The proposal that primate brains grew large to manage social relationships rather than to solve ecological problems, and that the size of a species' brain therefore predicts the size of its group. It generated one of the most widely known numbers in popular science and one of the most contested findings in comparative neuroscience.
Brains are extraordinarily expensive. Human brain tissue consumes roughly twenty percent of resting metabolism while making up about two percent of body mass, and the tissue must be maintained continuously. Anything that costly requires a proportionate return.
Ecological explanations were the default. Larger brains might be needed to remember the location of fruiting trees, to extract difficult foods, or to navigate large ranges. These predict correlations with diet and habitat, and the correlations found were weak.
Robin Dunbar argued from 1992 that the demanding problem for a primate is other primates.
Living in a group requires tracking not only individuals but relationships between them: who is allied with whom, who owes what, who can be relied on, who was recently in conflict with whom. The number of dyadic relationships rises with the square of group size, and the number of third-party relationships rises faster still. Social complexity therefore escalates sharply as groups grow, in a way that food-finding does not.
The prediction is that neocortex size, relative to the rest of the brain, should predict mean group size across primate species. Dunbar reported that correlation, and it was reasonably strong.
Extrapolating the primate regression to human neocortex ratio yields a group size of roughly 150.
Dunbar assembled supporting observations: the size of Neolithic farming villages, of Roman and modern army units, of Hutterite communities which split at around 150, of Christmas card lists, and of personal networks in several surveys. The figure entered general circulation and is now widely cited as the number of stable relationships a person can maintain.
Dunbar has consistently described it as an approximate range, roughly 100 to 250, rather than a precise value, and has proposed a layered structure of nested circles: about five intimates, fifteen close friends, fifty friends, 150 acquaintances, each layer roughly three times the last.
The comparative correlation has not held up as robustly as the early results suggested.
A 2021 analysis by Andrew Smaers and colleagues, using a larger dataset and modern phylogenetic methods, found that brain size scales with body size in ways that account for much of the pattern, and that the specific social correlation weakens substantially once evolutionary relatedness is handled properly. Closely related species are not independent data points, and early analyses treated them as though they were.
A 2021 study by Patrik Lindenfors and colleagues, using Bayesian methods across a large primate dataset, found no reliable relationship between neocortex size and group size, and concluded that the evidence did not support the hypothesis in its standard form.
Alternative explanations for large primate brains remain available and are not excluded: diet quality, particularly fruit eating, correlates with brain size in several analyses; extended developmental period and the cost of raising slow-maturing offspring may be the relevant constraint; and technical intelligence and cultural transmission have their own supporting evidence.
The 150 figure attracts separate criticism. A 2021 Swedish analysis argued that the confidence interval around any such extrapolation is enormous, spanning from around 5 to over 300, which makes a specific number unsupportable. Empirical support is also mixed: some studies of personal networks find values near 150, others considerably higher, and the definition of a stable relationship does most of the work.
The proposition that sociality is cognitively demanding is not in dispute. Primates track relationships, respond to third-party interactions, and behave in ways requiring representation of who stands where with whom. That much is well demonstrated experimentally.
What is contested is whether that demand is the principal cause of brain expansion, and whether any clean cross-species relationship exists to demonstrate it.
The hypothesis remains a hypothesis for a specific reason: the test available is a correlation across species, and correlations across species are confounded by shared ancestry, by body size, by diet, and by the difficulty of measuring group size in a way that means the same thing in a chimpanzee and a lemur. No experiment can be run. Better comparative methods have weakened the original result rather than confirming it, and the field has not converged.
