A perplexing paradox for aggression researchers is the marked sex difference in same-sex aggression and its absence in partner-directed aggression. Archer is well placed to take on the explanatory challenge, given his impressive output of meta-analyses in this area.

In discussing fear and risk-taking as candidate mediators of the sex difference in aggression, Archer notes that “there are currently no direct tests of these.” However, a colleague and I recently conducted such a study (Campbell & Muncer, in press). We developed a scale of “risky impulsivity” – defined as a tendency to act spontaneously and without deliberation, where the act has potentially dangerous consequences. We purposely omitted any reference to hostile interchanges. We also measured direct physical and verbal aggression, both of which involve potentially injurious retaliation by the victim, and two forms of anger expression that do not involve confrontation and are therefore unlikely to excite retaliation. One form was what we called “defusing” anger (taking actions that reduce anger intensity, such as retreating from the scene or discussing the incident with a third party) and the other was “explosive” anger (discharging acute physical or verbal anger when alone, e.g., by hitting walls or shouting). Sex differences in physical and verbal aggression were eliminated when risky impulsivity was controlled. This complete mediation of sex differences was restricted to the two direct forms of aggression, those carrying the risk of retaliation.

Archer uses the term “sexual selection” to describe the evolution of sex-differentiated traits, but there has recently been heated controversy about the scope and interpretation of this term (Kavanagh Reference Kavanagh2006; Roughgarden et al. Reference Roughgarden, Oisha and Akcay2006). Darwin himself (Reference Darwin1871/2004, p. 245) left the door open to such debate when, with disarming honesty, he acknowledged that, “in most cases … it is impossible to distinguish between the effects of natural and sexual selection.”

Evolutionary psychologists have mainly restricted “sexual selection” to intrasexual or intersexual competition for mating opportunities. When male variance in reproductive success exceeds that of females, there is effective polygyny and intensified male competition. Daly and Wilson's (Reference Daly and Wilson1988) account of sex differences in aggression clearly identifies these mating opportunities as the driver of male competition and, in extremis, aggression. Under monogamy, there is two-way selection and researchers have now begun to address women's intersexual competition for mates. One popular topic is male preference for female body shape, with debate focusing on the relative importance of waist-to-hip ratio and body mass index, and the universality of these male preferences (Tovee & Cornelissen Reference Tovee and Cornelissen1999; Yu & Shepard Reference Yu and Shepard1999).

But what if female waist-hip ratio evolved, not with men's preferences, but with women's reproductive success, in mind? Bipedal locomotion, combined with pregnancy and infant carrying, meant that a lower relative centre of body mass increased women's stability and this shift corresponds with a lower waist-hip ratio (Pawlowski & Grabarczyk Reference Pawlowski and Grabarczyk2003). Or a lower waist-hip ratio may have resulted from foetal developmental demands: The supply of long-chain polyunsaturated fatty acids needed for neurodevelopment is optimised where the mother's lower body fat exceeds upper body fat. Waist-hip ratio, a proxy for this fat distribution, is positively associated with children's cognitive test scores (Lassek & Gaulin Reference Lassek and Gaulin2008). This is not to deny men's preference for female body shapes but to suggest that the selective advantage for women was not obtaining a better mate but producing more surviving, high-quality children.

This suggests a second, more inclusive interpretation of sexual selection – “the advantage which certain individuals have over others of the same sex and species solely in respect of reproduction” (Darwin Reference Darwin1871/2004, p. 243). Because Darwin focused chiefly on males, with their typically lower parental investment, his notion of reproduction was largely restricted to mating opportunities. Yet, as Hrdy (Reference Hrdy1999, p. 81) emphasises, “Unless mating results in the production of offspring who themselves survive infancy and the juvenile years and position themselves so as to reproduce, sex is only so much sound and undulation signifying nothing.” And in the vast majority of mammals, mothers take this responsibility. It might therefore be argued that any female trait that confers an advantage over competitors in reproduction (in this broader sense beyond mate competition) should be considered a sexually selected trait. In my own proposal (Campbell Reference Campbell1999; Reference Campbell2002) for a psychological mediator of sex differences in aggression, this would include a lower threshold for experiencing fear.

In summarising my proposal, Archer accepts that it derives from unequal parental investment but describes it as an “alternative to the sexual selection view” of Daly and Wilson, thus implying the action of natural selection. And Darwin might agree with him:

Certainly from the infant's viewpoint, maternal care is about its own survival and hence about natural selection. But from the mother's viewpoint, her care is about increasing her reproductive success relative to her competitors.

So we can see heightened female fear in the service of offspring survival as a sexually selected trait defined in this more inclusive sense. But if sexual selection refers narrowly to competition for copulations it will predominantly apply to males. This means that traits that increase competitive ability are more likely to be attributed to males than females (Clutton-Brock Reference Clutton-Brock2007), and this narrow usage devalues women's parenting effort, which is so crucial to infant survival and female reproductive success. Carranza (Reference Carranza2009, p. 750) suggests the term “sex dependent selection” to capture “those natural selection forces that operate differently in males and females because of the different reproductive strategies of the two sexes.” Specifically, fear evolved in both sexes under natural selection but was hyper-selected in females because of its association with increased reproductive success in females but not males. In that sense, men's heightened competitive risk-taking and women's lower threshold for fear are both examples of sex-dependent selection.