Maternal Reports of Pregnancy, Genital, and Related Fantasies in Preschool and Kindergarten Children

Maternal Reports of Pregnancy, Genital, and Related Fantasies in Preschool and Kindergarten Children LINDA A. LINDAY, M.D.

ABSTRACT Objective: To test the hypothesis that there is no difference in the proportion of young boys with pregnancy fantasies versus young girls with penis fantasies and to explore the prevalence of pregnancy, genital, and related fantasies in 2through 6-year-old children. Method: The parents of 171 preschool and kindergarten children attending two private schools in New York City received questionnaires regarding the above fantasies. They were completed by the mothers of 31 girls and 34 boys (overall response rate of 38%). Results: There was no statistically significant difference in the proportion of girls reported to have penis/gender fantasies and the proportion of boys reported to have pregnancy fantasies. Reports of at least one fantasy were more common for girls (p < .006), as were reports of breast-feeding, other nurturing, and other reproductive fantasies (p < .03) and reports of wanting to urinate like a boy (p < .05). Reports of an "interest" in vaginas and breasts, as opposed to a fantasy about having them, were more common in boys (p < .03). Conclusion: These data support the "overinclusive phase" of Fast's differentiation model of gender identity. Developmentally appropriate androgynous fantasies must be distinguished from gender identity disorder, a psychiatric condition that can emerge in the same age group. J. Am. Acad. Child Ado/esc. Psychiatry, 1994, 33, 3:416-423. Key Words: gender identity, gender theory, development.

When he was between 3 and 3 Y2 years old, my son repeatedly stuffed a small doll under his shirt and said "I have baby in my 'room' [womb]." His father's secretary, whom he saw regularly, was pregnant at that time (I was not). This behavior was not repeated later, when 1 was pregnant (his sister was born when he was 4 years and 9 months old), nor was it repeated when his long-standing baby-sitter was pregnant (this baby was delivered when he was 5 years and 8 months old). Accepted August 24, 1993. Dr. Linday isAssistant AttendingPediatrician, St. Luke's/Roosevelt Hospital, New York, NY, and Assistant ClinicalProftssor of Pediatrics, The College of Physicians and Surgeons. This work wasperformed. in part, while the author was a Clinical Fellow in Child and Adolescent Psychiatry at the St. Luke's/ Roosevelt Hospital Center. The author thanks William Byne, M.D., Ph.D., Susan Coates, Ph.D., and Naomi Leiter, M.D., for their various review of the manuscript, instruction ongenderand gender identitydisorder, and inspiration for thisproject; Richard Shindledecker, M.A., for performing thestatistical analysis; PaulBarth,M.L.S., Terrence Gorman, and Patricia Tomasulo, M.L.S., for their help in locating bibliographic sources; and the administration, parents, and children of The Calhoun Lower School and The Rocking Horse Nursery School for making this study possible. Reprint requests to Dr. Linday, 340 West55th Street, Suite 9A, New York, NY 10019. 0890-8567/94/3303-0416$03.00/0©1994 by the American Academy of Child and Adolescent Psychiatry.

416

Discussions with mothers of young children confirmed that their children engage in similar behaviors. Experienced clinicians are aware that both young boys and young girls have "androgynous" fantasies. Brazelton and Cramer (1990, p. 11) state that a "pregnancy wish is already seen in the play of small children. Girls and boys try to portray a pregnancy by inflating their abdomens with pillows or by puffing out their stomachs." 1 hypothesized that the proportion of young boys with pregnancy fantasies would be no different from the proportion of young girls with penis fantasies. Ross (1974, p. 161) reported that almost half of the young boys (aged 3 years and 10 months through 5 years and 11 months) successfully interviewed in his study "offered positive and rich evidence of wishes to bear or rear babies in one form or another or, at the very least, of highly favorable attitudes and affects in relation to infants and reproductive processes." The existence of penis envy in young girls is well known (Freud, 1933), and the concept is crucial to Freud's conception of female sexualiry. Genital fantasies of young children are discussed by Gadpaille (1975), Galenson (1990), and Galenson and Roiphe (1971, 1976). However, 1

]. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 33:3, MARCH/APRIL 1994

CHILDREN'S PREGNANCY AND GENITAL FANTASIES

was unable to locate either the percentage of young girls with penis envy and/or fantasies, or a comparison of the prevalence of this fantasy with that of pregnancy fantasies in young boys. This study was performed to test the hypothesis that there would be no difference in the prevalence of these two phenomena in young children, as well as to explore the prevalence of pregnancy, genital, and related fantasies in this age group.

For statistical analysis, dichotomous variables (gender, presence of a fantasy, only-child status) were analyzed by X2 analysis and log-linear regression. The number of children with at least one fantasy in each category was used. Reported age of fantasy onset was analyzed by two-way analysis of variance. If a child exhibited more than one fantasy in any category, the youngest age of onset was used. Unless otherwise indicated, statistical analysis was performed with the SAS statistical package and two-tailed tests were used. The significance level was preset at p < .05.

RESULTS METHOD The administrations of two nonparochial, private schools in New York City consented to allow the parents of their students to be approached, in writing, regarding their interest in participating in this study. A questionnaire with a cover letter explaining the study and assuring confidentiality ofsubjects was sent to the parents of 171 preschool and kindergarten children attending the two schools. Those who did not respond to the first letter received a second mailing. A total of 65 questionnaires were returned, with a response rate of 38%. All questionnaires in the current study were reported to have been completed by the child's mother. Data were obtained regarding demographics, sibling status, and the presence or absence of fantasies in the following categories: pregnancy, breast-feeding (nursing), other nurturing, other reproductive, and genital fantasies. Each fantasy was clearly defined in the questionnaire; genital fantasies were defined separately for girls and boys. If a fantasy was present, the age at which it occurred and a description were requested. Inconsistencies and/or missing data were clarified by telephone contact. Psychopathology in the subjects and their families was not assessed, nor was it an exclusion criterion. A "pregnancy fantasy" was defined in the questionnaire as "your child showing you or telling you that he/she wants to be, or is pretending to be, pregnant." A "breast-feeding (nursing) fantasy" was defined as "your child showing you or telling you that he/ she wants to breast-feed a baby." "Other nurturing fantasies" were defined as "excluding pregnancy and breast-feeding, has your child ever shown you or told you that she/he wants to take care of (nurture) a baby? (Examples include bottle-feeding, bathing, diapering, and/or dressing a baby.)" "Other reproductive fantasies" were defined as "has your child ever had any other kind of reproductive fantasy?" "Genital fantasies for girls" were defined as "has your daughter ever shown you or told you that she would like to have, thinks she has, or thinks she will get a penis?" "Genital fantasies for boys" were defined as "excluding pregnancy and breast-feeding (see ... above) has your son ever shown you or told you that he would like to have, thinks he has, or will get a vagina, womb, or breasts?" Additional subcategories used for statistical analysis included (1) "interest": although not specifically requested, several mothers wrote that their children exhibited an interest in genitals without meeting the above definition of fantasy; this was counted separately as "interest." For example, one mother wrote that her daughter "in bathtub, pretended to push forskin [sic] back on herself to get brother to copy the behavior"; this was counted as "interest." Other subcategories were (2) reports of wanting to "pee like a boy/girl" and (3) "gender" fantasy, defined as a report of the statement "I want to be a boy/girl."

Demographics

Questionnaires were returned for 31 girls and 34 boys. Three mothers completed questionnaires for each of two sons, one of which was a set of fraternal twins. Subject ages were 2.6 to 6.7 years with a mean of 4.8 years. There was no significant difference between genders for age, race, or religion. Subjects were predominantly Caucasian (94%) and Jewish (45%); they were predominantly middle and upper-middle class. There was no difference between genders in the proportion of children with no siblings nor in the proportion of children who were the eldest child. Nongenital Fantasies

A mean of 2.7 nongenital fantasies were reported per girl and 1.2 were reported per boy (p < .0001). There was a tendency for more girls than boys to be reported to have pregnancy fantasies (p < .09; not significant) (Table 1). A greater proportion of girls than boys were reported to have nursing (breast-feeding), other nurturing, and other reproductive fantasies (p < .03) (Table 1). Ninety-seven percent of girls and 53% of boys were reported to have other nurturing fantasies. Some mothers commented that their sons did not have "other nurturing" fantasies with "babies" as described in the questionnaire (see "Method"), but added that their sons were nurturing with pets and younger children, although not with dolls. This study did not systematically assess nurturing with pets or younger children; nurturing activity in boys may therefore be underestimated. Other reproductive fantasies reported were predominantly birth fantasies; those reported for girls tended to be more realistic than the birth fantasies reported for boys. Only two children (one boy and one girl) were reported to have fantasies of going back into their mother's womb (or stomach). There were

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 33:3, MARCH/APRIL 1994

417

LINDAY

TABLE 1 Nongenital Fantasies Pregnancy Girls (N = 31) Boys (N = 34)

21 (68)} 15 (44)

t

Nursing 20 (64) } 2 (0.06) *

Other Nurturing

Other Reproductive

30 (97)}

13 (42)}

18 (53)

*

5 (15) *

Note: Values represent number (percent).

*p

<

.03; t not significant.

no written comments indicating that a girl was pretending to have or care for her father's baby. Genital/Gender Fantasies

One of 31 girls and 2 of 34 boys were reported to have gender fantasies ("I want to be a boy/girl"; not significant). The following categories were therefore combined for statistical analysis: (1) girls: penis and gender fantasies; (2) boys: vagina, breast, and gender fantasies. There was a tendency for more girls to be reported to have penis/gender fantasies (14/31 = 45%) than for boys to be reported to have vagina/breast! gender fantasies (8/34 = 24%; P < .08; not significant). Girls were more likely to be reported to want to urinate like boys than vice versa (6/31 girls versus 1/34 boys; P < .05). Of the 14 girls with penis/gender fantasies, only 2 were also reported to want to "pee like a boy," although two additional girls who wanted to "pee like a boy" had reports of penis/gender "interest." Reports of an "interest" (see "Method") in vaginas and breasts, as opposed to a fantasy about having them, were more common in boys (p < .03). When total "cross-gender" genital/gender activity (defined as a child with at least one report of a genital/gender fantasy and/or "interest" and/or wanting to "pee like a boy/ girl") was compared, there was no significant difference between genders and no trends were apparent. Two mothers commented on the intensity of their daughter's wish to have a penis, which occurred from 3 to 4 years of age. Pregnancy, nursing, and other nurturing fantasies were also reported for both girls, although these fantasies were not as intense. No girl was reported to deny or dislike her genitals or gender. The mother of one boy wrote that when her son was 3 to 3V2 years old "he repeated numerous times that he had a baby in his tummy and told me frequently that he had a "gina' and insisted that he did not have a penis and was going to grow up to be a girl. He liked to urinate sitting down on the toilet 'like a girl.'"

418

She noted that he "understood that girls have 'babies in their tummys [sic]' and vaginas so he wanted a vagina too." The mother reported that these fantasies "began when my daughter was about 5 months old and stopped when she was about 9 or 10 months old. He has not mentioned any of this in months". Other Comparisons

There was no statistically significant difference between the percentage of girls (14/31 = 45%) reported to have penis/gender fantasies and the percentage of boys (15/34 = 44%) reported to have pregnancy fantasies. All 14 girls who were reported to have a penis/ gender fantasy were also reported to have at least one nongenital fantasy. Similarly, all 19 girls who were reported to have a penis/gender fantasy and/or interest and/or wanting to "pee like a boy" were also reported to have at least one nongenital fantasy. There was no indication from mothers' written comments that these fantasies were temporally mutually exclusive. Girls were more likely to be reported to have at least one fantasy in any category than were boys (p < .006). Reported Age of Fantasy Onset

The mean reported age of onset of the earliest fantasy was 2.5 ::!: 0.7 (SD) years for girls and 3.0 ::!: 0.7 years for boys (p < .01). For girls, the reported age of onset of other nurturing fantasies (2.7 ::!: 0.8 years) was younger than that of other reproductive fantasies (3.3 ::!: 0.7 years; p < .04) (Table 2). For 11/14 (71%) of the girls reported to have a penis/gender fantasy, the reported age of onset of the penis/gender fantasy was equal to or older than the reported age of onset of the earliest nongenital fantasy (p < .03; one-tailed Sign test; Siegel and Castellan, 1988). For boys, there was no difference in the mean reported age of onset for the fantasies studied. The age of onset was unknown for one boy with a pregnancy fantasy, one boy with

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY. 33:3. MARCH/APRIL 1994

CHILDREN 'S PREGNANCY AND GENITAL FANTASIES

TABLE 2 Mean Reported Age of Fantasy Onset First Fantasy Girls Boys

2.5 ± 0.7* 3.0 ± 0.7*

Pregnancy 3.1 ± 0.7 3.6 ± 0.7

Nursing 3.0 ± 0.7 2.9 ± 1.2

Other Nurturing

Other Reproducti ve

Genital/Gender

2.7 ± 0.8** 3.2 ± 0.8

3.3 ± 0.7** 3.1 ± 0.2

3.0 ± 0.8 3.5 ± 1.0

Note : Values represent years ± SD . * P < .0 1; **P < .04.

an other nurturing fantasy, and one girl with a penis/ gender fantasy. Sibling Status

Logistic regression analysis was used to hold the variability of both gender and "only-child" status constant. This revealed an effect of being the eldest child on the report of both nursing (p < .03) and other reproductive fantasies (p < .02), but not on the other fantasies studied. DISCUSSION

Study Designs and Limitations

The current study is based on mothers ' responses questionnaires regarding their children 's fantasies; thus, maternal gender bias in reporting these fantasies could not be excluded and paternal contributions could not be assessed independently. An additional limitation is the overall response rate of 38%; this is discussed in detail below. The current study population was predominantly Caucasian and middle to upper-middle class. Gordon et al. (1990) reported that lower class children had less knowledge than middle and upperclass children about sexual body parts and pregnancy. Thus, the results of the current study may not be applicable to lower class children. In addition, although subjects were drawn from a nonclinica1 population, psychopathology in the subjects and their families was not assessed, nor was it an exclusion criterion. Finally, questions regarding the frequency and intensity of the fantasies studied were not included in the questionnaire. Within these limitations, the results of the current study support its original hypothesis. The overall response rate and the use of a questionnaire completed by subjects' mothers in the current study require additional discussion. In his study of nurturance and generativity in young boys, Ross (1974) to

successfully interviewed boys aged 3 years and 10 months through 5 years and 11 months from twO nonparochial, private nursery schools in New York City. Initially, 40% of the parents agreed to allow their children to be interviewed. However, Ross found it necessary to exclude subjects younger than 3 years and 10 months of age because he was unable to communicate with them in the interview conducted at school. Had such an age exclusion been applied in the current study, 15 of 65 or 23% of the subjects would have been excluded, and with them important early material. Interviews of children were not used in any of the five surveys regarding children's sexual (Friedrich et al., 1991) or gender behavior (Bates er al., 1973; Sandberg et al., 1993; Zucker et al., 1980; Zuger and Taylor, 1969) ident ified in the literature. Whereas Zuger and Taylor (1969) interviewed mothers with a structured protocol, questionnaires completed by one or both parents were used in the four other studies. Zucker et al. (1980) identified normal subjects from the files of a research institute, whereas Friedrich er al. (1991) recruited subjects from a pediatric clinic. In the three remaining studies, subjects were recruited on the basis of school attendance. The response rates in the above studies were 35% (Bates et al., 1973), less than 40% (Ross, 1974), 56% (Zuger and Taylor, 1969), 74% (Sandberg et al., 1993), and 82% plus an unknown response (Zucker et al., 1980). Friedrich et al. (1991) used an anonymous questionnaire and reported an initial response rate of 70.8%. However, only 80.6% of these questionnaires could be used; therefore, useful questionnaires were obtained from 57.1% of those initially contacted. In conclusion, the use of questionnaires completed by mothers as well as the overall response rate of 38% reported in the current study are consistent with prior research in the field. I postulate that low response rates

J. AM . ACAD. CHILD ADOLESC . PSYCHIATRY. 33:3. MARCH/APRIL 1994

419

LIN DAY

may reflect parental unwillingness to share sensitive information about the sexual and gender development of their young children, especially when study recruitment is based on school attendance. Theories of Gender Development

The classic work of Money and Ehrhardt (1972) revealed that despite biological abnormalities, the sex of rearing was the predominant factor in determining a child's gender identity. The early acquisition of sextyped activities is well documented in the developmentalliteratute (Huston, 1983). "Almost before they can label such items" it is clear that "trucks, trains, and active games are masculine while dolls, dishes and house-cleaning toys are feminine" (Huston, 1983, p. 449) . Berween the ages of 12 and 30 months, "children come to label themselves as boys or girls and begin to adopt behaviors which are sex-typed" (Fagot and Leinbach, 1985, p. 687); children who can correctly categorize girls and boys show more sex-typed behavior than those who cannot. Theories explaining the acquisition of sex-typed behavior include social learning theory and cognitive developmental theory (Fagot and Leinbach, 1985; Gadpaille, 1975; Huston, 1983; Person and Ovesey, 1983); Unger (1979) stressed the impact of the sociopolitical environment on the development of sex-typed behavior in young children. However, childhood fantasies of generativity, nurturing, and penis envy have historically been the purview of psychoanalysts (Chodorow, 1989; Fast, 1984; Freud, 1933; Galenson, 1990; Galenson and Roiphe, 1971, 1976; Horney, 1924, 1926; Person and Ovesey, 1983; Ross, 1974, 1975; Stoller, 1968, 1985). Freud (1933; Fast, 1984; Person and Ovesey, 1983; Stoller, 1985) maintained that both girls and boys are masculine in gender orientation in the first years of life (Fast, 1984), although he also proposed an innate bisexuality, with consequences for both normal and abnormal development (Stoller, 1985). Horney (Person and Ovesey, 1983) maintained that femininity and masculinity were innate , early (prephallic) and separately derived. In contrast to Freud, Stoller (1968, 1985; Person and Ovesey, 1983) proposed a first stage of primary femininity in both sexes, which he termed "protofemininity." The high prevalence of "cross-gender" fantasies reported at an early age (2 to 3 years) for both boys and girls (Tables 1 and 2) in this study is consistent

420

with and supportive of the "overinclusive phase" of Fast's (1984) differentiation model of gender development, in which she hypothesizes that both boys and girls have "early notions of undifferentiated sexual completeness" (p. 14). Fast proposes that both boys and girls are initially "overinclusive in their experience, not attuned to sex difference or aware of the limitations inherent in belonging to a particular sex" (p. 22). She notes that "both boys and girls must come to terms with their limits" (p. 22). Boys must come to terms with the fact that they cannot bear children; girls must come to terms with the fact that they do not have a penis. In the course of coming to terms with their limits, both boys and girls may envy the "sex and gender attributes" of the other sex. Both "may perceive the fact of not having those attributes as a loss or an incompleteness" (p. 22). For both boys and girls, success in coming to terms with the limits of one's gender requires focusing on the positive attributes of that gender while simultaneously deemphasizing the fact that one does not having the sexual attributes of the opposite gender. Fast adds that "there is ... no explicit suggestion in theory nor evidence in clinical experience that little girls imagine that having a penis would mean not having the capacity to bear children" (p. 13). The rwo girls reported to have intense penis fantasies in the current study were also reported to have had pregnancy, nursing, and other nurturing fantasies. All girls reported to have a penis/gender fantasy were also reported to have at least one nongenital fantasy. There was no indication from written comments that the fantasies were temporally mutually exclusive. Fast (1984) also hypothesized that "boys, too, must have notions that all sexual possibilities are open to them" and that they could "have male genitals and also have babies" (p. 14). This statement is supported by the finding that 44% (15/34; Table 1) of boys in the current study were reported to have pregnancy fantasies. Urination and Penis Envy

The girls in this study were more likely to be reported to want to urinate like boys than vice versa (p < .05). After toilet-training in our culture, girls sit both to urinate and to defecate, while boys stand to urinate and sit to defecate. However, if a boy plans to urinate and defecate at the same time, he will sit to do both. Thus, boys routinely have the experience of sitting to

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY . 33:3, MARCH/APRIL 19 94

CHILDREN'S PREGNANCY AND GENITAL FANTASIES

urinate, or "peeing like a girl," while girls have no routine experience standing to urinate. For the girl, then, "peeing like a boy" is new and different, whereas for the boy, "peeing like a girl" is a routine, mundane experience. Horney (1924, 1926; Person and Ovesey, 1983) noted the "narcissistic overestimation in which the excretory processes are held by children" (1924, p. 52). Horney (1926, p. 332) suggested that the term "primary" penis envy be applied to the childhood phenomenon, which she considered to be a "universal but transient stage in which the girl was said to be at a realistic disadvantage because she lacked the narcissistic, exhibitionistic components of urination" (Person and Ovesey, 1983, p. 210). Unlike Freud, she considered that this stage was "without substantive consequences for subsequent female development" (Person and Ovesey, 1983, p. 210). Fast's differentiation theory of gender development and the findings of the current study are compatible with the attitudes of early female psychoanalysts. As reported by Chodorow (1989), the latter "were more likely ... to argue that penis envy is a developmental stage rather than a major determinant of women's lives, that anyone can see that little girls envy the penis just as boys have castration anxiety, or that boys have womb envy just as girls have penis envy" (pp, 204-205). However, Galenson (1990) and Galenson and Roiphe (1971, 1976) emphasized that in her first year of life, threats to a girl's developing body or to her relationship with her mother place her at risk for developing a severe preoedipal castration reaction upon learning that she does not have a penis. Although such reactions were not reported by the mothers in the current study, they were not the primary focus of this research. Nurturing

The nurturing behavior in boys reported in the current study is compatible with the findings of Zucker et al. (1980) in their questionnaire survey of normative cross-gender behavior in very young boys. It is also equivalent to that reported by Ross (1974) in his interview study. Ross (1975, p. 812) also notes that '''maternal' or, better, parental aspirations are organic, indeed organizing components in the evolution of male identity and ... that there exists a substantial, though neglected history of psychoanalytic thought to this

effect." In the current study, some mothers commented that their sons were nurturing with pets and younger children, but not with babies or dolls. Nurturing with pets or younger children was not systematicallyassessed in the current study (see "Method"), which may therefore have underestimated nurturing behavior in young boys. These facts should be considered in the design of future studies; they are also clinically relevant for those wishing to encourage nurturing in young boys. As discussed by Jacklin (1989), the psychological anthropologists Whiting and Edwards have concluded that "we are the company we keep" (p. 131) and that if an "individual spends enough time with infants, he or she will become a nurturer" (p, 131). Whiting and Edwards note that in many cultures young girls are given child care responsibilities while young boys are not; as a result of these responsibilities, per their theory, girls are more likely than boys to become nurturers. However, the rationale for the initial assignment of child care responsibilities to girls rather than boys in these cultures remains unclear. As Jacklin (1989) notes, "this perspective deserves further exploration" (p, 131). The current study notes a high prevalence and early reported age of onset of nurturing fantasies in young girls, as well as a high prevalence of other nongenital fantasies in girls (Tables 1 and 2). This may represent an example of the early onset of sex-typed behavior (see above, "Theories of Gender Development"). Other

In the current study, only one girl (1/31 or 3%) and two boys (2/34 or 6%) were reported to have said that they wanted to be of the opposite gender. These data are consistent with the rarity of a wish to be of the opposite gender as documented in nonclinical populations for young boys (20 to 60 months of age; Zucker et al., 1980) and children in middle childhood (Meyer-Bahlburg, 1985; Sandberg et al., 1993). The current finding that boys were more likely to show an "interest" in vaginas and breasts, as opposed to a fantasy about having them, is compatible with the concepts of both Freud and Horney. As the latter noted (Horney, 1924), "a woman, because her genital organs are hidden, is ever the great riddle for man." An effect of being the eldest child was noted only for nursing and other reproductive fantasies. This finding suggests that the occurrence of these fantasies is more

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 33:3. MARCH/APRIL 1994

421

LINDAY

influenced by experience than the other fantasies evaluated in this study. Gender Identity Disorder

Developmentally appropriate androgynous fantasies of early childhood must be distinguished from "gender identity disorder" (GID), a psychiatric condition that may emerge between 2 and 4 years of age and occurs more commonly in boys (Coates, 1992; Green, 1985, 1987; Marantz and Coates, 1991; Zucker, 1985; Zucker and Green, 1992). Children with GID are distressed about their gender, exhibit chronic and pervasive wishes to be of the opposite gender, and manifest a predominant interest in activities and behavior that are typical of the opposite gender (Marantz and Coates, 1991). This includes an intense interest in dressing up in female clothes, playing with dolls, and playing a female role in fantasy play (Green, 1987; Zucker, 1985; Zucker and Green, 1992). GID "is defined by its intensiveness, pervasiveness and duration" (Coates, 1992, p. 246). It is usually associated with other psychopathology in both the child (Coates and Person, 1985) and the family (Marantz and Coates, 1991); it may emerge in the context of a traumatic experience; and it usually occurs in boys with certain temperamental characteristics, particularly the avoidance of rough-andtumble play (Coates, 1992). The fact that more boys than girls are referred for evaluation of a cross-gender complaint may be related to our society's greater tolerance of cross-gender behavior in girls than in boys (Meyer-Bahlburg, 1985). However, as noted by MeyerBahlburg (1985), "GID of childhood is not an all-ornothing phenomenon'.' (p, 682). In one case in the current study, a mother reported that her son denied that he had a penis and insisted that he would grow up to be a girl (see "Results"). This was in addition to frequently saying that he had a baby in his "tummy" and that he had a vagina. All of these fantasies were reported to be related to the birth of his younger sister, to be of approximately 5 months' duration, and to have resolved a few months before completion of the study questionnaire. Although there were no written comments by this child's mother indicating that he cross-dressed, wore makeup, or had an aversion to boys' games, questions regarding such behavior were not included in the study questionnaire. In addition, no information regarding family psychopathology is available. Had this boy continued to deny

422

that he had a penis and to insist that he would grow up to be a girl, a clinical evaluation would have been appropriate. Long-term follow-up of this child would be of interest. Additional Research

The current study provides documentation of androgynous fantasies in a nonclinical population of young children and supports the "overinclusive" phase of Fast's (1984) differentiation model of gender identity. However, this normal developmental phase must be distinguished from GID, a psychiatric condition that can emerge in the same age group. Additional study of developmentally appropriate "cross-gender" fantasies of early childhood, preferably longitudinal in nature, with evaluation of the relative frequency, intensity, duration, and interrelationship of the relevant fantasies, is warranted for both girls and boys. Such research would provide an empirical base for gender development theory; it would also help differentiate the spectrum of developmentally appropriate "crossgender" fantasies of early childhood from the spectrum of behaviors associated with GID.

REFERENCES Bates JE, Bender PM, Thompson SK (1973), Measurement of deviant gender development. Child Dev 44:591-598 Brazelton TB, Cramer BG (1990), The Earliest Relationship. Reading, MA: Addison-Wesley Chodorow NJ (1989), Feminism and Psychoanalytic Theory. New Haven, CT: Yale University Press Coates S (1992), The etiology of boyhood gender identity disorder: an integrative model. In: Interface ofPsychoanalysis and Psychology, Barron J, Eagle M, Wolitzky D, eds. Washington, DC: American Psychological Association, pp 245-265 Coates S, Person ES (1985), Extreme boyhood femininity: isolated behavior or pervasive disorder? JAm Acad Child Psychiatry 24:702-709 Fagot BI, Leinbach MD (1985), Gender identity: some thoughts on an old concept. JAm Acad Child Psychiatry 24:684-688 Fast I (1984), Gender Identity: A Diffirentiation Model. Hillsdale, NJ: Analytic Press Freud S (1933), Femininity. In: StandardEdition, Vol 22. London: Hogarth, 1964, pp 112-125 Friedrich WN, Grambsch P, Broughton D, Kuiper J, Beilke RL (1991), Normative sexual behavior in children. Pediatrics 88:456-464 Gadpaille WJ (1975), In: The Cycles of Sex, Freeman L, ed. New York: Charles Scribner's Sons Galenson E (1990), Observation of early infantile sexual and erotic development. In: Handbook ofSexology, Vol 7: Childhood andAdolescent Sexology, Perry ME, ed. New York: Elsevier Science Publishers BV Galenson E, Roiphe H (1971), The impact of early sexual discovery on mood, defensive organization, and symbolization. Psychoanal Study

Child 26:195-216 Galenson E, Roiphe H (1976), Some suggested revisions concerning early female development. JAm Psychoanal Assoc 24(suppl):29-57

j, AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 33:3, MARCH/APRIL 1994

CHILDR E N' S PREGNAN CY AND GENITAL FANTASIES Gordon BN , Schroeder CS, Abrams JM (1990), Age and social-class differences in children's knowledge of sexuality. Journal of Clinical

Child Psychology 19:33-4 3

Psychoanal Assoc 23:783--817

G reen R (1985), Gender identity in childhood and later sexual orientation: follow-up of 78 males. Am J Psychiatry 142:339- 341 Green R (1987), The "Sissy Boy Syndrome" and tb«Deuelopment ofHomosexuality. N ew H aven, Cf: Yale Un iversity Press Horney K (1924), On the genesis of the castration complex in women.

Int J Psychoanal 5:50--65 Horney K (1926), The flight from womanhood: the masculinity-compl ex in women, as viewed by men and by women . Int ] Psychoana17:324-339 Huston AC (1983 ), Sex-typing. In: Handbook ofChild Psychology, Mussen PH , ed. New York: John W iley & Sons, pp 387-467 Jacklin CN (1989), Female and male: issue s of gender. Am Psychol 44:127-133 Maranrz S, Coates S (1991), Moth ers of boys with gender identity disorder: a comparison of matched controls. J Am Acad Child Adolesc Psychia-

try 30:310-315 Meyer-Bahlburg HFL (1985 ), Gend er identity disorder of childhood.

J

Am Acad Child Psychiatry 24:684-688 Mon ey J, Ehrhardt AA (1972), Man and Woman, Boy and Girl Baltimo re, MD: John s Hopkins Un iversity Press Person ES, Ovesey L (1983) , Psychoanalytic theories of gender ident ity. J

Am Acad Psychoanalll:203-226 RossJ (1974), The children's children: a psychoanalytic study of generativity and nu rtu rance in boys. PhD dissertation, New York Un iversity (Unive rsity M icrofilms, Ann Arbor , MI)

J. AM.

Ross J (1975), The development of paternal identity: a critical review of the literature on nurturance and generativiry in boys and men. J Am Sandberg DE , Meyer-Bahlburg HF L, Ehrh ardt AA, Yager TJ (1993), The prevalence of gender-atypical behavior in elementary schoo l children.

J Am Acad Child Adolesc Psychiatry 32:306-3 14 Siegel S, Castellan NJ Jr (1988) , Nonparametric Statisticsfor the Behavioral Sciences; 2nd ed. New York: McGraw-Hili Sto ller RJ (1968), So: and Gender, Vol I. New York: Jason Aronson Stoller RJ (1985 ), Presentations of Gmder. New Haven, Cf: Yale Un iversity Press Unger RK (1979), Female and Male, Psychological Perspectives. N ew York: Harper & Row Zucker KJ (1985), Cross-gender-ident ified children. In: Gender Dysphoria: Development, Research, Management, Steiner BW, ed. New York: Plenum Press, pp 75-174 Zucker KJ, Bradley SJ, Corter CM , Doering RW, Finegan JK (1980 ), Cross-gender behavior in very you ng boys: a normative study. In: Childhood and Sexuality, Sampson J, ed. Monrreal: Editions Etudes Vivanres, pp 599-622 Zucker KJ, Green R (1992), Psychosexual disorders in children and adolescents. J Child Psychol Psychiatry33:107- 151 Zuger B, T aylor P (1969), Effeminate behavior present in boys from early childhood, II: comparison with similar symptoms in non-effeminate boys. Pediatrics44:37:5-38 0

ACAD. CHILD ADOLESC . PSYCHIATRY, 33:3, MARCH/APRIL 1994

423