02 October 2017

Dunbar and Brain Size and Triratna

One of my colleagues wrote something, a little vague about the importance of the number 150 in human society, and since I have a long fascination with this, I thought I would write a brief introduction.

Dunbar and Brain Size

In 1992 Robin Dunbar published a paper in which he compared the average neo-cortex-to-brain-volume ratio in wild primates with the size of their social groups. There was a linear relationship which enabled him to predict that the average human social group would be 150.

Given that many of us live in cities with millions of people, what does this mean? It means that we use the most recently evolved parts of our brain to keep track of relationships - and to imagine how other people see the world, especially how they view their own relationships. This is an essential skill for a social mammal.

For example, all social mammals understand and operate a system of reciprocity. Sharing food, resources, grooming, guard-duty, or mates etc creates obligations for other group members. If I share with you, you have a social obligation to share with me. And vice versa. In apes and humans, we also keep track of obligations that are between third parties. I may share with you, knowing that you share with Devadatta and that way come into indirect relationship with Devadatta. Devadatta will probably notice that I share with you, and my reputation with him increases. Ans so on.

Humans can routinely track these abstractions into 4th and 5th order. Shakespeare could imagine how his audience would feel that Othello would feel about Cassio, after being convinced that Iago thinks that Desdemona loves Cassio; while we also know that Iago is lying. Shakespeare could imagine our tension as the story progresses. What if Iago is found out? What if he is not? This is part of what makes Shakespeare a great story teller.

Keeping track of these social obligations takes brain power. The more of our brain given over to keeping track of such things, the more relationships we can keep track of.

The Magic Number 

Dunbar predicted that on average the maximum number of relations humans could keep track of in this way would be about 150. And it turns out that the average community size in the New Guinea highlands, the units of Roman armies, and the average village size in the Domesday book (and a whole range of other measures) was .... 150.

But 150 is not the whole story. 150 is the size of an intimate community where everyone knows everyone's business. But we are usually involved in both smaller and larger groupings. If 150 is a tribe, the a tribe is usually made up from several clans of about 50 members. Clans comprise several families of about 15 members. Each person has approximately 5 intimates. These groupings may overlap. On the other hand tribes may be part of larger groupings, of 500, 1500, and 5000 and so on. The smaller the grouping, the more intimate and detailed, the knowledge; and contrarily the larger the grouping the less intimate and detailing the knowledge. The limits seem to go roughly in multiples of three, starting with 5 as the smallest.

What we expect is that, in a society of 150 people who live together, relying on each other, each will know all of the others, and who is friends and relations with whom. They will be intimately familiar with trists and disputes. And they will know who has what status under what circumstances.

In larger groupings there may be people we don't know. Larger tribal grouping may adopt symbols of membership with which to recognise other members. For example they may hang a strip of white cloth around their neck. They know that anyone who has one of these white strips is a member of the tribe. They can expect to have some basic values and interests in common, and thus are open to each other socially in ways they might not be with complete strangers. It may even be the case where the tribe mandates certain levels of hospitality are required. Some cultures require this even for strangers, when travellers are particularly vulnerable (as in the desert).

In larger groupings there are a number of ways of ensuring that every gets a say in how things are run. But let's face it, beyond 7 ± 2 everyone having a say is unrealistic. This is another magic number (aka Miller's Number) and relates to the capacity of our working memory. Groups bigger than ca 9, tend to schism into separate conversations, unless formal procedures apply.


With respect to schism, the 150 level is the limit of a sense of knowing everyone in a society living together on a daily basis. Much beyond it, and some of the people are going to start seemingly like relative outsiders. We don't know who they are friends with for example. This may explain why when humans meet who are part of a larger less intimate grouping, they often exchange information that establishes *who* they know. It's likely that some above average connectors know many more people, and across social networks. They are the glue that hold larger groupings together.

There is no absolute requirement to schism at any number. Schisms happen in small groups and large. But primates feel more comfortable with groups where they know the others. Being surrounded by strangers is often quite stressful for a social primate because they have none of the knowledge they need to know how to relate to everyone. On the other hand, being experts at empathy, primates pick up this info very quickly.

What tends to happen is that we are comfortable being relatively informal members of several larger groups, but prioritise our most intimate relationships and family.

The Order

The Order is complicated because most members of the Order are still enmeshed in other groups, particularly family. Even if there were only 150 of us, we don't live together as one community, relying on each other to survive. It is already a somewhat looser grouping than that, so the fact that it has crossed several thresholds (in total membership) is not a clear cut indicator of anything. Those who were around in the early days do tend to reminisce about how good it felt when you knew everyone. I would expect nothing less. But they all still had friends and family outside the movement too.

The Dunbar Number describes the dynamics in close knit societies living together. Beyond 150, such communities do tend to split into more manageable groups.However, it doesn't really say anything about the Order because we are not that kind of society.

Note that the more plugged into other groups we are, the fewer relationships we can track in the Order. And vice versa. Note also that pair-bonding makes no difference to Dunbar's numbers. Primates adopt a wide variety of lifestyles and these are secondary.


Dunbar's original article rapidly became a classic of anthropology and evolutionary psychology (the latter being Dunbar's main subject of interest). His predictions became known as Dunbar Numbers while he was comparatively young (he is still alive and working at Oxford University). If there was a Nobel for evolutionary psychology, he'd have won it for this discovery.

As a final caveat I would insist that Dunbar's numbers are theoretical averages, albeit with considerable empirical support. There will be a bell-curve on which individuals sit. Some will easily cope with 300 relations, some will barely cope with 50. There will always be outliers, but the existence of outliers does not alter the theory or the supporting empirical evidence of accuracy.

For further reading on the Dunbar Numbers and other concepts mentioned above, I very highly recommend "Human Evolution" by Robin Dunbar, published by Pelican in 2016. Aimed at a general readership, and highly readable, it nonetheless takes a cutting edge look at human evolution by incorporating Dunbar and his group's research on group sizes and theory of mind. Dunbar explains how we went from being general purpose apes, to highly specialised humans. How we solved the energy gap required by our big brains and big social groups through cooking, dance, laughter, and religion.

No comments:

Post a Comment