Archer's article is a useful survey of the evidence in support of the role of sexual selection in male-female differences in the expression of physical violence and risk-taking, combined with a weak and inconsistent conceptual framework. Archer begins by opposing two views of gender differences in the expression of aggression, “sexual selection theory” and “social role/biosocial theory,” and, after reviewing the evidence, proclaims the first a hands-down winner. Then, he turns to conflicts between mates, notes a lesser sex difference in aggression, and shifts gears to propose an economics-style model of conflict between selfish maximizers in a dyadic relationship. Even then, he concedes that cultural variation leaves a large, unexplained residual.

All this conceptual confusion can be simply resolved by discarding the opposition between the two approaches, and adopting a unitary model of gene-culture coevolution. There is abundant evidence for the role of sexual selection in producing gender differences in behavior, including levels of aggression. But the same sexual selection model also explains sexually differentiated behavior between mates in a reproductive relationship. Indeed, that is the very core of sexual selection. No conceptual shift is needed at all. It is all one and the same.

Now for cultural variation: It, too, obviously exists, but by now it is clear that the circular explanation of culture in terms of culture is bankrupt. “They do it so because it is their custom” is not an alternative model of behavior. It merely describes; it does not explain. Cultural variation is not random; most of it is adaptive, and linked to adaptation by natural selection. Humans are on a double evolutionary track: genetic and cultural, complexly intertwined. A single coevolution model explicating variations in gene-culture interactions is called for (Alexander Reference Alexander1979; Boyd & Richerson Reference Boyd and Richerson1985; Chagnon & Irons Reference Chagnon and Irons1979; Lumsden & Wilson Reference Lumsden and Wilson1981). In some instances, cultural variation is near zero. In others, it is considerable. That range of variation is the explanandum.

Let us take the example of women's participation in warfare. It ranges from negligible to considerable (Adams Reference Adams1983; Lynn Reference Lynn2008). But a closer look reveals a near-zero variation in women's direct participation in combat, while, in highly militarized societies (such as Israel), women play a large supporting role in scores of military duties, except combat.

In short, nothing is gained by trying to show that sexual selection explains behavior better than “social role” or culture. Everything is gained by explaining the complex modalities of gene-culture coevolution in terms of a double adaptive track. To deny the importance of either is untenable.