Attackers move first

The most common form of intergroup violence in evolutionary history has been the raid, which, in part, is defined by its move order asymmetry. Specifically, attackers move first. When attackers move first, they are likely to choose D when: (1) defenders have already chosen C (however unwittingly; e.g., they are asleep, facilitating an ambush) or (2) defenders are weak. Similarly, defenders are unlikely to choose D unless attackers have already chosen D, because, as the authors say: “Unilateral defense is costly.” The above implies that attackers will choose D in a first move when: (1) the risks are sufficiently low (Lopez Reference Lopez, Weekes-Shackelford, Shackelford and Weekes-Shackelford2016; Tooby & Cosmides Reference Tooby and Cosmides1988; Wrangham Reference Wrangham1999) or (2) social institutions provide countervailing incentives to sustain coordinated behavior (Glowacki & Wrangham Reference Glowacki and Wrangham2013). Importantly, endogenizing move order is not only consistent with the analysis provided by D&G but also it is potentially more reflective of ancestral landscapes of intergroup violence (Gat Reference Gat2006; Wrangham & Glowacki Reference Wrangham and Glowacki2012).

Defenders can run, absorb, and retaliate

A value of the Attacker-Defender Game is its simplicity, which allows us to ask questions about what may be missing, whether what is missing is important, and how to integrate these elements. For example, one can see that both “defending” and “not defending” can have either active or passive manifestations. “Defending” can mean active physical denial, or it can mean a relatively passive sunk cost investment in defensive capabilities (e.g., walls & weapons). “Not defending” can mean anything from surrender, to being caught unaware, to running away. Indeed, the latter was likely a common and prudent choice in ancestral environments, although it is obviously and tellingly infeasible for modern nation states. In short, varieties of “not defending” may carry different payoffs; surrendering potentially means death, while running away allows, at a minimum, the possibility of survival, and at a maximum, opens the possibility of reorganization and retaliation. Thus, “running away to fight again another day” may also be properly understood as asymmetric defense, which suggests that defenders may profit just as much from cultivating asymmetries as their adversaries.

Specifically, under certain conditions, the defender preference for matching attacker strategy may be complemented by a defender preference for mismatching attacker tactics. One of these conditions is likely to be when defenders are weak. When attackers choose D, even weak defenders may prefer D > C when they have the option to run from or absorb an attack and retaliate. International relations scholars note that weaker adversaries are often able to win wars precisely when they can mismatch the conventional tactics of their attackers with unconventional guerilla warfare (Arreguin-Toft Reference Arreguin-Toft2005). Large nation states exploit a similar principle – the notion of “elastic defenses” – when they absorb an attack while preparing a counterattack, as the Russian Empire did to great effect against the advancing armies of Napoleonic France (Biddle Reference Biddle2006). Thus, Wrangham and Glowacki (Reference Wrangham and Glowacki2012) insightfully note that one of the elements that makes human coalitional violence so much riskier than chimpanzee violence is the likelihood of retaliation and revenge, which again works to undermine attacker success. This is particularly true in combination with the problems of coordination and motivation that D&G usefully discuss.

Sex differences exist

The authors present compelling evidence that defense benefits from rapid and spontaneous in-group support, while attack is vulnerable to failures of motivation and coordination. One additional wellspring of motivation the authors do not consider is sex differences in aggression (Daly & Wilson Reference Daly and Wilson1988; Goldstein Reference Goldstein2003; McDonald et al. Reference McDonald, Navarrete and Van Vugt2012; Van Vugt Reference Van Vugt2009). A growing body of evidence suggests that males and females likely possess distinct conditional psychologies of aggression (Brooks & Valentino Reference Brooks and Valentino2011; Ginges & Atran Reference Ginges and Atran2011; McDermott Reference McDermott2015), and it is likely that these psychologies find unique expression in attack and defense.

One of the key features distinguishing attack from defense is that the latter is more clearly a public goods problem (Lopez Reference Lopez2010; Rusch Reference Rusch2013; Reference Rusch2014a; Tooby & Cosmides Reference Tooby and Cosmides1988). This fact alone suggests that, all things equal, there should be no sex difference in support for investing in defense. What it does not suggest, however, is that females should be just as willing as males to physically fight. Indeed, experimental evidence indicates that, although males and females anticipate similar levels of benefit from successful defense, males profess a greater willingness to directly participate in the fighting. In contrast, males report greater anticipated benefit from and a greater willingness to participate in offensive coalitional aggression relative to females (Lopez Reference Lopez2017). In other words, although males are more physically aggressive across offensive and defensive scenarios, females are more likely to mirror male hawkishness in defense than in offense. Relatedly, Wrangham notes that “proactive aggression has been predicted to be associated with [larger] sex differences, consistent with the greater prevalence of psychopathy among males” (Reference Wrangham2018, p. 250). Taken together, we should expect sex differences in aggression to reflect an adaptive logic that functionally embodies the distinct challenges of offense and defense, which is consistent with both the “male warrior hypothesis” (McDonald et al. Reference McDonald, Navarrete and Van Vugt2012) and with observations of greater proactive aggression and psychopathy among males relative to females.

The argument for separate psychologies of attack and defense that D&G lay out is compelling; provides a useful example of the functional link between evolution, psychology, and behavior; and affords many lucrative opportunities for future study.