Archer's review is an important addition to the continuing debate around sex differences in human behavior, generally, and aggression, more specifically. Concerning the wider debate, a recent review (Hyde Reference Hyde2005) minimized the importance of sex differences in human behavior. One dimension of that argument was based on tallying the number of studies showing sex differences in relation to those not showing differences. That method presents an interesting issue: Do we judge the importance of differences based on the number of differences tallied across all observed behavior and traits, or should judgments be guided by testing for specific differences specified by a theory where differences between certain behaviors and traits are more important than others? My bias is with the latter approach.

Archer also takes this tack by evaluating the extant literature in relation to hypotheses for sex differences in aggression generated by Darwin's (Reference Darwin1871/1901) sexual selection theory and social role theory (Eagly Reference Eagly1997). This is a debate with a lively and healthy history; the very sort of debate that helps advance science. Focus on the degree to which aggression is intersexual, or primarily directed on members of the other sex, specifically males aggressing against females, or intrasexual, male-male or female- female, is crucial to the debate. Traditional versions of sexual selection theory (Darwin Reference Darwin1871/1901) as well as revised versions (e.g., Clutton-Brock Reference Clutton-Brock2009) posit that males' and females' competition should be primarily intrasexual; and that males, more than females, should compete with each other for mates using direct aggression, and females compete less directly with each other for resources. That within-sex aggression is greater than that between sexes, even among preschoolers (Pellegrini et al. Reference Pellegrini, Roseth, Mliner, Bohn, Van Ryzin, Vance, Cheatham and Tarullo2007b) and middle school youngsters (Pellegrini & Long Reference Pellegrini and Long2003), and among adults, as shown in this review, highlights the robustness of the finding.

Recently scholars have begun to recognize the importance of females' intrasexual competition (e.g., Campbell Reference Campbell1999), and this has lead to the reformulation of sexual selection theory (Carranza, Reference Carranza2009; Clutton-Brock Reference Clutton-Brock2009). In contrast to males, females' intrasexual competition should be aimed at accessing and maintaining resources related to raising and provisioning offspring and should result in the selection of a phenotype that maximizes social support, in the forms of alliance formation, manipulation, and indirect aggression (Hrdy Reference Hrdy, Hewlett and Lamb2005).

An important aspect of this reformulation is that the expression of any behavior is moderated by the ecological niche in which individuals are embedded, a position consistent with both Darwin (Reference Darwin1871/1901) and his predecessor John Hunter (Hunter Reference Hunter1780; cf. Clutton-Brock Reference Clutton-Brock2009). This is not to say that phylogenetic history is irrelevant. Instead, this history probably canalizes (Hinde & Stevenson-Hinde Reference Hinde and Stevenson-Hinde1973) the behaviors and strategies to be learned (during ontogeny) in the context of social groupings (Crook Reference Crook, Standen and Foley1989). Variations in different ecologies, in turn, result in individuals' developing alternative strategies in the course of ontogeny (Caro & Bateson Reference Caro and Bateson1986). Experiences early in ontogeny alert the organism to the environment in which they will function so as to maximize their adaptation to that niche. Different phenotypes develop during childhood in sexually segregated groups during childhood, to prepare boys and girls for adult roles. Briefly, ecologies labeled as “abundant” (Bateson & Martin Reference Bateson and Martin1999) should be characterized by polygynous mating patterns and males, relative to females, being more competitive and active. This high level of activity should result in segregated juvenile groups with males being characterized by physically activity and aggressive behavior and females by more sedentary and nurturing behavior (Pellegrini Reference Pellegrini2004). More severe ecologies should witness more “thrifty phenotypes” (Bateson & Martin Reference Bateson and Martin1999), resulting in monogamous or cooperative breeding and less difference between the sexes in competition and activity. Consequently, peer groups should be less segregated, resulting in fewer sex differences in aggression. Archer himself acknowledges the role of such variation, yet he does not go into sufficient detail to explain variation in different ecologies. For example, sexual selection theory predicts that sex differences in size and aggression should be most pronounced in polygynous societies. To what degree does this occur?

Traditional sexual selection theory also posits that more intense male-male competition should result in corresponding sex differences in body size, with males being physically larger than females. This review, however, has not explained the reliable sex difference is physical aggression during early and middle childhood when there are no differences in dimorphism for physical size. Why is that? One argument holds that the sex differences in physical activity (Pellegrini et al. Reference Pellegrini, Long, Roseth, Bohn and Van Ryzin2007a) and locomotor play (Pellegrini & Smith Reference Pellegrini and Smith1998) during early childhood reflect the differences necessary to shape males' and females' skeletal and muscular systems that differ in adulthood.

Further, contextual variation associated with males' uses of within-sex aggression to attract mates within an evolutionary framework has been addressed in a recent and very provocative paper by Buston and Emlen (Reference Buston and Emlen2003). They suggest that mate choice, at least among Cornell undergraduates, is not governed by mate “potential” attributes – for example, attractive women choosing strong, high-status males, and strong, high status men choosing attractive women. Instead, they found that individuals' choices were based on their perceptions of the possible duration of the relationships. When there was expectation of a long term, stable relationship, individuals chose mates similar to themselves (on factors such as physical attractiveness, status, and commitment to family). Attractive women chose attractive men; they did not choose strong, aggressive men. Situations in which costs associated with mate switching is high, for example, where divorce is economically disruptive, would result in individuals choosing mates who are similar to themselves (see Borgerhoff Mulder Reference Borgerhoff Mulder2004).

By contrast, aggressive men (again, where aggression is intrasexual) were chosen by physically attractive women in situations in which the quality of available mates is low. In this latter case, male-male aggression would be reinforced by female choice and females would “sample” widely from available males for a strong mate who also protects and provisions her and her offspring.

With these limitations stated, Archer's target article makes an important contribution toward the theoretical integration of a wide and disparate literature. He has done the field an important service.