A chattering class apart

The feel-good factor, that grail of all democratically elected politicians, has a special place in our primate ancestry. The bonds which bind most ape and monkey societies are based on the sense of well being which stems from contact, touch and the closeness arising from the repeated rhythms and caresses of grooming. Pet owners, for example, use such behaviour to form a social bond with another species. The amount of time we spend every day either patting the dog or stroking the cat to re-affirm these unspoken ties can often be considerable.

But think about the time management problem which faces, not the owner of 101 dalmatians, but a primate like a baboon in a group of 100 other baboons. Here grooming is a serious business since it binds individuals into social contracts that are vital in providing support during disputes over food and mating opportunities. The baboon, like all other primates, has to balance the demands on its time for sleeping, eating, moving to new feeding patches and socialising. The outcome is that the size of the baboon's immediate social world is limited even though they may live in large groups. Given the mechanism of one-to-one grooming, a baboon is unable to bind more than a very small number, from among a potentially much larger number, to its social agenda. Primate society is literally groomed into existence. It is a hands-on experience that allows any individual ape or monkey a degree of self-determination in building networks and organising support groups.

The problem Robin Dunbar tackles in this thought-provoking book is how we overcame the limits of our primate heritage, where society is built by grooming, to fashion much larger and more complex societies. What is novel about his approach is that rather than insisting on the obvious importance of language as the mechanism which made this possible, he instead provides the long-term evolutionary argument for its appearance and backs this up with fossil evidence. What emerges is a fascinating and well-rounded account of what was involved in becoming the language-using primate.

The evolutionary argument comes down to cost. Our big brains are expensive. They may add up to only 2 per cent of our body weight but they consume 20 per cent of all the energy we consume as food. During human evolution they could only grow in size if other internal tissues became proportionately smaller. Big heads led to smaller stomachs which had the knock-on effect of needing higher-grade foods to fuel the newly proportioned hominid. However, Dunbar's great insight lies in the correlation he produces between brain size and group size for all primate species. Basically the bigger the group the bigger the neocortex (literally the grey matter or thinking part of the brain) turns out to be. Since neocortex and total brain size are highly dependent upon each other it is possible to use measurements of fossil brain cases to estimate the amount of neocortex they once contained. During human evolution brain size increased steadily which implies that group size also expanded. Therefore, the fossils can be used to provide a timetable for the appearance of larger groups.

Why did human evolution go in this direction? Could defence have been the selection pressure which led to larger groups? Whatever the reason a new mechanism, other than grooming, was needed to ensure that they were coherent and lasting social units.

In Dunbar's scheme, language was the mechanism for individuals to manage other people in these larger groups. Alternative ideas that language stemmed from passing on hunting information or from ritual are dismissed. Instead it arose as a form of vocal grooming where bonds could be established and re- affirmed more quickly and to more than one person at a time. The social monkey which is limited by other demands to a maximum of 30 per cent of its time being devoted to grooming is outstripped by the talking human who can reach and influence a much larger audience in half the time.

The advantages of language over grooming are very well demonstrated. But what exactly are these larger groups predicted from the size of our neocortex? Dunbar presents evidence from ethnographies, address books, military training manuals and other sources to argue that there is a figure of 150 people whom we can know and interact with before the cognitive load of information overwhelms us. Such a high figure in comparison to any other primate is made possible by language, not as a means of symbolic communication, but rather as a means of passing on social information about the members in this group. These neocortex groups, supported by gossip, be they military companies or church congregations are the outcome of vocal grooming between three or four members in coffee bars, pubs and all the other places where social life is negotiated.

His book is interspersed by his own and his students' studies of such conversations; what was said, who they were looking at when talking and how many people were listening. They discovered, perhaps unsurprisingly, that social life is one big soap opera where gossip about others' interests us much more than politics or philosophy.

One problem I see with groups of this nature is that people often belong to more than one group. The soldier may be part of a company of 150 soldiers but will also have family, friends and other social contacts outside. In prehistory there is little suggestion that such numbers formed residential groups. Rather the 150 would have been scattered over the landscape, rarely if ever meeting as a single group. Individuals and smaller units would meet more regularly, suggesting that Dunbar's argument would be stronger if instead of groups he concentrated more on how people construct different types of network.

From a contemporary perspective we see that individuals build intimate, effective and extended networks using emotional, material and symbolic resources in ever-increasing proportions as the size of each network moves from approximately five to 25 to between 150 and 500 people. While language affects relationships at all three levels it is easy to see that physical contact, or grooming, necessarily declines for the same reasons facing the time-stressed baboon. That level of emotional intensity can only be supported among a small, intimate number. What has happened in human evolution has been the extension of other kinds of social relationship to ever-larger networks which had the consequence of producing group structures. This is a rather different view from the view that the group came first.

Here archaeology becomes essential to firm up Dunbar's argument based on the skulls alone. The timetable of 250,000 years for the appearance of brains of sufficient size to produce groups of 150 people is not reflected in any comparable changes in the archaeological record. While this does mean that Neanderthals vocally groomed each other with a gossip language, we have to wait another 200,000 years before any effects of social and geographical extension are found. This involves a dramatic burst of colonisation across oceans to Australia and the Pacific and through taiga and tundra into North America. It may be, as Dunbar argues, that at the same time we also see the emergence of a symbolic rather than a gossip language: figurative art, architecture and rich burials.

To make his case for language as the evolutionary means of binding together larger groups, Dunbar has played down the importance of material culture in human organisation. He also implies that social life is little more than living together. However, material culture is another bond which binds since technical acts are also social acts. Making things is also about making relationships. This is not an argument that bigger brains were driven by the need for better technology but rather that an emphasis on the evidence for past action indicates another mechanism by which social networks are created and maintained. Therefore the evolution of human society is not just a development from grooming by touch to grooming by language. Group size does not express the complexity of social relationships just as brain size does not adequately predict a hunters' camp site in the Kalahari versus a city the size of London. Dunbar has given us a framework and a clear indication that language and network size evolved slowly rather than in punctuated bursts. Now we can probe this structure for its information about past societies.

The great value of this book is that it strips the evolution of language of its uniqueness. By reminding us that language is about people and by making it understandable as a long-term evolutionary process he has demystified what many regard as off-limits to palaeoanthropology.

Dunbar's achievement is to show that we can learn about language from our closest animal relatives, who lack the facility, and from the fossil skulls, who can be made to talk if we put the question in the right way.

Clive Gamble is professor of archaeology, University of Southampton.