Robin Dunbar

Introduction

Robin Ian MacDonald Dunbar (born 28 June 1947) is a British biological anthropologist and evolutionary psychologist, and a specialist in primate behaviour. He is professor emeritus of evolutionary psychology in the Department of Experimental Psychology at the University of Oxford, where he leads the Social and Evolutionary Neuroscience Research Group. Trained first in psychology and philosophy and then in animal behaviour under Niko Tinbergen's influence at Oxford, he came to psychology from zoology and anthropology rather than the other way round, and his career has always sat across those boundaries.

Dunbar is best known for Dunbar's number, the claim that the size of the human neocortex sets a cognitive ceiling on the number of stable social relationships a person can maintain, usually put at about 150. The figure has passed into management literature, network design, and popular science, and it has also been seriously challenged on statistical grounds. His other major contributions are the social brain hypothesis and the argument that language evolved as a cheap substitute for physical grooming, an idea he developed in Grooming, Gossip, and the Evolution of Language.

Early Life and Education

Dunbar was born in Liverpool, the son of an engineer, and was educated at Magdalen College School, Brackley. He has written that he spent part of his childhood in East Africa during the last decades of the British Empire, an upbringing he later credited with giving him a lasting interest in how different societies organise themselves.

He went to Magdalen College, Oxford, where his teachers included the ethologist Niko Tinbergen, and took a B.A. in psychology and philosophy in 1969. He then moved to the Department of Psychology at the University of Bristol for doctoral work on the social organisation of the gelada, Theropithecus gelada, a close relative of baboons, completing his doctorate in the early 1970s, with sources giving either 1973 or 1974. Fieldwork in Ethiopia also left him with published expertise on the behavioural ecology of duiker and klipspringer. Like many scientists of his generation he then struggled to find a permanent post; he spent two years as a freelance science writer, and has said that he did not get his first real academic job until he was about forty.

Academic and Professional Career

Dunbar held a research fellowship at the University of Cambridge from 1977 to 1982, then a lectureship in primate biology at University College London's Department of Anthropology, where he rose to reader and professor between 1987 and 1994. He has described the UCL years as unexpectedly creative and as the turning point at which he switched from studying animals to studying humans, partly because funding for large-animal fieldwork had dried up and partly because anthropologists, unlike zoologists, were asking questions about people.

In 1994 he became professor of evolutionary psychology at the University of Liverpool, leaving in 2007 to become director of the Institute of Cognitive and Evolutionary Anthropology at Oxford. A European Research Council grant brought him into Oxford's Department of Experimental Psychology in 2012, where he has remained, now as professor emeritus. He was co-director of the British Academy Centenary Research Project "From Lucy to Language: The Archaeology of the Social Brain," and he holds a visiting professorship at Aalto University.

His honours include election as a Fellow of the British Academy in 1998, the Huxley Memorial Medal of the Royal Anthropological Institute, its highest award, and election as a foreign member of the Finnish Academy of Science and Letters in 2021. He is also a distinguished supporter of Humanists UK.

Major Contributions to Psychology

In the early 1990s Dunbar published two papers that made his reputation. The first reported a relationship across primate species between relative neocortex size and typical group size; the second extended the argument to humans, proposing that the neocortex ratio sets an upper bound on group size because keeping a group coherent requires tracking and defusing a rapidly growing number of relationships. Extrapolating the primate regression line to the human neocortex ratio produced a figure in the region of 150, and Dunbar argued that this matched observations of hunter-gatherer communities, military companies, business units, Neolithic and eighteenth-century villages, and, more playfully, the size of personal Christmas card networks.

The broader framework became known as the social brain hypothesis: primate cognition is expensive and evolved primarily to handle social complexity, with ecological problem-solving a secondary beneficiary. Dunbar also described the internal layering of human networks, in which roughly five intimates, fifteen close friends, fifty friends, and a hundred and fifty meaningful contacts are nested inside one another, each layer differing in the emotional closeness and the time investment required to maintain it.

Grooming, Gossip, and the Evolution of Language

A second line of work asks why primates groom. Physical grooming releases endorphins and cements bonds, Dunbar argued, but it is slow and one-to-one, so as group sizes grew it could not scale. Language, on this account, evolved as a form of grooming at a distance: vocal grooming lets one individual service several partners at once, and the content of early human conversation, he has suggested, is largely social gossip rather than the exchange of technical information. The argument was set out in Grooming, Gossip, and the Evolution of Language (1996) and has been developed ever since, including work on laughter, music, and feasting as mechanisms that release endorphins and bond larger groups than grooming can hold together.

Religion, Friendship, and Later Work

Dunbar has extended the social-brain perspective to religion, arguing that communal ritual performs the bonding work that small-scale grooming once did, and to friendship, where he and collaborators have shown that relationships decay quickly without contact and that maintaining a friendship requires a substantial investment of time. He has also written on the evolutionary background of pair bonding and on how modern communication technologies interact with ancient constraints on network size.

Major Publications

Reproductive Decisions: An Economic Analysis of Gelada Baboon Social Strategies (1984)
Primate Social Systems (1988)
The Trouble with Science (1995)
Grooming, Gossip, and the Evolution of Language (1996)
How Many Friends Does One Person Need? Dunbar's Number and Other Evolutionary Quirks (2010)
Human Evolution: A Pelican Introduction (2014, with Louise Barrett and John Lycett)

Impact and Legacy

Dunbar's number is one of the few quantitative claims from evolutionary psychology to reach a general audience, and it is routinely invoked in discussions of organisational design, online community size, and social media. Its very success is what makes the methodological criticism worth taking seriously.

The most substantial challenge came in 2021, when Patrik Lindenfors, Andreas Wartel, and Johan Lind of Stockholm University published "Dunbar's Number Deconstructed" in Biology Letters. Re-running the neocortex and group size analyses with phylogenetic comparative methods and updated brain data, they found point estimates well away from 150: Bayesian models gave average group sizes of roughly 69 to 109, and phylogenetic generalised least squares gave about 16 to 42. Their main point, however, was the width of the intervals rather than the location of the estimates. The 95 percent confidence intervals ran from about 4 to 520 and from about 2 to 336, ranges so wide that, as they put it, specifying any one number is futile and a cognitive limit on human group size cannot be derived in this way. They also noted that most supposed replications of the neocortex result draw on the same underlying brain data, that several studies fail to find a significant relationship at all, and that primate sociality is better predicted by diet and predation pressure than by brain measures.

Dunbar responded in a technical comment. He objected that the critics had used ordinary least squares rather than reduced major axis regression, and when his own preferred method was applied the estimate was not 150 either but about 290, with a confidence interval of roughly 226 to 372; restricting the analysis to monkeys and apes produced about 404, with an interval from roughly 301 to 543. Lindenfors and Wartel replied that reduced major axis regression over-corrects the slope, that the critics had reached the same conclusion with several methods, and that the deeper problem is that each new suggestion, a different brain measure, a different species sample, a different regression technique, generates a different Dunbar's number, so that the original estimate is silently revised again and again without being presented as a revision. They also raised conceptual objections: other primate brains do not process information the way human brains do, and human social network sizes vary enormously for cultural reasons, so a number fixed by brain architecture is unlikely to be recoverable in principle.

There is also a defence internal to Dunbar's own work, which he and his colleagues have pressed: the number was never meant as a precise constant, human networks vary widely around it, and the qualitative prediction of a limit and a layered structure is what matters, not the exact figure. Independent support is claimed from the recurring appearance of groupings in the hundred-to-two-hundred range across very different societies. Critics answer that a prediction compatible with almost any observed size is not a prediction, and that the anecdotal observations came after the number rather than before it.

Whatever the outcome, the episode has changed how the figure is used. Researchers now generally treat 150 as a rough central tendency with large individual variation rather than a hard cognitive ceiling, and the more durable part of Dunbar's legacy may turn out to be the social brain hypothesis and the comparative, quantitative approach to human sociality that produced it, rather than the specific number that made him famous.

Learn More

References

Dunbar, R. I. M. (1992). Neocortex size as a constraint on group size in primates. Journal of Human Evolution, 22(6), 469-493.