Debate regarding the origins or even existence of sex differences began with Darwin's (Reference Darwin1871/1901) seminal contribution and continues to this day. Denials of so-called biological influences on sex differences are less common than they once were, but arguments that such influences are trivial in relation to social-psychological ones are common (Hyde Reference Hyde2005). Evaluations of the relative influence of these mechanisms often pit “evolutionary” against “social learning” explanations (e.g., Wood & Eagly Reference Wood and Eagly2002). Archer proposes that, “the magnitude and nature of sex differences in aggression, their development, causation, and variability, can be better explained by sexual selection than by the alternative biosocial version of social role theory” (target article, Abstract; emphasis added). We argue that Archer's review, along with many previous contributions to this debate, assume, either implicitly or explicitly, that sexual selection and social learning are alternative explanations – but in fact, they are not necessarily so.

Progress in our understanding of evolutionary and social learning influences on expressed sex differences is hampered by mutually exclusive contrasts of these classes of theories. By focusing on proximate mechanisms and implicitly assuming these are alternatives to an ultimate mechanism, social learning researchers prevent themselves from integrating ultimate level influences on human behavior, including the capacity to socially learn, into their models. Understanding evolutionary influences on social learning can inform evolutionary and social learning researchers alike (e.g., Öhman & Mineka Reference Öhman and Mineka2001). Furthermore, evolutionary researchers who view social learning as an alternative to evolutionary theory might be missing many nuances in the ways in which evolved biases can be expressed in our species, and the evolved mechanisms that enable this variation in expression. We agree with many of Archer's concerns about the social roles model of sex differences in intrasexual aggression, and agree that sexual selection provides a very powerful and parsimonious explanation, and that the social roles model struggles on many dimensions. However, we ask Archer and others to reframe these arguments in terms of explicitly stated ultimate and proximate mechanisms. Sexual selection is necessary, in our view, for a complete understanding of the sex differences in intrasexual aggression but does not provide sufficient explanation for the variation in how men's competitive dominance- striving and behavioral aggression is expressed.

Intrasexual, male-male competition is found throughout the world, but the ways in which it is expressed can differ substantively from one culture or historical period to the next. Irons' (Reference Irons, Chagnon and Irons1979) concept of cultural success allows us to understand how ecology, cultural history, and current conditions influence how men express an evolved desire for status vis-à-vis other men. Pastoral raiders who steal another tribe's cattle to pay bride price and Wall Street raiders who seek hostile takeovers of competitor's companies may seem different on the surface, but they are not: Each of these activities is an expression of men's desire for control of the resources that affect their reproductive prospects and general well being in their culture. A Wall Street raider does not, of course, need that extra $10 million to attract a bride or live well, but as long as there are other raiders who make more than he does, our ambitious raider will continue the struggle.

This said, we agree with Archer, that male-on-male behavioral aggression is a manifestation of our evolutionary history, and reflects a motivation to achieve social dominance and cultural status at a proximate level. But even the clearest indicators of an evolutionary history of male-male competition – the sex differences in physical size, other physical traits, and behavioral aggression – are expressed in more ways than are found in other species (Geary Reference Geary1998).

Because they represent different levels of analysis, different types of data would be required to falsify hypotheses based on social learning and sexual selection. To falsify a social learning model, one would need to assess the proposed proximate mechanisms, not contrast the model with one that focuses on ultimate mechanisms. As one example, boys and girls who were not exposed to their respective “social roles” should not be as sex-typed as their same-sex peers who were exposed to these roles. One type of evidence comes from children of parents who discourage sex-typing. These children have less sex-typed explicit beliefs about sex roles, in keeping with a social learning component, but have the same toy and play preferences as other children, inconsistent with a causal link between this knowledge and behavioral sex differences (Weisner & Wilson-Mitchell Reference Weisner and Wilson-Mitchell1990). Male-typical behaviors in biological males raised as girls (Colapinto Reference Colapinto2001; Reiner & Gearhart Reference Reiner and Gearhart2004) are especially difficult to reconcile with a strict social learning model of gender development.

These results and others (e.g., Berenbaum & Hines Reference Berenbaum and Hines1992) suggest sex-typed activities are influenced by prenatal exposure to androgens, a proximate mechanism in the expression of sexually selected traits. Boys' attraction to karate and baseball are consistent with male-male competition, but the fact that they are culturally specific variations of one-on-one and coalitional male-male competition suggests some forms of proximate social leaning mechanisms are operating. Like Archer, we do not believe these mechanisms are the same as those identified in the social roles model. Rather, the activities that capture children's attention and that they wish to engage in, or not, are influenced by prenatal exposure to androgens, but the specifics of these activities (e.g., ice hockey) depend on exposure and the opportunity the activity affords for the expression of physical and social dominance and the formation of male coalitions (Geary et al. Reference Geary, Byrd-Craven, Hoard, Vigil and Numtee2003).

In short, we believe that Archer is correct in his conclusions that male-male aggression is well explained by sexual selection and poorly explained by social roles. Our point is that by framing the argument in terms of evolutionary mechanisms versus social leaning mechanisms, Archer and many others miss the opportunity to integrate these different levels of explanation. What are the proximate attentional, cognitive, motivational, and social learning mechanisms that enable boys and men to engage in sexually selected intrasexual competition in so many creative and varied ways?