Collective intelligence

One of the of soft skills commonly touted as of growing importance for organisational success is the capacity to work successfully with others.  ‘Teamwork’ is the most familiar label for this, but I think this sounds a little simplistic.  Brenda, one of my interviewees in The Paula Principle, put it this way:

Women are interested in getting stuff done;  they hate game-playing and internal politicking; the only build networks if they see them as a genuine vehicle for getting things done, and not for personal advancement.  Their language is about ‘we’, about looking for complementary skills in a team, recognising that they themselves cannot do it all.

I’ve just been reading Superminds by Thomas Malone. Much of this is about how we can use modern technologies to communicate better with each other and enhance our overall capacity.  But there’s a very relevant early passage on’ What makes a group smart?’ which explores the notion of collective intelligence.

Malone and his colleagues set out to look for some kind of collective equivalent of general IQ.  They found first that although there was some relationship between the average/maximum intelligence of group members and the group’s collective intelligence, this was not a strong one.  In other words, it’s not enough to get a group of intelligent people together.  You need, as Brenda said, to find complementarities.

So what was most closely related to the group’s collective intelligence?  Malone found three factors:

• the average social perceptiveness of the group’s members:  how well could they read other people’s states of mind;
• roughly equal participation in the conversations – ie no domination by one or two individuals, not exclusion of others;
• the proportion of women in the group.

The last and first factors were, unsurprisingly, strongly correlated.  In addition, the study found that the highest CI scores did not come from groups with equal proportions of men and women.  There is a direct, though not very strong, relation between proportion of women and CI level.  Malone says more work is needed on this, and doesn’t discuss whether there is a kind of tipping-point for the gender mix – in other words whether there is at least some admixture of men that is helpful (says I, pleadingly…).  But this is powerful stuff.

PS  Gender is of course not the only dimension when it comes to assessing collective intelligence.  I sent an earlier version of this post to a friend of mine Kerry Emmanuel, eminent climatologist and a colleague of Malone’s at MIT.  Kerry pointed out that for him there is a major divide between scholars working in the humanities and the sciences, with scientists showing more teamwork.  Of course this is partly because of the nature of their work – a scientist is unlikely to get far working on their own.  But it would be interesting to have a comparison of collective intelligence in scientific teams with different gender compositions, and then comparing them with similar teams in the humanities – and of course the social sciences…