Archer compares sexual selection (Darwin Reference Darwin1871/1901) and social role theory (Eagly Reference Eagly1987) as explanations for human sex differences in aggressiveness. Archer's empirical claims are grounded, but his case is weakened because he fails to emphasize two related meta-methodological points: the power of the comparative approach and the parsimony it provides. Greater male aggressiveness is neither uniquely human nor universal among animals. It exhibits a particular cross-species distribution, being present in some species and absent in others. That distribution strongly constrains its functional explanation. Unfortunately, social role theory neglects this powerful source of insight. It examines sex differences in aggression in a zoological vacuum, naively treating the human case as unique. As a result of its narrow focus, social role theory fails to shed new light on sex differences in human aggression. Like many models in the social sciences, it is a case of special pleading where none is required.

Darwin (Reference Darwin1871/1901) defined sexual selection as an evolutionary process that regularly produces differences between the females and males of a species, and he enumerated the reproductive differentials that drive it. Subsequently, twentieth-century biologists (Andersson Reference Andersson1994; Clutton-Brock & Vincent Reference Clutton-Brock and Vincent1991; Trivers Reference Trivers and Campbell1972) explained the forces that give sexual selection its polarity. It is this polarity – which sex is more aggressive – that both sexual selection and social role theory seek to explain.

Scientific theories explain relevant variance. Thus, social role theory would have some traction if there were cultures where women are more aggressive than men. For better or worse, no such cultures exist (Brown Reference Brown1991; Daly & Wilson Reference Daly and Wilson1988). So, where is the theory-testing variance? To find it one must escape anthropocentric Durkheimian biophobia and look across species.

Sexual selection theory provides a general explanation of sex differences that applies to all sexual species. It predicts the distribution of sex differences in aggression: which species will evolve greater male aggressiveness, which species will exhibit no such sex differences, and which will show greater female aggressiveness. According to sexual selection theory, aggressiveness is not a function of sex per se, but of sex differences in maximum reproductive rate (Clutton-Brock & Vincent Reference Clutton-Brock and Vincent1991), arising out of sex differences in parental investment (Trivers Reference Trivers and Campbell1972). To illustrate, because only female mammals gestate and lactate, a male could have many more offspring than a female. Every reproductive venture requires one male and one female; thus the slower sex is in short supply and worth competing for. (In this example reproductive physiology determines reproductive rates but aspects of the mating system may also be important.)

Sexual selection theory's accuracy in predicting the distribution of sex differences across species makes it logically prior to ad hoc explanations of sex differences in any particular species, unless that species fails to match its predictions. Thus, a baseline question for social role theorists is, do humans constitute an exception to sexual selection theory? Men have, and have had for thousands of generations, higher maximum reproductive rates than women. This implies that men will have found women in short supply and consequently evolved a suite of competitive tactics for acquiring mates, including aggression.

What remains for social role theory to explain? Its proponents might say “development.” But, whatever ontogenetic influences social scientists imagine for gender roles, their hypotheses will have to contend with a thick cross-species literature on the developmental effects of androgens. Wherever sexual selection has produced more aggressive males, androgens orchestrate the development of that sex difference. As a functional viewpoint would suggest, the very same hormones also shape the anatomical components of the male-competition complex. Compared to women, men are much stronger and more muscular in the upper body (the region most engaged in physical aggression).

The effect size for these sex differences is approximately 3.0, with 99.9% of women falling below the male mean; individual differences in muscle mass still reliably predict male mating success in the United States (Lassek & Gaulin, in press). Thus, sexual selection simultaneously explains both anatomical and behavioral sex differences and their joint hormonal mediation. Social role theory cannot approach this level of explanatory integration.

But the explanation gap is wider still. The male-competition complex has many evolved manifestations. Bimaturism, sexually differentiated mortality rates, and sex differences in navigational ability are well-described human traits; and again, the cross-species distribution of these traits strongly implicates sexual selection. Delayed sexual maturity of males is limited to species where they have higher reproductive rates than females (Leigh Reference Leigh1992, Leigh & Shea Reference Leigh and Shea1995). Controlled within-genus comparisons suggest that searching for mates drives the evolution of male navigational ability, but again, only where they have higher reproductive rates (Gaulin Reference Gaulin1992). Both bimaturism and navigational ability are developmentally linked to the same androgenic hormones that organize aggressive structures and behaviors. Many aspects of the male-competition complex entail costs reflected in higher male mortality rates: In both birds and mammals sex differences in mortality are not universal but proportionate to the intensity of sexual selection (Promislow Reference Promislow1992, Promislow et al. Reference Promislow, Montgomerie and Martin1992). These costs are not limited to combat-related mortality but include higher male susceptibility to infection that: (a) closely tracks the intensity of sexual selection across species (Moore & Wilson Reference Moore and Wilson2002), (b) manifests prominently in humans (Owens Reference Owens2002), and (c) is probably related to androgens' immunosuppressant effects (Folstad & Karter Reference Folstad and Karter1992) because castration removes the sex difference in infection rates and hormone replacement reinstates it (Zuk & McKean Reference Zuk and McKean1996). It is not the existence of these traits but their patterned, cross-species association that social role theory must confront.

In sum, humans exhibit a suite of traits – elevated male aggressiveness, greater male muscularity and strength, later male maturation, superior male navigational ability, and higher male mortality, all underpinned by an androgen-based developmental system – that they share only with species where male reproductive rates can significantly exceed those of females. The coherent distribution of these traits strongly suggests they were jointly produced by sexual selection. This patterning, revealed by cross-species comparison, supports the causal primacy of sexual selection. In the absence of a significant misfit with the predictions of sexual selection, any attempts to (re)explain men's greater aggressiveness in purely sociological terms constitute unparsimonious exceptionalism and have little scientific promise.