A couple of days ago Naumi had an excellent post on the diminishing returns of collaboration. He highlighted two areas where problems typically emerge – at an individual level (one person can only do so much) and at a project level (only so many people can do one thing). Included in the discussion was the mention of Dunbar’s number, which sets “a theoretical cognitive limit to the number of people with whom one can maintain stable social relationships” at 150. In terms of collaborative overhead, Dunbar speculates that “as much as 42% of the group’s time would have to be devoted to social grooming.”

As I was reading the post, it got me thinking about twitter (as has been happening a lot lately) – and the implications that it might have on various collaborative efforts within society. Many, many people scale their Twitter network well beyond the 150 person threshold, and many of them seem to be extremely active. This would seem to cause potential  problems at both the individual level (the brain getting distracted and the quality of the output suffering) and at the project level (if too many people in a network are responding to an ‘idea’, it could easily become a mess of distracting noise). One area we believe that this is playing out a bit is in relation to comments on blogs. Hypothesis: the quantity and quality of comments on blogs are dropping because of the time being allocated to tweeting and other such activities. The jury is out on whether this a good or bad thing.

This then brought me back to some earlier thinking I’d been doing about how Malcolm Gladwell’s Tipping Point theories might apply to Twitter (original post here, and the next step here), which makes sense given a lot of his theories were grounded in Dunbar’s work. The basic idea was that the way different people behave on Twitter might be analogous to the connector, maven, and salesmen definitions – and now I’m wondering whether each of these groups must be thought of differently in terms of things like Dunbar numbers.

I tried various different ways to classify people within the three categories, and while I by no means think that I’ve figured it out, what I came up with at the time (after making some adjustments based on some excellent comments) was that if you RT a lot, you are probably functioning mostly as a connector; if you @ reply a lot, you are probably functioning most as a salesman; and if you link to your own posts a lot, you’re primarily functioning as a maven. And depending on what your “function” is, it would seem easy to hypothesize that your “group size limits” are different.

The easiest comparison here is probably connectors and mavens. Connectors are primarily sharing the information of others – thus one would think that those that do it well can handle a very, very high dunbar number as the “effort per contribution” is relatively low. Mavens, on the other hand, are information specialists trying to come up with the new ideas and insights. I would think that their appropriate dunbar number would in turn be relatively low – and if it gets high their contributions suffer at both the individual and project level. Salespeople (which I used @ replies for, in terms of responding to particular problems / questions) would fall somewhere in the middle.

I could speculate a lot more on this, but as I think about it what continues to nag at me is a bigger question – does Dunbar’s number and the supporting arguments need to be modified due to the emergence of tools like Twitter, Facebook, etc.? After all, it was originally based on a limitation argument that was relative to the neocortext size, which in turn limits group size, and tied to studies of non-human primates and groups that have emerged over hundreds of thousands of years, mostly in pre-industrial villages and settlements in less developed countries.

We’ve now undergone a sudden and dramatic change in how easy groups can be formed and maintained – and I’m not so sure they need nearly as much social grooming as was required thousands, or even ten years ago. Dunbar’s work stated that only groups under intense survival pressure could ever sustain themselves at the 150 mark, and had to be physically close. Is that still true – or do we need to totally re-think our notions of groups and collaborative efforts thanks to the new technologies that allow them to emerge and maintain themselves over time without that much “grooming”?