In a book called Evolutionary Psychology: A Beginner’s Guide (2005), Robin Dunbar, Louise Barrett and John Lycett addressed this issue.  The book drew on additional research that can be found in Grooming, Gossip and the Evolution of Language, by Robin Dunbar (1997).

We don’t have limited numbers of friends and acquaintances merely because we choose to have such limited numbers.  Rather, as explained in these two works, physiological limitations constrain human social choices. We are limited in the number of acquaintances we can have because we are physiologically limited.  This is another example that those who claim to explain human animals without the benefit of careful science do so at their own risk.

Human societies are complex social environments.  Archaeologists have determined that pre-modern humans lived in small-scale hunter gatherer societies “characterized by very small, relatively unstable groups, often dispersed across a very large area.”  Only after agriculture was developed (10,000 years ago) did large permanent settlements become possible Living in groups gives members huge advantages such as reduced predation risk (we benefit from the “many eyes” advantage and large groups of individuals deter most predators).

Group living comes with costs, too.  We have conflicts over limited resources, such as food and mates.  Group living stresses immune systems too.  The menstrual cycles of female primates are disrupted.  In order to obtain necessary food, humans need to travel further each day. Associating with large groups of people also has a huge mental cost.  In order to live safely within large groups, we need to know who is who.  Who are our friends and enemies?  We need to know the individual personalities of the members of the group.  We need to keep track of favors owed.  We need to keep track of the cliques and alliances.  We need to strategize to identify and work with high-ranking members of the community.  Keeping track of other people’s relationships is very difficult work, indeed.  Not only do we need to keep track of relationships among people who are present.  We also need to keep track of relationships among absent individuals (“virtual individuals”).

It’s for these reasons that many have suggested the Machiavelli intelligence hypothesis: that “evolutionary pressure selecting for a large brain size and super intelligence in primates did seem to have something to do with the need to weld large groups together.”  (Grooming, page 60).  Also known as the “social brain hypothesis” the theory suggests that:

the demands of living in permanent social groups selected for a kind of intelligence was particularly adept at tracking the relationships that exist between oneself and all the other members of the group and, more importantly perhaps, keeping track of the relationships that the other animals in the group have with each other.

Robin Dunbar and his co-workers conducted a series of studies showing that the volume of the neocortex “correlates with various measures of social complexity across the primates.”  Primates that live in small social groups tend to have small neo-cortices.  Primates that live in larger groups have larger neo-cortices.  Until Dunbar conducted his studies, however, it was not well appreciated that there is an almost straight-line correlation between group size and neocortex volume within groups of primates.  It was “a remarkably good fit.”  This relationship between neocortex size and group size in primates can be extrapolated to predict that humans would have a group size of about 150 (Grooming, page 69).  Baboons and chimpanzees have the next largest group sizes (50-55) and neocortex volume. Dunbar suggests that “once the community exceeds 150 people, it becomes increasingly difficult to control its members by peer pressure alone” (Grooming, page 72).

This number (150) fits nicely with with the number of humans (100-200) who (according to archaeologists) collectively shared rights of access to the resources within their territory. 150 seems to be the largest grouping in which everyone knows everyone else, “in which they know not simply who is who but also how each one is related to the others” (grooming, page 71). This number roughly equates to the number of living descendents a person would expect an ancestral couple to produce after four generations (Grooming, page 71).  Various experiments have shown that 150 is a significant number in many sorts of human groupings.  A recent study of the Church of England showed that the ideal size for congregations was less than 200.  During World War II, a company has stabilized in size at around 170 soldiers.  A study of Christmas card distribution lists showed that a typical person who sends such cards sends them to approximately 154 individuals. 

The results of various studies show that

Human societies contain buried within them a natural grouping of around 150 people.  These groups do not have a specific function: in one society they may be used for one purpose, in another society for a different purpose.  Rather, they are a consequence of the fact that the human brain cannot sustain more than a certain number of relationships of a given strength at any one time.  The figure of 150 seems to represent the maximum number of individuals with whom we can have a genuinely social relationship, the kind of relationship that goes with knowing who they are and how they relate to us.  Putting it another way, it’s 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.

(Grooming, page 77).

Within our networks of 150 there are significant subgroupings.  For instance, there is strong evidence to suggest a sympathy group of about 12-15, which corresponds to the number of people we contact at least once each month.

According to Dunbar, the large size of human groups raises the interesting question of how humans maintain their social bonds.  Whereas grooming seems to be the main method of bonding in other primate groups, there are huge obstacles for using grooming (or any sort of touching) for human bonding, certainly so in groups of 150.  That led Dunbar to suggest that humans actually do groom, although they do so verbally.  According to Dunbar, then, gossip is a critically important bonding mechanism for humans. This makes good sense, given that gossip is a ubiquitous among human animals, as well as time consuming.

Dunbar’s findings regarding regularities in the sizes of human groupings and the importance of gossip and maintaining social bonds are critically important in that they substantiate an important use of language that has nothing to do with the conveyance of verbal information.  For a discussion of one such application of verbal grooming, see here.