On the Collective Nature of Human Intelligence

A fundamental assumption of cognitive science is that the individual is the correct unit of analysis for understanding human intelligence. I present evidence that this assumption may have limited utility, that the social networks containing the individuals are an important additional unit of analysis, and that this `network intelligence’ is significantly mediated by non-linguistic processes. Across a broad range of situations these network effects typically predict 40% or more of the variation in human behavior. The evidence presented strongly supports the position that humans must be understood as social animals as well as individuals, and that our behavior and thought processes should be understood as due to non-linguistic network interactions as well as individual properties and dynamics. These data therefore have the potential to broadly alter thinking in the cognitive sciences, since they undermine the assumption that the individual is the correct unit of analysis. These data could also unsettle parts of the social sciences, which have tended to treat culture as isolated from the properties of the individual.

Read