughout the school curriculum; frank discussion of all the options available for those who experience an unwanted pregnancy; and as much attention to adolescent boys as to adolescent girls.
Introduction
When Sigmund Freud pointed to work and love as two major factors necessary to a happy life, there is no evidence he was thinking in evolutionary terms. Yet work in the form of production of goods and services is the central means by which we survive, and love or some variant of sexual attraction is the major means that ensures species continuity through reproduction. I consider this point to any discussion of teenage sexuality and pregnancy, a reminder we need in Western societies because we have so clearly drawn a sharp distinction between recreational sex and reproduction. The impulses to grow, to learn, to make love, and to reproduce are not easily controlled, because they are built into our very nature as human animals.
Although sexual selection may affront our self-images as rational beings, it operates on us all; it is not restricted to peacocks and lions. Just as our food preferences, immune systems, and mechanisms to cope with summer heat by sweating or winter cold by shivering are the products of evolutionary adaptation, so too our inclinations in sexual attraction, desire, and mate selection represent adaptations that were necessary for the survival of the species. In the felicitous phrasing of David Buss, our evolutionary past “has grooved and scored our minds as much as our bodies, our strategies for mating as much as our strategies for survival” (Buss, 1994).
For the past two decades, my intellectual interests have centered on trying to bridge the gaps between evolutionary theory, biomedical knowledge of the physiological processes that underlie human behavior, the demographic shifts attending historic transitions from agricultural to industrial to postmodern information and service-oriented economies, and the social institutions in which we are embedded. My motivation in this effort at synthesis was to understand better human development across the life course and, in particular, to address the question of how stable or flexible differences between the sexes are. Drawing on the diverse disciplines that span historic structural changes at a macrolevel down to the inner workings of mind and body of contemporary men and women is an ambitious undertaking, subject to all the risks of partial understanding of any one discipline but holding the promise of a new synthesis that bypasses the turf boundaries disciplinary specialists tend to defend. But that is the essential goal of those of us who subscribe to a biopsychosocial perspective on human behavior.
In this brief paper, I hope to demonstrate the utility of that perspective as applied to understanding adolescent sexual behavior in our time and place. I believe the biopsychosocial perspective tempers the notion that programs to effect change in adolescent behavior have a high probability of success. This approach may be humbling, but it need not be pessimistic. Rather, it urges some caution against claims that a given program of education can radically change behavior or that the program can achieve dramatic success.
A basic point of my analysis can be simply stated: I believe there is a radical and unprecedented discordance between our biological natures that resulted from the processes of natural selection over eons of evolutionary time and the demands imposed on us by fundamental changes in modern societies. Let me give an example in an area remote from our concern with adolescent sexual behavior and its consequences. Anyone who has watched firsthand the tremendous drive human babies show in learning to get up on two feet and walk without adult support will understand how difficult it would be to curb this impulse in a healthy one- year-old. Imagine the social and physical restraints one would have to impose to prevent the baby from crawling and walking. Except by unthinkably cruel acts, our efforts would be either doomed to failure or require incredible crippling as the Chinese did for centuries of painful foot binding of young girls.
On comparable evolutionary grounds, I believe it is a lost cause to urge that adolescents abstain from sex for the very long span of time between sexual maturation at 12 and a “mature” age to mate and marry in the early to mid 20s. But more on this later. First, I want to sketch the mating strategies that have long differentiated males from females against which to assess our contemporary situation.
Sex-Differentiated Mating Strategies
For countless generations of human history, a basic biological difference between males and females predicted sex differences in mating strategies: a human female has a limited number of large-sized gametes in her ovaries compared with the millions of sperm in every male ejaculation. Given a conception, nothing further happens to the male progenitor, but the female is on a trajectory of a nine-month pregnancy and a decade and more of intense investment in child rearing. Further, assuming full consciousness or careful hospital procedures of identification at birth, a woman knows with full assurance that the infant born is her biological child. By contrast, a man has no assurance the child born is biologically his. Paternity depends on circumstantial evidence, even in cultures that require chastity at marriage, sequester women, and restrict their social contacts with males.
A male can maximize the transmission of his genes through multiple matings with a number of females. By contrast, the female has traditionally been more inclined to much more cautious behavior in the selection of a sex partner, driven by the search for a mate with sufficient skills and material resources to provide support for the offspring she produces. On these basic biological grounds alone, evolutionary theory predicts that the criteria guiding mate selection will differ between the sexes: to ensure the fertility of a female mate, and the stamina she will require to rear offspring to reproductive age, the male will seek a female who is young, attractive, and healthy; the female will seek a male who is healthy, dependable, with sufficient material resources (or the promise of future acquisition of such resources) to support her and the children she will produce through the years of their dependency.
Contemporary skeptics of evolutionary theory might concede that such sex differences probably did structure the relationship between men and women deep in the past of human history and further, that such theory helps explain the universal pattern in developing societies today that continues to buttress the socially legitimate mating pair with social supports through legal, economic, and cultural institutions designed to provide marital and lineage stability.
But many doubt that these long standing sex differences persist (or should persist) under the very different economic and social conditions of Western societies. This persistence is precisely what evolutionary psychologists have attempted to test in a number of fascinating studies in recent years. Studies of the criteria men and women seek in a mate from samples of adults 17 to 70 years of age, from 37 cultures on six continents (Buss, 1989, 1994) report a common finding: that women from all continents, all political systems, all racial and religious groups, and all systems of mating (intense polygyny to presumptive monogamy) place far greater value than men on good financial prospects in a potential mate. What type of resources are favored by women varies by culture or historical period: greater male strength in a hunting society; numerous cattle or sheep in a pastoral society; more land in a peasant society; and more education, higher social standing, or higher income in Western societies in our time. That women tend to marry men older than themselves is rooted in this same sex-differentiated mating strategy: age of the male is a proxy for greater demonstrated ability to acquire skills and resources (Kenrick & Keefe,1992). So too, male height is an index of likely command of resources. One of my research colleagues, Michael Marmot, who is conducting a large-scale longitudinal study of British civil servants, reports that height is significantly correlated in a linear fashion with grade level of men in the civil service: the higher the rank, the taller the man. Height shows no similar pattern in relation to grade level among women in the civil service. Even more persuasive evidence is Marmot’s finding that men who began their civil service career at lower grades and moved up the hierarchy were taller than their nonmobile peers in the lower grades.
The counterpart to female preference for men with resources to support them and their children is male preference for young, fertile, and attractive women. It does not sit well in our thinking to concede the role of physical and sexual attractiveness as a criterion for success in life; we prefer to believe that accomplishments and skills, not the good fortune of our genes, are what lead to productive and reproductive success. But a good deal of evidence accumulated over the past decade shows physical attractiveness as a powerful predictor of social and personal evaluations of ourselves and others. Youth and attractiveness in the female are proxies for fertility potential: a symmetrically shaped face, clear skin, shiny hair, slim waist, and full hips and breasts are indicators of good health and fertility potential, indicating as they do that the woman is free of parasite infection and has not given birth before (pregnancies tend to increase the female waistline), that she gives reasonable attention to personal cleanliness, and that she has enjoyed good nutrition. The waist-hip ratio is a particularly interesting indicator of a woman’s ‘attractiveness as perceived by both men and women. One outstanding study was conducted by Devendra Singh (1993), who designed an ingenious test for judgments of female physical attractiveness by drawing three types of female figures, identical except for weight (underweight, average, and overweight), and within each of these three weight classes, four figures that varied in their waist-hip ratio. The waist-hip ratio is a good indicator of sex hormone profile and fertility and the risk of such major diseases as hypertension, adult onset diabetes, cardiovascular and gallbladder diseases, and cancer of the breast, endometrium, and ovaries (National Academy of Sciences, 1991). Singh found that female waist-hip ratios were highly significant predictors of judgments of attractiveness of the female figures, overriding the role of body weight: the smaller the waist compared with the hips, the more attractive were the female figures judged to be. The waist-hip ratio can be readily perceived by eye and hence represents a good proxy for female health and fertility. Few men would explain their preferences in such terms, illustrating the point that mating strategies typically operate on an unconscious level. Nor are women and fashion designers consciously projecting fertility potential when they give visual reinforcement to low waist-hip ratios by the use of wide belts in one era, corsets and bustles in another, or midriff nudity in recent summertime dress.
This sketch of sex differences in mating strategies illustrates the relevance of an evolutionary perspective on human sexuality. To the question of how relevant such considerations are to today’s sexual and reproductive behavior, let me briefly mention a few illustrations from recent research:
One last consideration that emerges from comparative work in evolutionary biology concerns the high degree to which both female and male humans take pleasure from sex and have few physical or social moratoriums that preclude sexual behavior. An evolutionary assumption where behavior central to survival and reproduction is concerned is that the more important the behavior, the more multidetermined it will be. From this perspective, the human organism is designed to derive a great deal of pleasure from sexual activity of any kind, including our relative hairlessness, concealed ovulation of the female, lack of female estrus compared with other primates, greater capacity for female orgasm, permanently enlarged and erotically sensitive breasts, and enhanced tactile sensitivity, facilitated perhaps by our relative hairlessness compared with other primates (Alexander & Noonan, 1979).
The surplus sexual endowment of the human species can lead to considerable personal and social distress, not merely pleasure. We espouse monogamy, but because we are highly sexed, our sexual physiology and psychology predispose us to multiple matings. The most monogamous of all the primates are gibbons who are strongly pairbonded; but unlike humans and most other primates, gibbon pairs live in very large home ranges of some 100 acres, with few intruders; hence a gibbon pair has few temptations to sexual dalliance (Wright, 1994). By contrast, humans evolved in tight small groups, rife with genetically profitable alternatives to fidelity; and in our time, sheer population density and the anonymity of large urban centers set the stage for easy access to numerous sex encounters, marital fragility, extramarital affairs, and high divorce. Here then is one critical disjunction between the ancestral contexts within which we evolved and the social contexts within which we now live out our lives. For some, an internalized conscience, religious belief, or both may lead us to turn away from most sexual temptations, but we cannot turn off the deeper levels of our emotional responses, although they may be diverted into sexual fantasy rather than expressed in behavior. At this deeper level, we remain the creatures who evolved in a Pleistocene environment. As S. Boyd Eaton, a radiologist, recently put it to a reporter in Atlanta, “We’re Stone Agers in the fast lane.” Demographic and Normative Changes That Affect Adolescent Sexual Behavior
Trends in Sexual Maturation. In the course of sketching a bioevolutionary perspective on mating strategies, I have already identified one major factor that confronts adolescents coping with early adolescent sexual awakening: the drop in the age of sexual maturation, which in turn has affected the age of sexual initiation. There are several important consequences that flow from this century-long trend. On a macrolevel, the trend toward earlier sexual maturation might not have been associated with increasing personal and social problems were it the case that early adolescents were socially sequestered, subject to very close chaperonage, or married at very young ages. But as the 20th century unfolded, there have been solid reasons for postponing marriage as a consequence of the increasing need for higher levels of schooling to prepare for entry into a competitive labor market. For adolescents who do not go on to higher education, the economic uncertainty of obtaining or holding a secure job severely reduces their ability to support a family. In a future fraught with uncertainty, a poor young male has slim chances of obtaining sufficient resources to offer a young woman in marriage, and poor adolescent girls have little reason to be hopeful about finding and marrying a male with adequate resources to support her through the early stages of childbearing and child rearing.
If such poor young adolescents were embedded in stable intact families and living within kin and neighbor networks of supportive adults, the environment might go a long way toward encouraging adolescents to develop impulse control and learn the advantages of postponing childbearing, if not sexual gratification, to a more mature age. But this condition is precisely what is lacking in the hundreds of urban pockets of disorganized neighborhoods in our crowded urban habitats. Aspirations for a happy future life require a stable social setting with numerous examples of adults enjoying the payoff of hard work rewarded by steady, adequate income. But few such minority members remain in these deteriorating settings: as soon as they can afford to do so, they move out of inner cities to suburban homes with neighbors like themselves, with the consequence of escalating the rates of individual, family, and social pathologies in the communities they left behind (Wilson, 1987).
It is not widely enough recognized that growing up in a broken family or with a mother who had never married in the first place is itself conducive to early sexual maturation of young girls. The mean age at menarche may be 12.6 years, but there is considerable variance involved, from very early maturers under 10 years of age to 16 years or older. Some 30% of menarcheal timing is genetic: early-maturing women have early-maturing daughters and sons (Garn, 1980; Goldman & Schneider, 1987; Surbey, 1990).
Beyond the genetic factor, increasing evidence indicates that the presence of a biological father has a significant effect on the growth and development of his children. The presence of a biological father postpones sexual maturation of girls; the presence of stepfathers or cohabiting men living with the girls’ mothers tends to advance the age at maturation (Caspi & Moffitt, 1991; Moffitt et al., 1992; Surbey, 1990; Steinberg, 1988).
One recent study is particularly rich in identifying the effect of parent presence or absence on child development and sexual maturation (Flinn et al., 1996). In an eight-year study of children in a small village on the east coast of Dominica in the Lesser Antilles, Flinn and his associates studied children from birth on, with special attention to the effect of separation from father or mother on the child’s stress level, as indexed by cortisol and testosterone levels assayed from numerous saliva samples from the children. Cortisol is a key hormone produced in response to physical and psychosocial stressors. Without cortisol we would not be able to tolerate the ups and downs of daily life, because cortisol controls a wide range of somatic functions including energy release, immune activity, mental activity, growth, and reproductive functions. In other words, cortisol prepares the body to respond to unanticipated sudden changes in the environment. But when stress is a constant feature of life, it has negative consequences: an impaired immune system, a slowdown of cognitive development and mental activity, and a worsening of psychological adjustment. For that reason, prolonged stress associated with the death of a family member or separation of a child from the care-giving adult is so often followed by an elevation of infectious disease and other health and psychological problems.
In the Dominica study, the researchers found elevated cortisol levels in very young children when their mothers were away from home for more than a day, as they are in families with prolonged internal conflict between the parents, compared with the cortisol levels of children living in amiable families. The specific outcomes of persistent stress differ by sex: the boys show antisocial and delinquent behavior, whereas the girls show elevated anxiety and withdrawal behavior. Father absence had a particularly negative effect on boys. An evolutionary interpretation stresses the point that natural selection favors sensitive interaction with family care givers; for boys, fathers are particularly important as models of care, resources, and social status. For girls whose fathers are not present, Flinn and his associates believe their mothers may give them less consistent care because of economic stress in the absence of a marital partner or inattention to their daughters as they seek male partners, however temporary.
I have cited the Flinn study because it is unique in demonstrating the intervening physiological pathways between stress brought about by parental absence or family conflict and the outcome of behavioral and psychological disturbance in the children. Other studies in New Zealand (for example, the Caspi & Moffitt study referred to above) and Canada (Surbey, 1990) have also reported that girls mature at later ages in father-present homes and earlier ages when nonbiological males are present in the home. I consider the Dominica findings particularly important because families in these small villages are not isolated but embedded in extended families with much social life taking place outdoors. From a child’s perspective, the boundaries between family and unrelated neighbors may be blurred, and hence child supervision is difficult beyond the biological parents to other adult community members. Parent absence or conflict may have far more serious consequences in societies like our own because most families do not live in closely interdependent social groups but in isolated households in which a parent is of overwhelming psychological significance for the well-being of very young children.
Of particular interest are studies like Michelle Surbey’s in Toronto, which built on the thesis first espoused by Patricia Draper and Henry Harpending (1982), who argued that a society or social group within a society with unstable marital behavior and low paternal investment in the young would prompt women to develop a reproductive strategy that emphasized early childbearing, bypassing careful mate selection because men could not be depended on as reliable spouses or fathers. By contrast, in social groups in which men are perceived to be reliable providers who become actively involved in child rearing, women would postpone childbearing until they could make a judicious selection of mates. Jerome Barkow (1984) expanded on this idea by proposing that the two different reproductive strategies would predict an impact of father presence or father absence on the timing of sexual maturation of girls. Barkow drew on several sources of evidence in preliminary support for this thesis: studies of female adolescent promiscuity and teenage pregnancies, for example, often point to “broken homes” as a background factor associated with social deviance and early childbearing. The line of reasoning was therefore that disturbed families lead to early sexual maturation, followed by early sexual initiation and subsequent early childbearing, a development since confirmed by Richard Udry’s research (Udry, 1979; Udry & Campbell, 1994).
This was the background to Michelle Surbey’s study of the effect of father presence or absence on the timing of menarche in a sample of 1,247 daughters, among whom 16% (204 cases) were from families where father absence occurred before the girl’s menarche. Several major findings from this study follow:
It should be noted that father absence is a complex variable and may be a proxy for a number of social-psychological factors as well, including characteristics of the mother. As the saying goes, it takes two to tango, and it takes two to produce marital tension or to marry in the first place, even when pregnant. Women who divorce and rear children by themselves or who never married have histories of their own. In anticipation of such history, Surbey included a range of measures relating to the mothers, including their own menarcheal age, attitudes toward men, age they first dated, and number of children they bore; and from the daughters themselves, Surbey included measures of their attitudes toward men and scores on a lengthy life- events inventory.
Of particular relevance here, Surbey found that mothers in the father-absent families were younger when they first menstruated and first dated, had less positive views toward men and toward the family generally, and gave birth to more children. The daughters replicated their mother on many of the same measures: early menarche, early dating, more negative attitudes toward the family, and a highly significant elevation of negative life events. Such findings are consistent with Mavis Hetherington’s early studies (1972) that found that girls whose parents divorced were more likely to be sexually flirtatious and precocious than girls from intact families.
These studies have particular relevance to the situation of young black girls growing up in our urban ghettoes. Common findings are that black women have a much higher rate of early menarche, followed by early sexual initiation, and earlier and more births outside marriage. Both the genetic link and the fact of growing up in disorganized families and neighborhoods contribute to this profile and the similarity across the generations of many black women.
Conversely, an “upside” to this profile should be noted. In the lower social strata of a society, where males have limited economic opportunity and high rates of unemployment and imprisonment, men have little to offer to women as providers capable of investing in child rearing. From a woman’s perspective in the lower classes, men offer sociability, sexual gratification, and insemination. In terms of reproductive strategies, a black underclass woman has little to gain by postponing sex and childbearing; and from the point of view of pregnancy outcome, little is gained by bearing children at older ages, because lower-class women are subject to continuing inadequacy of good-quality food, experience prolonged social and economic stress, and have inadequate health care, with the result that the longer childbearing is postponed, the greater the toll of these factors on uterine and placental capability. In fact, infant mortality is lower among infants born to black underclass mothers during their teenage years than at older ages (Geronimus, 1987, 1991; Lancaster, 1986, 1994). In sum, postponement of childbearing does not necessarily improve a black underclass woman’s life chances or ensure infant and maternal health.
Changing Marriage Norms. A second factor implicit in an evolutionary perspective on mating strategies pinpoints a significant feature of modern societies that represents a radical historical departure from all we know about family structure and social norms concerning mate selection in the past. Until recent times, the choice of a marital partner was not left to the individual but to parental or lineage decisions, and they were made out of concern for the long-term best interests of the child or the lineage itself. Independent households of the newly paired couple were typically rare; more prevalent was the incorporation of the young couple into the existing family of the bride or more typically, of the groom (that is, patrilineal residence). In our time in Western societies, parents have only indirect control of the range of potential mates young people may acquire; for example, by choice of neighborhood or the school their children attend, parents can impose some restrictions on the range of likely other-sex friends their maturing children may meet and eventually marry. But the final choice is an individual one.
Furthermore, it is culturally permissible and technologically feasible (thanks to effective contraception) to distinguish between a sex partner and a marital partner, with a considerable difference between the criteria involved in such choices. Looks and social skills play a major role in choice of a sex partner, with little attention to material resources or care giving and fertility potential that are more relevant to the choice of a marital partner, as illustrated in countless well-known novels. The very language used to distinguish between pre-marital and marital sex has become archaic. Premarital used to mean having sex with the partner one eventually married. Now it is typical for young adults in Western societies to have several sex partners before marriage. As we saw, the most significant change in recent decades has been that it is no longer just men who enter marriage with considerable sexual experience behind them but women as well. Not only is sexual experimentation a common feature of adolescent life, but it is now often followed by cohabitation for some years in early adulthood before marriage is even considered.
In the political climate of the 1990s, we are inundated with media and political messages decrying the breakdown of “traditional family values.” Rarely do such invocations go beyond glib pieties to an analysis grounded in empirical facts; and there is much nostalgia in invoking such values (Coontz, 1992; Goldscheider & Waite, 1991). I have pointed out (Rossi, 1993, 1996; Rossi & Rossi, 1990) that no evidence suggests that intergenerational relations between parents and their children (especially mothers and their children) have undergone any profound disruptive change in the past half-century, either of normative obligations, affective mutual concern, or exchanges of help. There is no “war between the generations” among individuals, although friction on an aggregate political level may appear in the coming decade on entitlements enjoyed by the elderly.
There is strong evidence of changing social norms where marriage as an institution is concerned, a trend extremely difficult to reverse. With sex no longer confined to marriage, and increasing economic independence of women, marriage itself loses some of its appeal. It remains an open question what the social benefits were in the past or are today for women to enter or remain in marriages out of sheer economic dependence on men. Social acceptance of cohabitation is itself an index of an erosion of marriage norms, and this pattern is widespread in all Western societies (Bumpass, 1990, 1994). Surveys now report that an overwhelming majority of Americans no longer believe marriage or having children is necessary for a full, happy life for either men or women, and this view is spreading in newly industrialized countries like Korea and Japan as well (Bumpass & Choe, 1996). Together, these trends call into question any possibility that movements can successfully urge sexual abstinence, marital stability, or confining births to married couples who stay together.
It is nevertheless the case that the majority of young people do wish to marry and to have at least one child. I have come to believe that an important psychological function is served by simultaneously holding marriage and childbearing as a personal ideal, while at the same time subscribing to a backup contingent norm that says not all is lost if that ideal turns out to be impossible to attain.
I conclude from a variety of evidence that families are stable in some dimensions, that is, intergenerational bonds; flexible and adaptive to circumstances in other dimensions, that is, class and racial differences in economic opportunities that reduce the value of marriage to either men or women; and socially and psychologically problematic in still |
