A. Childhood Sexuality
DAVID L. WEIS*Within American culture, childhood sexuality remains an area that has been largely unexplored by researchers. Childhood is widely seen as a period of asexual innocence. Strong taboos continue concerning childhood eroticism, and childhood sexual expression and learning are still divisive social issues. This general ambience of anxiety associated with the sexuality of children is probably understandable, given the general history of sexuality in the U.S.A., with its focus on adult dyadic sex within committed intimate relationships and its opposition to other sexual expressions. This ambience remains, despite the fact that nearly a century has passed since Freud introduced his theory of psychosexual stages with an emphasis placed on the sexual character of childhood development. This reluctance to accept childhood sexuality is somewhat ironic, because Freudian theory, with its concepts of psychosexual stages (oral, anal, phallic, and latency), penis envy, the Oedipus/Electra complexes, repression, and the unconscious, has been immensely popular in the United States throughout much of the twentieth century. Yet, the general American public has been able to ignore the prominence given to childhood sexual development by Freudian theorists and to maintain its central belief that childhood is and ought to be devoid of sexuality.
* With input from Paul Okami.
Perhaps no area reviewed in this section has been the subject of less scientific research than this topic of childhood sexuality. To some extent, the paucity of research has been due to general social concerns about the ethical implications of studying children or assumptions about the possible harm to children that would result if they were to be included in sexuality research. Researchers have frequently had difficulty gaining the permission of legal guardians to ask children questions about their knowledge of sexuality. In this atmosphere, it would be exceedingly difficult to get permission to ask children about their sexual behavior. One consequence of this general social concern has been that most of the relevant research has been confined to asking adults or college students to report retrospectively about events that occurred in their childhood. There are rather clear and obvious limitations to this approach.
On the other hand, we should recognize that many American scientists themselves have been unwilling to study the sexuality of children. A recent review, Sexuality Research in the United Slates: An Assessment of the Social and Behavioral Sciences (di Mauro 1995), is notable for the fact that it never mentions childhood sexuality. It might be interesting to determine the extent to which American researchers accept the premise that scientific explorations of sexuality might be harmful to children. For example, the field of child development, a sizable branch of American psychology, has largely ignored the issue of sexuality in their work (Maccoby and Martin 1983; Mussen 1983). An examination of standard developmental texts or reviews of the child development research literature is striking for its omission of sexuality. Significant bodies of child-development research in such important areas as language acquisition, cognition, communication, social behavior, parent-child interaction, attachment (Allgeier and Allgeier 1988), parenting styles, and child compliance have emerged with scant attention to the possible sexual elements of these areas, or to the ways in which these areas might be related to sexual development (Mussen 1983). As just one example, Piaget never investigated the issue of children's sexual cognition, and there has been little subsequent research exploring the application of his theoretical model to sexual development. Similarly, the emergence of family systems theory has also largely ignored the sexuality of children - except to explain the occurrence of incest.
At the same time, it is just as true that sexuality researchers have largely ignored the work of child developmentalists and other scientific disciplines in their own work. They have speculated about how theories of psycho-analysis, social learning, cognition, attribution, social exchange, and symbolic interactionism might be applied to the sexuality of children or to the process of sexual development, but they have rarely tested such assertions empirically (see Allgeier and Allgeier 1988 and Martinson 1976 for examples). Moreover, sex researchers have largely failed to examine how the various processes studied by developmentalists relate to sexuality.
A third domain of this fractured American approach to child development is the fairly recent emergence of professional fields devoted solely to the issue of child sexual abuse. We present a review of child sexual abuse itself later in this chapter (see Section 8A2). Here, we wish to make the point that professional groups - e.g., social workers and family therapists devoted to the treatment of victims of child sexual abuse - have emerged, largely since the 1970s, with a corresponding body of work devoted to that concern. After having been largely neglected for much of the twentieth century, the treatment of child sexual abuse has become a sizable industry in recent years. Unfortunately, much of the work that has been done within this perspective has failed to consider existing data on normative childhood sexuality (Okami 1992, 1995). For example, it is frequently asserted that child sexual abuse has the negative consequence of sexualizing the child's world. We do not mean to claim that child sexual abuse is either harmless or nonexistent. However, the notion that a sexualized childhood is a tragic outcome of sexual abuse rests on the American premise that childhood should be devoid of sexuality. It assumes that childhood should not be sexual. From this perspective, the concept of child sexual abuse has been extended to include family nudity - a point certain to shock naturists in many countries around the world - parents bathing with their children, excessive displays of physical affection (such as kissing and hugging), and even children of the same age engaging in sex play (Okami 1992, 1995). Thus, we seem to have come full circle. Many professionals have come to accept the premise that childhood ought to be an innocent period, free of sexuality. The fact that this view ignores much of the existing data seems to have had little impact on either the American public or many professionals working with children.
Childhood Sexual Development and Expression
In reviewing the process of child sexual development and the phenomenon of child eroticism, it is crucial to consider the meanings that children attach to their experience. There is a tendency to interpret childhood experiences in terms of the meanings that adults have learned to attach to similar events. This ignores the reality that young children almost certainly do not assign the same meanings to sexual events as adults. They have yet to conceptualize a system of experiences, attitudes, and motives that adults label as sexual (Allgeier and Allgeier 1988; Gagnon and Simon 1973; Martinson 1976). A good example is provided by the case of childhood masturbation. Young children often discover that playing with their genitals is a pleasurable experience. However, this may well not be the same as masturbating. Masturbation, as adults understand that term, is a set of behaviors defined as sexual because they are recognized as producing sexual arousal and typically having orgasm or sexual climax as a goal. Young children have yet to construct this complex set of meanings. They know little more than that the experience is pleasurable; it feels good. In fact, it would be useful to see research that examines the process by which children eventually learn to label such self-pleasuring as a specifically sexual behavior called masturbation.
From this perspective, sexual development is, to a considerable extent, a process characterized by the gradual construction of a system of sexual meanings. Gagnon and Simon (1973) have provided a theoretical model of sexual scripting that examines how these meanings are assembled in a series of stages through social interaction with various socialization agents. In their discussion of the model, Gagnon and Simon stressed their intention that it would serve as an organizing framework for future research on the process of sexual development. Although we believe that the model does provide a potentially fruitful framework for thinking about the process of sexual development, and despite the fact that more than twenty years have passed since its original presentation, there is nearly as great a need for research of this type today as when they formulated the model.
One component of the model proposed by Gagnon and Simon (1973) was the concept of assemblies, by which they meant to convey their view that sexual development is actively constructed by humans rather than merely being an organic process. Among the major assemblies they identified were:
1. the emergence of a specific gender identity,Gagnon and Simon maintained that these assemblies were constructed through interactions with a variety of socialization agents, such as parents and family members, same-sex peers, cross-sex peers, and the mass media. To this list, we would suggest adding the church, the school, the neighborhood/community, and boyfriends/girlfriends as potentially important socialization agents. For Gagnon and Simon, the task for researchers was to examine and identify the associations between the activities of various socialization agents and the corresponding construction of specific sexual assemblies. Although a fair amount of research has been conducted on such associations among adolescents (see the following section), sadly there remains relatively little research along these lines for younger children. As such, we will not present a detailed discussion of the activities of each socialization agent here.
2. the learning of a sense of modesty,
3. the acquisition of a sexual vocabulary,
4. the internalization of mass-media messages about sexuality,
5. the learning of specific acts defined as sexual,
6. the learning of gender, family, and sexual roles,
7. the learning of the mechanisms and process of sexual arousal;
8. the development of sexual fantasies and imagery,
9. the development of a sexual value system,
10. the emergence of a sexual orientation, and
11. the adoption of an adult sexual lifestyle.
Lacking space to review each of the assemblies, we have had to be selective and have chosen to focus on the more explicitly erotic dimensions. However, we do wish to note that each is ultimately important to a full understanding of sexual development, and it is likely that each of these assemblies is related to the others. Although we do not have space to review the research on the development of gender roles and gender identity, it appears that most American children have formed a stable gender identity by the age of 2 or 3 (Maccoby and Martin 1983; Money and Ehrhardt 1972). It also seems likely that, as children acquire sexual information and experience, they filter what they learn in terms of what is appropriate for males and females. Since norms for male and female behavior, both sexual and nonsexual, tend to differ, this filtering process seems likely to lead to differences in the content of and processes of male and female sexual development.
On the other hand, we would caution the reader to resist the temptation to conclude that gender differences in sexuality are invariably large, or that they apply to all dimensions of sexuality. Recent reviews of existing research indicate that many aspects of sexuality are not characterized by male-female differences and that many differences are small in magnitude (Oliver and Hyde 1993). Ultimately, the issue is a matter for empirical investigation. Unfortunately, there has been relatively little empirical research attempting to link gender-role development (of which there has been a great deal of research in the last thirty years) with the processes of more overtly sexual development.
Childhood Sexual Eroticism and Expression. Martinson (1976) has drawn a distinction between what he calls reflexive and eroticized sexual experiences. Reflexive experience is pleasurable and may be a result of learning contingencies, but eroticized experience is characterized by self-conscious awareness and labeling of behavior as sexual. As a general guideline, younger and less-experienced children would seem more likely to react to sexual stimuli in a reflexive manner; older and more-experienced children are more likely to have learned erotic meanings and to define similar behaviors as sexual. However, there has been virtually no research detailing the process in which this transition occurs or identifying the factors associated with it.
Sexual Capacity and Autoerotic Play. It has been clear for several decades that infants are capable of reflexive sexual responses from birth. Male infants are capable of erections, and female infants are capable of vaginal lubrication (Allgeier and Allgeier 1988; Halverson 1940). Lewis (1965) observed pelvic thrusting movements in infants as early as eight months of age. Generally, these events appear to be reactions to spontaneous stimuli, such as touching or brushing of the genitals. However, the Kinsey research group (1953) did report several cases of infants less than 1 year of age who had been observed purposely stimulating their own genitals. In their cross-cultural survey, Ford and Beach (1951) reported that, in cultures with a permissive norm, both boys and girls progress from absent-minded fingering of their genitals in the first year of life to systematic masturbation by the age of 6 to 8.
With few exceptions, most research on childhood sexual experiences has asked adolescents or adults to describe events in their past. Males participating in such studies commonly report memories of what they call their first pleasurable erection at such ages as 6 and 9 (Martinson 1976), although, as we have just seen, studies of infants themselves document the occurrence of erections from birth. Kinsey and his associates (1953) did report that almost all boys could have orgasms without ejaculation three to five years before puberty, and more than one half could reach orgasm by age 3 or 4. Comparable data for females have not been presented. In addition, both boys and girls between the ages of 6 and 10 have reported becoming sexually aroused by thinking about sexual events (Langfeldt 1979).
Much has been made in the U.S.A. of the fact that sexual arousal in boys is readily visible (erections). A number of authors have argued that this increases the probability that young boys will discover their penis and are, thus, more likely to stimulate their own genitals than are girls. This idea has become part of the folklore of American culture. We know of no evidence that substantiates this idea. In fact, Galenson and Roiphe (1980) report that there are no gender differences in autoerotic play during the first year of life.
American culture does not encourage such childhood sex play and actively seeks to restrict it. In a study in the 1950s, only 2 percent of mothers reported that they were permissive about their own children's sex play (Sears, Maccoby, and Levin 1957). It is also interesting to note that the researchers in this study did not provide a response category that allowed mothers to indicate they supported or encouraged sex play. Martinson (1973) found this pattern extended well into the 1970s. In a later investigation of parental views toward masturbation, Gagnon (1985) found that the majority (86 percent) of this sample believed that their preadolescent children had masturbated. However, only 60 percent of the parents thought that this was acceptable, and only one third wanted their children to have a positive attitude about masturbation.
Sex Play with Other Children. The capacity to interact with another person in an eroticized manner and to experience sexual feelings, either homosexual or heterosexual, is clearly present by the age of 5 to 6. Langfeldt (1979) did observe both mounting and presenting behaviors in boys and girls at 2 years of age. He also observed that prepubertal boys who engaged in sex play with other children typically displayed penile erections during sex play. Ford and Beach (1951) found that children in cultures, unlike the U.S.A., who are able to observe adult sexual relations will engage in copulatory behaviors as early as 6 or 7 years of age. Moreover, in some cultures, adults actively instruct children in the techniques or practice of sexual relations (Ford and Beach 1951; Reiss 1986). This cross-cultural evidence appears to have had little impact on the way in which most Americans, including many sexuality professionals, think about childhood sociosexual interactions.
Again, most of the research in the U.S.A. has been based on recall data from adolescents or adults. Our impressions of childhood sexual interactions are biased toward periods that such older respondents can remember. A number of studies have examined the frequency of childhood sexual behaviors (Broderick 1965, 1966; Broderick and Fowler 1961; Goldman and Goldman 1982; Kinsey et al. 1948, 1953; Martinson 1973, 1976; Ramsey 1943). Taken together, these studies demonstrate that many American children develop and maintain an erotic interest in the other or same sex, and begin experiencing a wide range of sexual behaviors as early as age 5 to 6. It is not uncommon for Americans to report that they remember playing doctor or similar games that provide opportunities for observing and touching the genitals of other children, undressing other children, or displaying their own genitals to others. Many American children also acquire experience with kissing and deep kissing (what Americans call French kissing). In fact, generations of American children have played institutionalized kissing games, such as spin the bottle and post office. These studies also provide evidence that at least some American children experience sexual fondling, oral sex, anal sex, and intercourse prior to puberty. Many of these behaviors are experienced in either heterosexual or homosexual combinations or both.
We have purposely avoided reporting the specific frequencies of the childhood sociosexual experiences in these studies because each possesses severe limitations with respect to generalizability. Most have had small samples drawn from a narrow segment of the total population in a specific geographic region. As early as the 1960s, researchers found evidence of racial and community differences in the rate of such behaviors (Broderick 1965, 1966; Broderick and Fowler 1961). In addition, most have used volunteer samples, with respondents who were trying to recall events that had occurred ten or more years earlier. Moreover, these studies were conducted over a period of five decades, during which there would seem to be great potential for changes. Comparisons among these studies are virtually impossible. As a result, we would have little confidence in the specific accuracy of frequency estimates.
A review of a few of these studies illustrates this point. Interviewing a group of boys in a midwestern city in the early 1940s, Ramsey (1943) found that 85 percent had masturbated prior to age 13, one third had engaged in homosexual play, two thirds had engaged in heterosexual play, and one third had attempted or completed intercourse. The Kinsey group (1948), using a broader sample of adults, reported that 45 percent had masturbated by age 13, 30 percent had engaged in homosexual play, 40 percent had engaged in heterosexual play, and 20 percent had attempted intercourse. For girls, the Kinsey group (1953) reported that roughly 20 percent had masturbated prior to age 13, roughly one third had engaged in both heterosexual and homosexual play, and 17 percent had attempted intercourse. They also reported an actual decline in sexual behaviors after age 10 (Kinsey et al. 1948). The large differences between the Ramsey and Kinsey findings could be due to sample size, differences in geographic region or size of the city, differences in the time period of data collection, or differences in the age range 'of the samples. Here, it is interesting to note that the Kinsey group (1948) also interviewed a small sample of boys. Roughly 70 percent reported some form of child sex play, a figure that is much closer to Ramsey's findings. In the larger Kinsey sample, only 57 percent of adult males and 48 percent of adult females reported memories of childhood sex play, usually between the ages of 8 to 13 (Kinsey et al. 1948, 1953). It would seem possible, then, that studies with adult samples recalling their childhood experiences might well yield lower estimates than studies of children themselves.
John Money (1976) and Money and Ehrhardt (1972) argue that childhood sex play with other children is a necessary and valuable form of rehearsal and preparation for later adult sexual behavior. He has also suggested that such sex play may occur as part of a developmental stage in childhood. Certainly, this phenomenon has been observed in other primate species, such as the chimpanzee (DeWaal 1982). However, Kilpatrick (1986, 1987) found no differences in various ages of adult sexual functioning between persons who had childhood sexual experiences with other children and those who did not. Given the complexity of the model of sexual assemblies we have presented here, it is not surprising that the effects are not that simple.
Sibling Incest. We discuss incest and child sexual abuse more fully later in Section 8A on coercive sex. Here, we merely wish to note that, in one of the few studies of sibling incest with a nonclinical sample, Finkelhor (1980) found that 15 percent of female and 10 percent of male college students reported having a sexual experience with a brother or sister. Approximately 40 percent of these students had been under the age of 8 at the time of the sexual activity, and roughly 50 percent had been between the ages of 8 and 12. Three quarters of the experiences had been heterosexual. Some type of force had been used in one quarter of the experiences. The most common sexual activities were touching and fondling of the genitals. Only 12 percent of the students had ever told anyone about these sexual activities with a brother or a sister. Interestingly, most of the students reported that they did not have either strong positive or negative feelings about these experiences. Positive reactions were reported by 30 percent, and another 30 percent reported negative reactions. Positive reactions were associated with consensual activities (no force had been used) and an age difference of four or fewer years. For males, there were no correlations between prior sibling experiences and current sexual activity. Among females, those who had had sibling sexual experiences were more likely to be currently sexually active. Those women who had positive sibling experiences after age 9 had significantly higher sexual self-esteem, whereas those who had sexual experiences before age 9 with a sibling more than four years older had lower self-esteem.
Sexual Contacts with Adults. A recent national survey (Laumann et al. 1994) found that 12 percent of men and 17 percent of women reported they had been sexually touched by an older person while they were children. The offender was typically not a stranger, but a family friend or a relative, a finding that is comparable to more-limited samples. We present a more complete review of sexual contacts with adults later in Section 8A2 on child sexual abuse and incest. Relatively few studies of adult-child sexual contacts have been conducted with nonclinical samples. In general, they indicate that children experience a wide range of reactions, from highly negative or traumatic to highly positive, to such contacts in both the short term and long term (Kilpatrick 1986, 1987; Nelson 1986; Farrell 1990). Moreover, there do not appear to be any simple or direct correlations between such childhood experiences and later measures of adult sexual functioning. In her study of incest, Nelson (1986) found no correlation between affective outcomes and type of erotic activity, sexual orientation, or consanguinity. Kilpatrick (1986) did find that the use of force or abuse was significantly related to impaired adult sexual functioning in several areas.
Same-Sex Childhood Experiences. Our discussion to this point has not focused exclusively on heterosexual experience, but it is certainly fair to say that investigations of heterosexual child sex play have dominated existing research. One study of 4- to 14-year-old children found that more than one half of boys and one third of girls reported at least one homosexual experience (Elias and Gebhard 1969). Masturbation, touching of the genitals, and exhibition were the most common activities, although there were also some reports of oral and anal contacts. The fact that children have had such a homosexual experience does not appear to be related to adult sexual orientation (Bell, Weinberg, and Hammersmith 1981; Van Wyk and Geist 1984).
Storms (1981) has hypothesized that such experiences may be related to adult sexual orientation as a function of sexual maturation. He suggests that persons who become sexually mature during the period of homosocial networks (discussed below) may be more likely to romanticize and eroticize these childhood homosexual experiences and, thus, develop a later preference for sexual partners of the same gender. In effect, when sexual maturation, goal-directed masturbation, homosexual explorations, and eroticized fantasies are paired before heterosexual socialization occurs (typically at about age 13), they are more likely to lead to a homosexual orientation later. As far as we know, Storms's ideas have never been directly tested through research.
Childhood Social Networks. During middle childhood (roughly ages 6 to 12), both boys and girls in the U.S.A. tend to form networks of same-sex friends. A pattern of gender segregation, where boys and girls have separate friends and play groups, is central to the daily life of middle childhood. This pattern of homosocial networks is readily observable at elementary schools across the U.S.A. Girls and boys tend to cluster at school into separated, same-sex groups. At lunchtime, they frequently sit at separate girl's tables and boy's tables. On the playground, space and activities tend to be gendered. After school, children tend to associate and play in gender-segregated groupings. In fact, this pattern of gender separation may be more pronounced in middle childhood in the U.S.A. than the more-publicized racial segregation.
It should be acknowledged that these homosocial networks are not characterized by a total separation of the genders. There are some opportunities for heterosocial interactions and play, and children do vary with respect to the extent in which they associate with the other sex. As just one obvious example, some girls, who are known as tomboys, spend considerable time associating with boys. Still, to a large extent, the worlds of boys and girls in middle childhood in the U.S.A. are separated.
Maltz and Borker (1983) have suggested that these homosocial networks can be viewed as distinct male and female cultures. As cultures, each has its own set of patterns, norms, and rules of discourse. Boys tend to play in groups that are arranged in a hierarchy. They stress a norm of achievement (doing) and emphasize competitive, physical activities. Conflict is overt and is often resolved directly through physical Fighting. Differentiation between boys is made directly in terms of power and status within the group. Since boys belong to more than one such group, and because group memberships do change over time, each boy has an opportunity to occupy a range of positions within these hierarchies. Boys' groups also tend to be inclusive. New members are easily accommodated, even if they must begin their membership in a lower-status position. Courage and testing limits are prime values of boys' groups, and breaking rules is a valued form of bonding. In examining how these patterns influence male communication, Maltz and Borker (1983) report that males are more likely to interrupt others, they are more likely to ignore the previous statement made by another speaker, they are more likely to resist an interruption, and they are more likely to directly challenge statements by others.
Girls tend to associate in smaller groups or friendship pairs. Girls, for example, tend to be highly invested in establishing and maintaining a best friend relationship. They stress a norm of cooperation (sharing) and pursue activities that emphasize working together and being nice. They frequently play games that involve taking turns. Friendship is seen as requiring intimacy, equality, mutual commitment, and cooperation. However, girls' groups also tend to be exclusive. Membership is carefully reserved for those who have demonstrated they are good friends. Conflict tends to be covert, and it is highly disruptive, leading to a pattern of shifting alliances among associates. Differentiation between girls is not made in terms of power, but rather in relative closeness. Girls are more likely to affirm the value of rules, especially if they are seen as serving group cohesion or making things fair. Girls may break rules, but their gender group does not provide the intense encouragement and support for this behavior seen among boys. Maltz and Borker (1983) note that girls are more likely to ask questions to facilitate conversation, they are more likely to take turns talking, they are more likely to encourage others to speak, and they are more likely to feel quietly victimized when they have been interrupted.
These largely segregated gender networks in middle childhood serve as the contexts for learning about adolescent and adult sexual patterns, as well as for other areas of social life. There is, of course, a certain irony to the fact that homosocial networks serve as a principal learning context for heterosexuality in a culture with such strong taboos against homosexuality as the U.S.A. In fact, Martinson (1973) has argued that these gender networks and this period serve as the settings for a fair amount of homosexual exploration and activity. In one sense, it is almost certainly true that some homosexual activity results from these patterns of social organization. However, this assertion is largely undocumented, and we are not aware of any studies that compare the level of homosexual activity in cultures with homosocial networks with cultures having some other form of childhood networks.
Thorne and Luria (1986) have used this concept of gendered cultures to examine the process of sexual learning in middle childhood. They found that talking dirty is a common format for the rule-breaking that characterizes boys' groups. They noted that talking dirty serves to define boys as apart from adults, and that boys get visibly excited while engaging in such talk. Boys also often share pornography with each other and take great care to avoid detection and confiscation by adults. These processes provide knowledge about what is sexually arousing, and they also create a hidden, forbidden, and arousing world shared with other boys, apart from adults and girls. Miller and Simon (1981) have argued that the importance attached to rule violations creates a sense of excitement and fervor about sexual activity and accomplishment.
One other feature of boys' groups is that they serve as a setting for learning both homoeroticism and homophobia. Boys learn to engage in what Thorne and Luria call fag talk. That is, they learn to insult other boys by calling them names, like faggot and queer. Eventually, they learn that homosexuality is disapproved by the male peer group. Boys at age 5 to 6 can be observed touching each other frequently. By age 11 to 12, touching is less frequent and reduced to ritual gestures like poking each other. On the other hand, much of the time spent with other boys is spent talking about sex. This serves to maintain a high level of arousal within the group. Moreover, the sanctioning of rule-breaking leads to some homosexual experimentation that is kept hidden from the group. Homosexual experiences may become one more form of breaking the rules and one more feature of the secret, forbidden world of sexuality.
In contrast, girls are more likely to focus on their own and their friends' physical appearance. They monitor one another's emotions. They share secrets and become mutually vulnerable through self-disclosure. They have giggling sessions with their friends, with sex often being the source of amusement. Their talks with other girls tend to focus less on physical activities and more on relationships and romance. They also plot together how to get particular boys and girls together in a relationship.
These sexual patterns are largely consistent with the norms of the respective gender cultures. Males tend to focus on physical activities; females on cooperation and sharing. They are also quite consistent with patterns that will become firmly established in adolescent sociosexual patterns. Thus, male and female peer groups become the launching pads for heterosexual coupling as boys and girls begin to go together. Finally, they serve to heighten the romantic/erotic component of interactions with the other gender.
Professional and Social Issues of Childhood Sexuality
As we stated at the beginning of this section and as should be apparent from the review of sex education in the U.S.A., there are a number of issues concerning childhood sexuality that have been controversial for decades. Moreover, several new issues have become points of social conflict in recent years. We can only briefly mention four here.
The Oedipus and Electra Complexes. The Goldmans' (1982) multinational study of children and sexual learning, including a sizable American sample, raises questions about these complexes. Freud's thesis about castration anxiety and its resolution (typically by the age of 5) would presumably require some awareness of genital differences between males and females, unless one wishes to interpret Freud's terminology strictly as metaphorical. In the Goldman study, the majority of English-speaking children did not understand these differences until they were 7 to 9 years old. Interestingly, a majority of the Swedish children could accurately describe these differences by the age of 5.
Is There a Latency Period? The notion of a latency period, roughly from ages 6 to 11, has had great appeal in American culture. This may be due to the impression that the homosocial networks of middle childhood reflect a lack of sexual interest, and to the fact that many Americans prefer to believe that childhood is a period of sexual innocence. Freud (1938) originally proposed in 1905 that middle childhood is characterized by relative sexual disinterest and inactivity, something like a dormant period. Freud also maintained that latency was more pronounced among boys than girls. The review above should certainly dispel the notion that childhood, at any point, is essentially characterized by sexual disinterest.
In addition, Broderick (1965, 1966) not only provided evidence of active sex play during middle childhood, but also demonstrated that most children indicate they wish to marry as an adult, and that most of these children are actively involved in a process of increasing heterosocial interaction and love involvements during childhood, A majority said they had had a boyfriend or girlfriend and had been in love, and 32 percent had dated by age 13. If anything, we would expect that the age norms for many of these behaviors have actually decreased since that time. Interestingly, those children who indicated that they did not wish to marry eventually were substantially less likely to report any of these activities.
Parental Nudity. Experts have disagreed over the years as to the impact of parental nudity on children (Okami 1995). Some have argued that childhood exposure to parental/adult nudity is potentially traumatic - largely because of the large size of adult organs. Others have insisted that strong taboos on family nudity may lead to a view that the body is unacceptable or shameful. This group has argued that a relaxed attitude toward nudity can help children develop positive feelings about sexuality. Similar concerns have been expressed about the primal scene and sleeping in the parental bed. In a survey of 500 psychiatrists, 48 percent indicated that they believe that children who witness their parents engaging in intercourse do suffer psychological effects (Pankhurst 1979). American experts appear to overlook the fact that most families throughout the world sleep in one-room dwellings. In one study of these issues, Lewis and Janda (1988) asked 200 college students to report their childhood experiences. Exposure to parental nudity for ages zero to 5 and 6 to 11 was generally unrelated to a series of measures of adult sexual adjustment. Sleeping in the parental bed yielded several small, but significant correlations. Persons who had slept in their parents' bed as children had higher self-esteem, greater comfort about sexuality, reduced sexual guilt and anxiety, greater frequency of sex, greater comfort with affection, and a higher acceptance of casual sex as college students.
Okami (1995) reviewed the literature in these same three areas. His review provides a thorough summary of clinical opinions in each area, as well as an assessment of the empirical evidence. Despite the growing number of clinical professionals who label such acts as sexual abuse, there is virtually no empirical evidence of harm. In fact, the only variable found to be associated with harm is cosleeping, which has been found to be associated with sleep disturbances. However, Okami notes that these sleep disturbances may well have preceded and precipitated the cosleeping, rather than vice versa.
Female Genital Cutting. In December 1996, the Center for Disease Control and Prevention (CDC) estimated that more than 150,000 women and girls of African origin or ancestry in the United States were at risk in 1995 of being subjected to genital cutting or had already been cut. This estimate was based on 1990 Census Bureau data gathered before the recent increase in refugees and immigrants from the 28 countries that span Africa's mid-section where female genital cutting varies widely in prevalence and severity (Dugger 1996ab). A second source cites a different estimate from the CDC using data on how much circumcision is practiced in immigrants' homelands and, making assumptions about sex and age, that about 270,00 African females in the United States were circumcised in their home country or are at risk here (Hamm 1996)
In 1996, Congress adopted a dual strategy to combat the practice here. In April 1996, Congress passed a bill requiring the Immigration and Naturalization Service to inform new arrivals of U.S. laws against genital cutting. It also mandated the Department of Health and Human Services to educate immigrants about the harm of genital cutting and to educate medical professionals about treating circumcised women. A law, which went into effect March 29, 1997, also criminalizes the practice, making it punishable by up to five years in prison and a fine of up to $250,000 for individuals and $500,000 for organizations such as hospitals. Enforcement of the law, however, is problematic for several reasons. First, no one is sure how the law will apply to those immigrants who take their daughters out of the country for the rite. Second, doctors who spot cases of genital mutilation are reluctant to report it for fear of breaking up tight-knit families. Also, when the wounds are healed, it is impossible to ascertain whether the rite was performed here or before arrival in the United States. Finally, there is the secretiveness surrounding this rite of passage, which many African cultures consider essential, and also the hidden nature of the wounds and scars. Sierra Leoneans, for instance, who consider genital cutting part of an elaborate, highly secret initiation rite, view questions about it as a profound invasion of their privacy (Dugger 1996ab).
A government's prevention program focuses on educating both old and recent immigrants in how to survive and assimilate in American society while maintaining their own culture and religion. To this purpose, the U.S. Department of Health and Human Services has organized meetings with advocates for refugees and nonprofit groups that work closely with Africans to develop strategies for combating this practice. Muslim religious leaders, for instance, are invited to explain that the Koran does not require this practice. However, lack of a specific budget hampers this effort.
In one attempt to ameliorate this clash of cultural values, doctors at Harborview Medical Center in Seattle, Washington, persuaded Somali mothers to be satisfied with nicking the clitoral hood without removing any tissue. The ritual usually involves removing the clitoris and sewing the labia closed. The compromise was abandoned in December of 1996 when the hospital was inundated with hundreds of complaints protesting even this compromise, even though the nicking of the clitoral hood has no short-or long-term negative consequences. The massive objection to this compromise raises serious questions of ethnocentrism on the part of the Americans who protested it. It seems somewhat ironic that such complaints would be made in a culture where we routinely circumcise penises. Although some maintained that the compromise of nicking may violate the letter of the law, it remains to be seen what kind of solution will be achieved in this matter (Dugger 1996b).
Child Pornography. It is widely believed, and the Federal Bureau of Investigation (FBI) perpetuates the notion, that child pornography is pervasive and increasing. Several state and federal laws have been enacted in the last twenty years to combat this perceived social problem. The mere possession of a photograph of a naked child has been criminalized in some states. Yet, it is virtually impossible to find any commercial child pornography in the U.S.A. In fact, most of the materials seized by the FBI are private photographs of naked children - with no adults appearing in the photos and no sexual behaviors depicted (Klein 1994; Stanley 1989). Efforts to raid child-pornography businesses have routinely failed to seize any child pornography. FBI sting operations may well have arisen from the corresponding frustrations of government agencies to find any child pornography. One recent legend now circulating is the claim that the U.S. government is now the largest producer of child pornography in the world. This claim is unsubstantiated as far as we know, but, again, it reflects the anxiety of American culture over the sexuality of its children.
B. Adolescent Sexuality
DAVID L. WEISCourtship, Dating, and Premarital Sex
In stark contrast to the relative inattention given to childhood sexuality in the U.S.A., Americans have been fascinated by the sexual behavior of adolescents throughout the twentieth century. One is tempted to describe the interest as an obsession. Perhaps no area of sexuality has received as much scrutiny, by both the general public and professionals, as the sexual practices of American teenagers. There have been literally hundreds of scientific studies attempting to determine the rate of adolescent premarital coitus, as well as other aspects of adolescent sexuality. The easy availability of populations to study is only one of the more-obvious reasons for this extensive research.
Since more than 90 percent of Americans ultimately do marry, investigations of adolescent sexual development and premarital sexual practices largely overlap. General trends have been well documented, compared to other areas of sexuality. Given the vast scope of this research, we can review only the highlights here. (For more extensive reviews of research on adolescent and premarital sexuality, see Cannon and Long 1971; Clayton and Bokemeier 1980; and Miller and Moore 1990.)
The issue of premarital sexuality (hereafter PS) and virginity has been a focus of considerable social conflict and concern throughout this century, and remains so to this day. Beginning in the early years of this century, a large literature documents the continuing concern of American adults about the increasing number of teenagers who have experienced sexual intercourse prior to marriage. Interestingly, each successive birth cohort of American adults in this century has been concerned about the tendency of their offspring to exceed their own rate of premarital coitus.
Much of the professional literature has reflected these same concerns. Through much of the twentieth century, the tone of most professional writings has been moralistic. Adults in the U.S.A., including most sexuality researchers, have tended to view adolescent premarital sexual intercourse, PS, as a deviant behavior, as a violation of existing social norms, and as a growing social problem (Spanier 1975). Research has tended to parallel this perspective by emphasizing the costs or negative consequences of adolescent sexuality, such as sexually transmitted disease (venereal disease), illegitimate pregnancy, and loss of reputation (Reiss 1960). This tone may have shifted to a less-judgmental, more-analytic perspective in the 1960s and 1970s (Clayton and Bokemeier 1980). However, with the emergence of AIDS and the rise of out-of-wedlock pregnancies in the early 1980s, the general tone has reverted in recent years, with studies of risk-taking behavior, at-risk youth, and portrayals of adolescent sexuality as a form of delinquency (Miller and Moore 1990).
Trends in Adolescent Sexuality
Despite these adult concerns, it would be fair to suggest that premarital virginity has largely disappeared in the U.S.A., both as a reality and as a social ideal. As we approach the end of the century, the overwhelming majority of Americans now have sexual intercourse prior to marriage, and they begin at younger ages than in the past. Love has largely replaced marital status as the most valued criteria for evaluating sexual experience (Reiss 1960, 1967, 1980). Virtually all Americans believe that intimate relationships (like marriage) should be based on love, that love justifies sexual activity, and that sex with love is a more-fulfilling human experience. This view has not only been used to justify PS activity between loving partners, but has also become a criteria for evaluating marital sexuality itself and justifying a pattern of divorce and remarriage.
Premarital Sexual Behavior. These trends may not be quite as dramatic as most Americans imagine. A study of marriages in Groton, Massachusetts, from 1761 to 1775 found that one third of the women were pregnant at the time of their weddings (cited in Reiss 1980), demonstrating that PS was already fairly common in the colonial period (see discussion of bundling in Section 1A). Several early sexuality surveys also document that PS occurred among some groups prior to the twentieth century. Terman (1938) compared groups who were born in different cohorts around the beginning of the century. Of those born before 1890, 50 percent of the men and only 13 percent of the women had premarital coitus. Two thirds of the men who had PS did so with someone other than their future spouse, whereas two thirds of the women who had PS did so only with their future spouse. For those born after 1900, two thirds of the men and nearly half of the women had PS. The relative percentage having PS with their fiance also increased. Fully half of the men and 47 percent of the women had sexual relations with their fiance prior to marriage.
The Kinsey team (1953) found that one quarter of the women born before 1900 reported they had PS, whereas one half of those born after 1900 said they had PS. Like the Terman study, the major change was an increase in the percentage of women born after 1900 who had PS with their fiance. The Kinsey study also indicated that the period of most-rapid change was from 1918 to 1930 - the Roaring Twenties. Burgess and Wallin (1953) reported similar findings for a birth cohort born between 1910 and 1919. These studies indicated that roughly two thirds of the men born after 1900 had PS. The Kinsey studies also found that there had been comparable increases in female masturbation and petting behavior as well.
It is important to note that the growth of PS in the first half of the century occurred primarily within the context of ongoing, intimate relationships. It appears that the percentage of males and females having PS remained fairly stable through the 1950s and early 1960s. In a study of college students during the 1950s, Ehrmann (1959) found rates similar to the Kinsey figures cited above. Ehrmann found that males tended to have greater sexual experience with females from a social class lower than their own, but they tended to marry women from their own social class. Males who were going steady were the least likely to be having intercourse. In contrast, females who were going steady were the most likely to be having intercourse. In a study comparing college students in Scandinavia, Indiana, and Utah (predominantly Mormon), Christensen (1962) and Christensen and Carpenter (1962) found that rates of PS vary by the norms of the culture and that guilt is most likely to occur when PS is discrepant with those norms.
A second wave of increases in PS seems to have occurred in the period from 1965 to 1980. A number of studies of college students through this period indicated increasing percentages of males and females having premarital coitus (Bauman and Wilson 1974; Bell and Chaskes 1968; Christensen and Gregg 1970; Robinson, King, and Balswick 1972; Simon, Berger, and Gagnon 1972; Vener and Stewart 1974). For example, Bauman and Wilson (1974) found that, for men, the rate having PS increased from 56 percent in 1968 to 73 percent in 1972. For women, the increase was from 46 percent to 73 percent. There was no significant change in the number of sexual partners for either gender. Several of these studies indicate that the increases were still moderate by 1970 (Bell and Chaskes 1968; Simon et al. 1972). In an unusual study of male college students attending an eastern university in the 1940s, 1960s, and 1970s, Finger (1975) found that 45 percent had PS in 1943-44, 62 percent in 1967-68, and 75 percent in 1969-73.
Subsequent studies have indicated that this pattern of increasing PS characterized American youth in general. In a study of urban samples in the mid-1970s, Udry, Bauman, and Morris (1975) found that 45 percent of white teenage women had intercourse by age 20, and 80 percent of black women did. Roughly 10 percent of whites had PS by age 15 and 20 percent of blacks did Zelnik and Kantner found similar percentages in their studies in 1971 and 1976 (Udry, Bauman, and Morris 1975; Zelnik, Kantner, and Ford 1981).
Reports of increasing sexual activity among adolescents have not been limited to coitus. A number of researchers have reported similar increases in the rate of heavy petting (manual caressing of the genitals) through the late 1960s and 1970s (Clayton and Bokemeier 1980; Vener and Stewart 1974). There have also been reports of increasing levels of oral sex among adolescents (Haas 1979; Newcomer and Udry 1985). In some studies, teenage girls have been more likely to have participated in oral sex than intercourse, and between 16 percent to 25 percent of teens who have never had intercourse have had oral sex (Newcomer and Udry 1985). Weis (1983) has noted that this group may be involved in a transition from virginity to nonvirginity, at least among whites.
Perhaps the single best indicator of the trends occurring from 1965 to 1980 is the series of studies by Zelnik and Kantner in 1971, 1976, and 1979 (Zelnik et al. 1981). These studies, known as the National Surveys of Young Women, investigated the sexual histories of 15- to 19-year-old women. The 1971 and 1976 studies were full national probability studies while the 1979 study focused on women living in metropolitan areas. The Zelnik and Kantner research shows a dramatic rise in sexual activity for both black and white women from 1971 to 1976. The pattern of increases continued for white women through 1979, but PS rates for black women remained stable from 1976 to 1979. Among metropolitan women, PS rose from 30.4 percent in 1971 to 49.8 percent in 1979. For blacks, the rate moved from 53.7 percent in 1971 to 66.3 percent in 1976, and was 66.2 percent in 1979. The 1979 study also showed that 70 percent of males had PS intercourse; the figure for black men was 75 percent (Zelnik and Shah 1983; Zelnik et al. 1983).
In a review of these trends, Hofferth, Kahn, and Baldwin (1987) noted that females in the 1980s became sexually active at younger ages and that fewer teenagers married. As a result, the rate of PS increased. The proportion of women at risk of premarital pregnancy increased dramatically from 1965 to the 1980s. The out-of-wedlock pregnancy rate among teenagers increased for both blacks and whites from 1971 to 1976. This trend continued for whites through 1982, but remained level for blacks after 1976. Finally, they noted that, for women born between 1938 and 1940, 33.3 percent had PS by age 20. For women born between 1953 and 1955, the figure was 65.5 percent.
Despite recent claims in some quarters of a return to chastity and abstinence in the late 1980s and 1990s (McCleary 1992), there is no evidence of a decline in PS behavior. National data from 1988 indicate that one quarter of females have PS intercourse by age 15; 60 percent do so by age 19. About one third of United States males have PS intercourse by age 15, and 86 percent by age 19 (Miller and Moore 1990). In fact, a random telephone survey of 100 students attending a midwestern state university in 1994 found that 92 percent had had sexual intercourse; only 8 percent said they were still virgins. Nearly two thirds (63 percent) said that they had participated in what the survey described as a one-night stand. With respect to their most recent sexual intercourse, 42 percent reported using something to protect themselves. Of these, 84 percent reported using condoms; 16 percent said they used the pill (Turco 1994). If anything, the trends that have been well established throughout this century appear to be continuing. Given the continuation of patterns that have been frequently cited as leading to increasing rates of PS, such as industrialization, rapid transportation, dating, and going steady, we would not expect a reversal in what is now a century-long trend.
Premarital Sexual Attitudes (Permissiveness). There has also been a substantial number of studies examining the attitudes of Americans toward PS, although systematic research in this area began later than research on PS behavior. Reiss (1960) used the term permissiveness to describe the extent to which the attitudes of an individual or a social group approved PS in various circumstances. In general, research has found that PS attitudes have become progressively more permissive throughout this century, roughly parallel to the increases in PS behavior (Bell and Chaskes 1970; Cannon and Long 1971; Christensen and Gregg 1970; Clayton and Bokemeier 1980; Glenn and Weaver 1979; Vener and Stewart 1974). Reiss (1967) developed what has come to be called Autonomy Theory to explain this process. According to Reiss, PS permissiveness will increase in cultures where the adolescent system of courtship becomes autonomous with respect to adult institutions of social control, such as the church, parents, and the school. This appears to have happened in the U.S.A. and most other industrialized nations in the twentieth century.
By far, the biggest change has been the growth of a standard that Reiss (1960, 1967, 1980) called permissiveness with affection, in which PS is seen as acceptable for couples who have mutually affectionate relationships. This standard has grown in popularity in the U.S.A. as the double standard - the view that PS is acceptable for males but not females - has declined (Clayton and Bokemeier 1980; Reiss 1967, 1980). By 1980, a majority of adults as well as young people in the U.S.A. believed that PS is appropriate for couples involved together in a serious relationship (Glenn and Weaver 1979). Moreover, although there has been a historical tendency for males to be more permissive about PS than females, these gender differences have been diminishing in recent decades (Clayton and Bokemeier 1980).
Circumstances of Adolescent Sexual Experiences
Most research on adolescent sexuality has tended to focus on whether or not teenagers or college students have had PS intercourse. Although this allows us to provide reasonable estimates of the percentages of Americans who have had PS in various time periods and to track trends in the rate of virginity and nonvirginity, this same focus has frequently led researchers to ignore the circumstances in which adolescent sexuality occurs (Miller and Moore 1990). As a consequence, we cannot be as confident about the trends in several related areas, and many questions about the specific nature of adolescent sexual experiences and relationships remain to be explored.
First Intercourse. A good example of this lack of perspective is provided by the evidence concerning age at first intercourse. The available research indicates that the average age of first intercourse has been declining since 1970. It seems likely that this trend extends back prior to 1970, but the paucity of relevant data from earlier time periods makes such a conclusion highly tentative. As late as that year, only about one quarter of the males and 7 percent of the females who attended college had intercourse prior to age 18 (Simon et al. 1972). In the Zelnik and Kantner studies, the average age for females dropped from 16.5 in 1971 to 16.2 in 1976 (Zelnik et al. 1981). By 1979, the average age of first intercourse for women was 16.2; for males, it was 15.7. Blacks of both genders tended to experience sexarche at slightly younger ages than whites. Females had first partners who were nearly three years older; whereas males had first partners who were about one year older than they (Zelnik and Shah 1983).
In a study of college females in the 1980s, Weis (1983) found the average age of sexarche to be 16.2. A later study of college students found that the average age was 16.5 (Sprecher, Barbee, and Schwartz 1995). It should be noted, however, that persons who attend college may well be more likely to postpone sexual activity. It is conceivable that a trend of declining age at first intercourse is still occurring among populations that do not attend college, and it is possible that teenagers in the 1990s (who have yet to reach the age of college) may also be having intercourse at younger ages.
Intercourse appears to be, at least among whites, the culmination of a sequence of increasing and expanding experiences with kissing, petting, and possibly oral sex (Spanier 1975; Weis 1983). There is some evidence that women who have rehearsed these noncoital activities extensively, and thus gradually learned the processes of sexual interaction, are more likely to report positive reactions to their first intercourse (Weis 1983). Weis (1983) found that there is great variation as to when people go through these stages and how quickly.
Most authors have stressed the negative aspects of first intercourse for females by citing the finding that females are significantly more likely to report negative affective reactions to their first intercourse than males (Koch 1988; Sprecher et al. 1995). However, the available data strongly suggest that the differences between males and females may not be large in magnitude. It is clear that females report a wide range of affect, from strongly positive to strongly negative (Koch 1988; Schwartz 1993; Weis 1983), but it is also clear that many males report experiencing negative reactions as well. In a study of college students, the males were more likely to report experiencing high levels of anxiety, the females were less likely to report experiencing high levels of subjective pleasure, while sizable numbers of both genders reported experiencing guilt (Sprecher et al. 1995). Positive reactions to first intercourse have been found to be related to prior experience with noncoital sexual activities, having an orgasm in that first intercourse encounter, descriptions of the partner as gentle and caring (for females), involvement with the first partner for more than one month prior to first intercourse, continued involvement with the partner following the first intercourse, and situational factors, such as the consumption of alcohol (Schwartz 1993; Sprecher et al. 1995; Weis 1983). Several researchers have reported that age is associated with affective reactions, but Weis (1983) found that age was not as strongly or directly related as the level of prior noncoital experience. Schwartz (1993) also reported that Scandinavian teenagers were more likely to report positive reactions than a group of American adolescents.
Over the past three decades, a convergence of male and female PS behavior has been identified, with females reporting less emotional attachment to their first coital partners than in the past (Hopkins 1977; Kallen and Stephenson 1982; Koch 1988). Yet, there is still a significant difference between the genders, with males reporting more casual relationships and females more intimate relationships with their first partners (Koch 1988).
In the only national study of first intercourse, Zelnik and Shah (1983) found that more than 60 percent of the females were going with or engaged to their first partner. Another third described their first partner as a friend. Roughly a third of the males described their first partner as a friend, and 40 percent were going with or engaged to their first partner. The males were twice as likely to have their first intercourse with someone they had just met, although few males or females did this (Zelnik and Shah 1983).
Relationship factors have been reported to be associated with affective reactions to the first intercourse. However, the precise nature of this association remains unclear. There is some evidence that involvement with a partner for longer than one month, and continuing involvement following the first intercourse, are associated with positive affective reactions (Sprecher et al. 1995). There is some evidence that females who are going with or engaged to their first partner are more likely to experience positive affect (Weis 1983). However, Weis (1983) also found that attributions that the first partner was caring, considerate, and gentle were more strongly related to affective reactions. Moreover, many women who were going with or engaged to their first partner, nonetheless, described their partners as uncaring and inconsiderate. It should be noted that each of these studies found so few participants who were married at the time of their first intercourse that no analyses could be done for that relationship category. For example, not one woman in the Weis (1983) study was married at the time of her first intercourse.
Adolescents appear to have many reasons for becoming involved in PS behavior. Motivations most frequently mentioned by a group of college women for becoming involved in their First intercourse experience included (rank-ordered by declining frequency): love-caring, partner pressure, curiosity, both wanted to, alcohol or other drugs, and sexual arousal (Koch 1988). The comparable rank-ordering of motivations by a group of college men included: both wanted to, curiosity, love-caring, sexual arousal, to get laid, and alcohol/drug use. Women were four times more likely to report partner pressure than men, whereas men were seven times as likely to say they were looking to get laid and twice as likely to report sexual arousal as a motivation for sexarche (Koch 1988).
Most American teenagers describe their first intercourse as an un-planned, spontaneous event. Only 17 percent of the females and one quarter of the males in a national study said they had planned their first intercourse (Zelnik and Shah 1983). In the same study, less than one half of the males and females used a contraceptive. Those who had their first intercourse at age 18 or older were more likely to use a contraceptive. White women were more likely to have used some form of contraception, but black women were more likely to use a medically prescribed method. Women who described their first intercourse as planned were more likely to have used a contraceptive - fully three quarters of these women did. However, more than two thirds of these women relied on their partners to use a condom or withdrawal. Black women were more likely to use a contraceptive themselves, rather than rely on their partner.
Finally, various aspects of sexarche have been found to be significantly related to later sexual functioning among college students (Koch 1988). Women who had experienced first coitus at an earlier age had less difficulty reaching orgasm during later sexual interactions than did women who had sexarche at a later age. Men with earlier sexarche had less difficulty in keeping an erection during later sexual interactions than men who had been older at sexarche. Also, women who had reported negative reactions to their first intercourse were subsequently more likely than those who felt more positively to experience: lack of sexual interest, sexual repulsion, inability to reach orgasm, or genital discomfort, pain, or vaginal spasms. Men who reacted negatively to their first intercourse were more likely to ejaculate too quickly during later sexual experiences than men who had positive reactions. Both men and women were more likely to experience subsequent sexual functioning concerns when they were pressured by a close partner to engage in intercourse for the first time.
Number of Premarital Sexual Partners. It is difficult to provide good estimates on the number of PS partners prior to 1950, simply because researchers failed to ask such a question. On the other hand, it does seem clear that the increase in the percentage of American women who reported they had ever had PS after 1900 was due primarily to an increase in the percentage of women who reported they had PS only with their fiance (Kinsey et al. 1953; Terman 1938). In contrast, there is abundant evidence of a significant increase in the number of PS coital partners for females from the late 1960s through the late 1980s (Cannon and Long 1971; Clayton and Bokemeier 1980; Miller and Moore 1990; Vener and Stewart 1974; Zelnik et al. 1983). This finding is, however, potentially misleading. A close inspection of the results of pertinent studies reveals that most of the increase is explained by a shift from zero to one partner and from one to two partners. There were no increases in the percentage with seven or more partners.
Among males, there is some evidence that adolescent boys of recent decades are less likely to use the services of a prostitute than in the past (Cannon and Long 1971). In a unique study of males attending the same eastern university from the 1940s through the 1970s, Finger (1975) actually reported a decline in the number of PS partners with a corresponding increase in the frequency of sexual relations. This was primarily due to an increase in the percentage of men who had PS only with their girlfriends. Finger also reported a decline in the percentage of males reporting they ever had a homosexual experience. However, among those who had a homosexual experience, the frequency of such encounters had increased.
Although there appears to be consistent evidence that there have been significant increases in the number of PS partners throughout this century, at least for females, it should be stressed that, as late as 1990, the majority of American teens had had zero or one PS partner. Only 4 percent of white females, 6 percent of black females, 11 percent of white males, and 23 percent of black males reported six or more partners (Miller and Moore 1990). Thus, the widely held idea that large percentages of American adolescents are now promiscuous is greatly exaggerated.
Rates of Teen Pregnancy and Birth. In an examination of how the trends we have been reviewing are related to trends in adolescent pregnancy and birth, it is important to bear in mind that, as late as 1965, several states in the U.S.A. prohibited the sale of contraceptives to married couples. Such laws banning the sale of contraceptives to teenagers and/or single persons were common until 1977 (see Section 9A). Details on out-of-wedlock births, contraception, and abortion are presented later. Here, we want to note that the birthrate among unmarried women has been increasing since 1965, with a notable surge in the rate during the 1980s (Baldwin 1980; Forrest and Fordyce 1988; Miller and Moore 1990). Throughout this period, the percentage of unmarried, adolescent women exposed to the risk of pregnancy has been increasing. One principal reason for this is, of course, the increasing percentage of unmarried persons having PS in the U.S.A. (Forrest and Fordyce 1988).
However, there are several interesting twists among these trends, many of which do not fit with the conventional wisdom in the U.S.A. First, much of the increase since 1980 is attributable to women 20 years of age or older. In fact, the adolescent birthrate has actually been declining since the early 1970s (Baldwin 1980; Forrest and Fordyce 1988). Second, the overall birthrate for adolescent women increased through the late 1940s and 1950s, remained stable in the 1960s, increased in the early 1970s, and has been declining since (Baldwin 1980). The misperception, widespread through the U.S.A., that teen-pregnancy rates have been rising is largely due to two factors: (1) the increasing number of such pregnancies, but not the rate, when the children of the baby-boomer generation began having children, and (2) the fact that, as the average age at first marriage has been increasing, adolescent pregnancies are more likely to occur with unmarried women (Baldwin 1980; Miller and Moore 1990). Finally, the perception that adolescent pregnancy has become a recent social problem has emerged as the out-of-wedlock birthrate has increased more dramatically among white women in the last two decades (Baldwin 1980; Miller and Moore 1990).
Contraceptive Use. To most Americans, an increase in the rate of adolescent pregnancy (widely assumed, though not true) would seem to be an inevitable result of increases in PS activity. However, research in many European countries demonstrates that high rates of adolescent sexual activity can be associated with low rates of adolescent pregnancy, when contraceptives are used widely, consistently, and effectively (Jones et al. 1985). There seems little doubt that the U.S.A. has one of the highest adolescent-pregnancy rates among developed nations, largely because of inconsistent contraceptive use (Forrest and Fordyce 1988; Miller and Moore 1990).
It appears that roughly one half of adolescent women use no contraceptive during their first intercourse (Miller and Moore 1990), and most of the women reporting the use of some contraceptive during their first intercourse note that their partner used a condom (Weis 1983). Moreover, most adolescent girls who seek contraceptive services have been having sexual intercourse for some time, many for more than a year before they seek services (Miller and Moore 1990; Settlage, Baroff, and Cooper 1973). After this delay, it appears that roughly two thirds of American teenagers now use some form of contraceptive (Miller and Moore 1990).
Although these figures certainly indicate that large numbers of American youths continue to experience sexual intercourse with no contraceptive protection, they nonetheless represent an increase in contraceptive use over the last several decades. Research in the early 1970s indicated that two thirds to three quarters of American teens rarely or never used contraceptives (Sorensen 1973; Zelnik et al. 1981). Forrest and Fordyce (1988) report that overall use of medically sound contraceptives remained stable through the 1980s. Of those women age 20 or less who sought family-planning services in 1980, nearly three quarters used the pill. By 1990, this had dropped to 52 percent. In 1980, 14 percent had used no contraceptive at all (Eckard 1982).
By 1990, Peterson (1995) reported that 31.5 percent of 15- to 19-year-old women consistently used some form of contraceptive; 24.3 percent of 15-to 17-year-olds did so, as did 41.2 percent of 18- and 19-year-olds. This behavior appears to be unrelated to social class (Settlage et al. 1973). Among women of childbearing age (15 to 44), Peterson (1995) found that 52.2 percent of Hispanic, 60.5 percent of white non-Hispanic, and 58.7 percent of black non-Hispanic women reported using some form of contraceptive (see Table 6 in Section 9A under current contraceptive behavior).
Despite the popularity of the idea that adolescent pregnancy is a result of poor sexual knowledge, knowledge of one's sexuality or birth control has not been shown to be a strong predictor of contraceptive behavior among teenagers (Byrne and Fisher 1983). No relationship was found between contraceptive use and early sex education by family, or a congruence between attitudes and behavior. Reiss, Banwart, and Foreman (1975), however, reported that contraceptive use among teenagers is correlated with endorsement of sexual choice (permissiveness), self-confidence about desirability, and involvement in an intimate relationship.
Explanations of Adolescent Sexuality
Of course, researchers are not content to provide descriptions of social trends. Instead, they seek to provide theoretically useful explanations of the factors underlying those trends. The essence of scientific analysis is the identification and testing of potential correlates of those trends. There have been thousands of studies of adolescent sexuality testing possible correlates. We cannot review them all here. We will, however, briefly identify several different approaches that have been used to explain the trends we have described above. We have tried to select perspectives that have enjoyed some popularity among sexuality professionals at some point. We have also tried to include explanatory models that represent the diversity of professional opinions about adolescent sexuality.
Changes in Social Institutions. By far, the most common approach to explaining the growing acceptance of PS within American culture and the increasing tendency of adolescents to have PS has been a sociological perspective that locates these trends as part of a series of social changes occurring in response to industrialization and urbanization. (Much of this explanation was presented in Section 1, where we reviewed the sexual history of the U.S.A.) As patterns of residence and community relations changed in the late nineteenth and early twentieth centuries, changes began to occur in most social institutions. These included changes in male-female roles, a lengthening of the period of formal education, and the emergence of new forms of heterosexual courtship (Ehrmann 1964; Reiss 1967, 1976). One example of the complex web of social changes that have occurred in the last century is the increasing average age of first marriage (Surra 1990). In one century, the average age at first marriage has shifted from the late teens to the mid-20s. Combined with the earlier age at which American adolescents reach puberty, this has led to a much longer period between physical maturation and marriage, thus, greatly expanding the probability that sexual activity will occur prior to marriage.
As social institutions changed in response to the growing industrial character of American society and the increasingly urban pattern of residence, new forms of adolescent courtship emerged. The custom of dating appeared in the 1920s following World War I, and the practice of going steady emerged in the 1940s following World War II (Reiss 1980). By the 1990s, the practice of going together has become so universally common that few American young people can conceive of other courtship forms. Dating provided a forum for adolescents to pursue male-female relationships independent of adult supervision and control. The appearance of modern transportation, such as the automobile, and the development of urban recreational businesses allowed adolescents to interact with each other away from home. Increasingly, decisions about appropriate sexual behavior were made by adolescents themselves. The practice of going steady placed adolescents into a relationship with many of the features of marriage. Steady relationships were defined as monogamous and exclusive with respect to sexuality and intimacy. As such, they carried high potential for intimacy, commitment, and feelings of love. Together, the increased independence and greater potential for intimacy led to increased rates of PS behavior (D'Emilio and Freedman 1988; Kinsey et al. 1948, 1953; Seidman 1991). There is evidence that this general pattern has occurred in other countries as a consequence of industrialization as well (Jones et al. 1985).
Reiss (1960, 1967) developed the Autonomy Theory of Premarital Permissiveness, mentioned earlier, to explain the association between social institutions and premarital sexual permissiveness. Essentially, Reiss maintained that, as adolescent courtship institutions (dating and going steady) become independent of adult institutions of social control (parental supervision, the schools, and the church), the level of premarital permissiveness in a culture increases. There has been considerable research testing the specific propositions of the theory since Reiss proposed it (Cannon and Long 1971; Clayton and Bokemeier 1980; Miller and Moore 1990). Generally, research from this perspective has tended to presume that PS has become normative within American culture.
Sources of Sexual Information and Sexual Knowledge. Several other explanations of PS behavior have been more likely to view it as a social problem and more likely to focus on the individual character of PS attitudes and behavior. One of the more popular and enduring ideas within American culture about adolescent sexual activity is the belief that sexual behavior and pregnancy risk are influenced by knowledge about sexuality and its consequences. In fact, advocates of sex education in the schools have argued for more than a century that American teens typically possess inadequate and inaccurate sexual knowledge. Some have maintained that sex education could solve such social problems as out-of-wedlock pregnancy and sexually transmitted disease by providing thorough and accurate information about sexuality. Embedded in these assertions is an underlying presumption that sexual decision making and behavior are primarily cognitive processes. Operating from this perspective, there have been dozens of studies of the sources of sexual information for children and adolescents in the U.S.A. Generally, these studies have found that young people in the U.S.A. are more likely to receive sexual information from their peers or the mass media than from adult sources, such as parents or the school (Spanier 1975; Wilson 1994). These studies have been used to conclude that peers are a poor source of sexual information, and that such inaccurate information leads directly to unwanted pregnancies and disease. We should note here that few studies of sexual information have sought to demonstrate a correlation between source of information and sexual decisions or outcomes. That connection has typically been assumed. (See also Section 3, which deals with formal and informal sources of sexual knowledge and education.)
However, in a national probability study of American college students, Spanier (1975; 1978) found no differences in premarital sexual behavior between those students who had ever had a sex-education course and those who had not - regardless of who taught the course, when it was offered, or what material was included. Moreover, a number of studies have found a weak correlation between sexual knowledge and sexual behavior or contraceptive use (Byrne and Fisher 1983). More generally, researchers have consistently found a low correlation between knowledge level and a variety of health-related behaviors, such as smoking, drug use, and eating patterns (Kirby 1985).
Cognitive Development. A somewhat similar focus on cognitive processes has been the basis for an argument that adolescents typically lack a sufficient level of cognitive development required for effective sexual decisions. A number of authors have argued that adolescence is characterized by a cognitive level that is inconsistent with sound sexual decision-making and contraceptive use (Cobliner 1974; Cvetkovich, Grote, Bjorseth, and Sarkissian 1975). Within this perspective, it has become common to describe adolescents as having an unreal sense of infallibility that leads them to underestimate the actual risks of sexual experience (Miller and Moore 1990).
Although references to the works of Jean Piaget have been common in this realm, actual empirical tests of a correlation between Piaget's stages of cognitive development and sexual decisions remain to be conducted. Moreover, this explanation has failed to incorporate the cross-cultural evidence that adolescents in many other nations establish high rates of sexual frequency, maintain consistent contraceptive use, and experience low rates of adolescent pregnancy (Jones et al. 1985).
Interaction of Hormonal and Social Determinants. Udry (1990) has attempted to examine how pubertal development, hormones, and social processes may interact to affect the sexual behavior of adolescents. Hormonal studies seem to indicate that androgenic hormones at puberty directly contribute to explaining sexual motivation and noncoital sexual behaviors in Caucasian male and female adolescents (Udry and Billy 1987; Udry et al. 1985, 1986). Because of the differing social encouragement versus constraints for young white males and females, initiation of coitus seems to be strongly hormone dependent for males, whereas for females it seems to be strongly influenced by a wide variety of social sources with no identifiable hormone predictors. The interaction of hormonal and social determinants is unclear for African-American youth and does not fit the models for white youth that emphasize the importance of sociocultural context on sexual behavior.
Delinquency Models. Perhaps the zenith of models which regard adolescent sexuality as a social problem is the emergence of frameworks that explicitly define adolescent sexual behavior as a form of juvenile delinquency (Jessor and Jessor 1977; Miller and Moore 1990). Vener and Stewart (1974) reported that sexual behavior by 15- and 16-year-olds was correlated with the use of cigarettes, alcohol, and illicit drugs, and with less approval for traditional institutions like the police, the school, and religion.
In a subsequent study using this perspective, Jessor and Jessor (1977) conceptualized sexual behavior as a problem behavior if it occurred prior to age-appropriate norms. In other words, intercourse was characterized as deviant and delinquent if it occurred prior to the mean age (roughly 17 years of age at the time of the study). Jessor and Jessor found that such early sexual behavior was correlated with other problem behaviors such as alcohol use, illicit-drug consumption, and political protest. They concluded that these associations demonstrated that adolescents tend to exhibit multiple forms of delinquency.
By the 1990s, Miller and Moore (1990) reported that a number of studies have found that early sexual behavior is associated with a variety of criminal behaviors such as those described above. Some authors have overlooked the fact that these studies have found this association with delinquent behaviors only for early sexual behavior and have tended to characterize all adolescent sexual behavior as delinquent. These studies do suggest the possibility that developmental issues may be relevant to these findings.
Sexual Affect. A different approach has been taken by a group of researchers interested in examining the role of affective reactions to sexual stimulation, both as a factor that may influence sexual decisions and behavior and as an outcome of sexual experience. Sorensen (1973) reported that 71 percent of teenagers agreed with the view that using the birth-control pill indicates that a girl is planning to have sex. This has been offered as evidence that adolescents are unwilling or unable to accept responsibility for contraceptive use, and thus lack cognitive development. However, affective theorists would argue that it is just as likely that sexual guilt, fear, or embarrassment prevent such a decision.
In the early 1960s, Christensen (1962) conceptualized sexual guilt as a variable response to sexual experience. He found that adolescents are more likely to report experiencing guilt in cultures with restrictive PS norms. He called this a value-behavior discrepancy. Schwartz (1973) found that persons with high sex guilt retain less information in a birth-control lecture, especially when aroused by a sexually stimulating condition. In the Schwartz study, females retained more information than males across all conditions.
Donn Byrne and his associates have maintained that individuals can be placed on a continuum ranging from erotophilic, reacting to sexual stimuli with strongly positive emotions, to erotophobic, reacting to sexual stimuli with strongly negative emotions. Erotophobic persons have been shown to be less likely to seek contraceptive information, to have lower levels of contraceptive knowledge, and to be less likely to purchase contraceptives or use those contraceptive methods that require them to touch themselves (Byrne and Fisher 1983; Goldfarb, Gerrard, Gibbons, and Plante 1988). However, they are no less likely to retain information about contraceptives, even though they become more sexually aroused by a lecture (Goldfarb et al. 1988).
There is a need for much future research on the association between adolescent sexuality and affective variables. However, the studies just mentioned suggest that affective variables may prove to be a fruitful way of explaining adolescent sexual behavior and its consequences. This approach seems particularly suited to examining the variety of ways that adolescents behave and the diverse consequences of such behavior.
Reference Group. Yet another approach to explaining adolescent sexuality has been the attempt to identify persons or groups who have influenced teenagers. Perhaps the most developed theoretical perspective of this type is known as Reference Group Theory. There is some evidence that, as adolescents progress from age 12 to 16, they shift their primary reference-group identification from their parents to their peers. Peer orientation has been shown to be related to sexual intercourse. Moreover, association with peers who are seen as approving PS is correlated with PS permissiveness and PS behavior (Cannon and Long 1971; Clayton and Bokemeier 1980; Floyd and South 1972; Reiss 1967; Teevan 1972). Similarly, Fisher (1986) found that the correlation between the attitudes of teenagers and their parents decreased as adolescence progressed. However, females who cited their mothers as their major source of sexual information were less likely to engage in intercourse and more likely to use contraceptives when they did.
These results should not be interpreted to mean that parents or families do not or cannot exert influence on the sexuality of adolescents. There have been relatively few scientific studies of the influence of differing parental styles and the PS behavior of children. One study (Miller, McCoy, Olson, and Wallace 1986) found that adolescents were least likely to have PS or to approve of PS when their parents were moderately strict. Teenagers who described their parents as very strict or not at all strict were more likely to have had PS. This correlation also held when parents were asked to describe the rules they set for their children. There is some evidence that the age of a mother's first intercourse is related to the age of her daughter's first intercourse (Miller and Moore 1990). Miller and Moore (1990) also showed that girls from single-parent families tend to have sex at younger ages.
Thus, there appears to be two conflicting sets of empirical findings. One set of studies finds evidence that adolescent sexuality is most strongly related to peer influences, especially as age increases. Another set of studies provides evidence that families and parents can exert influence in various ways. Obviously, important questions remain to be resolved.
Rehearsal. A more direct perspective views adolescent sexuality as a developmental process, in which intercourse is seen as the culmination of a sequence of progressively sexual behaviors (Miller and Moore 1990; Simon et al. 1972; Weis 1983). Adolescents appear to move through a series of stages, from kissing to petting of the female's breasts to genital petting to intercourse. There is evidence that, among white adolescents, this pattern is strongly consistent. White adolescents appear to take an average of two years to move through this sequence (Miller and Moore 1990; Weis 1983). In contrast, blacks appear to move through the stages more quickly, and there is greater variability in the actual sequence of behaviors (Miller and Moore 1990). Within this perspective, each subsequent sexual behavior can be viewed as a rehearsal for the next behavior in the sequence.
Not only is there evidence that adolescent sexual experience is acquired in a process that produces an escalating and expanding repertoire of sexual behaviors, but dating and going steady appear to serve as the key social contexts in which this process occurs (Clayton and Bokemeier 1980; Reiss 1967; Spanier 1975). The age of onset of dating and the frequency of dating appear to be major factors in the emergence of sexual behavior (Spanier 1975). In fact, adolescent experiences with intimate relationships (dating and going steady) and the sequencing of sexual behaviors have been shown to be more influential in predicting PS intercourse than general social background variables, parental conservatism or liberalism, or religiosity (Herold and Goodwin 1981; Spanier 1975).
As dating frequency and noncoital experiences increase, exposure to eroticism, sexual knowledge, and interest in sex are all likely to increase concomitantly. Male behavior appears to be more strongly related to the sequencing of behaviors. In contrast, female behavior seems to be more a result of involvement in affectionate relationships. Increased dating interaction and frequency increase sexual intimacy, since opportunities and desire increase. This process is likely to overshadow the influence of prior religious, parental, or peer influences. Thus, adolescent courtship provides the context for the general process of sexual interaction. As Reiss (1967, 1980) has noted, such adolescent courtship also serves as a rehearsal experience for adult patterns of intimate involvement. It is also possible that such adolescent rehearsal experiences are a more powerful and direct explanation of adolescent sexual behavior (Spanier 1975; Weis 1983).
Multivariate Causal Models. An important trend in American research on adolescent sexuality has been the growing recognition that several of the factors reviewed here will eventually need to be included in a sound theory of adolescent sexual development and expression. Reiss (1967) was one of the first to test competing hypotheses in an attempt to identify the strongest predictors of PS permissiveness. Since then, a number of researchers have used multivariate techniques to examine the relative strength of PS correlates (Byrne and Fisher 1983; Christopher and Cate 1988; DeLamater and MacCorquodale 1979; Herold and Goodwin 1981; Reiss et al. 1975; Udry 1990; Udry, Tolbert, and Morris 1986; Weis 1983).
A few examples should illustrate the potential usefulness of this multivariate approach. Herold and Goodwin (1981) found that the best predictors of the transition from virginity to nonvirginity for females were perceived peer experience with PS, involvement in a steady, committed relationship, and religiosity. In contrast, parental education, grade-point average, sex education, and dating frequency failed to enter the multivariate equation.
Udry and his associates (1990; Udry et al. 1986) have investigated the relative influence of hormonal and social variables in explaining adolescent sexual behavior. Several studies demonstrate that androgenic hormones present at puberty directly contribute to the sexual motivation and precoital sexual behavior of white males. For white males, the initiation of coitus seems to be strongly related to androgen levels. Female initiation of coitus seems, on the other hand, to be strongly related to a series of social variables, but not to any hormonal predictors. Udry has argued that these results reflect the differing social encouragement versus constraints placed on males and females respectively. Interestingly, the behavior of African-American youth does not appear to fit with these same explanations, so that the exact interaction between social factors and hormonal variables remains unclear.
Adolescent Sexual Relationships: The Neglected Research
Before moving to the issue of adult heterosexuality, we wish to make a few comments about the nature of intimacy in adolescent sexual relationships and the process of relationship formation. Most of the research on adolescent sexuality reviewed here has tended to focus on the specifically and explicitly sexual elements of such experiences and to ignore the broader relational aspects. In one sense, this is understandable, given the fact that Americans have generally viewed adolescent sexuality, especially its premarital forms, as a social problem. Consistent with this perspective, Americans have tended to deny the possibility that any genuine intimacy occurs in sexual experiences involving adolescents. This is unfortunate in at least two respects. First, it tends to ignore the fact that most adolescent sexual encounters in the U.S.A. occur within the context of what the participants define as a meaningful, intimate relationship. It also ignores the reality that sexual expression within loving, intimate relationships (rather than marital status) has become the dominant attitudinal standard for Americans of all ages. Second, the tendency to ignore the relational character of adolescent sexuality means that researchers have tended to overlook the reality that patterns of sexual and intimate interactions are largely learned within the context of adolescent experiences, and these are likely to be extended well into adulthood. Thus, the failure to investigate these larger relational questions probably impairs our ability to fully understand adult intimate relationships as well. This is not meant to denigrate other forms of sexual expression or to deny that other forms of expression do occur, both in adolescence and later. Rather, it is to suggest that one strong characteristic of American sexuality is the tendency to associate love and sexuality. Any attempt to understand or explain American sexual expression must acknowledge that it generally occurs within the context of ongoing, intimate relationships. This is as true for adolescents as for adults.
The separation of sexuality and relational concerns is well reflected by the emergence of two independent bodies of research within the American academy. On the one hand, there is a well-established field of research on the formation of adolescent intimate relationships, dating and courtship, and mate selection. This tradition extends back to the 1920s and has largely been explored by family sociologists. Social exchange theory has become the dominant perspective in this tradition in recent decades. Surra (1990) provides an excellent review of such research through the 1980s. However, this tradition has largely failed to consider sexuality as an issue in courtship and mate selection, although it ought to be apparent that sexual dynamics and processes are key components of adolescent attraction, dating, courtship, and mate selection. Sexuality carries the potential both for increasing intimacy between teenagers or young adults and for creating intense relationship conflict and, possibly, termination. Yet, Surra's (1990) review is notable precisely for the fact that there is not one single citation of a study including sexuality variables. This is not an indictment of Surra per se. Her goal was to review the field of mate selection as it stood at the beginning of the 1990s. Her assessment serves to document that researchers in this area continue to ignore the role of sexuality in adolescent relationship processes after seven decades of empirical research.
This tendency to ignore sexuality within the courtship process is unfortunate, because of the growing evidence that one of the major influences on PS behavior is the intimate relationship in which most adolescent sexual activity occurs. Being involved in a loving and caring relationship increases the probability of a decision to engage in intercourse (Christopher and Cate 1985) and contributes to sustained activity once it begins (DeLamater and MacCorquodale 1979; Peplau, Rubin, and Hill 1977). In fact, most adolescent sexual experiences in the U.S.A., especially for females, occur within the context of an ongoing intimate relationship. It does appear, however, that as the general rates of PS have increased and as the average age of first intercourse have declined throughout this century, intercourse has tended to occur at earlier stages in a relationship (Bell and Chaskes 1970; Christensen and Carpenter 1962; Christensen and Gregg 1970). With respect to attitudes, Americans are more likely to approve of PS in the context of a relationship. This permissiveness-with-affection-and/or-commitment standard has increasingly become the norm for both adults and young people (Christensen and Carpenter 1962; Christensen and Gregg 1970; Reiss 1960, 1967).
A second body of research examining the formation of sexual relationships has begun to emerge in recent decades. Much of this work has been done by biologists or evolutionary social psychologists and extends a model of mammalian mating first presented by Beach (1976). We discuss it here because it also reflects the separation of the sexual and intimate domains of relationships, and because much of the pertinent human research has been done with samples of college students. Essentially, this body of work forms the foundation for what might be called female selection theory.
The traditional view had always been that males are the aggressors and initiators of sexual involvement. From this perspective, females were seen as sexual gatekeepers. Their role supposedly was to regulate male access by accepting or rejecting male advances (Perper 1985; Perper and Weis 1987). Beginning with Beach (1976), a growing number of researchers have provided evidence that this traditional view is highly flawed. Instead, females select desirable partners and initiate sexual interaction by proceptively signaling selected males (Fisher 1992; Givens 1978; Moore 1985; Moore and Butler 1989; Perper 1985; Perper and Weis 1987). Males, in turn, respond to these proceptive signals. Moore (1985; Moore and Butler 1989) has demonstrated that, not only do women use such signaling, but that men are more likely to approach women who do. Perper (1985; Perper and Weis 1987) has provided evidence that American women employ a variety of complex strategies to arouse male interest and response. Finally, Jesser (1978) has provided some evidence that males are just as likely to accept direct initiations from women as they are to respond to more covert strategies, although females tend to believe that men are turned off by female sexual assertiveness.
This new line of research raises fundamental questions about the roles of males and females in the formation and maintenance of sexual relationships - for both adolescents and adults. It indicates a need for research that is focused on the dynamics within and the processes of sexual relationships themselves. As just one example, Christopher and Cate (1988) found that, early in a relationship, the level of conflict was positively related to a greater likelihood of intercourse. As the relationship progressed, love and relationship satisfaction eventually became significant predictors of sexual involvement. In the case of adolescence, we need to move beyond social bookkeeping, counting the number of American teenagers who have PS, to examine what actually happens in their relationships with each other.
C. Adult Heterosexuality
DAVID L. WEISThe National Health and Social Life Survey
Strangely, there has been considerably more research on the sexual conduct of American adolescents than of adults, and much of the existing research on adults has tended to focus on sexual problems such as extramarital sex (ES) and sexual dysfunction (see Section 12 on sex dysfunctions and therapies). There has been little research on the patterns of sexual interactions within nonclinical marital relationships. This is striking, precisely because of the fact that marriage is the most widely accepted setting for sexual relations in the U.S.A. and because more than 90 percent of Americans do marry. Taken together, the preponderance of research on adolescent sexuality, ES, and dysfunction indicates the tendency of American sexuality professionals to focus on sexual behaviors that have been defined as social problems, rather than on normal sexuality.
In October 1994, a national survey of adult sexual practices was released with great media fanfare (Laumann, Gagnon, Michael, and Michaels 1994). The survey, titled the National Health and Social Life Survey (NHSLS), randomly sampled 3,432 persons, aged 18 to 50. It was touted as the most comprehensive American sex survey ever, and the first national study of adult sexuality. However, Reiss (1995) has noted that this claim is misleading, as there have been more than a dozen national surveys of a more-limited scope. Given our interest in reviewing the nature of American sexuality research, it is interesting to note that the survey was originally planned and approved as a government-sponsored project. Funding was denied for this project and a similar study of teens (the Udry study) when conservatives in the U.S. Congress objected to the studies. Conservatives argued that the government should not use taxpayer money to study private matters like oral sex - clearly rejecting the significance of the health concerns involved. The researchers found private funding instead. Also interesting is the fact that conservatives hailed the findings when the study was released (Peterson 1994).
There is little doubt that the NHSLS is the most comprehensive study of adult sexuality to date, with literally hundreds of variables assessed. Among the key findings are the following:
· Most Americans report that they are satisfied with their sex life - even those who rarely have sex. Among married persons, 87 percent reported they were satisfied with their sex life.The NHSLS has sparked considerable controversy among sexuality professionals. Questions have been raised, primarily about the legitimacy of the prevalence estimates for such behaviors as number of sexual partners, homosexual experience, and ES. In general, the NHSLS estimates tend to be lower than those found in most prior sex research - including prior national studies (Billy, Tanfer, Grady, and Klepinger 1993). It should be noted that the NHSLS estimates are remarkably similar to Findings in a series of studies conducted by the National Opinion Research Center using similar national probability samples (Davis and Smith 1994; Greeley et al., 1990; Smith 1990, 1991). These national samples have been carefully constructed to be representative of gender, age, race, education, marital status, size of city of residence, and religion in the U.S.A. The NHSLS did obtain a 79 percent response rate, probably because participants were financially reimbursed. Few prior studies have had comparable response rates, and few have reimbursed participants. Questions about how this impacted the results are a legitimate matter for future research.
· For the entire sample, 30 percent of men and 26 percent of women have sex two or three times a week; 36 percent of men and 37 percent of women have sex a few times a month; and 27 percent of men and 30 percent of women have sex a few times a year. Married persons have sex more often than single people, and persons who are cohabiting have sex more often than marrieds.
· Approximately 80 percent of married women and 65 percent of married men have never had ES. The majority of those who are cohabiting also have never cheated. The group most likely to have extradyadic sex is unmarried men, aged 42 to 51, who have lived with a woman for three years or less (32 percent).
· There has been a slight increase in the number of lifetime sexual partners, largely because people now have intercourse earlier, marry later, and are more likely to get divorced.
· Among marrieds, 94 percent had sex only with their spouse in the last year; 75 percent of cohabiting persons had sex only with their partner in the last year. About 80 percent of American adults have had either one or no sexual partners in the last year. Only 3 percent have had five or more partners in the last year. About 50 percent of men and 30 percent of women have had 5 or more partners since age 18.
· Most Americans have a fairly limited sexual menu of activities. Roughly 80 percent of both men and women reported that sexual intercourse is very appealing; only 50 percent of men and 33 percent of women find receiving oral sex appealing; 37 percent of men and 19 percent of women describe giving oral sex as appealing. About 25 percent of both men and women have tried anal sex at least once.
· People who already have an active sex life with a current sexual partner are more likely to masturbate. Among married people, 57 percent of husbands and 37 percent of wives have masturbated in the last year.
· About 2.8 percent of men and 1.4 percent of women identified themselves as homosexual or bisexual. Only 9 percent of men and 4 percent of women reported ever having a homosexual experience. These rates are considerably higher in the twelve largest U.S. cities.
· Most heterosexuals are not at risk of contracting AIDS, because they are not part of social networks with high risk.
In a review of the NHSLS, Reiss (1995) credits the study for its comprehensiveness, the richness of the data generated, the theoretical nature of the investigation, and the high quality of the sampling techniques. However, he also raises several questions that may influence the validity of the findings. Here, we will focus on a few of the more serious. One concerns the fact that 21 percent of the respondents were interviewed with someone else present during the interview. As Reiss notes, a person with an intimate partner or a family member present may well have answered questions differently for obvious reasons. For example, only 5 percent of persons interviewed with another person present reported that they had two or more sexual partners in the last year. In contrast, 17 percent of those interviewed with no one else present reported two or more partners in the last year. This is a sizable difference, and it raises questions about the validity of responses to many questions in the survey. Similarly, the NHSLS asked respondents to report the number of sexual partners they have had since age 18. Most previous studies asked respondents to report their lifetime number of sexual partners. Here, one half of the sample did have sexual relations prior to age 18. This reduced estimates for lifetime number of partners. The NHSLS reported a median number of six sexual partners for men and two for women. Reiss notes that these estimates are lower than comparable studies (Billy et al., 1993), and that this reported gender difference cannot possibly be true in the real world.
To this critique, we can add that it is possible that prevalence estimates have been inflated by the volunteer bias of most sex research. There are unexamined questions about the effects of volunteer bias and response rates. Paul Gebhard (1993), a member of the original Kinsey research team, has argued that estimates of lifetime prevalence rates for homosexual behavior have been remarkably similar when adjusted for sampling weaknesses. Gebhard also criticized the NORC and NHSLS studies for failing to use trained sex researchers to conduct their interviews, and for their own sampling flaws that overrepresented rural populations. In fairness, it is appropriate to note that several of the volunteer samples overrepresent urban populations, and there is evidence that urban-rural differences in sexual attitudes remain substantial (Weis and Jurich 1985). Finally, although there is a general consensus that persons who agree to participate in sex research are more permissive and more sexually experienced, two recent studies strongly suggest that persons who decline to answer particular items in a sex survey are attempting to hide behavior in which they have engaged (Wiederman 1993; Wiederman, Weis, and Allgeier 1994).
Although these questions will require considerable future research to resolve, it should be acknowledged that the NHSLS is a major contribution to the field of sex research in the U.S.A. It is a landmark study with important new information about the sexual practices of the vast and diverse American adult population, and it will set the parameters for questions yet to be explored. Finally, it provides important data on each of the topics we will explore further in this section.
Sexuality and Single Adults
Practically every American spends at least a portion of his or her adult life unmarried. At any one point in time, more than 20 percent of the U.S. population is single, and this percentage has been increasing for several decades (Francoeur 1991; Shostak 1987). The chief reasons for this are the greater tendency to postpone marriage (median age is now in the late 20s), the increasing divorce rate (5 per 1,000 by the 1980s and fairly stable thereafter), and the increasing rate of cohabitation (which tripled since 1960), both as an alternative to marriage and as a form of courtship prior to marriage (Glick 1984; Norton and Moorman 1987; Shostak 1987). Glick (1984) has speculated that the prolongation of formal education, the increasing acceptability of premarital sexuality, the growing independence of women, and the earlier mortality of males may also be factors promoting the growth of singlehood.
Actually, the single adult population contains three groups who may share little in common: Those who have never married, those who have divorced, and those who are widowed. Persons within each group may or may not have chosen to be single, and they may or may not intend to remain single. Also, persons in each group may be living alone, may be living with roommates who are not intimate or sexual partners, or may be cohabiting with an intimate partner. By 1980, it was estimated that close to 2 percent of the adult U.S. population was cohabiting (Glick and Norton 1977; Yllo 1978). Of course, some single persons are gay or lesbian, although they are not typically included in estimates of cohabitation, even when they live with their partners.
It should be stressed that the population of single adults is a fluid one. The U.S.A. has high rates of marriage, divorce, and remarriage (Glick 1984; Norton and Moorman 1987). Most of those who are classified as having never married at any one point will eventually marry. This is especially true for the growing group who have remained unmarried well past the age of 20. Approximately three quarters of women who get divorced, and more men, eventually remarry (Glick 1984; Norton and Moorman 1987). Thus, the composition of the single population is always shifting as some marry and others divorce or are widowed. We are not aware of any research examining the impact of this shifting character on the sexual lifestyles of single persons. Some singles become involved in intimate relationships that lead to cohabitation or marriage, although we know little about whether these processes are similar to adolescent courtship. For those singles who are not involved in an ongoing intimate relationship, it is possible that finding sexual partners can be problematic.
It is popularly believed that being single in adulthood has become more acceptable in the U.S.A. today. There is, however, some evidence that married couples continue to associate primarily with other couples. Certainly, it is more acceptable to be sexually active while single today. Singles have greater social and sexual freedom than ever before to pursue a variety of lifestyles. In fact, the labeling of a category of single adults may serve to obscure the fact that the range of sexual and intimate lifestyle options is just as wide as for married persons.
Despite the large number of single adults in the U.S.A., there has been virtually no research on the sexual practices or attitudes of these groups. The NHSLS (Laumann et al. 1994) did distinguish between single and cohabiting respondents, an important distinction. As we discussed earlier, the NHSLS did find that single persons had sex less frequently than married persons, and that cohabiting persons had sex more often than married persons.
The Never Married. We know of no research that has focused on the population of never-married adults who are not cohabiting. Of course, this group does include persons in their early 20s who have yet to marry. A portion of that group is included in many of the studies of premarital sexuality, although that group is not isolated for separate analysis. There is virtually no scientific information on how never-married persons find or meet sexual partners, establish sexual encounters, or maintain sexual relationships.
Divorced (Postmarital Sex). Divorce has increased in the U.S.A. dramatically throughout the twentieth century (Berscheid 1983). The rate has leveled since 1980 (Current Population Reports 1985; Glick 1984; Norton and Moorman 1987; Shostak 1987). Of the roughly 40 percent of the American population that gets divorced, about 70 percent eventually remarry, often within a few years (Glick 1984; Norton and Moorman 1987).
Again, there has been little research on this group. It appears that about 80 percent of women, and nearly all men, remain sexually active following a divorce (Gebhard 1968; Hunt 1974). Most persons have sex with a new partner within the first year following a divorce (Hunt, 1974). In the 1970s, Hunt (1974) reported that divorced women averaged four sexual partners a year, and had a higher frequency of orgasm in their postmarital sex than they had had in their marriage. Men averaged nearly eight partners a year.
Again, there has been little research on the process by which divorced persons form or maintain sexual relationships. However, it is fair to suggest that, as the title of an American novel and corresponding movie implies, most divorced persons find that they must start over. After a period of marriage, they find themselves in the position of dating and courting again. Some have anecdotally reported that they find this anxiety-provoking, whereas others find it exhilarating.
Widowed. This process of starting over may be relevant to those persons who are widowed as well. Our review of the research literature identified only one study of the sexual practices of widowed persons. Nearly three decades ago, Gebhard (1968) reported that widowed persons were less likely to have sexual experiences than divorced persons. Francoeur (1991) has suggested that this may be due, in part, to a sense of loyalty to the former spouse or to perceived and real pressure from kin members.
By far, the most common adult sexual lifestyle in the U.S.A. is legal marriage, and marriage is the context for the overwhelming majority of sexual experiences in the U.S.A. In fact, marriage is the only context in which sexuality is universally approved. Despite this, researchers have investigated marital sexuality less than nonmarital forms of sexual expression. Greenblat (1983) has suggested that sex within marriage is more likely to be the object of jokes than of scientific investigation. Strong and DeVault (1994) report that only nine of 553 articles on sexuality that appeared in scholarly journals between 1987 and 1992 were devoted to marital sexuality.
This pattern of research is somewhat odd in light of the widespread belief that effective sexual functioning is indispensable to a good marriage (Frank and Anderson 1979). In this regard, it is striking that much of the research conducted on couples has utilized clients in sex therapy. Here we review works on nonclinical samples.
Sexual Frequency and Practices. Most of the research on sexual relations within marriage has assessed the frequency of sexual relations. Many of these studies have also examined how that frequency is related to marital satisfaction. Americans seem to be fascinated with comparing their own frequency to other couples. Until recently, this research was based on volunteer samples, which typically were also quite small.
Perhaps the first sex survey ever conducted in the U.S.A. was done by Clelia Duel Mosher (1980), who investigated the sexual practices and attitudes of forty-five women between 1890 and 1920. Most of these women reported that they found sex to be pleasurable and believed that it was necessary for both men and women. The women who were interviewed before 1900 were less likely to describe sex as important or enjoyable, and they were less likely to associate sex with the expression of love. The Mosher survey documents the first signs of a shift to a post-Victorian culture.
In a study of more than 1,000 men and women, Dickinson and Bean (1932) reported that sexual dissatisfaction was more important in explaining marital difficulties than disputes over work, money, and children. Davis (1929) drew similar conclusions in her study of 2,200 women. Sexual satisfaction within marriage had clearly become a norm in the U.S.A. by the early twentieth century. Somewhat later, Hamilton (1948) interviewed a hundred married men and women and concluded that an unsatisfactory sex life is the principal cause of marital dysfunction. Without addressing the validity of that particular claim, the Hamilton data do demonstrate that, in the small sample surveyed in the 1930s and 1940s, sex was considered to be an important part of a marriage.
The Kinsey group (1953) reported that married couples in the 1940s had sex an average of two times a week in the early years of marriage, declining to about once a week after ten years of marriage. By comparing those born before 1900 and those born after 1900, they found that the frequency of marital coitus had remained the same. However, virtually every other aspect of marital sex had changed. Couples born after 1900 engaged in more and longer foreplay, used more coital positions, were more likely to have oral sex, were more likely to use French (deep) kissing and manual caressing of genitals, and had sex more often naked.
More-recent studies have tended to fit two patterns. Small samples with volunteers have found a general average of three to four times a week in early marriage with a decline to twice a week in later years. However, studies with national samples have tended to get lower figures more like Kinsey's (Bell and Bell 1972; Blumstein and Schwartz, 1983; Call, Sprecher, and Schwartz 1995; Hite 1976; 1983; Hunt 1974; Sarrel and Sarrel 1980; Tavris and Sadd 1974; Trussell and Westoff 1980; Udry 1980; Westoff 1974). Interestingly, married women tend to report lower frequencies than married men (Call et al. 1996).
A few researchers have asked respondents to report their ideal or preferred frequency. Hite (1976) found that one third of married women would like to have sex at least daily, another third wanted it two to five times a week, and a final third less often.
(1) Changes Throughout Marriage. The evidence of a decrease over time or length of marriage is strong and consistent (Blumstein and Schwartz 1983; Edwards and Booth 1976; Greeley 1991; Hunt 1974; Kinsey et al. 1953; Michael et al. 1994; Trussell and Westoff 1980; Westoff 1980). Longitudinal studies of the same couples over time have also documented this pattern (James 1981; Udry 1980), as have retrospective studies of couples looking back over the course of their marriage (Greenblat 1983).
In a national study of the 1988 National Survey of Families and Households (Call et al. 1995), frequency decreases over the length of marriage were correlated with biological aging, diminished health, and habituation. In a multivariate analysis, age was most strongly related to frequency, followed by marital happiness, and factors that reduce the opportunity for sex (such as pregnancy and small children). Couples who had not cohabited prior to marriage and who were still in their first marriage had less-frequent sex than cohabitors, married persons who had cohabited prior to marriage, and those who were in their second or later marriage.
These findings are largely consistent with prior research. Decreasing frequency of marital sex has been found to relate to age-related reductions in the biological capacity for sex, including declines in male motivation and physical ability, declines in women's testosterone levels, and increases in illness (Greenblat 1983; Hengeveld 1991; James 1983; Udry, Deven, and Coleman 1982). Negative social attitudes about sex and the elderly may also lead some to believe that their interest and capacity should decline (Masters and Johnson 1970; Riportella-Muller 1989). However, these aging factors do not explain the decline in frequency that occurs within the first several years of marriage (Jasso 1985; Kahn and Udry 1986). James (1981) found that the coital rate dropped by one half during the first year of marriage. Some have suggested that there is a honeymoon effect early in the marriage. As the honeymoon period ends, habituation occurs and frequency declines (Blumstein and Schwartz 1983; Doddridge, Schumm, and Berger 1987). Habituation may be seen as a decreased interest in sex that occurs with the increased accessibility of a regular sexual partner and the routine predictability of behavior with that partner over time (Call et al. 1995).
Other reasons that have been cited as influencing a decrease in frequency include fatigue, work demands, child care, and management of complex schedules (Michael et al. 1994).
(2) Effects of Children. A few comments on the effects of children are worth special note. There is some evidence that sexual frequency declines by the third trimester of pregnancy - prior to the actual birth of a child (Kumar, Brant, and Robson 1981). The birth of a child introduces parental roles into the marital relationship. The child increases fatigue, reduces time alone together for the couple, and decreases time in situations that are conducive to sexual encounters (Blumstein and Schwartz, 1983; Doddridge, et al. 1987; Greenblat 1983).
(3) Association with Sexual and Marital Satisfaction. A majority of Americans report that they are satisfied with their marital sex life (Hunt 1974; Lauman et al. 1994). In general, researchers have not found frequency to be related to sexual or marital satisfaction (Blumstein and Schwartz 1983; Frank, Anderson, and Rubinstein 1978). However, there is evidence that the congruence between ideal and actual frequency is related (Frank and Anderson 1979). There is some evidence that sexual problems are likely to occur fairly early in a marriage (Brayshaw 1962; Murphy et al. 1980).
Some studies have found social factors associated with relationship satisfaction. Rainwater (1964) found, in a study of couples in poverty in four different cultures, that lower-class couples were more likely to have highly gender-segregated role relationships (traditional gender roles); they were less likely to have close sexual relationships, and the wife was not likely to view sex with her husband as gratifying.
Several studies have found that sexual satisfaction is related to both sexual and nonsexual aspects of the marriage. The Kinsey group (1953) found that divorce was related to decreases in the wife's orgasm rate. Hunt (1974) reported a strong correlation between marital closeness and sexual satisfaction. He found that the most important predictor was the extent to which couples share similar sexual desire. Thornton (1977) found that couples who spend more time having sex than they do fighting tend to have happier marriages. Sarrel and Sarrel (1980) found that couples who talk with each other about sex often, who rate their communication about sex as good, where the wife likes oral sex, and where the man believes the women's movement has been good for women tend to have more satisfying sexual relationships.
Hite (1976) asked women to identify what aspect of their marital sex gave them the greatest satisfaction. Responses given by 20 percent or more included closeness, orgasm, coitus, and foreplay. In response to what they liked least, more than 10 percent said oral or anal sex, lack of orgasm, the messiness following sex, excessive or rough foreplay, and the routine nature of their activities.
In the Redbook magazine surveys (Tavris and Sadd 1975; Tavris 1978), marital satisfaction did not decline with length of marriage or age. The majority reported enjoying oral sex. Most respondents believed that good communication is an important ingredient of marital and sexual happiness. The most common complaint was that they had sex too infrequently. For women, religiosity was related to a happier sex life and marital satisfaction.
In an unusual study of a hundred mostly white and well-educated couples who were happily married (selected because none had ever had ES or been in therapy), Frank and Anderson (1979) found that 85 percent described themselves as sexually satisfied. One half of the wives reported they had difficulty becoming aroused or reaching orgasm. Roughly 10 percent of the husbands reported they had experienced erectile difficulties. One third of the couples expressed complaints about such things as anxiety, too little foreplay, and low sexual desire. There was no correlation between sexual dysfunctions and marital satisfaction, but complaints by the wife were associated with reduced marital happiness.
(4) Unexplored Issues. This review of research on marital sexuality serves to confirm the narrow range of the questions researchers have investigated. We know little about the dynamics of sexual relationships in marriage - about the ways couples interact sexually, about how they transact or negotiate sexual encounters, or about how they initiate and terminate encounters. Little is known about how sexuality in marriage is affected by power dynamics between the couple. There has been little study of sexual coercion in marriage. Perhaps it is time to end the focus on counting episodes and begin to examine what happens within marital sexual relationships.
Extramarital Sexual Relationships. Researchers have been studying ES for decades, although the range of the questions they have examined has been fairly narrow (For more-thorough reviews of ES research and nonexclusive lifestyles, see Macklin 1980; Thompson 1983; Weis 1983).
(1) ES Attitudes. One focus of concern has been the degree of normative consensus reflected by ES attitudes. A series of national surveys indicate that ES has consistently been disapproved by 75-85 percent of the adult American population (Glenn and Weaver 1979; Greeley, Michael, and Smith 1990; Reiss, Anderson, and Sponaugle 1980; Weis and Jurich 1985). Weis and Jurich (1985) found that nearly one third of residents in the twelve largest cities found ES acceptable, the only locations in the U.S.A. where as many as 20 percent approved. In small towns and rural areas, fewer than 10 percent approved. The norm of sexual exclusivity within marriage is so widespread in American culture that few question it.
Approval of ES has been found to be related to (1) being male, (2) young age, (3) being nonwhite, (4) living in a large city, (5) high levels of education, (6) low religiosity, and (7) being unmarried (Glenn and Weaver 1979; Reiss et al. 1980; Weis and Jurich 1985; Weis and Slosnerick 1981). Although a number of researchers have reported that approval of ES is related to lower levels of marital happiness, Weis and Jurich (1985) found that marital happiness was less strongly related to ES attitudes than several of these other variables.
(2) ES Incidence/Prevalence. A second major concern of researchers has been the attempt to establish estimates of the prevalence and/or incidence of ES behavior. Generally, this has taken the form of asking respondents to indicate whether or not they have ever had ES. Authors have regularly claimed that roughly one half of married persons in the U.S.A. have had at least one ES experience, citing the Kinsey research (1948, 1953) as the basis for this claim. Although the point is often ignored, the Kinsey team actually found that 33 percent of husbands and 26 percent of wives reported having ES. Because of suspicions of underreporting, they raised the estimate for male - but not female - ES to 50 percent. Several researchers have reported that the figures for husbands have remained fairly stable since then, but that the rate for wives has increased to approximately that of husbands (Blumstein and Schwartz 1983; Hunt 1974; Levin 1975). Researchers have reported lifetime prevalence rates from as low as 20 percent (Johnson 1970) to nearly 75 percent (Hite 1981).
Several recent studies by the National Opinion Research Center (Smith 1990; 1991; Greeley et al. 1990) have found that only 2 to 3 percent of American married men and women have ES each year. Further, they reported that 65 percent of wives and 30 percent of husbands have the same number of lifetime sexual partners as spouses. According to these researchers, the increases in premarital sex and cohabitation, the rising rate of divorce, and the later age at first marriage that have characterized the last forty years have resulted in less sexual exclusivity among the unmarried, but no such trend has occurred among married persons in the U.S.A. The Greeley group concluded that Americans are overwhelmingly monogamous [sic] and that rates of ES have been overestimated by previous researchers. The National Health and Social Life Survey (Laumann et al. 1994), also conducted by the NORC, found that only 35 percent of men and 20 percent of women reported ever having ES, and 94 percent had sex only with their spouse in the last year.
As we have already discussed, making comparisons between the results of the NORC national probability samples and previous studies is most difficult. Most previous studies have reported lifetime prevalence rates. The NORC studies have generally reported annual incidence rates. It seems likely that the conditions surrounding the collection of data and the greater representation of rural respondents in the NORC studies led to low estimates. On the other hand, the volunteer nature of most previous studies and their greater inclusion of urban respondents may well have led to high estimates. For the time being, we must conclude that questions about the incidence and prevalence of ES in the U.S.A. remain largely unanswered.
(3) Marital Happiness. The third major focus of ES research has been the attempt to demonstrate an association between ES behavior and marital happiness/satisfaction. By far, this has been the most frequently tested hypothesis. As a consequence, there has been little research exploring the circumstances or conditions surrounding ES behavior itself or testing alternative hypotheses. A number of researchers have found that ES behavior is significantly related to lower levels of marital happiness (Bell et al. 1975; Edwards and Booth 1976; Glass and Wright 1977; 1985; Prins, Buunk, and Van Yperen 1993; Saunders and Edwards 1984). Lower marital happiness has also been found to be related to ES attitudes (Reiss et al. 1980; Weis and Jurich 1985).
However, the association may not be as strong as these findings imply. The research by Glass and Wright (1977, 1985) suggests that the actual association between ES and marital happiness may be quite complex. In their earlier study, Glass and Wright (1977) found that husbands who had ES in the early years of marriage did have lower marital satisfaction. However, there were no differences in marital satisfaction between husbands who had never had ES and those who began ES later in their marriages. Interestingly, exactly the reverse was true for wives, There were no differences in marital satisfaction between wives who had never had ES and those who began it early in their marriages. Yet, wives who began their ES experiences later in marriage did have significantly lower marital satisfaction. In their later study, Glass and Wright (1985) found that ES was related to lower marital happiness only for wives. They concluded that male ES is likely to be more strongly associated with individual factors, rather than marital issues.
The Glass and Wright studies represent a level of complexity that has rarely been seen in ES research. Few studies have examined the possibility that marital happiness might relate to different types of ES experiences. As just one example, we can take the case of consensual ES. In one of the few comparisons of couples who had made an agreement to include ES in their marriage with couples who did not have this agreement and had a sexually exclusive relationship, there were no significant differences in marital stability, marital happiness, or level of jealousy (Rubin and Adams 1986). Similarly, Gilmartin (1978) found no differences in marital happiness between a group of couples who participated in swinging and a control group of nonswinging couples.
Moreover, Ellis (1969) has made the obvious point, substantiated by all the studies cited here, that some people who have ES also report high marital satisfaction. In fact, although the two variables have been consistently found to be significantly related, the proportion of ES variance explained by marital quality variables has tended to be rather small. This may be due, in part, to the tendency to dichotomize ES into ever versus never categories, thus ignoring the diversity of ES types. This treatment of ES as a simplistic construct that uniformly reflects poor marital dynamics may reduce our ability to establish better explanations of ES. For example, Weis and Jurich (1985) did report that ES attitudes and marital happiness were significantly related in a series of national probability samples, but they also found that marital happiness was more weakly related to ES attitudes than several background variables.
(4) Exploring the Diversity of ES Experience. This failure to recognize the diversity of ES experience may be the single greatest obstacle to the development of sound research and theory. ES experiences are, in fact, a class of relationship types, every bit as complex as other relationship forms. With few exceptions, American researchers have failed to recognize the historical and cross-cultural evidence that male and female ES behavior is universal, despite the strong normative traditions and sanctions against it. They have also largely ignored the cross-cultural evidence that amply demonstrates a wide variety of ES patterns and normative responses to it (Buss 1994; Fisher 1992; Ford and Beach 1951; Frayser 1985; Murdock 1949).
(5) Specific Aspects of ES. Ultimately, a full understanding of ES will require more-thorough investigation of the myriad ways in which ES experiences vary. Several factors require additional research. These include:
· Specific Sexual Behaviors Involved. ES can range from flirting, kissing, and petting to intercourse (Glass and Wright 1985; Hurlbert 1992; Kinsey et al. 1948; 1953).(6) Gender Issues. Before discussing theoretical factors for ES, we want to note that the available evidence strongly suggests that researchers explore the possibility of separate predictive models for men and women. There is evidence that men are more likely to have ES than women and to have more numerous ES encounters (Buss 1994; Glass and Wright 1985), more likely to report ES relationships with limited involvement (Glass and Wright 1985; Spanier and Margolis 1983), and tend to have more partners (Buss 1994; Thompson 1983). Men and women may also experience different outcomes. There is some evidence that women are more likely to report experiencing guilt as a result of ES (Spanier and Margolis 1983). It is possible that women, as a group, are more likely to be motivated to engage in ES activities by marital factors and may be more likely to seek intimacy as a primary goal in ES (Reibstein and Richards 1993). Several studies have found that marital variables are more strongly related to ES for women than for men (Glass and Wright 1985; Saunders and Edwards 1984). All of these findings indicate that the ES experiences of men and women may differ substantially.
· Specific Relationship Behaviors Involved. ES relationships vary from those in which sexual interaction is nearly the sum total of the relationship to those where sexuality is a minimal component (Hurlbert 1992; Richardson 1985; Thompson 1983, 1984).
· Number of ES Partners. In general, the scant evidence available suggests that most Americans have a small number of ES partners (Bell et al. 1975; Greeley et al. 1990; Kinsey et al. 1953; Pietropinto and Simenauer 1977).
· Length of ES Relationship. It appears that most, but certainly not all, ES relationships are of relatively short duration and entail less than ten actual sexual encounters, with some evidence that females tend to be involved for longer periods (Bell et al. 1975; Gagnon 1977; Hall 1987; Hunt 1974; Hurlbert 1992; Kinsey et al. 1953; Pietropinto and Simenauer 1977).
· Level of Involvement. ES ranges from single sexual encounters in which partners know little of each other to highly intimate affairs with characteristics that are quite similar to intimate marriages.
· Consensual Versus Secretive. Although most ES is secretive or clandestine (Gagnon 1977; Hunt 1974), it is important to recognize that some spouses do know about their partner's ES activities and expressly agree to permit ES (see section below on alternatives to traditional marriage) (Blumstein and Schwartz 1983; Thompson 1983; Weis 1983).
· Motives and Meanings. There are dozens of motives for ES. Weis and Slosnerick (1981) demonstrated that a distinction between individual motives (such as adventure, variety, romance, or pleasure) and marital motives (such as revenge against a spouse, marital hostility, marital sex problems, or as an alternative marriage form) was useful in explaining differences in ES attitudes.
· Bisexual/Homosexual. ES has usually been assumed to be heterosexual, but there is evidence that, at least some ES is homosexual (D. Dixon 1985; J. K. Dixon 1984).
(7) Building Theoretical Models. Edwards and Booth (1976) have argued that the context of marital interaction is more important than background factors in explaining the process leading to ES involvement. However, Weis and Slosnerick (1981) have maintained that individuals enter marriage with internalized scripts for sex, love, and marriage. Ultimately, the scripts of married persons stem from an interaction of marital dynamics and background factors. Each of these, in turn, is likely to be influenced by one's position within the social structure.
As just noted, there is evidence of a significant correlation between marital happiness and both dichotomous measures of ES experience and ES attitudes, although this association has not always been a strong or robust one. In a study of ES attitudes (approval), Weis and Slosnerick (1981) isolated two orthogonal factors of justifications for ES. One was a set of motivations for ES that mentioned aspects of the marital relationship. The other was a set of individual motives for ES. Both factors were significantly related to approval of ES, but the individual motivations were more strongly related than the marital motivations.
These findings suggest two possible paths for future research that seeks to elaborate the complex nature of the association between ES and marital satisfaction. One is to contrast the types of ES experiences that persons with individual versus marital motivations tend to have and to explore how these relate to marital satisfaction and, perhaps, to outcomes of ES relationships. The other is to separate happily and unhappily married persons and to investigate the types of ES experiences and outcomes for each group. It seems reasonable to expect that the two groups might well pursue different kinds of ES experiences under different circumstances, with different outcomes.
A second theoretical factor may be background variables. A number of researchers have reported that premarital sexual attitudes and behavior are related to ES attitudes and behavior, several arguing that it is the best predictor of ES involvement (Bukstel et al. 1978; Christensen 1962, 1973; Glenn and Weaver 1979; Medora and Burton 1981; Reiss et al. 1980; Singh et al. 1976; Thompson 1983; Weis and Jurich 1985; Weis and Slosnerick 1981). ES variables have been found to correlate with premarital sexual permissiveness, number of premarital sexual partners, and early premarital sexual experience (low age). Weis and Jurich (1985) found premarital sexual permissiveness was the strongest and most consistent predictor of ES attitudes in a series of regression analyses with national probability samples throughout the 1970s.
Several questions remain to be explored. Do these Findings suggest that there is something particular about premarital sexual interactions with partners that is associated with ES, or are measures of premarital sex merely indicative of a broader interest in and history of sexual pleasure in various forms? Which of these will prove to be more useful in explaining various types of ES activities? For example, Joan Dixon (1984) found that female swingers tend to have early and continuing histories of heterosexual involvement, but that they also tend to have early and continuing histories of masturbation and high current sexual frequencies with partners. Gilmartin (1978) also found that swingers tend to have early heterosexual experiences and high sexual frequencies with their spouses. One might conceivably argue that such persons like sex, and ES is an extension of a broader orientation to pleasure.
A third factor has been suggested by Cazenave (1979), who has criticized work in the area of alternative lifestyles for its emphasis on ideological preference and its failure to explore how structural variables (such as age, gender, and race) may impose external constraints. In fact, there is evidence that ES behavior and ES permissiveness (attitudes) are related to (1) young age, (2) being nonwhite, (3) low education for behavior and high education for attitudes, (4) low religiosity, and (5) residence in a large city (Fisher 1992; Greeley et al. 1990; Smith 1990, 1991). Several of these associations may, in fact, be quite complex. For example, the Kinsey group (1948, 1953) found that blue-collar males tend to have ES in their 20s and their behavior diminishes by their 40s. White-collar males with college educations tended to have little ES in their 20s. This rate gradually increased to an average of once a week by age 50. In contrast, female ES peaked in the late 30s and early 40s. Finally, there is a need for research that explores the role of such American social trends as the increasing age at first marriage, the growing divorce rate, the unbalanced gender ratio, and greater mobility and travel in ES behavior.
(8) Unexplored Issues. There has been little research to this point on the process of ES relationships. For example, there has been little investigation of how opportunities for ES involvement occur in a culture with strong prohibitions against ES. Cross-sex friendships and interactions have been frequently cited as creating the opportunity for ES (Johnson 1970; Saunders and Edwards 1984; Weis and Slosnerick 1981), although this has not been empirically tested. The matter is somewhat complicated by the evidence that friendships outside of marriage are associated with higher levels of marital satisfaction (Weis and Slosnerick 1981). Wellman (1985, 1992) has documented how the friendship networks of men have shifted from public spaces (bars, cafés, and clubs) to private homes. This has led to a narrowing of the concept of friendship to emotional support and companionship. Husbands' and wives' networks are now both based in private, domestic space, and many wives actively maintain their husbands' ties to friends and kin. Men get much of their emotional support from women, as well as men, and women get almost all of their support from women. Wellman argues that marriage may impose constraints on men's ability to spend time and be intimate with other men or women. Whether this is related to ES remains to be explored.
Similarly, little is known about the outcomes of ES involvement. Generally, it is assumed that ES relationships are short in duration, exploitive in character, and tragic in outcome. For example, it is generally assumed that ES and cross-sex friendships will be a source of jealousy in a marriage. Although there is a growing body of evidence about jealousy, little research has specifically investigated jealousy in the context of ES (Bringle, 1991; Bringle & Boebinger, 1990; Buunk, 1981; 1982; Denfeld, 1974; Jenks, 1985).
Alternatives to Traditional Marriage. Although most ES is secretive, some couples do pursue lifestyles that permit ES (Blumstein and Schwartz 1983; Thompson 1983; Weis 1983). There is some evidence that consensual ES is unrelated to marital satisfaction (Gilmartin 1978; Ramey 1976; Rubin and Adams 1986; Wachowiak and Bragg 1980), suggesting there might be different outcomes for the consensual and nonconsensual forms of ES.
A number of models for consensual ES have been proposed, particularly during the 1970s. These include swinging (recreational and shared ES) (Bartell 1971; Gilmartin 1978; Jenks 1985), comarital sex (Smith and Smith 1974), open marriage (O'Neill and O'Neill 1972), intimate friendship networks (ES within context of friendship) (Francoeur and Francoeur 1974; Ramey 1976), and group marriage (Constantine and Constantine 1973; Rimmer 1966). Certainly, there are differences among these various nonexclusive lifestyles. We do not have the space to review fully the distinctions among them here (see Libby and Whitehurst 1977; Weis 1983). What unites them for the discussion here is that they all represent a consensual agreement to allow multilateral sexual involvement. As such, ES is assigned a different set of meanings from betrayal.
Consensual agreements can vary in terms of the degree of sexual involvement desired, the degree of intimate involvement desired, the degree of openness with the spouse, and the amount of time spent with the ES partner (Sprenkle and Weis 1978). Buunk (1980) studied the strategies couples employ in establishing ground rules for sexually open marriages. The five most common were: (1) primary value placed on maintaining the marriage, (2) limiting the intensity of ES involvements, (3) keeping the spouse fully informed of ES relationships, (4) approving ES only if it involves mate exchange, and (5) tolerating ES if it is invisible to the spouse. It would be useful to see research on the association between the types of strategies employed and outcomes of ES.
Interestingly, husbands tend to initiate swinging (Bartell 1971; Weis 1983). There is some evidence that most couples swing for a few years, rather than pursuing it for a lifetime (Weis 1983). Dropouts from swinging report problems with jealousy, guilt, emotional attachment, and perceived threat to the marriage (Denfeld 1974). As far as we know, there have been no studies comparing dropouts and those who enjoy and continue swinging.
The Constantine study (1973) is virtually the only source of data on group marriage in contemporary America. They report that the typical relationship includes four adults. Most enter a group with their spouses, and if the group dissolves, most of the original pair bonds survive. In fact, the original pair bonds retain some primacy after the formation of the group, and this may be a factor working against the success of the group. Jealousy between male partners appears to be a common problem.
Studies of marital models that permit ES have tended to employ small, volunteer samples with no control or contrast groups for comparison. There is no basis for a firm estimate of the incidence or prevalence of such alternative lifestyles, although Blumstein and Schwartz (1983) suggested that as many as one of seven marriages in the U.S.A. may have some agreement allowing ES. Despite the vast attention given to these alternative lifestyles in the 1970s, and despite the more recent claims that Americans are returning to traditional models of monogamous marriage, there is no scientific basis for concluding that these patterns increased in popularity earlier or that they have become less common in the 1980s and 1990s.
Sexuality and People with Physical and Developmental Disabilities
MITCHELL S. TEPPERGovernment Policies Affecting Sexuality and Disability. Over the past twenty years, pivotal legislation has been enacted in the United States that enables people with disabilities to gain their rightful place as equal members of American society. These changes have been led by spirited people with disabilities and their advocates. The Rehabilitation Act of 1973, the 1975 Education for All Handicapped Children Act (Public Law 94-142), and the Americans with Disabilities Act passed in 1990 have all added opportunities for inclusion and integration into the community for people of all abilities. With inclusion and integration have come greater opportunities for social interaction and sexual expression. The same spirit that has raised disability-rights issues to a national priority is now demanding that people with disabilities be recognized as sexual beings with a right to sexual education, sexual health care, and sexual expression afforded under the law.
Demands for the sexual rights of people with disabilities have resulted in a resurgence of research interest in the area of sexuality and disability in the 1990s. Notably, the National Center for Medical Rehabilitation Research (NCMRR) of the National Institute of Child Health and Human Development under the National Institutes of Health has identified sexuality as a priority issue that impacts the quality of life of people with disabilities. It subsequently issued a Request for Applications on Reproductive Function in People with Physical Disabilities in February of 1992. The purpose of the request was to develop new knowledge in the areas of reproductive physiology, anatomy, and behavior that are common to people with disabilities with the goal of restoring, improving, or enhancing reproductive function lost as a consequence of injury, disease, or congenital disorder. The request for applications included a specific objective to characterize the effect of impairments of sexual function on psychosocial adaptation, emotional state, and establishment of intimate relationships. Special focus was placed on research with women and minorities who have disabilities. NCMRR has funded six studies on sexuality and disability over the last three years. Two of the studies were with women who have spinal cord injury, and a third was a study of women with a variety of disabilities.
Consumers with Disabilities Leading the Way. Research, education, and advocacy efforts in the area of sexuality and disability are being led by people with disabilities (consumers). A review of the most recent annotated bibliography on sexuality and disability published by the Sexuality Information and Education Council of the United States (SIECUS, 1995) reveals a growing number of books, newsletters, special issues of publications, and curricula on sexuality and disability written by people with disabilities. In addition, national consumer-based organizations like the National Spinal Cord Injury Association, the National Multiple Sclerosis Foundation, and the Arthritis Foundation are beginning to publish self-help brochures on the specific effects of particular disabilities on sexuality. Most recently, self-help groups have been appearing on the Internet, computer bulletin-board services, and commercial computer services like American Online.
Health-Care Professionals Involved in Sexuality and Disability. In addition to the work by people with disabilities and nonprofessional advocates, health-care professionals are also taking an increased interest in sexuality and disability. The American Association of Physical Medicine and Rehabilitation has a Sexuality Task Force; the American Association of Sexuality Educators, Counselors, and Therapists has a special-interest group that focuses on educating medical and allied help professionals in the area of sexuality and disability; the Society for the Scientific Study of Sexuality includes presentations and workshops in the area of sexuality and disability for its members; and Planned Parenthood agencies around the country have increased education and services in the area of sexual health care to people with disabilities. More rehabilitation hospitals are including privacy rooms to give patients an opportunity to experiment sexually while still in the hospital, and many are adding specialty programs in the area of fertility and erectile function for men, obstetric and gynecological care for women, and parenting for both men and women with disabilities.
Portrayals of Sexuality and Disability in the Popular Media. The portrayal of people with disabilities as sexual beings has improved over time in the popular media. Movies that include a focus on the sexuality and relationships of people with disabilities, such as Forest Gump, Passion Fish, Water Dance, Regarding Henry, My Left Foot, Children of a Lesser God, and Born on the Fourth of July have dealt with the issue of sexuality and disability with varying degrees of sensitivity, and have enjoyed success at both the box office and in video stores. TV shows have also included people with disabilities and sexuality themes. One show, LA Law, where one of the stars portrayed a person with a developmental disability who had a sexual relationship with another person with a developmental disability, was honored by the Coalition of Sexuality and Disability for the positive portrayal of sexuality and disability in the media. There has also been an increase in TV commercials that include people with disabilities in relationships or with children. Popular magazines ranging from Bride to Penthouse and Playboy are also beginning to include feature articles on sexuality and disability. Efforts to portray people with disabilities as part of everyday life in the media are slowly helping to explode the myth that people with disabilities are asexual.
Problems, Controversies, and Hurdles. Two of the most serious sexual problems facing people with disabilities are (1) the high rate of sexual abuse, exploitation, and unwanted sexual activity, especially among women with physical disabilities and all people with developmental disabilities, and (2) the risk of STDs, including HIV, among people with cognitive impairments who are sexually active. Two leading areas of controversy are (1) the issue of what constitutes informed consent for sexual activity in people with serious cognitive impairments, and (2) the area of reproductive rights, eugenics, abortion, and prenatal testing for disabilities. As far as hurdles, there is still a need for greater access to information and educational material that affirms the sexuality of people of all abilities, including people with early- and late-onset disabilities, physical, sensory, and mental disabilities, and disabilities that hinder learning. Despite the positive current trends in sexuality and disability, we still have a long way to go in increasing the number of sexuality education and training programs for teachers, health-care workers, and family members to help them understand and support the normal sexual development and behavior of persons with disabilities. A goal is that all social agencies and health-care delivery systems develop policies and procedures that will insure sexual-health services and benefits are provided on an equal basis to all persons without discrimination because of disability.
Sexuality and Older Persons
ROBERT T. FRANCOEURIn 1860, over half of the American population was under 20 years of age and only 13 percent over age 45. In 1990, less than a third were under age 20 and 21 percent were over age 45. The so-called Baby Boomers born between 1945 and 1965 are now in their middle years. With the birthrate less than 15 per 1,000, America has become a graying society.
Although Americans over age 50 are the fastest-growing segment of our population, research on their lifestyles and patterns of intimacy has been almost exclusively limited to studies of the chronically ill, the socially isolated, and the poor. Edward Brecher (1984) was one of the First to study older healthy Americans. His sample of 4,246 persons between ages 40 and 92 was largely white and affluent, although he did include a low-income group. His overall conclusion was that the sexual interests and activities of older persons are the best-kept secrets in America. Although there is a common belief that the elderly are no longer interested in sexual intimacy, older persons were just as affected as young people by the social turmoil and changing attitudes of the 1960s and 1970s.
Brecher found that healthy, older person today are enormously different from the older person of forty or fifty years ago, and very much interested in intimacy and sexual relations. Not one of Brecher's 4,246 respondents was sexually inactive, although masturbation was the most common sexual outlet. Forty-four percent rated their sexual satisfaction as most enjoyable; less than 1 percent rated their sexual activity as not enjoyable (Table 4). Poor health was a major determinant in hindering older persons from maintaining an active sexual life.
Table 4: Sexual Activity Among 4,246 Americans, Ages 45 to 92, in the Brecher 1984 Survey
Orgasms while asleep or awakening
Women who masturbate
Masturbation frequency for women who masturbate
Wives having sex with husband
Frequency of marital sex
Orgasms while asleep or awakening
Men who masturbate
Masturbation frequency for men who masturbate
Men having sex with wife
Frequency of marital sex
In another study of healthy, upper-middle-class men and women, ages 80 to 102 living in residential retirement communities, 14 percent of the men and 29 percent of the women were still married. Sexual touching and caressing, followed by masturbation and then intercourse were the most common sexual activities. Of these outlets, only touching and caressing declined with age, a decline more evident in men than in women. Those who had been sexually active earlier in life tended to remain sexually active in their 80s and 90s, although the frequency of sexual intercourse was sometimes limited by their current physical health and by social circumstances including the lack of an available partner (Bretschneider and McCoy 1988).
The Starr-Weiner Report on Sex and Sexuality in the Mature Years (1981) examined the sexual lives and attitudes of 800 persons, aged 60 to 91, from four regions of the country. When the sexual activities of these 60- to 90-year-olds were compared with the 40-year-olds Kinsey studied thirty-five years earlier, there was no significant decline when opportunities for sexual activity existed. Sex remains pretty much the same unless some outside event intrudes, such as a health problem, the loss of a spouse, impotence, or boredom. A reliable predictor of the sexually active life of older persons is their acceptance or rejection of the social stereotype of the dependent, sickly older person. Older persons who maintain an active participation in life in general tend to be more sexually active in their later years.
Starr and Weiner also identified two major problems with no easy remedy. First is the tendency for older men to become asexual when they encounter an occasional erection or orgasmic problem. Instead of exploring noncoital pleasuring, many older men simply give up all interest in sex. The second problem is the ever-growing number of older women who are without sexual partners and, thus, deprived, against their will, of sexual intimacy and pleasure. (See Section 6B below on sexuality among older homosexual men and women.)
A Closing Comment
Throughout this section, we have noted the tendency of sexuality researchers in the U.S.A. to focus on the incidence and/or frequency of sexual behaviors in various lifestyles. There has been little corresponding research on the process of sexual relationships or the dynamics within them. This is precisely the same point we made in summarizing the section on adolescent sexuality. Suffice it to say that American researchers need to move beyond asking how many people do it and how often they do it to more fully investigate the contexts surrounding adult sexual lifestyles, and to identify the social, psychological, and biological factors associated with sexual practice.