1442323972-featured.jpeg
Dave Snowden

Reflections on atomism

RSS Feed

I mentioned in yesterday’s post that I was starting to think about the range of the work I and others have been engaged in for some years. Is it an instance of complexity science or is it something wider? So I have long argued that complexity in human systems is something qualitatively different from birds flocking and termites building nests. That means incorporating insights from the cognitive sciences, anthropology and so on. In parallel with that thinking the development of SenseMaker® is based on creating a quantitative approach in what has traditionally been a qualitative domain. I’ve been looking for a scientific approach, against purist (I’m not sure if that is the right qualifier but it is intended to avoid total rejection) forms of post-modernism and social constructivism.

Now that has resulted in a series of quandaries as to what it is that my colleagues and I are engaged in. We live in a world of knowledge silos and those of us who seek to integrate and/or synthesis can encounter problems. To be clear I am not suggested a theory of everything, that is a dark path that leads to dark places. One of the reasons I plan a series of posts to differentiate Cynefin from other complexity approaches is to explore that space; and I hope the authors of those other approaches are not too precious as sometimes you have to draw harder distinctions in writing than you do in conversation in order to make progress. So what to call the field, how to be authentic to the various sources and ideas is a matter of some import.

But it is also a wider issue, namely that of scientific method in social systems. If we go back to the foundations of sociology with Saint-Simon and then Comte we see it taking part in the context of Newtonianism. That tradition which continues to the present day seeks to reduce things to elements and rules, It looks for repeatability, hypothesis following by testing using the méthode scientifique. Now I don’t want to challenge that, but I think atomism has its limits and obviously scientific method is not confined to that. But oh boy is it popular, from the crude linear causality of Dawkin’s selfish gene to the more interesting work on fundamental particles in physics.

But somehow or other this has always seemed problematic in human systems. The hypothesis based method, the assumption of repeatability does not seem adequate to the sheer complexities (sic) of social systems. So you can see where the various forms and inheritors of post-modernism seek solace, they define themselves by what they are not which is often necessary but problematic. A form of waiting for the synthesis to emerge to take a Hegelian perspective. Related we have the issue of the depth of specialisation within the various knowledge silos which can make it problematic for the generalist as you are always skirting over partially understood theories. Also we have fewer generalists, Max Boisot was one of the last and we all miss him still. So it can be difficult to gain the attention of the silo specialists to a wider programme.

Now I am starting to see a way forward here. at the Hay festival this year I had a chance to spend half a day looking at Constructor Theory in physics. This seeks to shift away from atomistic to whole systems approach. Crudely (and I know it is far more than this) you look at what is possible with the bounds of what is known within the natural sciences and then within those boundaries ask what is the lowest energy cost of replication. That opens up new possibilities not just within physics but within social sciences. It allows us to look at what we know from siloed scientific disciplines, and what we know to be false (both are important and different) then explore replication energy levels. That might be through narrative landscapes, through safe-to-fail experiments or through trans disciplinary exchanges at various levels. Interestingly it also allows us to make statement about the current and future state and to provide an explanation of the past which is non-linear in nature.

So what is the right name for this type of study? Is it genuinely novel? Those and other related questions currently concern and are driving my own reading and experimental work.

Top