Dave Snowden

Inherent uncertainty is our natural habitat

RSS Feed

Yesterday I used Paul Cilliers list of characteristics of a complex adaptive systems to introduce the idea that complexity in human systems – anthro-complexity – has many distinguishing characteristics. I plan to run through ten of them starting today and will do so in a discursive manner. This is an exploration of the subject not a series of definitive statements so those who want less ambiguity may want to skip the series and wait for the posts I will make around the book chapters during my eight day writing period at the end of the month.

The first area I want to explore is how we evolved to manage uncertainty. As a species we faced constant uncertainty and threats from more physically capable creatures than ourselves. Probably the most significant mechanism that evolved was various social groupings – the family, the clan & the tribe (or from time to time the horde). Within those units we have diversity of experience, cognitive diversity, different levels of intelligence and so on. However the collective has sensory capability, psychological support and response ability that cannot be possessed by the individual, or the some of the individuals. Social process and experience over time constitute a modification of the interactions so that the individuals have less significance. Dunbars number, relates to grooming and the number of acquaints we can maintain, it appears to be a neo-cortext limit and in more recent writing Dunbar has suggested that the cortex is the size it is as 150 is the optimal size for a hunter-gatherer community. The Bernard–Killworth suggestion which roughly doubles this is based on field studies in a modern society where city states, remote education and identity structures change the field a little, a subject to which we will return in a future post. Other studies show a number between 15 and 30 as a natural level of trust and we have Millers number on short term memory.

Without getting in the efficiency of individual numbers here, the point is that within a limited number mix we are capable of far more than as individuals. And its not just about assembling individual over short time horizons (something that the modern HR professional is too prone to). In hunter-communities training and interaction take place over extended periods of time: surviving to fifty or more means you have wisdom to pass on but should probably stop leading the hunt, so teach the children or provide advise. Communities with deep knowledge acquired over time can easily accommodate and exploit cognitive differences, something a suddenly assembled team cannot.

Within this we have another very human capability, the emergence of generations of knowledge. As we develop a capability, for example medicine, the body of knowledge and practice emerges over generations. Knowledge is built into processes that are not fully understood when they are created and rarely understood by future generations. But they are part and parcel of managing uncertainty. Long term employment, apprentice systems and the like all create capability that cannot be modified or reduced to competences or the like.

Knowledge of process, understanding built of capability but not fully articulated, the aesthetic experience of sharing physical and spiritual experiences, the ritual of transition and so on are all key elements of managing uncertainty. The reduction of humanity to individuals and collections of individuals, the reduction of institutions to explicit process all fatally undermine our capacity to manage uncertainty. One problem, in consequence, is that we try and increate the explicit nature of the process, further define idealistic models of the individual this making things worse rather than better.

So understanding how to manage interactions between differing types and instantiations of experience and education are key to anthropology-complexity. Its not about agents and simple rules ….


PS: readers may note a lot of back posting here. I’ve been down with flu over the holiday period and not really been up to writing posts but I have been keeping notes and I’m now catching up. So posts will be logged over the twelve days regardless of posting date.