Other Mental Models · OM-08
There's a cognitive ceiling — roughly 150 people — on how many stable social relationships a single human being can actually maintain, no matter how much technology promises to expand our social reach.
A proposed cognitive limit, roughly 150, on the number of stable social relationships a person can maintain at any one time — relationships in which one knows who each person is and how each person relates to every other person in the group — derived from a correlation between primate brain (specifically neocortex) size and observed group size across primate species.
Proposed by British anthropologist Robin Dunbar in the early 1990s, based on his research correlating neocortex size with typical social group size across primate species, then extrapolating the resulting ratio to predict a human cognitive limit.
The Mechanism
A cognitive ceiling on stable relationships, not an arbitrary social convention
Dunbar's original research found a consistent ratio between neocortex size and typical group size across dozens of primate species, and extrapolating that same ratio to the human neocortex predicted a group size of roughly 150 — a number subsequently found to recur across surprisingly varied real-world human groupings, from hunter-gatherer band sizes documented anthropologically to the typical size of a military company and certain historical village and clan sizes.
01 · THE LIMIT IS COGNITIVE, NOT SOCIAL OR CULTURAL, IN ORIGIN
The proposed mechanism is a hard constraint on brain processing capacity for social relationships
Dunbar's underlying claim is that maintaining a stable relationship — tracking a person's identity, history, and their relationships to everyone else in the group — imposes a real cognitive processing cost, and that the specific size of the human neocortex sets an upper bound on how many such relationships can be simultaneously tracked, independent of any particular culture's social norms or technology.
02 · SOCIAL MEDIA AND DIGITAL COMMUNICATION DO NOT APPEAR TO RAISE THE NUMBER
The proposed cognitive ceiling appears to persist despite technology that expands reach
Despite the vastly larger number of 'connections' or 'followers' possible on digital platforms, subsequent research (including Dunbar's own later studies of social media usage) has generally found that people's networks of genuinely close, stable, actively maintained relationships remain clustered around the same rough scale — technology appears to expand the number of weak-tie contacts far more than it expands the number of relationships that meet Dunbar's stricter definition of stable and known-in-relation-to-others.
03 · THE NUMBER, AND THE UNDERLYING METHODOLOGY, REMAIN GENUINELY CONTESTED IN THE RESEARCH LITERATURE
This is a real scientific debate, not settled consensus
Subsequent researchers have both found supporting evidence in various group-size datasets and separately raised substantive methodological critiques of the original neocortex-ratio extrapolation, including studies suggesting the true number could plausibly range more widely depending on definition and dataset — Dunbar's number is a genuinely influential and widely cited but not universally accepted precise figure.
Where It Fails / Inversion
Where it fails / inversion
The precise number 150 should be treated as an approximate, contested order-of-magnitude claim rather than an exact universal constant — actual observed group sizes vary considerably by context and definition of 'relationship,' and the underlying research methodology (extrapolating from primate neocortex ratios) has been genuinely challenged, not merely a settled scientific fact.
How To Use It
Worked example · structuring an organization's team sizes
An organization scaling past roughly 150 people in a single functional unit, without deliberately introducing sub-team structures, may find that informal, trust-based coordination that worked at smaller scale (everyone roughly knows everyone and their role) breaks down — deliberately structuring the organization into smaller sub-units of a size closer to Dunbar's proposed limits, each with clearer internal relationships, is a commonly cited practical application, though the specific optimal size for any given organization should be validated against its own actual coordination needs rather than assumed from the number alone.
How to use it
When an organization, community, or social structure seems to lose cohesion past a certain size, consider whether it has exceeded the rough cognitive limits on stable relationship-tracking that Dunbar's research points to — restructuring into smaller sub-groups, each within that rough range, is a frequently effective, if not universally validated, response.
See Also