Tuesday, 31 May 2011

Depth Psychology and Agent-based Modelling

It would seem to be common sense to say that no social or economic model which didn't take into account the psyche of individuals could possibly have any realistic hope of authenticity. Yet economics glosses over this omission: it would seem that it simply is too difficult. Keynes, to be fair, does spend a lot of time discussing the psychology of consumer: the idea of the 'propensity to consume', 'propensity to save' and 'liquidity preference' underpin his theory, although there is little detail as to how these might manifest in individuals beyond idealised reasoning about personal economic well-being.

However, Kahneman and others have started to unpick this rational model: people do strange things, particularly in the face of risk and anxiety. Kahneman's assertion is that people tend to overreact to small probability events, but underreact to medium and large probabilities. I think Axelrod's experiments with the Prisoners Dilemma echo these findings too. But losing rationality from the equation is just where our problems really start...

The real issue I think is that people are not the same, and certainly do not behave similarly in similar situations. However, the crowd dynamics (in situations where crowds are a factor) can create unusual social movements in response to small initial conditions (a la complexity theory). But even that doesn't allow us to think of people being the same. The real issue for me is that clearly we are different. How are we different? 

On the one hand, we behave rationally: we communicate, observe the behaviour of others, judge what might be in our best interests. But on the other hand, each of us hides vulnerabilities, dreams and desires that, even if they are not unique on a social level, appear to us to be utterly unique and fundamental to our identity. This sensual side and the rational side seem to co-exist. Personally, I believe that what keeps them working together is 'property'; or put another way a relationship to property is a way of balancing the needs of rational discourse with the sensual aspects of the personality.

In my agent-based model of communication and wellbeing, I am starting to consider these two mechanisms working together. But there are other problems which I am thinking about. For the sensual and vulnerable self has dimensions and depths which lie well beyond the making of rational discourse. The Freudian level of id, or Jung's shadow lurk behind the scenes often sticking a spanner in the works with regard to conscious behaviour.

This has led me to think about possible ways in which the subconscious might be modelled. One possibility is to model it as a sort of 'internal conversation': i.e. a conversation with an imaginary 'other', with the results of the internal conversation eventually moving into conscious communication after a process of selection (as in Luhmann). Another possibility is to elaborate Jung's archetypes as regulating levels in a subsystem of the 'conscious' (i.e. communicating) system (much as I discussed here: http://dailyimprovisation.blogspot.com/2011/05/luhmann-beer-harre-and-jungian.html)

Each of these are effectively internal reflexive 'mirrors' which turn a conversation in on itself for a period of time before that conversation emerges into the social system. It's a bit like a laser exciting photons before they have enough energy to be emitted in the beam. There are experiments to try: each configuration of 'mirrors' will produce different effects. I might hope that some of those effects have some bearing on what we actually see happening!

No comments: