Dunbar (1992, 1993)

Citation: Dunbar, R. I. M. (1992). “Neocortex size as a constraint on group size in primates.” Journal of Human Evolution.
Dunbar, R. I. M. (1993). “Coevolution of neocortical size, group size and language in humans.” Behavioral and Brain Sciences.

Robin Dunbar’s work proposes a cognitive limit on the number of stable social relationships humans can maintain, commonly referred to as Dunbar’s number. His research connects brain size (specifically neocortex size) to social group size in primates and then extrapolates to humans.

In the 1992 paper, Dunbar analyzes primate data and finds a strong correlation between neocortex ratio and average group size across species. The interpretation is that social complexity requires cognitive capacity; larger brains can manage larger social networks. Using this relationship, Dunbar predicts an upper bound for human social group size.

In the 1993 paper, he extends the argument to human evolution and suggests that language and cultural tools emerged partly to manage social coordination in larger groups. Language is seen as a mechanism that allows humans to maintain group cohesion at scales that would otherwise exceed cognitive limits.

The widely cited number—around 150 stable relationships—is not meant to be a precise constant but a statistical average. Dunbar also emphasizes layered social networks: smaller inner circles (e.g., 5 intimate ties, 15 close friends, 50 good friends) and larger outer circles (around 150 stable relationships). These layers reflect different levels of emotional intensity and cognitive investment.

A key implication is that relationship maintenance has real cognitive costs. People can maintain large numbers of weak ties, but strong trust‑based relationships are scarce and time‑consuming. This has practical consequences in domains where trust and reputation matter, such as business, finance, and professional networks.

Dunbar’s work has been supported by later empirical studies showing that even in large digital networks, active relationships remain limited and layered. This suggests the constraint is cognitive rather than purely technological.

Social group size is constrained by cognitive capacity. Humans can sustain only a limited number of stable, trust‑based relationships, and this limit shapes how social and economic systems organize themselves. Any system built on human trust and relationship management faces a hard scaling constraint—unless new tools relax the cognitive burden.