The Dunbar number: the reason my family consists of just one hundred and fifty people

I was at a marvellous family event in Bangalore recently: an uncle and aunt's fiftieth wedding anniversary. It was a evening of laughter, music and food, brimming over with emotion, triggered in part by family photographs assembled from the past seventy-five years. When I came home, I started to explain to my daughters how all the people at that evening are related to us, and struggled. I still need to call my mother to clarify exactly how I'm related to my relatives.

I know exactly how I'm related to my first cousins, and for most part, to my second cousins. I'm hazy about genealogical links beyond that. Regardless, there are several members of my family, important people who I know and care about, who are in that hazy-genealogical-link zone. There are also several people who operate as members of the family despite not being specifically related either by blood or by marriage.

While my family is large and colourful, it is not unique. I know other Indian families which operate the same way. Mediterranean (and therefore Latino) cultures have a similarly capacious, rambling sense of family. Maybe family is the wrong word: clan, kinsfolk or the Welsh cymri might be closer to the mark. But what is it, this circle of kindred spirits? If it is not delineated by ancestry or marriage, is it infinitely large?

Professor Robin Dunbar thinks my clan isn't infinitely large, it consists of no more than one hundred and fifty people.

Dunbar's theory is that a stable social group is one in which "an individual knows who each person is, and how each person relates to every other person". Keeping track of this complex web of relationships is cognitively difficult, and gets ever more so as the size of the group increases. Dunbar's observation is that this web of relationships gets unmanageably complex, or too computationally intensive for the human brain, when the group size approaches one hundred and fifty. Empirically, we've lived through most of recorded history in settlements of about one hundred and fifty people.

This is sometimes mis-interpreted as "you can't have more than one hundred and fifty friends". I don't think Dunbar's number makes any such claim.

For instance, I play tennis with an Oxford-educated Asian-Brit landscape-architect called Jason. I know Jason as a tennis buddy, and in that limited context, he is a good buddy. I don't know anything about the other spheres of Jason's life - his Oxonian tribe, or his landscape-architect tribe - and therefore my friendship with Jason is not computationally intensive. It is one dimensional, rather than a matrix. I don't think Dunbar's concept in any way limits the number of simple, or one-dimensional, friendships I can have. The Dunbar number is about the number of people in a self-regulating matrix of relationships, not about simple one-to-one friendships.

I also don't think the Dunbar number limits the number of distinct tribes I can belong to. I belong to Dunbar-tribes from family, school, college and from work. Each tribe is manageable because it is no more than one hundred and fifty people. If it were all one network, it would be overwhelming. Facebook's new "smart lists" feature is an attempt to capture this reality.

Understood this way, my family is a constantly shifting network of about one hundred and fifty, with people, or entire branches, coming in and peeling off as time passes and circumstances evolve. At one time, say one hundred and fifty years ago, a large family like mine would have operated to strict command-and-control protocols. Family would have defined every aspect of identity. Now, the family, or the clan, operates as an opt-in network, in which location on the family tree is not irrelevant but is not defining either. Welcome to the brave new world of open source kinship.