Monday 2 May 2011

Axelrod, Cultural Diffusion and Rationality

There's something fascinating about Axelrod's model of cultural diffusion - particularly if you select the 'bubble mode' in the applet below (which I stuck into the Wookie widget server). It has a particularly realistic biological quality which I've always missed in most cellular automata models. However, there are still a great many questions where I think Luhmann and Beer's thinking might provide some deeper insights.

Some of Axelrod's most significant work stems from his examination of the Prisoner's Dilemma. He famously challenged a number of computer scientists to come up with algorithms which could deal with the repeated prisoners' dilemma. He identified that ultimately, cooperative 'altruistic' strategies seemed to be more effective in the long term, and that an evolved approach might lead to these sorts of strategies. Whilst I find this fascinating, I'm tempted to remark that a strategy isn't 'forgiving' or 'altruistic'; people are. And what goes on in people is rather more than the calculation of next moves in the prisoners dilemma. Tooth ache or an annoying twitch is more likely to sway things one way or another.

The fundamental issue relates to the locus of rationality. If a strategy is rational then it's locus is in someone's head, and that head causes the strategy to be rational because the head processes the available data and produces a logical conclusion. However, rationality may not be quite like this. With a distributed cognition approach, rationality (or what we think rationality is) exists in communications. The extent to which communications can be successfully made, we can deem these to be rational. In this sense, it is not whether someone wins or loses the prisoner's dilemma, but whether the move that a person makes can be deemed 'within the game'. Attributing mental causes to that rationality is inevitably an act of conjecture.

If rationality is simply about making successful communications (communications which are understood), then the question is not about the causes of rational behaviour, but about the causes of communication. With regard to communication, and thinking about Luhmann, I think there is a secondary mechanism at work which relates to sensuality, and this works hand-in-hand with communication processes. When we consider the full complexity of this, rational strategies for the prisoners' dilemma (altruistic, greedy, etc) seem a bit shallow.

The cultural diffusion model has, I think, gone down a similar path to identifying mental causes for strategy. The boundaries in the model reflect degrees of cultural affinity. But cultural affinity depends on the extent to which people can have conversations, and that depends on their ability to make communications at a number of levels. Are cultural boundaries in the world or are they in people? I think this is where we need Positioning Theory!

No comments: