I argue 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. Thus, sex differences in physical aggression increase with the degree of risk, occur early in life, peak in young adulthood, and are likely to be mediated by greater male impulsiveness, and greater female fear of physical danger. Male variability in physical aggression is consistent with an alternative life history perspective, and context-dependent variability with responses to reproductive competition, although some variability follows the internal and external influences of social roles. Other sex differences, in variance in reproductive output, threat displays, size and strength, maturation rates, and mortality and conception rates, all indicate that male aggression is part of a sexually selected adaptive complex. Physical aggression between partners can be explained using different evolutionary principles, arising from the conflicts of interest between males and females entering a reproductive alliance, combined with variability following differences in societal gender roles. In this case, social roles are particularly important since they enable both the relatively equality in physical aggression between partners from Western nations, and the considerable cross-national variability, to be explained.

We agree with Archer that human sex differences in aggression are well explained by sexual selection, but note that “social learning” explanations of human behaviors are not logically mutually exclusive from “evolutionary” explanations and therefore should not be framed as such. We discuss why this type of framing hinders the development of both social learning and evolutionary theories of human behavior.

Archer provides seemingly compelling evidence for his claim that sexual selection explains sex differences in human aggression better than social role theory. I challenge Archer's interpretation of some of this evidence. I argue that the same evidence could be used to support the claim that what has been selected for is the ability to curb aggression and discuss implications for Archer's theory.

Archer presents a traditional view of intrasexual competition. Knowledge of a species' social structure provides a more complete picture. Human males compete against individuals with whom they may cooperate later in inter-group aggression. By contrast, females compete against individuals for a mate's continued support. Females' aggression may aim at eliminating the competition, whereas males simply may attempt to dominate others.

Archer examines sex differences in aggression, and argues that these differences may be better explained by sexual selection theory than by social role theory. This commentary examines sex differences in the developmental antecedents of aggression and violence, and presents a preliminary framework for examining whether the observed sex differences amongst these developmental antecedents can also be accounted for by sexual selection theory.

Sex differences in aggressive and risk-taking behaviors have practical implications for sexual integration of military combat units. The social-role theory implies that female soldiers will adapt to their role and display the same aggressive and risk-taking propensities as their male comrades. If sex differences reflect evolved propensities, however, adoption of the soldier's role is unlikely to eliminate those differences.

Human psychology contains adaptations to deploy aggression as one solution to many distinct adaptive problems. These include expropriating resources, defending against incursions, establishing encroachment-deterring reputations, inflicting costs on rivals, ascending dominance hierarchies, dissuading partner defection, eliminating fitness-draining offspring, and obtaining new mates. Aggression is not a singular strategy. Comprehensive theories must identify the “design features” of multiple adaptations for aggression.

Supporting a mediating role for fear in inhibiting female aggression, a recent study shows that aversion to “risky” impulsivity completely mediates the sex difference in direct aggression but not in angry acts where dangerous retaliation is unlikely. A more inclusive use of the term “sexual selection” to encompass reproductive advantage would recognise females' crucial role in nurturing and protecting offspring.

The target article claims that evolutionary theory predicts the emergence of sex differences in aggression in early childhood, and that there will be no sex difference in anger. It also finds an absence of sex differences in spousal abuse in Western societies. All three are puzzling from an evolutionary perspective and warrant further discussion.

We draw a distinction between defensive and predatory forms of aggression, and how these forms relate to basic neuropsychological systems, especially the Fight-Flight-Freeze-System (FFFS; putatively related to defensive aggression), and the Behavioural Approach System (BAS; putatively related to predatory aggression). These systems may help further to account for proximal brain processes and personality influences in the context of sex differences.

We add to Archer's review with mention of sexual orientation differences in aggression and empathy, which suggest a biological basis for the mediating role of empathy. We also note that Archer's view of sex differences will illuminate discussion of conduct disorder, which can only be of help to researchers in this field.

Our social role/biosocial theory provides a more adequate account of aggression sex differences than does Archer's sexual selection theory. In our theory, these sex differences arise flexibly from sociocultural and ecological forces in interaction with humans' biology, as defined by female and male physical attributes and reproductive activities. Our comments elaborate our theory's explanations for the varied phenomena that Archer presents.

We agree that sexual selection is a more comprehensive explanation for sex differences in direct aggression than social role theory, which is an unparsimonious and vestigial remnant of human exceptionalism. Nevertheless, Archer misses several opportunities to put the theoretical predictions made by himself and by others into direct competition in a way that would further the interests of strong inference.

According to I3 Theory, individuals enact aggressive behaviors when (a) instigating triggers are severe, (b) impelling forces are strong, and/or (c) inhibiting forces are weak. Archer's analysis of human sex differences in aggression could be bolstered by a careful analysis of male-female discrepancies in reactivity (or exposure) to instigating triggers, proneness toward impelling forces, and/or proneness toward inhibiting forces.

With respect to aggressiveness it is not enough to say that humans are “like other mammals.” We resemble only those species where males have higher maximum reproductive rates than females. In such species males evolve a set of hormonally mediated competitive traits via sexual selection. Because humans match the predictions of this general evolutionary model, attempts to (re)explain men's aggressiveness in sociological terms are superfluous and misleading.

Human aggression has two important dimensions: within-group aggression and between-group aggression. Archer offers an excellent treatment of the former only. A full explanation of sex differences in aggression will fail without accounting for our history of inter-group aggression, which has deep evolutionary roots and specific psychological adaptations. The causes and consequences of inter-group aggression are dramatically different for males and females.

Archer's argument regarding sex differences in partner violence rests on a general account of between-sex differences in reproductive strategies and in social roles. However, men's partner-directed violence often is predicted by perceived risk of female infidelity. We hypothesize that men's partner-directed violence is produced by psychological mechanisms evolved to solve the adaptive problem of paternity uncertainty.

Quantitative genetic studies of human aggressive behavior only partly support the claim of social role theory that individual differences in aggressive behavior are learnt rather than innate. As to its heritable component, future studies on the genetic architecture of aggressive behavior across different contexts could shed more light on the evolutionary origins of male-female versus male-male aggression.

Given the strength of Archer's case for a sexual selection account, he is too accommodating of the social roles alternative. We present data on historical changes in violent crime contradicting that perspective, and discuss recent evidence showing how an evolutionary perspective predicts sex similarities and differences responding in a flexible and functional manner to adaptively relevant triggers across different domains.

Archer recognizes that sexual selection theory is sensitive to the effects of ecologies on sex differences, yet he does not explain the impact of such variation. For example, to what degree are there sex differences in aggression in polygynous and monogamous societies? I demonstrate how differences in mating perceptions affect the traditional dichotomy that males compete for and females choose mates.