Friday, 18 May 2012

Symmetry, Experience and Abstraction

I've had some fascinating discussions with Oleg over the last two days, discussing our book on "Educational Cybernetics" and negotiations with publishers (which are sometimes frustrating), but also dealing with new developments in cybernetics and their possible implications. The most significant of these is the emerging deep critique of 'information' coming from physics and biology. This is currently affecting my thoughts as I write a paper with Leydesdorff on the VSM and Luhmann's social systems theory (I think there's a matter vs. information thing going on there), and my thinking around papers I intend to give at the American Society for Cybernetics on abstraction and time, and a similarly themed paper on abstraction, mechanisms and critical realist economic theory in Marseille next month.

Oleg's view is that Bateson had these same insights into the nature of information a long time ago. My initial reaction is that I wouldn't be at all surprised. But it's useful to do a deep reading of Bateson's arguments to unpick what he means. First of all, the issue of abstraction and experience he's documents very clearly with this diagram:

There is an oscillation between abstraction and experience. From experience emerges a description of the experience, and from the description a form is produced within which new experiences are framed, and then described, and so on. An interesting comparison can be made to a similar oscillating diagram a few pages later which shows the relationship between 'calibration and feedback'.

This I think is significant for the processes of economic control, because in order to calibrate, some knowledge of the ontological mechanisms of the processes must feed into knowledge about what interventions to make. But what's in a mechanism? Mechanism is everywhere, and everywhere it is underpinned by a concept of successionism (which is the topic of my economics paper). Bateson says A->B->A->B,, and within A we describe x->y->z in language. In describing x->y->z, a typology of x', y' and z' emerge at B, with new descriptions of mechanisms x'->y'->z', and so on. But is this it?

What I think is important is the abstraction of successionism: what's implicit in the "->". I remember having a discussion with a mathematician years ago about what "=" meant. For it is one thing to say it means "tautology", but acutally, tautology is itself a process. There is a mental process by which we recognise that a tautology is taking place. And there is succession in the mental process. And there is tautology in the succession. Here I think we get to Von Foerster's Eigenform. In this way, maths is epistemology, not calculation (as Kauffman said on the CYBCOM list today). In its epistemology, there is tied-up, ontology.

But how does symmetry fit with all this? The issue is, I believe, form. In Bateson's paper "A re-examination of Bateson's Rule"  (where Bateson's rule is the biological theory proposed by Gregory's father William, who first coined the term 'genetics'), Bateson makes the now-famous definition of 'information' as "a difference which makes a difference  in some later event". Bateson is interested in the radial and bilateral symmetry of organisms (for example Beetles). He suggests that the process of loss of symmetry - from radial symmetry, to bilateral symmetry to assymetry, is the result of additional information in the environment, since, he argues, it is information which attenuates the self-organising development of the organism.

Bateson develops this idea to expand on his concept of negative causation and constrain. Instead of talking of information necessary to determine asymmetry, he instead talks of information necessary to prohibit bilateral symmetry. In this way, the information environment of the developing organism is an environment of constraint, squeezing the possibilities for development of the organism.

How does this help me understand the nature of abstraction and its relationship to symmetry. Abstraction is information and consequently it is constraint. Experience occurs within constraints. What is generated is new information which constrains future development. The abstraction of time, like the abstraction of abstractions generates information within which experience occurs. But what of the form or symmetry of experience?

On the one hand, we can't grasp it without abstracting; however, technology may allow us to manipulate informational constraints within which our experience is formed. This may be useful in social systems because the relationship between calibration and feedback in society is so nebulous. The free participatory manipulation of informational constraints may reveal shared expectations and means of collective constraints (through collaborative thinking about data for example) which may in turn help in the processes of shared decision-making.

I have to think more about this, but it is the shared decision-making within a complex and ever-changing environment of informational constraints that is the central social and organisational problem we are facing today.

No comments: