Came across a Dave Snowden keynote to the KMWorld 2010 conferences in which he spoke about resilient organisations, how to build them, and what impact that has on knowledge management and IT systems. You can find the podcast and matching slides here.

A couple of the hightlights – early corporate hierarchies were modeled on the military, as one of the few pursuits at the time which organised large numbers of people and coordinated their activities. This is often derided as a model for the 21st century business, but Snowden points out that the error (if there was one at the time) was to model it on peace-time military, rather than war-time … where resilience of the military effort is conditional on ceding control to the people in the field.

Snowden talks of three management paradigms, two past, one emergent. He sees management as moving from scientific management through systems dynamics to (now) cognitive complexity as part of a complex adaptive system, where the participants and the system continually change in interaction with one another. Perhaps coincidentally, the shift from one paradigm to the next has coincided with periods of economic recession, and with reasonably significant improvements in IT capabilities – and points out the possibilities that are available with basing our management on natural science and taking advantage of pervasive computing.

I liked the counterpoint to “best practice” – that we learn better from “worst practice” – in an evolutionary sense learning from failure has been more effective, so we should move away from risk-averse “fail-safe” systems, and towards “safe-fail” experimentation.

And (music to an enterprise architect’s ears) at about 1:05:30 into the podcast, he points out that moving from software applications to an architectural approach improves resilience. Being less dependent on monolithic, slow-changing do-it-all applications in favour of using loosely-coupled “objects” (including people) was more adaptive, flexible and therefore more resilient in times of change and uncertainty.

Do yourself a favour and listen to the whole thing …