Sunday, 15 January 2017

Bio-Entropy and Learning

Gregory Bateson had a word for the underlying process of living things. In contrast to the idea of entropy in physics, where everything eventually runs down and where order gives way to disorder, Bateson argued that a different kind of process was at work in living things: Bio-Entropy, which he saw as a continual struggle to maintain a median point between extremes of rigidity and flexibility. Rigid organisms die because they are unable to adapt. Organisms which are almost amorphous (the extreme of flexibility) die because they cannot organise to exploit the resources available to them which would help them to survive. To live is to exist in the middle, to oscillate between periods of semi-rigidity or semi-amorphousness.

In information theory, information - or neg-entropy - is seen as the countervailing force to physical entropy, but derived its formulation (by Shannon) from the statistical representation of physical entropy created by Boltzmann. Negentropy creates surprise and difference, which can stimulate complexification in response to entropy's run-down. Terry Deacon talks of Orthograde and Contragrade forces in biological systems. In Stafford Beer's work on management cybernetics, there is a similar relationship between vertical forces of top-down management (Beer calls this "metasystemic intervention"), and the horizontal forces of self-organisation: the trick in organisations (and government) is to get the balance right. Ecologist Robert Ulanowicz, himself deeply influenced by Bateson's ideas, uses information calculus as a way of gauging the ways in which the orthograde and contragrade forces operate in ecologies by measuring the amount of surprisingness generated in a system through taking various information measures from the components of an ecosystem. Information theory has been useful to ecologists for many years, but Ulanowicz is also aware of the deep confusion which lies inherent in its apparently simple formulation.

I find Bio-entropy the most powerful idea when I think of education, learning, organisation, love, art and music. It is, fundamentally, a different kind of way of expressing dialectics. In things like music, we experience it in action in a most direct way. The challenge to us is to find a better way of measuring it.

Shannon's formulation of information was intended to address a specific engineering problem. We do not talk with each other in the way that computers exchange messages. Our talking is more like a dance, or perhaps a game, within which we can become incredibly happy, angry, sad, bored, and so on. Out of our talking, politics arises. One person might realise a power to make another feel a certain way, and then to do certain things, and to exploit this to their own ends. Capitalism begins (and possibly, ends) in conversation. But if it's a game, what kind of a game is it? How does it arise? How are the rules decided? What is it to play or not to play? What is it to win or lose?

Another of Bateson's aphorisms concerns the "multiple descriptions of the world": cognition is connotative in the sense that it feeds its way through differing descriptions of the same thing. We have two eyes which give a slightly different description of the world. When we talk, our words describe one thing, whereas our body language describes something else. Sometimes the words and the body language describe different things, leading to what Bateson famously called a "Double-bind".

Great artists and great teachers are similar in the sense that they are masterful in their manipulation of descriptions. A maximal ability to generate descriptions for a variety of different circumstances (some of which might challenge the viability of other people) whilst maintaining a coherent centre for those descriptions is the true mark of the expert, not the dogmatic insistence on a single description (we have misunderstood expertise for a long time!). It is this ultimate flexibility of description-making that great teachers seek to inculcate in their students.

A utterance by a great teacher or an artist will contain within it many descriptions. There might be the statement of a concept, or an observation of something they find interesting, coupled with their tone of voice, posture, body language, etc. People might hear the utterance, but they also sense the other descriptions that are presented to them. If the utterance is the central focus (which it usually is), then all the other features envelop it:
((((Utterance) voice) body language) posture) environment
One utterance will be followed by another, and another set of descriptions is produced. Sometimes the context of these descriptions will be  different - maybe it's only expressed in text:
((Utterance) text medium) online environment
A learner is really a different kind of environment for a great teacher. The teacher realises they can change this environment (the learner) by making the right kind of descriptions. The response to the descriptions that the teacher makes tells the teacher what kind of an environment the learner is. The process of teaching is a process of providing the learner with greater flexibility to make their own descriptions across the different environments which they inhabit.

Maintaining the balance of Bio-entropy is, I think, related to the process of creating the maximal number of descriptions. If a formalisation of this is at all possible (and I'm not sure), I think it looks rather different from Shannon information theory. It's not surprise we are looking for; it's the coordination of constraint.

No comments: