Tuesday, 24 January 2017

Brains and Descriptions - some Agent-based modelling

Imagine that a brain is a set of constraints which produce multiple descriptions. The constraints organise themselves as a binary tree: with enough layers of recursion it quickly becomes very complex:
In the diagram (left) produced in NetLogo, they are multi-coloured. After 2 levels of growth, the resulting four nodes are coloured red, green, yellow and blue. These colours are then inherited by subsequent levels.

The question is how does the brain decide which descriptions to make? If red, yellow, green or blue are the choices for different kinds of action, how does it select which action it should perform?

The answer to this question which I'm currently exploring is that it is the action which carries the maximum number of possible descriptions. Some descriptions are very powerful for us in being able to describe a very wide range of phenomena. These descriptions, and the actions associated with them, are the ones we select. This can explain why religious descriptions, for example, can have so much power because they can be expressed in so many ways.

But the choice as to action and different descriptions fluctuates. To simulate the fluctuating levels of different kinds of description, I've created a kind of "brain eating" algorithm. Basically this simulates a kind of mental atrophy - a gradual process of losing richness of description. Since the process is random, different kinds of actions are selected because the balance of "maximum descriptions" shifts from one moment to the next.

However, brains do not die like this. Knowledge grows through communication. The actions (red, yellow, green or blue) might be communications with another brain. The result on the other brain is to stimulate it into making descriptions... and indeed to stimulate reflexive processes which in turn can lead to mental development.

In the communicating process, there is a fundamental question: what is it one brain seeks from another? The answer, I think, is a kind of symbiosis: brains need each other to grow. It is in the "interests" of one brain that another brain acquires rich powers of description.

This is interesting when thinking about teaching. The teacher's brain makes rich descriptions, and in response to these, the learner's brain reveals the patterning of its constraints. The teacher's brain then needs to work on these constraints - often by revealing their own constraints - and transform them so that the learner is able to make a richer set of descriptions. Deep reflexivity in the teacher - that is, deep levels of recursion - help to generate the powerful descriptions which stimulate growth in the learner and help transform their own capacity for making descriptions. 

No comments: