Dunbar’s and Wellman’s Numbers

In my previous post I provided a temporal analysis of the main discussion group. I also said activity might have peaked because of Dunbar’s Numbers. What are Dunbar’s Numbers I hear you say?

Professor Robin Dunbar has researched the size and structure of human social networks. This research, combined with his work on primates, provides some interesting insights into our cognitive cache (to use a machine metaphor). To generalise his work, seemingly we have circles of about 5, 15, 35, 80 and 150 people, which correspond to our family, our close friends, our colleagues and acquaintances, our club and business affiliations, and finally our village or neighbourhood. He also suggests there are mega-band numbers of around 500, and an upper limit of about 1,500 people. (For more information on these network circles review the papers on Dunbar’s website).

A recent paper by Professor Barry Wellman, a noted SNA expert, suggests that in modern Western cultures Dunbar’s Number’s may be too low, especially for the outer layers. He found that 90% of the American adult population “know” between 250 and 1,710 people, with 50% “knowing” between 400 and 800 people.
Both Dunbar’s and Wellman’s numbers have implications for the optimal size of an organisation and the KM4Dev online community. It’s interesting therefore to examine the graph below.

Now we need to be careful. First, note there are two Y axes with the same interval. Second, neither scholar relates network size to time, so that’s a bit artificial. Third, note that posts can be interpreted as “Smith answered Jones”, or one-on-one relationships. With these cautions in mind the correlation between posts and Dunbar’s Numbers up until 2006 is remarkable. It’s also remarkable how close it is in 2010. But be careful, to some degree we are comparing apples and oranges. The graph below provides a better comparison.

Note the period between 2006 and 2007. The active people in the main KM4Dev discussion group had more or less settled around Dunbar’s Number. Then note the peak in 2009. I posed some questions on this in yesterday’s post. Finally, note that in 2010 and 2011 the active participants have “stabilised” at just below Wellman’s Number, which it had almost reached in 2008 just before the peak of 2009. Is this pure coincidence? Has KM4Dev reached its optimal number? I look forward to your comments and thoughts.

In my next post I will provide the results of the early SNA.

Regards Graham

1.  Dunbar, R 2010, How many friends does one person need? Dunbar's number and other evolutionary quirks., Faber and Faber, London.
2. Wellman, B 2011, 'Is Dunbar's Number up?', British Journal of Psychology, pp. 1-3