Like others before them, Gowdy & Krall (G&K) marvel at human sociality. On the one hand, humans cooperate in ways like some other animals on the planet. On the other, humans are unrivalled in the complexity and scale of the societies we build. How should we explain these similarities and differences? The authors suggest that the defining aspects of human sociality are derived from selective forces operating on the group level during the agricultural transition. They point to similarities with other agricultural and social animals and to historical trends to support their argument. Unfortunately, their premises are mistaken and their conclusion false. They have missed the agricultural trees for the forest.
While leaf cutter ants and humans both practice agriculture and create complex societies, the psychological mechanisms that underlie those abstractly similar behaviors are devastatingly different. Seeing this requires stepping beyond the math models of multilevel selection and engaging with the real cognitive problems involved in actually behaving successfully in a complex social environment. While leaf cutter ant society and agriculture are complicated, they are nowhere near the scale and complexity of human society. This difference in outcome exposes the difference in cognitive processing problem each species must solve. Appreciating the complexity of the cognitive processing problems that humans solve so deftly and intuitively to create and participate in societies of billions forces your attention to the questions of what cognitive mechanisms could solve these problems, and how and when they evolved.
G&K grant that hunter-gatherer populations possessed certain preconditions for the shift to ultrasociality when the economic forces of agriculture took hold. Yet, they ascribe the similarities of otherwise independent cultures to convergent evolution by group selection. This is not the only theory available to explain these data, and without any consideration of alternative theories, it is entirely premature to conclude this theory is correct. Let us consider one alternative now.
Decades of research in evolutionary psychology reveal that the human mind contains a rich social psychology for small-scale group living, including specialized mechanisms for: inferring kinship and cooperating with kin; estimating the value of resources to the self and others; apprehending opportunities for mutual gains in trade; inferring opportunities to exploit and be exploited; perceiving, building, and maintaining coalitions and alliances; identifying free riders on collective actions; achieving and managing coordination; and on, and on, and on (e.g., Cosmides & Tooby Reference Cosmides and Tooby1989; Delton et al. Reference Delton, Cosmides, Guemo, Robertson and Tooby2012; Krasnow et al. Reference Krasnow, Delton, Cosmides and Tooby2015; Kurzban & Neuberg Reference Kurzban, Neuberg and Buss2005; Lieberman et al. Reference Lieberman, Tooby and Cosmides2007; Thomas et al. Reference Thomas, DeScioli, Haque and Pinker2014; Tooby & Cosmides Reference Tooby and Cosmides1996). Whenever we've looked, this psychology is not unique to the agricultural or industrialized world, but rather appears to be universal in our species (e.g., Sugiyama et al. Reference Sugiyama, Tooby and Cosmides2002). This universality should not be surprising when the ancestral social ecology is considered. The social world of our hunting and gathering ancestors was complex, presenting many complicated adaptive problems that the mechanisms above and other adaptations are solutions to. This fact alone – that much of the human social psychological architecture is reliably developing even in the absence of an agricultural context – presents a sizable and likely fatal barrier to G&K's argument.
Further, because environments have always being variable, it should be expected that these mechanisms can be facultatively responsive or otherwise calibratable by relevant environmental parameters (Tooby & Cosmides Reference Tooby and Cosmides1990). This inference has two important implications. First, it gives the expectation that different human cultures in different ecological circumstances will have different norms, behaviors, and social patterns despite having universal cognitive mechanisms. Second, because the ecology can be changed over time by the organisms that occupy it (who can change themselves in response), it is expected that even independent populations that discover a common agricultural niche in parallel will evidence similar evoked cultures. In other words, a clear alternative to the authors' view is simply that human psychology operating in a given ecology, plus time for cultural change are sufficient factors to account for the major patterns of human ultrasociality. It is possible that this alternative is not correct, but G&K do nothing to consider even this most basic alternative hypothesis.
Why are G&K motivated to pursue a group selection (or multilevel selection) argument in the first place? The authors argue that many of the behaviors humans engaged in on the way to and since agriculture appear to degrade our quality of life. They point out the many hazards that emerge as people congregate into larger communities and cities. The point of this detail is for the authors to counterpoint these apparent individual-level costs with group-wide benefits in terms of production, economies of scale, and intergroup competition. G&K use this pattern of costs and benefits to motivate their group selection argument. However, this analysis is fundamentally flawed. The currency of natural selection is reproductive fitness. Selection does not operate on quality of life. Selection can shape motivational mechanisms in the service of reproductive fitness that may attend to facets of quality of life, but only to the extent that those facets reliably predicted reproductive fitness over ancestral environments. If the motivational machinery of agricultural humans pushed them into cities that degraded their quality of life, this by itself is no evidence of group selection in action. Our motivational machinery for diet choice pushes us towards Big Macs and heart attacks, but this is similarly no evidence of group selection in action. You don't need group selection for individuals to act suboptimally; therefore, evidence of suboptimal performance is not evidence for group selection.
Humans are remarkably social; this is not controversial. Multilevel selection is a coherent way of understanding evolution; this is not controversial. Human culture has changed over time and the agricultural revolution marked a major transition in our history; this is not controversial. However, the data do not support the argument that multilevel selection operating on competing post-agricultural groups is responsible for human ultrasociality.
Like others before them, Gowdy & Krall (G&K) marvel at human sociality. On the one hand, humans cooperate in ways like some other animals on the planet. On the other, humans are unrivalled in the complexity and scale of the societies we build. How should we explain these similarities and differences? The authors suggest that the defining aspects of human sociality are derived from selective forces operating on the group level during the agricultural transition. They point to similarities with other agricultural and social animals and to historical trends to support their argument. Unfortunately, their premises are mistaken and their conclusion false. They have missed the agricultural trees for the forest.
While leaf cutter ants and humans both practice agriculture and create complex societies, the psychological mechanisms that underlie those abstractly similar behaviors are devastatingly different. Seeing this requires stepping beyond the math models of multilevel selection and engaging with the real cognitive problems involved in actually behaving successfully in a complex social environment. While leaf cutter ant society and agriculture are complicated, they are nowhere near the scale and complexity of human society. This difference in outcome exposes the difference in cognitive processing problem each species must solve. Appreciating the complexity of the cognitive processing problems that humans solve so deftly and intuitively to create and participate in societies of billions forces your attention to the questions of what cognitive mechanisms could solve these problems, and how and when they evolved.
G&K grant that hunter-gatherer populations possessed certain preconditions for the shift to ultrasociality when the economic forces of agriculture took hold. Yet, they ascribe the similarities of otherwise independent cultures to convergent evolution by group selection. This is not the only theory available to explain these data, and without any consideration of alternative theories, it is entirely premature to conclude this theory is correct. Let us consider one alternative now.
Decades of research in evolutionary psychology reveal that the human mind contains a rich social psychology for small-scale group living, including specialized mechanisms for: inferring kinship and cooperating with kin; estimating the value of resources to the self and others; apprehending opportunities for mutual gains in trade; inferring opportunities to exploit and be exploited; perceiving, building, and maintaining coalitions and alliances; identifying free riders on collective actions; achieving and managing coordination; and on, and on, and on (e.g., Cosmides & Tooby Reference Cosmides and Tooby1989; Delton et al. Reference Delton, Cosmides, Guemo, Robertson and Tooby2012; Krasnow et al. Reference Krasnow, Delton, Cosmides and Tooby2015; Kurzban & Neuberg Reference Kurzban, Neuberg and Buss2005; Lieberman et al. Reference Lieberman, Tooby and Cosmides2007; Thomas et al. Reference Thomas, DeScioli, Haque and Pinker2014; Tooby & Cosmides Reference Tooby and Cosmides1996). Whenever we've looked, this psychology is not unique to the agricultural or industrialized world, but rather appears to be universal in our species (e.g., Sugiyama et al. Reference Sugiyama, Tooby and Cosmides2002). This universality should not be surprising when the ancestral social ecology is considered. The social world of our hunting and gathering ancestors was complex, presenting many complicated adaptive problems that the mechanisms above and other adaptations are solutions to. This fact alone – that much of the human social psychological architecture is reliably developing even in the absence of an agricultural context – presents a sizable and likely fatal barrier to G&K's argument.
Further, because environments have always being variable, it should be expected that these mechanisms can be facultatively responsive or otherwise calibratable by relevant environmental parameters (Tooby & Cosmides Reference Tooby and Cosmides1990). This inference has two important implications. First, it gives the expectation that different human cultures in different ecological circumstances will have different norms, behaviors, and social patterns despite having universal cognitive mechanisms. Second, because the ecology can be changed over time by the organisms that occupy it (who can change themselves in response), it is expected that even independent populations that discover a common agricultural niche in parallel will evidence similar evoked cultures. In other words, a clear alternative to the authors' view is simply that human psychology operating in a given ecology, plus time for cultural change are sufficient factors to account for the major patterns of human ultrasociality. It is possible that this alternative is not correct, but G&K do nothing to consider even this most basic alternative hypothesis.
Why are G&K motivated to pursue a group selection (or multilevel selection) argument in the first place? The authors argue that many of the behaviors humans engaged in on the way to and since agriculture appear to degrade our quality of life. They point out the many hazards that emerge as people congregate into larger communities and cities. The point of this detail is for the authors to counterpoint these apparent individual-level costs with group-wide benefits in terms of production, economies of scale, and intergroup competition. G&K use this pattern of costs and benefits to motivate their group selection argument. However, this analysis is fundamentally flawed. The currency of natural selection is reproductive fitness. Selection does not operate on quality of life. Selection can shape motivational mechanisms in the service of reproductive fitness that may attend to facets of quality of life, but only to the extent that those facets reliably predicted reproductive fitness over ancestral environments. If the motivational machinery of agricultural humans pushed them into cities that degraded their quality of life, this by itself is no evidence of group selection in action. Our motivational machinery for diet choice pushes us towards Big Macs and heart attacks, but this is similarly no evidence of group selection in action. You don't need group selection for individuals to act suboptimally; therefore, evidence of suboptimal performance is not evidence for group selection.
Humans are remarkably social; this is not controversial. Multilevel selection is a coherent way of understanding evolution; this is not controversial. Human culture has changed over time and the agricultural revolution marked a major transition in our history; this is not controversial. However, the data do not support the argument that multilevel selection operating on competing post-agricultural groups is responsible for human ultrasociality.