« On Hurricanes and Politics | Gene Expression Front Page | The Honor of a Lady »
September 14, 2004

Grooming => Language ~ Larger Social Groups

In the post below I begin to explore Robin Dunbar's assertion that 150 is a crucial threshold in human social organization. I believe this central fact is really what Grooming, Gossip and the Evolution of Language focuses on, language itself is almost an afterthought. Dunbar sums up his most important theses in the final chapter:


  1. "...among primates, social group size appears to be limited by the size of the species' neocortex...."
  2. "...the size of human social networks appears to be limited for similar reasons to a value of around 150...."
  3. "...the time devoted to social grooming by primates is directly related to group size because it plays a crucial role in bonding...."
  4. "...it is suggested that language evolved among humans to replace social grooming because the grooming time required by our large groups made impossible demands on our time...."

As you can see, group is the ultimate, while grooming and language are proximates, adaptations to deal with the reality of large groups.

One can imagine all sorts of reasons why individuals embedded in larger coalitions of hominids would be more reproductively fit than those who were members of smaller groups. As I suggested in my previous post, I suspect these units of 150 are "basic" or "atomic" units, not necessarily the smallest, but operationally crucial in forming a reductionistic theory of sociology. Obviously there is a great deal of intragroup competition, and the "free rider" is always looming on the horizon. This is the reason that humans have a well developed social intelligence, to perpetuate a stable group equilibrium. Earlier I made the analogy to the individual: even though individuals are emergent from smaller units, the organism can be treated as a unitary whole. But, if you read books like The Cooperative Gene, you will note that there are sometimes cracks in the intraorganismic genetic "ceasefire."

Common sense tells us that groups of 150 are not as unitary as mulitcellular organisms, the intuitive human capacity for social living is a "good enough" solution, especially judging by the commonality of individuals who lack social skills. Dunbar notes that conversations can only truly flow together in groups of less than five, so the 150 limit does not imply a quorum of engaging equals, but a coalition of smaller semi-transient intersecting networks of intimates (~10). In this fashion enough exchange of information occurs so that penalties can be efficiently imposed on free riders who would degrade the coherency of larger groups.

Even though the book is 7 years old, I think it is an important (and short!) read for those interested in evolutionary psychology. Dunbar does not come to the number 150 simply through surveying history or taking on faith what comes out of his line-of-best-fit when he plots neocortex ratio vs. mean group size:

He has researched what modern humans report about their social networks, and again the ~150 number shows up, as if we really are souped up cavemen who remain caged in by the social lives of our ancestors. Additionally, Dunbar offers that there are physiological & anatomical constraints in the EEA which would have prevented our neocortexes from increasing in size (the metabolic cost of a large brain resulting in a smaller gut and so a more energy rich diet as well as the stress that the infant's large cranium fitting through the pelvic opening induces). There is also the problem that the complexity of a social network does not increase linearly. At a certain point one would need a brain the size of an elephant to handle all the possible permutations of social interaction.

Dunbar's case for the importance of 150 is clearly convincing for me. I don't know if I buy his argument about the importance of language and its relationship to 150. I find it plausible, and don't have an alternative thesis, but there are too many question marks between the development of archaic Homo sapiens and the putative "Great Leap" of cultural creativity ~50,000 years ago in my mind. There are issues where the 1997 date shows, Dunbar wrote before revisions in "The Great Leap" theory of human biocultural evolution began to creep in, and he also did not know about the language gene, FOXP2. I don't know how he would revise his argument in light of these facts, but I do think that his interlocking cascade of conjectures would be altered.

E.O. Wilson once said of Communism, "Great idea, wrong species." In the generality Wilson is correct, but there are elements of "communism" prebuilt into human societies. Familial units often share property communally, and sometimes extended family clans also do this. The Hutterites, who live in groups of 150, practice some level of communal property. Wilson's quip might be reformulated as "Great idea wrong level of social organization [1]." Marx and his successors who propogated "scientific socialism" lived under the delusion that sociology and the affinal human sciences were in fact peers to the physical sciences in their ability to reduce the object of study to a few variables and infer predictions. As Wilson and others from the natural sciences have asserted, social science has no theory [2]. Social science often involves broad descriptions of trends and patterns without attempting to really reduce the variables down to a level where predictions can work back up the ladder of complexity to validate a hypothesis. Obviously scholars realize that humans often organize at discrete levels of complexity, from family, clan, tribe and so forth, but a a scientific and multi-pronged examination of these terms can help us consider which ones are truly useful beyond the stage of verbal description, that is, which ones are elemental in the construction of higher levels and the shape they take.

Individual level selectionism has reshaped evolutionary biology. It has been a fruitful avenue of inquiry which has yielded a great number of testable hypothesis (some confirmed, some falsified). Today, social scientists and historians have an avalanche of facts but little rhyme or reason to their organization. This results in faux "theories" spun out of Post Modernism or artful but subjective Just-So factual fantasy like The Decline of the West or A Study of History. A basic understanding of the biological basis of the human organism can augment the method of the human sciences. Older historical theories like "diffusionism," which seemed to posit each cultural invention as a sui generis miracle can have been critiqued in light of the fact that the human brain is not a blank slate, but is subject to universal constraints and common biases shaped by an ancient environment that our species shared. Genghis Khan need not have settled on the number of 100 for one of his basic military units because he borrowed from the Roman example (via some other culture or state), rather, it emerged out of commonality of human experience derived from the limitations of our mind shaped by the forces of evolution!

To understand Newtonian Mechanics you do not have to reduce down to the level of quarks. To understand ecology you do not have to reduce down to the level of the biomolecule. Perhaps to understand the human sciences you do not have to reduce down to the level of the individual, rather, perhaps 150 is some elemental unit. We really don't know yet, the theory needs to be tested and fleshed out. Human choice and the random vicissitudes of history and environment shape how human societies develop, but it seems quite likely that there are universal constraints and variables out of which this contingent complexity develops.

Related: Great post from Life With Alacrity, much more thorough than my entry....

[1] There are a lot of issues of accuracy here, as families and communal units are often in flux, and some concept of personal property obviously exists. The more important point is that in groups of under 150 one could conceive of a social situation where some degree of communism might be tenable, usually augmented by real or fictive kinship (I could religious as a form of fictive kinship). This is a utopian fantasy at higher levels sans totalitarian methods of control, and even that is futile (as we know).

[2] I know economists might reject this characterization.

Posted by razib at 01:39 PM