How Many Friends Does One Person Need?: Dunbar’s Number and Other Evolutionary Quirks

This is brief write up of the Robin Dunbar book “How many friends does one person need?”.  I read the book as a balance to the way in which its very easy to start thinking of networks mathematically and systemically even though its incredibly important to remember the fact that social networks are made up of people and as a result will behave in messy, emotional and human ways.  Its also an important contrast for me with the enormous rationalist thinking of the age of enlightenment approach of Habermas’ Public Sphere for example.

Its very easy to ignore the ‘social’ bit of ‘social networks’ – when you are working online there is a pull towards a technological / engineering mindset which is difficult to resist as you build your environment.  Online space is created both by code and by the narratives and content that use that code – but the fact is that the code comes first and that often results in an environment that ultimately wants to resolve to a whole series of zeros and ones.  One of the exciting things for me in terms of technology development over the last few years is the way in which User Experience and user centred design has grown as a field – acknowledging the fact that we don’t all need to earn our ‘right’ to be online by hand coding at the command line.

Dunbar’s work looks at how groups work, not as networks or learning sets but as biological and social imperatives.  His research discusses the evidence base that shows our social behaviour is innate and our forming of a groups a necessary part of our humanity.  He is a longstanding science journalist as well as a researcher and he is best know for coining the term the ‘Dunbar number’ which he defines as:

“….as the set of people who, if you saw them in the transit lounge during a 3 a.m. stopover at Hong Kong airport, you wouldn’t feel embarrassed about going up to them and saying: `Hi! How are you? Haven’t seen you in ages!’ In fact, they would probably be a bit miffed if you didn’t. You wouldn’t need to introduce yourself because they would know where you stood in their social world, and you would know where they stood in yours. And, if push really came to shove, they would be more likely than not to agree to lend you a fiver if you asked.”

He’s describing the number of people you can trust and have an emotional affinity to – and the number got considerable press as at the time it was very close to the average number of ‘friends’ each person had on Facebook.

I just want to pause at this point and be clear that we are really talking about two different definitions of ‘friend’ here.  Dunbar is using the term to refer to people who we have an emotional attachment to which meets certain criteria.  In the context of Facebook ‘friend’ is used far more loosely – would you lend £5 to everyone who you are friends with on Facebook?

Dunbar talks about the ‘Dunbar number’ as being an evolutionary limit – basically our brains can’t handle more connections than this.  There are also similarities with clan group sizes in tribal cultures – groups of this size as being discernible within overall clan sizes of between 500 and 1500 people.

This is not to say that he thinking all social grouping larger than that are doomed – but at that point we need different social structures in order to provide the sense of mutual obligation and reciprocity that communities need in order to function – it can’t just be done on the basis of that emotional bond.

Dunbar also describes patterns in these larger social grouping which work on broadly a factor of three.  This builds up from what social psychologists call a ‘sympathy group’ of 12-15 people whose death would leave you distraught , to groupings of 50 which are found in aboriginal society, through to a group size of 1500.  Interestingly Plato went one further than that when he suggested 5300 as the ideal size for a democracy.

He also talks about the need for physical contact – the chimps grooming each of other – and the fact that language provides another kind of social grooming with gossip having a role in providing the glue that keeps communities together – passing information about each other indicates you are part of the same group.

One of his overarching points is the fact that our brains evolve a lot more slowly than the pace of technological and social development – the aphorism of ‘stone age brains in a space age universe’ – and that we need to remember this when we consider our reactions to change around us and take into account the fact that our first reaction may be an emotional rather than a rational one.

This is a challenge to the thinking of sociologists like Habermas who approach the question of society and democracy with an age of enlightenment rationality which often fails to take into account that human reactions are often irrational and based on our feelings of belonging rather than an idealised view of how we want the world to me.  Successful orators tap into this emotional response and any effective communicator is appealing to you emotionally as well as rationally.  What’s more Dunbar explains that judgements about morality and practical utilitarian decisions are not located in the same part of the brain and are not necessarily called on at the same time – Hume was right when he talked about the idea that morality is born of our emotional response and not our rationality – we are not always thinking when we act.

So what is it that forms that social glue that turns friendship groups of 150 into communities of 5000?  Dunbar talks about gossip and conversational grooming.  He also points out the role of religion in creating this common framework.   Laughter and music also generate the endorphins that we need to ‘bond’ with other people.  What is clear is that he is suggesting shared activities and opportunities to connect to people beyond your immediate circle of friends.

Relating this back to civic space and hyperlocal communities – it would be interesting to look at the numbers of participants to see how they relate to Dunbar’s groupings and it would also be interesting to look at how these numbers relate to the types of tools and interactions that work well with different numbers in the groups.  There is also little work as yet as to whether or not the Dunbar number and the tribal based group behaviour patterns translate in any way to online interactions.

Dunbar (and evolutionary biologists) provides a useful balance to the far more rational approach taken by many of the authors looking at the network society and provide an important reminder that the our emotional needs are as important – or more important – than our intellectual needs.  As we contemplate our to re-connect people to their local communities – either online or offline – its important that we remember this perspective and design for hearts as well as heads.