Van Lange et al. propose that the CLASH model will facilitate understanding of “differences within and between countries in aggression and violence in terms of differences in climate” (Abstract). Although the authors cite several studies to support their contention that differences in aggression over time and within region are associated with temperature, their argument that “countries closer to the equator are generally more violent” (sect. 2, para. 1) lacks empirical support. This misrepresentation of the literature on temperature and aggression, paired with the authors' exclusion of historical context in theory building, betrays what Harrison (Reference Harrison1995, p. 50) has identified as “essentialist and universalized ‘biomoral’ assumptions concerning the natural history of human variation.” We contend that in this presentation, Van Lange et al. fail to adequately disentangle empirical findings from culturally constructed assumptions. The result is a model that is eerily reminiscent of 18th-century European imperialist ethos (Cohen Reference Cohen2003):

In this commentary we identify unsubstantiated claims in the article and express our concern that the CLASH model uses the language of science to lend false credibility to problematic and damaging sociocultural discourse.

First, the article lacks evidence of a relationship between latitude and a cultural propensity for aggression. To support the claim that an aggression-climate pattern holds such that countries closer to the equator are more violent, Van Lange et al. cite a brief summary of a 1989 crime survey published by the Australian Institute of Criminology examining crime rates in Australia, the United States, Canada, and other countries in northern or western Europe (Walker et al. Reference Walker, Wilson, Chappell and Weatherburn1990). They then provide an unsystematic selection of violent crime statistics for broad groupings of countries (e.g., “Central America,” “Middle Africa”) selected from the 2013 Global Study on Homicide, glossing over relevant counterexamples that suggest a more complex relationship between climate and conflict. The authors also extensively cite a recent meta-analysis by Burke et al. (Reference Burke, Hsiang and Miguel2015). Yet Burke et al. (Reference Burke, Hsiang and Miguel2015) limit their analyses to within-population (e.g., within the same country, municipality, province, state, site, or district), and they explicitly question the validity of comparisons between groups that do not share culture and history: “It seems implausible that the conditions needed for causal inference are met in this setting” (p. 580). Thus, there appears to be little empirical support for Van Lange et al.'s central claim that “countries closer to the equator are generally more violent.”

Similarly, although the authors' linking of Life History Theory and climate is an interesting idea that may overlap with some contemporary economic theories regarding climate and economic development (Easterly & Levine Reference Easterly and Levine2003; Sachs Reference Sachs2001), they provide little empirical support for the inclusion of impulsivity or time orientation in the model. No research is cited to directly support the assertion that cultures closer to the equator are more broadly “impulsive” than more northern cultures. With respect to time orientation, the authors fail to acknowledge the body of research linking time urgency (i.e., “clock time”) with hostility and aggression (e.g., Glass et al. Reference Glass, Snyder and Hollis1974; Karlberg et al. Reference Karlberg, Krakau and Undén1998).

Finally, the authors' inconsistent definitions of aggression result in an ambiguous construct that conflates reactive and proactive aggression. Even if we accept the premise that “countries and regions closer to the equator tend to have higher levels of aggression,” Van Lange et al. are imprecise and inconsistent in defining the type of “aggression” explained by the model. For example, they state that their “focus is not limited to acts of interpersonal aggression and violence. We also include acts … such as political violence, wars, and riots” (sect. 1, para. 5). This broad inclusion of war and political violence within the purview of “reactive aggression” is inconsistent with the way this construct was defined by Dodge and Coie (Reference Dodge and Coie1987), the authors that were cited. Further, some of the greatest violence witnessed within the last centuries has been initiated and perpetrated in a planned and systematic (i.e., “proactive”) fashion by societies situated quite far from the equator (e.g., Nazi Germany's systematic process of occupying countries and implementing genocide; the Stalinist Soviet Union's Great Purge and forced labor camps). Based on the evidence presented, Van Lange et al.'s proposed model does not appear to have direct relevance to war and/or political violence. If the authors wish to argue for this point, then a careful historical analysis of the relationship between war and latitude would be needed.

Ultimately, the CLASH model as presented here does not sufficiently acknowledge the influence of historical, cultural, and contextual variables that might also be significant contributors to aggression and violence. Although the authors claim that the model is “organized around calls for the development of more interdisciplinary theories” (sect. 3, para. 2), they fail to appropriately acknowledge the ways in which sociopolitical factors such as income inequality, political instability, and scarcity or resources greatly influence aggression, irrespective of latitude (Krug et al. Reference Krug, Mercy, Dahlberg and Zwi2002). Many countries closer to the equator have a history of colonization and exploitation by northern countries, which some scholars contend has led to poverty, lack of basic necessities, and limited educational resources (Ghandi Reference Ghandi1998). Furthermore, given the strong history of using psychology and brain science to support eugenicist movements in Europe and the United States (Fienberg & Resnick Reference Fienberg, Resnick, Devlin, Fienberg, Resnick and Roeder1997), we argue that scientists in these fields have a duty to monitor our work for remnants of this legacy. We suggest that researchers maintain a cautious skepticism of this and other models that purport to explain population-level behavioral differences, particularly when the constructs under consideration, such as aggression, are fraught with moral overtones.