Why are humans so intelligent?
– by Jonathan Egeland, Leif Edward Ottesen Kennair, and Thomas Haarklau Kleppestø
A central puzzle in the evolutionary behavioral sciences, studied since Darwin, is how to explain our uniquely human intelligence. Why did it evolve? What made it useful enough to be favored by selection? And what is the proximate explanation for how it works? Our recent paper in Evolution and Human Behavior argues that reciprocal cooperation holds the key to answering these questions.
Over the second half of the twentieth century, many researchers came to think that human intelligence has to be understood in light of the social environments our ancestors lived in. Humans evolved in groups, and surviving in those groups meant dealing with a great many kinds of social interaction throughout development. That basic idea is widely accepted. The harder part has been turning it into a clear, plausible, and testable explanation.
Our paper was an attempt to make progress on this issue. We focus both on what intelligence is for (its function), and on how this kind of explanation can be made consistent with relevant findings from personality research, psychometrics, and comparative psychology. We began by highlighting a simple coordination problem: how can the pursuit of individual fitness-related advantages be made compatible with social cooperation between members of the same coalition or group? In many other primate species, social life is characterized by dominance hierarchies. But in humans, cooperation is often the norm. A lot of research suggests that the coordination problem was solved by instituting a system of social exchange where individuals gain status by providing benefits to other group members. This is a form of reciprocity in which cooperation is established and maintained by repeated mutually beneficial interactions. In such a system, status will typically take the form of prestige rather than dominance, as voluntary deference is shown to individuals with traits or competencies indicative of what we call “coalitional value” (i.e., social partner value).
Our main claim is that intelligence—or its behavioral effects, to be more specific—is one of the traits that signals this kind of value. Our Coalitional Intelligence Hypothesis proposes that the behavioral effects of intelligence reliably signal an individual’s ability to benefit others. Intelligence allows people to solve problems, generate useful ideas, coordinate activities, and provide other kinds of benefit to their group, and it grants prestige in return. This leads to several predictions, one of which is that intelligence should be positively related to prestige, and this relationship should be mediated by the extent to which individuals actually benefit their social partners. In slightly technical terms, intelligence enables reciprocal exchanges of status and cognitive-computational services.
Although there are many studies whose results are consistent with this prediction, few, if any, provide direct tests of it. For example, research from psychometrics and personality psychology shows that intelligent individuals achieve higher status, in part because they provide social benefits. However, status in these studies is operationalized in terms of socio-economic status (i.e., income, education, and occupation), rather than prestige. On the other hand, there are many findings from anthropology linking the relevant variables, but intelligence is usually measured subjectively and not in terms of actual cognitive performance. Future tests of the hypothesis should therefore focus on whether cognitive performance leads to higher prestige, mediated by ability to benefit one’s social group.
In addition to discussing what the hypothesis predicts and how it compares to other, established hypotheses from the literature, we also focused on whether human intelligence can be studied meaningfully as an evolved trait. Some researchers have argued that there is a single intelligence mechanism or system, while others see it as a loose collection of many different cognitive mechanisms; and the debate between these camps goes back more than a hundred years. However, we suggest that this may be a false dichotomy. At the level of underlying mechanisms, there is clearly a lot of heterogeneity. Human intelligence involves many different systems—there is no such thing as the intelligence algorithm—and claims to the contrary risk becoming incoherent. However, at the same time, research from comparative psychology and animal cognition suggests that human intelligence is made possible by increased computational capacity across a variety of underlying cognitive mechanisms. From this we draw two inferences: first, that intelligence can be “many” at the mechanistic level, while still for certain purposes being meaningfully conceptualized as “one” functional, higher-order trait; and second, it may be plausibly conjectured that individual differences in intelligence also are reflective of patterns of covariance in computational capacity across different cognitive mechanisms.
Taken together, our paper suggests that uniquely human intelligence evolved, at least in part, because it enabled reciprocal exchanges of status and social benefits. If people voluntarily defer to social partners providing them with useful services, then there can be selection for traits and competencies indicative of social partner value. As Richard Alexander once noted, ancestral humans living in such social environments had to “cooperate to compete”. One way of competing in this manner is to provide cognitive services to others, which intelligent individuals are better able to do.
