In 1992, a British anthropologist named Robin Dunbar noticed something curious about primates. The more complex a species's social life, the larger its neocortex — the outer region of the brain associated with higher cognition. Dunbar plotted the relationship across 38 primate species and ran the regression. The line was tight enough to predict, given a primate's brain, roughly how big its social group should be.
Then he plugged in the human neocortex.
The number that came out was roughly 150.
That number, give or take a wide margin, has come to be known as Dunbar's Number — and despite some genuine controversy in recent years, it has been one of the most influential ideas at the intersection of evolutionary biology, sociology, and organizational design.
What the Number Actually Refers To
Dunbar's Number is often quoted casually as "the number of friends you can have." That is not what Dunbar meant. The number refers to the size of a stable social group — people whose existence, history, and relationships you can keep track of well enough to maintain a meaningful connection. In Dunbar's phrase, it is "the number of people you would not feel embarrassed about joining uninvited for a drink if you happened to bump into them in a bar."
It is not your closest friends. It is not all the names you can remember. It is the layer where you genuinely know who someone is in relation to you — their family, their work, recent events in their life, the texture of the relationship.
Dunbar later refined the model into a series of nested layers, each roughly three times the size of the previous:
- ~5 — closest intimates. The people you call when something terrible happens.
- ~15 — sympathy group. Close friends and family.
- ~50 — good friends. People you'd invite to a significant event.
- ~150 — meaningful contacts. Dunbar's Number proper.
- ~500 — acquaintances. Faces you recognize, names you sort of remember.
- ~1,500 — recognizable but unknown. People whose names ring a bell.
Each layer absorbs more cognitive resources. The closest layers require frequent, high-quality interaction; the outer layers can be maintained with much less.
Why 150?
Dunbar's hypothesis is that maintaining a stable relationship is cognitively expensive. You have to remember who someone is, who they are connected to, what their history is with you, what they prefer, what they have promised, what you owe them. Multiply that across a social group, and you quickly hit a ceiling on how many people any given brain can keep track of.
Why does the ceiling sit roughly at 150? That is where the math from primate brain size lands. Dunbar argued that human group size historically reflected this constraint. He pointed to a striking range of examples:
- Hunter-gatherer band sizes — typically 100–200 across cultures.
- Neolithic village sizes — many archaeological sites cluster in the same range.
- Roman maniples — the basic infantry unit, ~120–160 men.
- Modern military companies — typically 100–200 soldiers.
- Hutterite colonies — split when population exceeds about 150, by long-standing practice.
- Average Christmas card lists in 1990s UK studies — roughly 150 names.
The convergence is suggestive. It seems to argue that something about human social cognition keeps gravitating to the same scale.
Where the Idea Got Misapplied
Pop business writing turned Dunbar's Number into a management slogan: organizations should stay under 150 people; community apps should target groups of 150; if your team grows past it, you need to subdivide. Dunbar himself has said this oversimplifies the research.
The number is not a hard cap. It is a central tendency with substantial variance — different studies put the range anywhere from 100 to 250. It describes a constraint on the cognitive maintenance of relationships, not a magic number for productivity. A 200-person company can function fine if it does not require everyone to know everyone.
The number also says nothing about which 150 people you can keep track of. The set is dynamic. Move to a new city, and the names slowly turn over. Joining a new church, a new team, or a new community shifts the composition without changing the ceiling.
The Recent Controversy
In 2021, a study by Patrik Lindenfors and colleagues in Biology Letters re-ran Dunbar's primate analyses with newer data and broader statistical methods. They found that the relationship between brain size and group size held in some forms but produced dramatically different point estimates — anywhere from 4 to 520 people, depending on how the regression was specified. Their conclusion: the precise number 150 is not statistically defensible from primate data alone.
Dunbar responded that the broader pattern remains, even if the specific number is debatable, and that the convergent ethnographic and historical evidence supports the order of magnitude (100s, not 1,000s). The honest scientific position now is that the layered structure of human social networks is real, the cognitive constraint on close relationships is real, but the exact "Dunbar Number" is more rule of thumb than law.
Why It Still Matters
Even with the controversy, the underlying observation has held up across decades of research: human social networks are organized into nested layers; closer relationships require more time and effort; and there is a real ceiling on how many people any one person can know meaningfully.
That has practical consequences. Time is the main currency that maintains relationships. If you spend most of your time on your top 5–15, you will have less attention for the next layer out. Social investment is zero-sum, even if friendship feels like it should be infinite. The closest layers do not survive without contact.
It also has consequences for technology. A phone that lets you stay loosely connected with 1,500 people does not mean you have 1,500 friends; it means you have many weak ties and possibly fewer strong ones. Volume of contact is not the same as depth of contact, and the brain still has to do the hard work of knowing who someone really is.
What to Do With It
Dunbar himself, asked in interviews for practical takeaways, suggests two:
Tend the inner layers first. The 5 and the 15 are the people whose presence in your life is doing the most for your wellbeing. Keep up actual time with them.
Be honest about the cost of the outer layers. You cannot maintain 500 close friends. Trying to do so usually means doing the closer ones poorly. Recognizing the constraint is the first step in spending attention well.
The number may not be exactly 150. The pattern is closer to a small number of intimates, a slightly larger group of close ties, and a bounded ring of meaningful contacts. That is the architecture of a human social life. Building on it deliberately is one of the better uses of self-knowledge research has to offer.
