Dunbar's Number
A reader's summary of Robin Dunbar's 1992 estimate that roughly 150 is the ceiling on stable human relationships — the primate brain-size data behind it, the Gore-Tex factory anecdote that carried it into mainstream use, and the 2021 reanalysis that challenges the precision of the number itself.
The number at a glance
Dunbar's number is the proposed cognitive limit on how many people a person can maintain as genuine, stable social relationships — ties involving a real sense of mutual obligation and context, not just recognition of a face or a name. The commonly cited figure is about 150, derived from a correlation between brain structure and group size observed across primate species and then projected onto humans, rather than measured directly from human social networks.
Origin
British anthropologist and evolutionary psychologist Robin Dunbar introduced the argument in a 1992 paper, “Neocortex Size as a Constraint on Group Size in Primates,” published in the Journal of Human Evolution. Dunbar plotted neocortex size relative to overall brain size against typical social-group size across dozens of primate species, found a consistent relationship, and extrapolated the resulting curve to the human neocortex ratio to arrive at the approximately 150 figure.
History and context
The number circulated mainly within anthropology and evolutionary psychology for most of the 1990s before Malcolm Gladwell's 2000 book The Tipping Point carried it into general audiences, pairing the abstract cognitive claim with a concrete corporate anecdote: manufacturer W. L. Gore & Associates' long-standing practice of capping individual plant headcounts near 150 and splitting any site that grew larger, on the reasoning that informal, trust-based coordination broke down past that scale. Dunbar himself later extended the idea into a fuller model of concentric social circles at roughly 5, 15, 50, 150, 500, and 1,500 — layers of decreasing intimacy rather than one hard boundary.
Main ideas
A cognitive ceiling, not a headcount rule
Dunbar's number is a claim about the maximum number of relationships a person can maintain well enough to know how each individual relates to every other individual in the group, not simply the number of people someone can recognize or has met. The commonly cited figure is approximately 150, but Dunbar always presented it as a rough estimate with a wide plausible range, not a precise cutoff.
Built from primate neocortex ratios, then extrapolated to humans
British anthropologist Robin Dunbar published the core argument in a 1992 paper correlating neocortex size relative to total brain size across primate species with the typical size of their social groups, then extrapolated the resulting curve to the human neocortex ratio to estimate a natural human group-size ceiling.
It comes with a full set of nested circles, not just one number
Dunbar's fuller model describes concentric layers of decreasing intimacy: roughly 5 intimate confidants, 15 close friends, 50 good friends, 150 meaningful contacts, 500 acquaintances, and 1,500 recognizable faces — a structure that anticipates the site's own circles-of-interest framing of social life as layered rather than flat.
Gladwell's 2000 bestseller carried it into mainstream vocabulary
Malcolm Gladwell's The Tipping Point popularized the number far beyond anthropology, retelling W. L. Gore & Associates' factory policy of splitting any plant that grew past roughly 150 employees into two separate units to preserve the informal, trust-based coordination the company credited for its output — a real corporate anecdote that did much of the work of making 150 a household figure.
The precise value has been directly challenged
A 2021 reanalysis by Swedish researchers, applying updated statistical methods to Dunbar's original primate dataset, argued the confidence interval around the human estimate is far wider than 150 alone suggests — spanning roughly from the mid-single digits to over 500 — concluding the data can't support a single precise number with the confidence the popular version implies.
Group size also varies with subsistence style and culture
Anthropological critiques note that observed human group sizes — hunter-gatherer bands, historical military units, villages — cluster loosely around figures in the same broad range Dunbar cites, but vary enough by subsistence pattern, technology, and culture that treating 150 as a fixed universal constant overstates how tightly the neocortex correlation actually constrains real social organization.
Critique
- A 2021 reanalysis widened the error bars sharply. Applying modern statistical methods to Dunbar's original primate dataset, researchers at Stockholm University argued the human estimate's genuine confidence interval runs from roughly the mid-single digits to over 500, concluding the data can't support a single precise number with anything like the confidence popular use assigns to 150.
- Extrapolation across species is doing a lot of work. The core relationship was measured across non-human primates with far smaller and simpler social groups than humans maintain; project it onto a single human data point and small changes in modeling assumptions swing the resulting estimate substantially.
- Real group sizes vary by culture and subsistence style. Historical and ethnographic group sizes — hunter-gatherer bands, military units, villages — cluster loosely in a similar broad range but shift enough with technology, culture, and way of life that treating 150 as a fixed universal constant overstates how tightly it actually constrains human social organization.
Impact
The number is now a routine reference point in commentary on social platforms with friend and follower counts running into the thousands or millions, used to argue that such counts measure reach rather than anything like genuine relationship capacity, and it regularly comes up in debates over how large an online community, Discord server, or Slack workspace can grow before it needs to fragment to preserve real cohesion. It pairs naturally with the circles of interest, which frames social and political life as the same kind of layered, concentric structure Dunbar later built around his original number.