Please cite as: Brooks, Ross (2004) The desire and pursuit of the whole: Testosterone, perceived dominance, and erotic preference in men. <>.

© Ross Brooks 2004-2006. Worldwide rights reserved.

Please contact The Symposium with comments, questions, updates, and corrections.


Download in PDF format for printing.


The Desire and Pursuit of the Whole:

Testosterone, Perceived Dominance, and

Erotic Preference in Men


Ross Brooks



In nature, the yang and the yin interweave everywhere; nature is inconceivable

without the incessant and joyful intermingling of receptivity and initiative,

curiously mingled in all snail shells and oak trees, all tigers, all mountains, all bees.

Robert Bly, Iron John (1990, p. 221)


ABSTRACT. Scientific research into a proposed neuroendocrine etiology of human erotic preferences may in fact be observing another physiological phenomenon only indirectly related to eroticism. Studies demonstrate that the prenatal environment is effective in organizing men’s and possibly women’s somatic features along a non-dominant (passive) – dominant (active) continuum, an evolutionary strategy that may be designed to minimize intrasex competition and infanticide. Following the principle of complementary attraction, a model of men’s erotic preference is presented that highlights the importance of the passive – active duality. As well as accommodating psychosocial dynamics associated with dominance morphology, the model respects the totality of an individual's experiences, variables associated with societal norms, and also embraces the potential for the exercise of conscious choice in determining erotic pathways. The model offers hope of a new paradigm in erotic preference research based on the cooperation of biological and psychosocial theorists.

KEYWORDS. Human eroticism; men’s erotic preferences; prenatal androgens, perceived dominance, evolutionary psychology, erotic role.




Despite its failure at producing consistent evidence, scientific research into human erotic preferences continues to focus on the Steinach / Dörner (S/D) paradigm that implicates genetic, neural, and hormonal mechanisms as a direct causal etiology of homo- and heteroerotic desires and behaviors (Herrn, 1995). Largely based on experiments with laboratory animals, the paradigm assumes two modal pathways for the development of erotic interests; one “male” and one “female.” Predominant heteroerotic preferences are assumed to result when an individual follows the developmental pathway that is assumed to correlate with their physical gender. Predominant homoerotic preferences are assumed to result when an individual follows the pathway assumed to correlate with the opposite physical gender to their own. The model is highly deterministic leaving little room for psychosocial influences or the exercise of personal choice in the development of erotic interests. A recent exposition of the S/D model, a book titled Born Gay, stated it this way; “Modern scientific research indicates that sexual orientation is largely determined by the time of birth, partly by genetics, but more specifically by hormonal activity in the womb arising from various sources” (Wilson & Rahman, 2005, p. 10 – authors’ italics).

Although research into psychosocial influences on the formation of erotic interests has been severely sidelined by the scientific community in recent decades, the strict determinism implied by the S/D model is nonetheless considered wholly unrealistic by many researchers who, although at least prepared to consider a role for physiological systems, continue to highlight the importance of learned psychosocial, particularly peer-related, variables in the development of adult erotic preferences (e.g., Friedman & Downey, 2002; Woodson, 2002). This academic impasse between “nature” and “nurture” has characterized and marred erotic preference research since the time of Plato and Aristotle. In recent years, attempts have been made to build bridges between these approaches (Byne & Parsons, 1993; Bem, 1996; Kauth, 2000) but, so far, they have failed to break the stranglehold of the S/D paradigm. The following section will briefly review the current status of scientific research and then a suggestion will be made to break the deadlock of the ages and pave a way toward a new, dynamic paradigm in erotic preference research worthy of modern scientific enquiry.




For historical and social reasons, scientists researching the development of erotic preferences have focused their attention almost entirely on the development of homoeroticism. Although attempts are no longer made to manipulate biological variables in order to enforce a standard erotic pathway in humans, development of homoerotic desires is still largely assumed to be a transsex phenomenon caused by fault, mistake, or disturbance to standardized (usually referred to as “normal”) developmental pathways. In particular, a large amount of research has been conducted in order to identify psychological and somatic characteristics usually identified as “masculine” in women who identify as lesbian and characteristics usually identified as “feminine” in the personalities, behavior, and physiology of men who identify as gay.

In particular, these studies have focused on the neuroendocrine system and its relationship to adult homoeroticism. Scientists have long discovered that adult hormone levels do not differ on average between individuals of differing erotic preferences (Meyer-Bahlburg, 1984). Research continues into the possibility that hormone levels in the prenatal environment influence the differentiation of the fetal nervous system, directly organizing later physical, psychological, and behavioral patterns such as “masculinity,” “femininity,” “erotization,” and “sexual mentality” (Steinach & Loebel, 1940, p. 19) in a mechanical, predictive fashion (for review and references see Rahman & Wilson, 2003a).

Evidence for the assertion, however, has been patchy at best. Lines of enquiry into a human prenatal stress syndrome, an abnormal LH response, naturally occurring examples of individuals exposed to atypical levels of prenatal hormones, and observations of somatic markers assumed to correlate with intrauterine hormonal variations (so far; dermatoglyphics, 2D:4D ratios, auditory system emissions, cerebral lateralization, and penis size) have all yielded “provocative but inconsistent results” (Muscarella et al., 2001, p. 393).Positive correlations have been offset by null findings and methodological weaknesses and it remains unproven that the neuroendocrine system has a direct causal relationship to adult erotic preferences in this manner (the conclusions of Banks & Gartrell, 1995 remain largely applicable today).

By way of example, the research linking erotic preference to finger-length ratios typifies the inconsistency that characterizes the wider literature. The ratio of the index finger (2D) to the ring finger (4D) is sexually dimorphic, with men generally showing reduced ratio and women generally greater. The difference is more pronounced on the right hand and is thought to be influenced by androgen levels that a fetus is exposed to in the womb (e.g., Neave et al., 2003). Researchers have thus looked to finger-length ratios to shed some light on the relationship between the prenatal environment and adult erotic preferences. Williams et al. (2000) reported that lesbian subjects’ finger length ratios tended to be shorter when compared to heterosexual female subjects. Overall, gay and straight male subjects were not significantly different until the researchers looked at those gay male subjects who had several older brothers. When segregated by birth order, it was reported that later-born gay male subjects with two or more older brothers had a lower 2D:4D ratio than those with one or no older brothers. Gay male subjects without older brothers had a 2D:4D ratio indistinguishable from straight first born sons. The researchers reported that their findings are consistent with at least some gay-identified men being “hyper-androgenized.”

Subsequent finger-length ratio studies have produced highly ambiguous results. Three studies have reported lower finger-length ratios in gay compared to straight male subjects, and this was unrelated to birth order (Robinson & Manning, 2000; Rahman & Wilson, 2003b; Rahman, 2005). Two other studies have reported higher ratios for gay male subjects (McFadden & Shubel, 2002; Lippa, 2003). One report found no significant differences in 2D:4D ratios between gay and straight male subjects at all (Voracek et al., 2005). Manning and Robinson (2003) introduced an interesting variable of cross-cultural variations in 2D:4D ratio. They reported; “Normative values of male mean 2D:4D ratio vary between 0.94 to 1.00 across populations. However, mean 2D:4D ratio is relatively constant among male homosexuals from different populations, with a mean of 0.96 to 0.97. Therefore homosexuals appear to be more androgenised (lower 2D:4D) than heterosexuals in some populations, while in others they appear to be less androgenised (higher 2D:4D) than heterosexuals” (p. 303). They continue; “If it turns out that we are correct it is not easy to see the underlying mechanism which links a particular world-wide value of 2D:4D with homosexuality” (p. 306). The current status of this research thus remains highly confused and can be manipulated to support cases both for and against a direct neuroendocrinological etiology of erotic preference.

In summary, almost a century’s worth of research has yet to demonstrate a direct causal relationship between the neuroendocrine system and adult erotic preferences. The prenatal hormone environment appears to play a role but cannot be shown to directly govern the psychosexual development of either hetero- or homoerotic preferences. Current research remains “a product of our social values rather than of objective appraisal of the evidence” (Bancroft, 1989, p. 300), and is “riddled with inconsistencies” (Rahman & Wilson, 2003b, p. 289).




In contrast, studies have found much more conclusive evidence that the prenatal environment, particularly levels of T, organize facial and other somatic features along a non-dominant (passive) – dominant (active) continuum in male fetuses. Men’s faces are known to signal messages that are stable over time, including attractiveness as a potential mate (Barber, 1995) and dominance (Neave et al., 2003; Swaddle & Reierson, 2002; Mazur & Mueller, 1996; Mazur & Booth, 1998; Collins & Zebrowitz, 1995; Mueller & Mazur, 1997). Facial dominance is positively correlated to dominance behaviors (Mazur & Booth, 1998) and social success (Mueller & Mazur, 1997; Mazur et al., 1994). Circulating levels of T are not related to perceived dominance, suggesting that the effect reflects organizational rather than activational processes (Neave et al., 2003; Campbell et al., 1998). There are, however, indications that dominance morphology changes across the life span (Zebrowitz & Montepare, 1992).

Early exposure to higher levels of T produces dominant characteristics and fewer non-dominant characteristics whereas exposure to lower levels of T exacts the reverse (Mazur & Booth, 1998). Neave et al., (2003, p. 2167) state:

These pre- and perinatal hormone effects are regarded as organizing the architecture of the body and brain. When male testosterone increases later in life during puberty, it is thought to activate pre-existing structures; for example, in males higher androgen serum levels at puberty together with a higher androgen receptor expression at certain skeletal sites, may contribute to sex differences in facial morphology . . . . In males, a high testosterone-to-oestrogen ratio facilitates the lateral growth of the cheekbones, mandibles and chin, the forward growth of the bones of the eyebrow ridges, and the lengthening of the lower facial bone, all of which are considered masculine facial features.

Estrogen inhibits this growth, producing non-dominant characteristics and less dominant characteristics that Neave et al. describe as “high eyebrows, more gracile jaw and fuller lips.” Although these researchers use gendered terminology (e.g., “masculine” and “feminized”) to describe these somatic features, this may reflect the prevailing paradigm that unadvisedly equates non-dominant features with femininity. It may more accurate to describe the development of non-dominant features in terms delayed maturation; i.e., the retention of juvenile features into adulthood (sometimes termed neoteny – for discussion see Bromhall, 2003).

Researchers investigating human dominance effects have struggled to fit the phenomenon into a heteroerotic framework. It has been suggested that perceived dominance may be an honest indicator of male fitness as high-T can impose an immunocompetence handicap on the bearer that only the best males can carry. This suggestion is offset, however, by the discovery that dominance morphology is not correlated with attractiveness, health, or reproductive success (for discussion see Swaddle & Reierson, 2002). Mueller and Mazur (1997) therefore propose that the effects of facial dominance and attractiveness should be considered separately by statistical analysis. Neave et al. (2003, p. 2167) support this conclusion, stating that “the actual role of perceived facial dominance still remains equivocal.”

Physical markers of dominance are not just limited to facial characteristics (Dabbs, 2000). Receptors for T developed in the fetus largely govern the degree to which the body and brain will respond to T throughout life. Developmental changes in adolescence are thus orchestrated in part from what was fixed before birth and in part from an increase in sex hormones during adolescence. The prenatal environment therefore has a strong influence over the physicality of the adult since T affects growth in all parts of the body during adolescence, especially the reproductive system, thyroid gland, blood, bones, skin, skull, and brain. T gives men more muscle, along with more red blood cells to carry oxygen to the muscle. High-T creates a strong physique characterized by broad shoulders and heavy muscles in the arms and legs as well as a capacity to store fat around the stomach. T also effects voice production, deepening the voice in male adolescents. Most of these characteristics are instantly recognizable as markers of the strength, charisma, and dominance of a man. By manipulating T levels in the womb, a mother’s body is therefore influencing her son’s social performance in a way that will be experienced throughout his lifetime.

The extent to which T directly affects psychological characteristics and behavior is poorly understood and subject to much guesswork. In humans, it is most likely that fetal levels of T organize basic brain lateralization and that other (activational) effects linked to T result from psychosocial variables interacting with that basic patterning. Geschwind and Galaburda (1987) have speculated that fetal T affects the growth rate of the two hemispheres of the brain. High-T increases the rate at which the right hemisphere develops and, correspondingly, slows the rate at which the left hemisphere develops. Baron-Cohen (2003) has developed this idea and identified three basic brain types; those hardwired for empathizing more than systematizing (left-dominant – well adapted for language, communication, and social skills); those hardwired for systematizing more than empathizing (right-dominant – brains well geared for establishing dominance hierarchies and interpreting complex systems); and a balanced brain in which systematizing and empathizing are both equally strong but without specializing in either.

Although the literature on cerebral lateralization is suffused with references to male and female brains – a trend wholly attributable to the S/D paradigm – it is clear, as Baron-Cohen (2003, p. 8) states, that “your sex does not dictate your brain type”. It is more accurate to describe cerebral lateralization along a non-gender based axis such as non-dominance – dominance or empathizing – systematizing as it facilitates a value-free analysis of intrasex variation, a subject of increasing interest to scientists (e.g., Rhen & Crews, 2002). It must also be recognized that T levels are highly heritable from both parents and that physical and social conditions produce temporary changes in average levels. So too does learning, as Dabbs (2000, p. 10) states; “if we manage to sit through school and listen to our elders, the lessons we learn restrain and guide the effects of testosterone.” The organizing effects of T are, then, themselves subject to a multitude of biological and environmental influences.

Deplorably, there has been no significant research conducted into the perceived dominance effects of women. There are some indications that similar dynamics are at work for women as well as men (e.g., Brown et al., 2002; Singh et al., 1999) although these effects may be less tied to the influence of T. The dearth of information about the specific nature of women’s dominance morphology, combined with observations that women’s eroticism operates significantly differently to that of men (e.g., Peplau et al., 1998; Baumeister, 2000), means that the model proposed in this paper must currently be restricted to men.




The prenatal environment is active in organizing dominance morphology, not later erotic preferences. Dominance morphology (including cerebral lateralization) is organized partly in response to cues from the environment into which the young male is being born. The degree of likely hostility and competition between men within the immediate social group, measured by the number of previous male births and in other ways, is a crucial factor. Somatic cues that signal non-dominance / dominance in sons are therefore primarily cues to other males. Rather than fighting for social position, dominance cues facilitate the relatively peaceful establishment of social systems. Such systems create strong, harmonious, and therefore more successful social groupings and reduce the risk of implosion through infighting and infanticide. This, in turn, increases rates of survival and enables higher states of psychological and cultural evolution.

Erotic bonding based on the dynamics of social dominance strengthens these hierarchical systems. Neuronal patterning correlating to adult erotic preferences develops postnatally allowing for even greater adaptation to the demands of the social environment (thereby maximizing an individual’s chance of survival). Individual erotic preferences develop largely, but not exclusively, as psychosocial variables – experiences – interact with dominance cues, shaping and refining an undifferentiated / bipotential infantile eroticism. Many of these variables are themselves influenced by an individual’s dominance morphology and the attitudes and behaviors it provokes in others. By interweaving eroticism with social hierarchy, evolution has equipped the higher species with strong and dynamic systems to protect themselves from the deleterious effects of infighting for social position, thereby bestowing selective advantage. In short, make love, not war is a natural imperative for social animals.

Individual psychosexual differentiation largely follows the erotic directive of complementary attraction. Based on this directive, the pure form of the model predicts that strongly defined homoerotic preferences can be expected at the two extremes of the dominance continuum (i.e., among extreme non-dominant individuals and extreme dominant individuals) as they will be experienced as different and therefore attractive. Such fantasies and behaviors will be characterized by the same passive – active duality that gives rise to their definition and need not exclude the development of heteroerotic interests. However, extreme non-dominant men are more likely to develop exclusive homoerotic thoughts and desires (in our own culture, identify as gay) as, generally speaking, it is more difficult for a woman to satisfy their passive erotic interests.

Homosocial and homoerotic bonding between the two masculine extremes serves as a powerful emotional glue that binds the extreme dominant (active) males to the extreme non-dominant (passive) males. This creates a strong passive-active link system (PALS) that prevents dominance becoming violence / dictatorship and non-dominance becoming apathy / exclusion. Working in harmony, PALS ensure that the gap between the two masculine extremes is kept within relatively safe and functional parameters. This dynamic is enshrined in all the great mythological traditions of the world (e.g., Zeus and Ganymede) and is most vividly witnessed today in environments where men must live together in restricted conditions (e.g., prison).

Strongly defined heteroerotic interests will become defined where females are experienced as different and will also be subject to influence from the passive – active continuum (in both sexes). Men who are balanced for dominance (i.e., those men in the middle of the passive – active duality) are most likely to retain their infantile bipotentiality and demonstrate a high degree of versatility in erotic relations. They are, however, more likely to be molded by familial, peer, and societal pressures toward particular erotic pathways. Such developmental patterns are by no means standard and rely heavily on individual experience, gender identity, societal norms, and ecological influences as well as the ability and willingness to exercise conscious choice.

Genetic and neuroendocrine mechanisms underlie the development of dominance morphology which in turn exerts a strong influence over the crystallization of adult erotic preferences. Dominance morphology therefore provides domain specific markers from which it is possible to observe the vicissitudes of human erotic preference but variations from these basic markers should be expected and celebrated. The inherent plasticity of the model is, in fact, its guiding principle enabling adaptation to the highly volatile environments in which our primate and hominid ancestors evolved and which still provides adaptive advantages in environments that men find themselves living in today. The plasticity of the model accounts for the suggestive but elusive relationship between the neuroendocrine system and adult erotic preferences noted previously.




It is the purpose of this paper to forward a new perspective on a large and complex topic. An extensive analysis of all its implications would be neither practical nor desirable at this point. It is, however, considered appropriate to allude to two important issues of greatest pertinence to the proposed model; the question of evidence, and the validity of the theory of complementary attraction.




It has already been demonstrated that at least one body of research used to support a direct prenatal endocrine etiology of adult erotic preferences, finger-length ratios, is insufficient in demonstrating the premise. A closer look at the wider context of this body of research suggests that researchers in this field have in fact been indirectly observing dominance effects – effects that are only subsequently and partially influential in the development of erotic preferences.

Research into the relationship between skeletal structure and eroticism was originally reported by Schlegel in the mid-1970s (reported by Eysenck & Wilson, 1979) who implicated prenatal androgens in determining the shape of the pelvis. Generally speaking, females and males have different shaped pelvises; males tend to have a funnel-shaped pelvis that is narrow at the bottom while females tend to have a tube-shaped pelvis that is relatively broad at the bottom. The shape of the pelvis differs from person to person within each sex. Thus there are many males with tube-type pelvises and many women with funnel-type pelvises. Schlegel reported that men and women with funnel-type pelvises tended to behave in a similar manner; the shape of their pelvis correlated with leadership, an active sexual role, dominance, and preference for younger sex partners. Tube-type pelvises, in men and women alike, correlated with empathy, suggestibility, and compliance, as well as preference for older sex partners. In other words, dominance-related behavior in both sexes seems to be largely determined by the same hormonal factors (prenatal androgens) that produced the skeletal features of the pelvis. Schlegel found that the same phenomenon held true for cows too. He reported that men and women concordant for pelvic shape (i.e., men with funnel-type pelvis, women with tube-type pelvis) were much less likely to be divorced than men and women with discordant for pelvic shape. Schlegel also reported that “homosexual males” tended to have the tube-type pelvises although few details are provided.

Building on Schlegel’s research, Wilson (1983) determined that another, more readily accessible, skeletal feature – finger-length ratios – could be used to assess the correlation between body type and what he termed “masculinity-femininity social behaviour” in women (and, therefore, the degree to which social behavior is determined by prenatal hormones). This hypothesis was confirmed by the finding that women with shorter finger ratios were more assertive and competitive than other women.

As previously stated, some researchers have subsequently looked to finger-length ratios as a way of finding a direct causal link between the prenatal environment and adult erotic preferences, research that Wilson recognizes as producing “somewhat confusing results” (Wilson & Rahman, 2005, p. 79). In contrast to the ambiguity of the findings of this group of studies, another group of studies have continued to demonstrate a consistent relationship between finger-length ratios and dominance-related traits. For example, 2D:4D ratios predict male ability in sport. Compared to men with a high 2D:4D ratio, men with a low ratio are reported as being more competitive and attaining higher success in a range of sports. Manning and Taylor (2001) report that professional footballers have lower 2D:4D ratios than controls while Manning (2002, p. 446) goes so far too suggest that the 2D:4D ratio “could be used to help identify young skiers who have potential to reach high levels of performance.” Neave et al. (2003) also provide empirical support the connection between 2D:4D and male assertiveness and dominance, reporting that men with lower 2D:4D ratios are perceived as being more “masculine” and dominant by female observers. Bailey and Hurd (2005) reported significant correlation between 2D:4D and physical aggression in men.

In summary, the only uncontested evidence to emerge from this branch of enquiry is that skeletal structure is strongly correlated to levels of assertiveness, competitiveness, and physical aggression – all dominance-related traits. There are suggestions that homoerotic preferences are following this basic blueprint, appearing at both the extreme dominant / competitive and extreme non-dominant / non-competitive ends of the spectrum. However, the pattern is not as robust with a great many variations to the pattern. The introduction of cross-cultural variation suggests that psychosocial variables are playing a large part in the development of erotic preferences but not in dominance-related traits.

Other bodies of research that have suggested a relationship between prenatal hormones and adult erotic preferences also betray a much stronger relationship between the prenatal environment and dominance-related traits. For example, in a classic study Yalom et al. (1973) reported on twenty 16-year-old sons of diabetic mothers who had received estrogen or progesterone during pregnancy. The researchers reported that there was a trend (though not statistically significant) for the experimental subjects to have had less heteroerotic experience and fewer “masculine” interests than control groups. Of the ten subjects who were evaluated as being the least developed heteroerotically, seven were experimental and three were controls. The only subject who had actually engaged in significant homoerotic activity was a control. Of the five subjects who had engaged in repeated cross-dressing, three were controls and two were experimentals. Crucially, much more decisive results were evident when researchers measured for aggression and assertion (i.e., dominance-related traits), with the experimental subjects demonstrated as being significantly less aggressive-assertive than the controls. The finding on aggressivity was “clear-cut”, “striking”, and occurred “regardless of the instrument or perspective employed”. It is suggested here that dominance morphology was organized by estrogen effects (known to inhibit T) in utero and that this had important implications for the development of erotic preferences postnatally but did not pattern them per se.

The strong implication from both these lines of enquiry is that erotic preferences largely follow somatic templates established in utero but are not, in and of themselves, patterned prenatally. Many other bodies of research that purport to find somatic differences between gay and straight subjects can be shown to simply be identifying differences between non-dominant and dominant individuals. Further evidence would undoubtedly ensue from more focused studies on dominance morphology and erotic preferences. It is predicted that in both bodies of research, dominance and non-dominance will continue to emerge as more robust categories than the more precariously constructed categories of heteroerotic (“straight”) and homoerotic (“gay”) can ever afford in scientific literature (a situation already identified by Muscarella, 1999, p. 8-9). Studies will also benefit from controlling for the erotic role preferences of subjects, a move that is currently being made in studies of women’s eroticism with fascinating results (e.g., Csathó et al., 2003). So too must scientists in this field stop kidding themselves that bisexuality is rare and therefore ignorable (e.g., Wilson & Rahman, 2005), a wholly delusional premise.


The Principle of Complementary Attraction

Many observers have noted that when it comes to matters of erotic interest, people tend to seek out qualities that they perceive themselves as lacking. The idea has endured throughout history. Plato’s famous account of the origins of erotic desire in The Symposium is founded in the principles of complementary souls uniting in order to satisfy a fundamental human drive that he termed “the desire and pursuit of the whole” (Plato, 1951, p. 64). A wider look at Greek mythology, indeed all the great ancient mythological traditions, finds love expressed as a union of opposites, perhaps most strikingly represented in the figure of Hermaphrodite (the double-sexed prodigy of Hermes and Aphrodite). In India, the archetype is represented by the Hindu couple Purusha-Prakriti (“as big as a man and woman enclasped”); in Chinese Taoism as the Yin and Yang symbol, the two principles held in balance. As in Plato’s account, the union of opposites can equally occur within a sex. Tales of older, dominant males seeking union with younger, non-dominant males suffuse all the ancient traditions. Even the almighty Zeus became enchanted by the beautiful youth Ganymede and whisked him up to Mount Olympus to become his cup bearer.

Before the hegemony of the S/D model of the development of erotic preference, Karl Heinrich Ulrichs, the first modern theorist of erotic attraction, developed a model of eroticism that stressed the importance of polarity in the erotic interests of Urnings (gay men). Citing Schopenhauer’s assertion that “people love what they themselves are lacking”, Ulrichs wrote in 1864; "Our lover is every bit our complement, our “acid” or our “alkali.”” (Ulrichs, 1994, p. 68). He declared that a Mannling (a gay man with active desires) naturally demands a Weibling (a gay man with passive desires) for a sexual partner. Ulrichs’ ideas may have been more fully developed had it not been for the Nazi’s persecution of the German gay rights movement and the ascendancy of a desire in the wider scientific community to reduce homoeroticism down to a single biological phenomenon in order to eliminate it (e.g., Fitzherbert, 1967).

In more recent times, the idea of complementary attraction as a primary motivation in human erotic attraction has re-emerged in various guises. Tripp (1975, p. 60) states; “It is the gap, the optimal distance between partners, that pervades all sexual attractions and without which they cannot exist. The fact that sexual motivations founder or simply prove impossible once a fairly critical distance between partners is violated in either direction turns out to be the key to many a locked-up secret in human relations. . . . Sexual zest does not arise from the comforts of similarity and agreement. Even after it finds its natural origins it is put to sleep by the music of high accord and, in that atmosphere, local injections of novelty and newness do not arouse it for long. With no outside source of resistance, sexual attraction requires the sharper zap of clash or of foreignness – a note of antagonism set in a disparity of outlook or of rank.”

Stoller (1979, p. 18) includes mystery, risk, and hostility as the generators and enhancers of sexual excitement. He states; “Buried in excitement may be another awareness, that of contrast. Between the known and the unknown, the familiar and the mysterious, the overt and the hidden (these pairs are not quite synonymous), there can be a dynamic of pseudo-risk (controlled surprise) that drives the process to the level sensed as excitement.” Bell (1982, p. 2) also notes that “persons perceived as essentially different from ourselves become the chief candidates for our early romantic and, later, erotic investments.”

Feierman (1990) has suggested that patterns of erotic differentiation of the brain can be associated with particular patterns of preferred partner characteristics based on evaluation of features of the target relative to self. Although Feierman bases his observations on the SHD-based “feminized – masculinized” model of brain differentiation, the idea can be equally applied to non-dominant – dominant / juvenile – mature models of erotic preference patterning. Thus, in their pure forms, dominant men are more likely to seek to non-dominant erotic objects and vice versa; and older men are more likely to seek out younger erotic objects and vice versa.

Bem (1996; 1998; 2000) also presents a variation of complementary attraction as a basis for a general model of human erotic preferences. In common with the current model, Bem sees early neurohormonal patterning as governing basic dispositions only (he cites aggression and activity). Subsequent erotic development is dependent upon psychosocial influences that interface with these temperaments under a guiding principle of “exotic becomes erotic” (EBE). In common with Feierman, however, Bem failed to recognize the importance of intrasex dynamics and could only therefore envisage a single modal path of homoerotic attraction whereby gender non-conformity was a “causal antecedent” of homoeroticism.

The current model differs essentially from Bem’s in the emphasis on multitudinous and non-exclusive modal paths in the development of erotic desires, in the identification of dominance morphology as the domain specific markers observed in erotic preference research, in highlighting the importance of accounting for active, versatile, and passive erotic roles, and in the rejection of rhetoric that dismisses non-dominant characteristics in men as aspects of gender non-conformity. In other ways, the current model follows through on Bem’s theory. For example, Bem failed to recognize that under the principle of EBE boys who are not sidelined for having a non-competitive or non-aggressive nature will become attracted to those that are. The current model suggests that this a primary motivation for the development of strongly defined active homoerotic thoughts and desires. Bem’s study does contain a useful analysis of some psychological processes that transform psychological needs into erotic preferences. These include the extrinsic arousal effect, the opponent process, and imprinting (a further discussion here is not deemed necessary). Also, Bem’s observations of “gender-atypical” boys may remain an applicable model for the development of the erotic interests of boys who later identify as transsexual.

Bem (1998) and Peplau et al. (1998) locked horns on evidence for the complementary attraction / EBE principle. Subsequent studies have supported Bem on this point although the picture that is emerging is a complex one. Although no empirical studies of men’s preferred partner characteristics have accounted for erotic role preference, they have nonetheless reported evidence that supports complementary attraction as an erotic directive for men. For example, studies report that men who identify as gay are significantly more attracted to “masculine” looking men than to men with “non-masculine” appearance (Bailey et al., 1997; Mealey, 1997) and to men more “masculine” than themselves in some respects but not others (Muscarella, 2002; Muscarella et al., 2004). Muscarella (2002; following the lead of Feierman, 1990) comes to the intriguing conclusion that erotic preferences change during the lifespan in response to the developing “masculinization” (increased dominance rating? maturity?) of the brain. Muscarella et al. (2004) similarly reported a complex relationship between erotic self-identity and preferred partner characteristics (including a “meaningful difference” between self-identified butch and femme lesbians). The results supported a view of brain differentiation forwarded by Kauth (2000) and Woodson and Gorski (2000) whereby any brain differentiation involved in erotic preferences does not involve the manifestation of discrete categories but rather a continuum that will lead to a variety of patterns of erotic expressions and, by extension, preferred partner characteristics. Dominance morphology offers just such a continuum, particularly as it has been suggested that dominance morphology changes through the lifespan (Zebrowitz & Montepare, 1992).

Rind (2000) also supports the principle of complementary attraction in reporting that adolescent male subjects who had sex with adult men were strongly attracted to older males and sought them out or took advantage of invitations for sex. Some subsequent studies have supported the finding (Muscarella, 2002; Muscarella et al., 2004) but other studies have reported that “gay men” prefer younger partners relative to self (e.g., Silverthorne & Quinsey, 2000; Hayes, 1995). The key to understanding this apparent discrepancy lies, of course, in the age of the participants. For example, the average age of Muscarella et al.’s (2004) sample of gay-identified men was 20.2 years. They preferred partners “slightly but significantly older than self.” The average age of gay-identified participants in Hayes’ (1995) study was 31.3 years. They showed a preference for younger partners. As well as suggesting that these scientists should get together and establish a lucrative dating service, it is also overwhelmingly clear (from all the studies on preferred age of partner) that a principle of complementary attraction is robust for age. Intriguingly, Silverthorne and Quinsey further report that age preference and sex preference operate independently. It may therefore be more applicable to refer to the principle of complementary attraction as a set of directives rather than as a single entity. Much more focused research needs to be conducted in this area before further conclusions can be drawn. A longitudinal study that charts individual erotic preferences in different contexts and over the life span would be enlightening.

There are no absolutes when it comes to human eroticism. The principle of complementary attraction is only one set of forces that shape erotic desire that is discernable by looking at a wide variety of human experiences. Many individual histories will counter the trend for many different reasons and this diversity should be appreciated rather than negated. Bem rightly points out that there are many partnerships (homo and hetero) that are apparently based on similarity rather than opposition. Although it is impossible for a scientist to truly know the secrets of the bedroom, other behavioral templates (especially those involving societal expectations and the exercise of free will) are clearly at work and which interface with dominance effects to create a high degree of variety in erotic attraction.

The observations made in the current paper strongly support behavioral scaling as a model of sexual preference. Behavioral scaling (a.k.a. behavioral plasticity) is exhibited in many behaviors by many species (Barash, 1982) and has been suggested as a model for mammalian erotic preferences by Muscarella et al. (2001), Bagemihl (1999, pp. 250-251) and Eaton (1978, pp. 55-56). Behavioral scaling does not assume a dichotomous relationship between homo- and heteroerotic preference, rather that expression of either or both preferences in any given individual will depend on a variety of circumstantial factors, especially the degree to which each preference has been developed / allowed to develop. It should be understood that people differ in their ability to express behavioral plasticity and this has profound implications for the exercise of sexual versatility. Several factors inhibit an individual’s ability to express sexual variety; familial and societal expectations being the most powerful. Repetitive thinking patterns, endemic in modern bureaucratic societies, create strong monistic neural patterns that mean only aspects of our brains’ potential are developed rather than its totality. Such a process has given rise to the modern myth of “sexual orientation” (De Cecco & Parker, 1995; Bem, 1996), a concept that is alien, indeed laughable, to many indigenous cultures and other societies whose diverse experiences of life (and particularly of the natural world) have led them to understand that that human erotic responses are essentially bipotential and dynamic.

Finally, the current model finds harmony between the world of the evolutionary biologist and the world of the psychologist and allows for a great deal of variation and plasticity that is a necessary part of both perspectives of life. Most importantly, it finds no friction between “nature” and “nurture” but rather finds a creative, interactive relationship between the two approaches – something that is not only instinctively satisfying but is in keeping with the highly plastic nature of the human brain (e.g., Breedlove & Jordan, 2001) and of human eroticism. It is time that scientists involved in erotic preference research embraced this reality.





Studies that have been seeking to establish a neuroendocrine etiology of human erotic preferences have been indirectly observing the effects of intrasex variation along a non-dominant (passive) – dominant (active) continuum. There is strong pre-existing evidence to show that this is true for men and it appears highly likely that a similar situation exists for women. Although men’s erotic preferences are inclined to follow patterns inherent in this continuum, particularly in the dynamic identified as PALS, this is by no means a given. Many other variables come into play and influence erotic choices in concordance with different natural templates and also, in some individuals, through conscious awareness. A new paradigm in erotic preference research is called for that abandons the labeling of male homoeroticism as having essentially feminine or gender-atypical attributes and embraces the fact that masculinity is a multi-dimensional phenomenon that does not just include the homoerotic but owes its survival to it.




Bagemihl, B. (1999) Biological Exuberance: Animal Homosexuality and Natural Diversity, London: Profile Books.

Bailey, A. A., & Hurd, P. L. (2005). Finger length ratio (2D:4D) correlates with physical aggression in men but not in women. Biological Psychology, 68, 215-222. [Abstract]

Bailey, J. M., Kim, P. Y., Hills, A., & Linsenmeier, J. A. (1997). Butch, femme, or straight acting? Partner preferences of gay men and lesbians. Journal of Personality and Social Psychology, 73, 960-973. [Abstract]

Bancroft, J. (1989). Human Sexuality and its Problems. (2nd ed.). Edinburgh: Churchill Livingstone.

Banks, A., & Gartrell, N. K. (1995). Hormones and sexual orientation: A questionable link. Co-published simultaneously in Journal of Homosexuality, 28, 247-268 [Abstract]; and J. P. De Cecco & D. A. Parker (Eds.) Sex, Cells, and Same-Sex Desire: The Biology of Sexual Preference (pp. 247-68). New York: Haworth Press .

Barash, D. P. (1982) Sociobiology and Behavior, 2nd ed., New York: Elsevier.

Barber, N. (1995). The evolutionary psychology of physical attractiveness: Sexual selection and human morphology. Ethology and Sociobiology, 16, 395-424.

Baron-Cohen, S. (2003). The Essential Difference. London: Allen Lane.

Baumeister, R. F. (2000). Gender differences in erotic plasticity: The female sex drive as socially flexible and responsive. Psychological Bulletin, 126, 347-374. [Abstract]

Bell, A. P. (1982). Sexual preference: A postscript. Siecus Report, 11, 1-3.

Bem, D. J. (1996). Exotic becomes erotic: A developmental theory of sexual orientation. Psychological Review, 103, 320-335. [Available online - n.b. PDF document]

Bem, D. J. (1998). Is EBE theory supported by the evidence? Is it androcentric? A reply to Peplau et al. (1998). Psychological Review, 105, 395-398. [Abstract]

Bem, D. J. (2000). Exotic becomes erotic: Interpreting the biological correlates of sexual orientation. Archives of Sexual Behavior, 29, 531-548. [Abstract]

Breedlove, S. M., & Jordan, C. L. (2001). The increasingly plastic, hormone-responsive adult brain. Proceedings of the National Academy of Sciences USA, 98, 2956-2957. [Available online]

Bromhall, C. (2003). The Eternal Child: How Evolution has Made Children of Us All. London: Ebury Press.

Brown, W. M., Finn, C. J., Cooke, B. M., & Breedlove, S. M. (2002). Differences in finger length ratios between self-identified “butch” and “femme” lesbians. Archives of Sexual Behavior, 31, 123-127. [Abstract]

Byne, W., & Parsons, B. (1993). Human sexual orientation: The biologic theories reappraised. Archives of General Psychiatry, 50, 228-239. [Abstract]

Campbell, A., Muncer, S., & Odber, J. (1998). Primacy of organizing effects of testosterone. Behavioral and Brain Sciences, 21, 365.

Collins, M., & Zebrowitz, L. (1995). The contributions of appearance to occupational outcomes in civilian and military settings. Journal of Applied Social Psychology, 25, 29-163.

Csathó, A., Osváth, A., Bicsák, E., Karádi, K., Manning, J., & Kállai, J. (2003). Sex role identity related to the ratio of second to fourth digit length in women. Biological Psychology, 62, 147-156. [Abstract]

Dabbs, J. M. (2000). Heroes, Rogues, and Lovers: Testosterone and Behavior. New York: McGraw-Hill.

De Cecco, J. P. & Parker, D. A. (1995). The biology of homosexuality: Sexual orientation or sexual preference? Co-published simultaneously in Journal of Homosexuality, 28, 1-27 [Abstract]; and J. P. De Cecco & D. A. Parker (Eds.) Sex, Cells, and Same-Sex Desire: The Biology of Sexual Preference (pp. 1-27). New York: Haworth Press.

Eaton, G. G. (1978). Longitudinal studies of sexual behavior in the Oregon troop of Japanese macaques. In T. E. McGill, D. A. Dewsbury, & B. D. Sachs (Eds.) Sex and Behavior: Status and Prospectus (pp. 35-59). New York: Plenum Press.

Eysenck, H. J., & Wilson, G. (1979). The Psychology of Sex. London: J. M. Dent.

Feierman, J. R. (1990). A biosocial overview of adult human sexual behavior with children and adolescents. In J. R. Feierman (Ed.), Pedophilia: Biosocial dimensions (pp. 8-68). New York: Springer-Verlag.

Fitzherbert, J. (1967). Some suggestions concerning homosexuality. British Journal of Psychiatry, 113, 446.

Friedman, R. C., & Downey, J. I. (2002). Sexual Orientation and Psychoanalysis: Sexual Science and Clinical Practice. New York: Columbia University Press.

Geschwind, N., & Galaburda, A. M. (1987). Cerebral Lateralization. Cambridge, Mass: MIT Press.

Hayes, A. F. (1995). Age preferences for same- and opposite-sex partners. Journal of Social Psychology, 135, 125-133. [Abstract]

Herrn, R. (1995). On the history of biological theories of homosexuality. Co-published simultaneously in Journal of Homosexuality, 28, 31-56 [Abstract]; and J. P. De Cecco & D. A. Parker (Eds.) Sex, Cells, and Same-Sex Desire: The Biology of Sexual Preference (pp. 31-56). New York: Haworth Press.

Kauth, M. R. (2000). True Nature: A Theory of Sexual Attraction. New York: Kluwer Academic.

Lippa, R. A. (2003). Are 2D:4D finger-length ratios related to sexual orientation? Yes for men, no for women. Journal of Personality and Social Psychology, 85, 179-188. [Abstract]

Manning, J. T. (2002). The ratio of 2nd to 4th digit length and performance in skiing. Journal of Sports Medicine and Physical Fitness, 42, 446-450. [Abstract]

Manning, J. T., & Robinson, S. J. (2003). 2nd to 4th digit ratio and a universal mean for prenatal testosterone in homosexual men. Medical Hypotheses, 61, 303-306. [Abstract]

Manning, J. T., & Taylor, R. P. (2001). Second to fourth digit ratio and male ability in sport: Implications for sexual selection in humans. Evolution and Human Behavior, 22, 61-69. [Abstract]

Mazur, A., & Booth, A. (1998). Testosterone and dominance in men. Behavioral and Brain Sciences, 21, 353-363. [Abstract]

Mazur, A., Halpern, C., & Udry, J. (1994). Dominant looking male teenagers copulate earlier. Ethology and Sociobiology, 15, 87-94.

Mazur, A., & Mueller, U. (1996). Facial dominance. In A. Somit, & S. Peterson (Eds.), Research in Biopolitics Volume 4 (pp. 99-111). London: JAI Press. [Available online]

McFadden, D., & Shubel, E. (2002). Relative lengths of fingers and toes in human males and females. Hormones and Behavior, 42, 492-500. [Abstract]

Mealey, L. (1997). Bulking up: The roles of sex and sexual orientation on attempts to manipulate physical attractiveness. Journal of Sex Research, 34, 223-228. [Available online]

Meyer-Bahlburg, H. F. L. (1984). Psychoendocrine research on sexual orientation. Current status and future options. Progress in Brain Research, 61, 375-398.

Mueller, U., & Mazur, A. (1997). Facial dominance in Homo sapiens as honest signal of male quality. Behavioral Ecology, 8, 569-579.

Muscarella, F. (1999). The homoerotic behavior that never evolved. Journal of Homosexuality, 37, 1-18. [Abstract]

Muscarella, F. (2000). The evolution of homoerotic behavior in humans. Journal of Homosexuality, 40, 51-77. [Abstract]

Muscarella, F. (2002). Preferred partner characteristics in homosexual men in relation to speculated patterns of brain differentiation. Neuroendocrinology Letters, 23, 299-302. [Abstract]

Muscarella, F., Elias, V. A., & Szuchman, L. T. (2004). Brain differentiation and preferred partner characteristics in heterosexual and homosexual men and women. Neuroendocrinology Letters, 25, 297-301. [Abstract]

Muscarella, F., Fink, B., Grammer, K., & Kirk-Smith, M. (2001). Homosexual orientation in males: Evolutionary and ethological aspects. Neuroendocrinology Letters, 22, 393-400. [Abstract]

Neave, N., Laing, S., Fink, B., & Manning, J. T. (2003). Second to fourth digit ratio, testosterone and perceived male dominance. Proceedings of the Royal Society of London Series B, 270, 2167-2172. [Abstract]

Peplau, L. A., Garnets, L. D., Spalding, L. R., Conley, T. D., & Veniegas, R. C. (1998). A critique of Bem’s “exotic becomes erotic” theory of sexual orientation. Psychological Review, 105, 387-394. [Abstract]

Plato (1951). The Symposium. Trans, W. Hamilton. Harmondsworth, Middlesex: Penguin.

Rahman, Q. (2005). Fluctuating asymmetry, second to fourth finger length ratios and human sexual orientation. Psychoneuroendocrinology, 30, 382-391. [Abstract]

Rahman, Q., & Wilson, G. D. (2003a). Born gay? The psychobiology of human sexual orientation. Personality and Individual Differences, 34, 1337-1382.

Rahman, Q., & Wilson, G. D. (2003b). Sexual orientation and the 2nd to 4th finger length ratio: Evidence for organising effects of sex hormones or developmental instability? Psychoneuroendocrinology, 28, 288-303. [Abstract]

Rhen, T., & Crews, D. (2002). Variation in reproductive behaviour within a sex: Neural systems and endocrine activation. Journal of Neuroendocrinology, 14, 517-531. [Abstract]

Rind, B. (2001). Gay and bisexual adolescent boys’ sexual experiences with men: An empirical examination of psychological correlates in a nonclinical sample. Archives of Sexual Behavior, 30, 345-368. [Abstract]

Robinson, S. J., & Manning, J. T. (2000). The ratio of 2nd to 4th digit length and male homosexuality. Evolution and Human Behavior, 21, 333-345. [Abstract]

Silverthorne, Z. A., & Quinsey, V. L. (2000). Sexual partner age preferences of homosexual and heterosexual men and women. Archives of Sexual Behavior, 29, 67-76. [Abstract]

Singh, D., Vidaurri, M., Zambarano, R. J., & Dabbs, J. M., Jr., (1999). Lesbian erotic role identification: Behavioral, morphological, and hormonal correlates. Journal of Personality and Social Psychology, 76, 1035-1049. [Abstract]

Steinach, E., & Loebel, J. (1940). Sex and life. London: Faber and Faber.

Stoller, R. J. (1979). Sexual Excitement: Dynamics of Erotic Life. New York: Pantheon.

Swaddle, J. P., & Reierson, G. W. (2002). Testosterone increases perceived dominance but not attractiveness in human males. Proceedings of the Royal Society of London Series B, 269, 2285-2289. [Abstract]

Tripp, C. A. (1975). The Homosexual Matrix. New York: McGraw-Hill.

Ulrichs, K. H. (1994). The Riddle of “Man-Manly” Love: The Pioneering Work on Male Homosexuality. Trans, Lombardi-Nash. Volume 1. Buffalo: Prometheus Books.

Voracek, M., Manning, J. T., & Ponocny, I. (2005). Digit ratio (2D:4D) in homosexual and heterosexual men from Austria. Archives of Sexual Behavior, 34, 335-340. [Abstract]

Williams, T. J., Pepitone, M. E., Christensen, S. E., Cooke, B. M., Huberman, A. D., Breedlove, N. J., Breedlove, T. J., Jordan, C. L., & Breedlove, S. M. (2000). Finger-length ratios and sexual orientation. Nature, 404, 455-456.

Wilson, G. D. (1983). Finger-length as an index of assertiveness in women. Personality and Individual Differences, 4, 111-112.

Wilson, G., & Rahman, Q. (2005). Born Gay: The Psychobiology of Sex Orientation. London: Peter Owen.

Woodson, J. C. (2002). Including ‘learned sexuality’ in the organization of sexual behavior. Neuroscience and Biobehavioral Reviews, 26, 69-80. [Abstract]

Woodson, J. C., & Gorski, R. A. (2000). Structural sex differences in the mammalian brain: Reconsidering the male / female dichotomy. In A. Matsumoto (Ed.), Sexual Differentiation of the Brain (pp.229-255). Boca Raton: CRC Press.

Yalom, I. D., Green, R., & Fisk, N. (1973). Prenatal exposure to female hormones: Effect on psychosexual development in boys. Archives of General Psychiatry, 28, 554-561.

Zebrowitz, L. A., & Montepare, J. (1992). Impressions of baby-faceness and attractiveness across the life span. Developmental Psychology, 28, 1143-1152.