Dunbar’s Number is an estimate of the sizes of human and primate social networks. Dunbar discovered that the size of the neocortex is strongly correleated to group sizes.

The average maximum network size for humans is 147.8. Some individual humans may have a larger or smaller maximum social network size. The mean clique size is just under 12. There are cognitive limitations on group sizing, so very large networks have shallow relationships compared to the intimate network.

Humans are one species of Apes. All ape species travel in troops of different sizes and it’s interesting to discover possible reasons why. Social networks and cohesion are related to intelligence – this also shows that there are cognitive limitations on human group size.

Here’s an article measuring the Neocortex Size and Social Network Size in Primates by H. Kudo and R.I.M. Dunbar.

They first measure the ratio of neocortex size to overall brain size. The neocortex governs advanced reasoning and intelligence in mammals. The larger the neocortex ratio, the larger the social networks.

The authors make a distinction between cliques and networks:

We consider two aspects of the substructuring
of social relationships within primate groups. One is the
number of primary social partners that each animal has
(which we term its clique size); the other (which we term
network size) is the total number of individuals who are
linked into a single chain such that it is possible to pass
from one member to any other member either directly or
through a series of intermediates, each of whom is linked
to the next member of the chain as a primary social

Chimpanzees and Bonobo have the largest neocortex ratios excepting humans. Their neocortex ratio is 3.22 and their mean clique size is 3.07 and 3.2 respectively. The Human ratio is 4.10 with a mean clique size of 11.8 (Dunbar)


“The reason why neocortex size, group size and clique size are all interrelated is probably that coalitions allow animals to minimize the costs of living in groups. The costs of grouping are considerable for primates, and include not only the conventional indirect costs (increased day journey length, added feeding and travel time costs) but also significant direct costs in terms of competition over resources (van Schaik & Janson 1988; Dunbar 1988) and reproductive suppression induced by harassment and stress (Bowman et al. 1978; Abbott et al.1986). The latter costs are particularly significant for females, who may risk considerable reductions in lifetime reproductive output if they remain low ranking in large groups (e.g. gelada: Dunbar 1987).”

Human social networks are somewhat more difficult to study than primates. Human networks constantly change in composition and tend to overlap with one another. Humans may be in contact with a larger number of other individuals, but the relationships are very shallow, distant and interchangeable. The goal is measuring the number of valued members of a social network.

There are different methods of measuring intimate network size: frequency of contact, depth of knowledge of another’s personal affairs, ability to ask for personal favors, people to count on in emergencies, etc.

Dunbar (and many others) have researched the average social network sizes in human societies, from hunter-gatherer tribes to modern industrialized cities. In modern Britain, the same pattern holds.
Social Network Size in Humans By R.A. Hill and R.I.M. Dunbar

Here’s the Abstract:

“Maximum network size averaged 153.5 individuals, with a mean network size of 124.9 for those individuals explicitly contacted; these values are remarkably close to the group size of 150 predicted for humans on the basis of the size of their neocortex. Age, household type, and the relationship to the individual influence network structure, although the proportion of kin remained relatively constant at around 21%. Frequency of contact between network members was primarily determined by two classes of variable: passive factors (distance, work colleague, overseas) and active factors (emotional closeness, genetic relatedness). Controlling for the influence of passive factors on contact rates allowed the hierarchical structure of human social groups to be delimited. These findings suggest that there may be cognitive constraints on network size.”

Even eliminating factors like geographic separation cannot explain why mean group size does not increase over 150 and mean clique size over 12.

In discussing the Evolution of Language, Dunbar speculates that language replaced grooming. In larger social groups, grooming consumes up to 42% of the group’s time in order to maintain social cohesiveness. Language is less time consuming and more efficient.

Only in rare cases and under extreme pressure to survive will humans maintain groups of 150. Dunbar points out that this is the case for military units and small tribes living on the margins of society.

The number 150 should be considered a maximum limit for cohesive networks. Beyond that number, there is greater fragmentation and conflicts of interests between members which cause the group to split.

Many group sizes are considerably lower than 150. Cliques of 12 work reasonably well for some activities and groupings of 70-100 are frequent for other activities.

Beyond a human’s network of 150 individuals, the person can only maintain shallow and temporary relationships due to cognitive limitations.