Library of the History of Autism Research, Behaviorism & Psychiatry

What Type of Sex-Role Behavior Should Behavior Modifiers Promote?

Journal of Applied Behavior Analysis, 1977, 10:549-552, Number 3, Fall 1977

.pdf version

Rekers and Lovaas (1974) reported a study using reinforcement procedures aimed at modifying the behavior of a 5-yr-old who engaged in cross-gender behavior. The use of traditional sex-role concepts in forming target behaviors is criticized and androgynous behavior is suggested as an empirically based alternative. The pathological nature of the behavior that the study was designed to prevent is also questioned, as is the ability to predict sex-role behavior from a 5-yr-old child's current behavior. These issues are raised in the context of the more general question, whom should the therapist serve?
DESCRIPTORS: behavior therapy, ethics, androgyny, homosexuality, preschool children

Rekers and Lovaas (1974) reported a study in which reinforcement procedures were used to treat a 5-yr-old boy, Kraig, who engaged in marked cross-gender behavior. A number of questions are raised by the way behavioral goals were defined in the study, questions similar to those raised by Winett and Winkler (1972) in this journal and by Davison (1976), in his presidential address to the Association for the Advancement of Behavior Therapy.

The behaviors that were designated desirable, i.e., indicative of therapeutic success or worthy of reinforcement were: (a) dressing in "football helmet", "army 'fatigue' shirt with stripes and other military decorations", "army belt with hatchet holder", "sea captain's hat"; (h) what the authors called masculine aggression: playing with dart gun, rubber knife, handcuffs, and cowboys and Indians; (c) playing with a toy submachine gun, toy soldiers and airplanes, road scraper, race car, and dump truck.

Behaviors designated as undesirable, i.e., indicative of need for treatment or not worthy of reinforcement were: (a) dressing in girls' clothes; (b) what the authors called maternal nurturance: playing with baby doll, baby bottle, crib, baby powder; (c) under the heading, feminine behaviors: playing with girls, dolls, "feminine gestures", and female role-play in games. Kraig, according to the authors "seemed almost compulsive or 'rigid' in the extent to which he insisted on being a girl".

The study raises a fundamental question, to whom does the therapist owe first allegiance: to the client (or in this case the client's parents), to the therapist's own values, or to prevailing relevant social norms?

In many cases, there is little conflict between these perspectives. The client believes what power structures, such as schools or media, inculcate and the therapist has been selected and taught according to the same beliefs. Rekers and Lovaas appear, through their paper, to be in such a situation. Schools, churches, media, and families have generally taught that males must rut "masculine sex roles" and females, "feminine sex roles". Boys and girls are therefore taught from an early age that such a situation is desirable and that negative sanctions will befall them if they do not "fit" society's prescribed sex roles. There is an extensive literature on sex typing and sex-role socialization (Mischel, 1970). A number of authors have, however, commented that sex-role socialization research appears to assume that traditional sex roles are "natural", rather than questioning whether these sex roles will permit the maximal development of individual potential (Levy, 1972). Rekers and Lovaas' paper, in justifying their work, appears to be subject to a similar problem.

Rekers and Lovaas present their work as designed to (1) prevent future sexual deviance: transsexualism, transvestism and some forms of homosexuality and the predicted need for treatment of these behaviors, and (2) reduce current aversiveness of the child's behavior to the parents and the child's peers (Rekers and Lovaas, 1974, p. 188).

For the first goal to be acceptable, evidence must be produced on two counts: (a) that the authors can predict from a 5-yr-old child's sex-role behavior that he will be transsexual, transvestite, or homosexual, (b) that this is cause for therapy. This second question is only partly an empirical matter. The evidence relevant to the first point is weak not surprisingly, since in most areas of behavior, there are few prospective, follow-through studies of development. Rekers and Lovaas refer to studies indicating that most adult transsexuals, transvestites, and some homosexuals report that their cross-gender behavior began in early childhood. There is no evidence as to how many children with early cross-gender behavior did not continue into adulthood to become transsexuals, transvestites, or homosexuals. In reference to the second point, the authors suggest that being an adult transsexual in today's society is not very pleasant (depression, self mutilation, prison). However, they do not cite evidence pertinent to the other adult behaviors they claim to be predicting and therefore preventing, i.e., transvestism and homosexuality. Evidence does not allow them to predict differentially between the three possible outcomes they mention, so evidence of adult adjustment in these two areas is pertinent.

Studies of adult homosexuality in clinical and penal populations (e.g.. Eysenck and Eysenck, 1964), have found homosexuals to be more neurotic than heterosexuals. However, it is invalid to generalize from patient or prison populations to nonpatient, nonprison populations. There is now considerable, replicated evidence from surveys of nonpatient homosexuals that homosexuals are not more abnormal or less well-adjusted than heterosexuals (e.g., Siegelman, 1972). A similar point applies to the few studies of transvestites and transsexuals, most of which have been carried our using patient populations.

The authors make no mention of the evidence of the changing attitudes to homosexuality and other sexual behavior labelled deviant, evidence such as changing laws, gay liberation movements, and psychiatric opinion. Where "pathology" is associated with sexual deviance, much of it, if not all, can be regarded as a function of social attitudes to sexual behavior (Davison, 1976). As attitudes change, it becomes increasingly presumptuous to guess about the type of adult life a child with cross-gender behavior will lead.

The second set of arguments Rekers and Lovaas uses to justify their goals involve the current parental and peer rejection of the child's behavior. The seriousness of parental and peer concerns can be accepted by the therapist, yet different goals may be used in defining target behaviors. It can be argued that the target behaviors Rekers and Lovaas work toward are traditionally defined sex-role behaviors that are not in the best long-term interests of society. This, of course, is a value-judgement, just as it is a value-judgement to formulate target behaviors as Rekers and Lovaas have done. Rekers and Lovaas appear to assume that adjustment to the sex-role status quo brings the greatest "happiness" or psychological "adjustment". Bem (1975) reviewed research pertinent to this difference of opinion and concluded:

"A high level of sex-typing may not be desirable. For example, high femininity in females has consistently been correlated with high anxiety, low self-esteem and low social acceptance; and although high masculinity in males has been correlated during adolescence with better psychological adjustment, it has been correlated during adulthood with high anxiety, high neuroticism, and low self acceptance. In addition, greater intellectual development has been correlated quite consistently with cross-sextyping.... Boys and girls who are more sex-typed have been found to have lower overall intelligence, lower spatial ability, and lower creativity" (Bem, 1975, p. 3, fully referenced in original).

Thus, it can be argued that Rekers and Lovaas, by using traditionally defined sex roles, may be preparing children for less than optimal adult roles.

If clinicians are to accede to parental (social) pressure to deal with cross-gender behavior, an alternative approach might be to develop a list of androgynous target behaviors as an alternative to both the existing behavior and the target behavior used by Rekers and Lovaas.

Bern's androgyny scale suggests such behavior (Bern, 1974). The Bern Sex-Role Inventory is made up of masculinity, femininity, and social desirability scales (Bern, 1974). An androgyny score is defined as the difference between an individual's masculinity and femininity scores normalized with respect to the standard deviations of his or her masculinity and femininity scores. The lower the score, the higher the androgyny score. The Inventory might he used to assess a variety of behavior-change programs with adults (e.g., assertive training, social-skills development, sexual therapy) as well as a guideline to determining target behaviors. For example, a high androgyny scorer might be: assertive, willing to take a stand, forceful, able to defend own beliefs, and independent as well as, affectionate, gentle, sympathetic, compassionate, and cheerful, to take items from each of the masculinity and femininity scales respectively. Behavior-change programs with these broader goals would provide people with a more flexible and appropriate response to a larger range of situations. Ability to behave in both "masculine" and "feminine" ways according to the demands of different situations would seem a more desirable goal than strengthening only one type of sex-role behavior. People seeking assertive, social-skill, sexual, and sex-role programs are often deficient in both "masculine" and "feminine" behaviors. Behaviors relating to sympathy. compassion, and cheerfulness require as much care in their definition and development as do the better known behaviors associated with such programs. This may be particularly so for males.

Such an approach might help to reduce the "rigidity" the authors impute to Kraig's behavior. With adolescents or adults, a further alternative might involve accepting the child's existing cross-gender behavior, modifying the parents' lack of acceptance and teaching the child, by assertive training and social reinforcement techniques to modify the behavior of those who do not reinforce his cross-gender behavior. Such an approach has been used by Graubard, Rosenberg, and Miller (1971) with children in special-education classes and by Russell and Winkler (1976) with adult homosexuals.

Research suggesting that generally held notions of sex-role behavior are not necessarily optimal for individual development raises potential conflict between a desire to respond to parental (or client) concerns and a desire to create normal psychological "adjustment". Parents concerned about social rejection, particularly with school approaching, as in the Rekers and Lovaas case, make judgements that the therapist has to decide whether to serve, or define as a target of change. In either case, the therapist' has to make a value judgement, hopefully guided by a full knowledge of relevant information. It appears that Rekers and Lovaas have not attended to research indicating that the popular mythology about sex roles may be misleading, and therefore fail to see a discrepancy between conforming to parental wishes and promoting social adjustment in the psychological sense that Bern (1975) uses. In clinical practice, Rekers and Lovaas may be more sensitive to these issues than their paper suggests, and may use approaches that represent a mixture of the alternatives discussed here. One of the purposes of this comment is to stimulate discussion of arguments Rekers and Lovaas have almost certainly experienced in personal contacts, but which are not apparent in their paper.


    Bern, S. L. The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 1974, 42, 155-162. Bern, S. L. Sex-role adaptability: one consequence of psychological androgyny. Journal of Personal and Social Psychology, 1975, 31, 634-643. Davison, G. C. Homosexuality: the ethical challenge. Journal of Consulting and Clinical Psychology, 1976, 44, 157-162. Eysenck, H. J. and Eysenck, S. B. G.. Manual of the Eysenck Personality Inventory, 1964. Graubard, P. S., Rosenberg, H., and Miller, M. B. An ecological approach to social deviance. In E. H. Ramp and B. L. Hopkins (Eds), A new direction for education: behavior analysis. Lawrence: University of Kansas Press, 1971, 80-101. Levy, B. The school's role in the sex-role stereotyping of girls: a feminist review of the literature. Feminist Studies, 1972, 1. Mischel, W. Sex-typing and socialization. In P. Mussen (Ed), Carmichael's manual of child psychology, 1970. Pp. 3-72. Rekers, G. A. and Lovaas, O. I. Behavioral treatment of deviant sex-role behaviors in a male child. Journal of Applied Behavior Analysis, 1974, 7, 173-190. Russell, A. and Winkler, R. Evaluation of assertive training and homosexual guidance service groups designed to improve homosexual functioning Journal of Consulting and Clinical Psychology, 1976, (in press). Begelman, M. Adjustment of homosexual and heterosexual women. British Journal of Psychiatry, 1972, 120, 477-481. Winett, R. A. and Winkler, R. C. Behavior modification in the classroom: be still, be quiet, be docile. Journal of Applied Behavior Analysis, 1972, 5, 499-504.

Reprints may be obtained from R. C. Winkler, Department of Psychology, University of Western Australia, Nedlands, W.A., 6009, Australia.

Received 14 April 1975.
(Final acceptance 20 October 1976.)
Page contents copyright © copyright holder.
Reproduced without permission of the author, for educational purposes only, according to the Fair Use doctrine.
Website copyright © 2004-2006, Kathleen Seidel. All rights reserved.