Dunbar's number of 150 friends is disputed, not proven

Dunbar’s number is the claim, made by Oxford anthropologist Robin Dunbar, that humans can maintain only about 150 meaningful relationships at once: people you know personally and whose connections to each other you can keep track of. It is a well-known idea, but whether the brain really sets that limit is still disputed.

The number comes from primates. In a 1992 paper Dunbar showed that primate species with a larger neocortex relative to the rest of the brain live in larger groups, and extrapolating that line to the human brain gives about 150. He argues that relationships come in nested layers, each roughly three times the size of the one inside it, from about five intimate friends out to some 1,500 people whose names we know. He points to Christmas-card lists, military fighting units and villages in the Domesday Book as groups that land near 150.

Why is Dunbar’s number contested?

Because re-running the analysis gives wildly different answers. In 2021, researchers at Stockholm University re-ran it on several updated primate datasets. Different methods produced average estimates ranging from the teens to over a hundred, with 95% confidence intervals as wide as 4 to 520, and they concluded that a cognitive limit on group size cannot be derived this way.

Dunbar replied that the critics used the wrong kind of regression. On the same data, he says, ordinary least-squares regression predicts 71, while the reduced-major-axis method he used predicts 158. So much of the argument is about statistical technique, not about how many friends anyone has.

Is the 150 limit used in the real world?

Yes; the idea has already escaped the lab. It featured in Malcolm Gladwell’s The Tipping Point, and the 2021 paper notes that the Swedish Tax Authority reorganised its offices to stay within the 150-person limit.

The most defensible reading is that 150 is a rough, debated estimate, not a measured ceiling.