Dave Snowden

The domain of disorder (iii)

RSS Feed

I made the point yesterday that the disordered domain within Cynefin is one of false confidence, assuming you know how to act without first making an ontological assessment. It is the opposite of chaos in terms of the individual response as to the decision maker nothing seems to be other than as expected. That aspect of disorder is in effect the sub-domain of inauthentic disorder coloured in red on the illustration. But there is also the use of disorder in terms of dynamic transition between domains, authentic disorder and coloured green. I keep thinking about changing the authentic word but I haven’t found anything better so it stands for the moment. In this final post (for the moment) on disorder I want to look at those transitions.

Upfront transitions between directly adjacent domains do not need to use disorder in other than one set of circumstances more fully described in the first disordered dynamic below and that is the lowest risk, mostly easily managed anyway. To move between Complex and Complicated you increase the constraints to see if you get repetition and/or predictability; if so the transition is achieved, if not try again. Relaxing constraints will shift from Complicated to Complex. Complicated to Obvious and back again is simply a question of the level of enforcement and the degree of flexibility afforded to those with competence. Complex to Chaotic is forced innovation, Ordered to Chaotic is a catastrophic failure induced by complacency.

The shift through disorder is needed for non-adjacent domains. I’ve identified three here and have deliberately not invaded Complex to Obvious as (at the moment) I can’t think of any legitimate need for that dynamic. All movements through authentic disorder always carry a danger, so the trajectory needs to be controlled and have sufficient momentum to prevent falling into the red centre. Someone once suggested that disorder was like an open cast mine. Trucks driving around the edge in danger of falling in. So what are the three?

  1. A shift from ordered to complicated should be able to be achieved directly, but if you have entrained patterns of thinking, a refusal to accept some variation then a disruptive approach may be needed.  So if we have a set of bureaucratic rules which need revision, but that revision is disputed by those responsible we are going to have to take a different approach.  The shift is shown as close to the boundaries as it is low risk.  You simply take the existing process and run parallel, contained safe-to-fail experiments to test alternatives.  That is by its nature a complex activity but easy to move over to complicated on an evidence base.  Its all a boundary shift and a shallow  use of disorder.
  2. More serious is where the rigidity of the Obvious domain has reached a level of complacency.  Informal networks are having to work harder and harder to make the system work despite itself.  It is all to frequent in a post BPR, post Sick Stigma world where compliance has become more important than delivery.  The process has become entrained and cannot easily be disrupted.  To change this situation requires a radical shift, moving people to the edge of chaos before shifting into the complex to allow a more sustainable and resilient system change as a whole.   I normally use the illustration of a project I ran many years ago when we exposed the partners of an accounting firm to the full bureaucratic horror of their expenses procedures by making then operational actors in the weekly horror of currency variation calculations, categories splitting and online reconciliation.  The result storm of fury resulted in racially, internally driven change.
  3. The final transition is a innovation one.  Here we have a complicated approach which is over constrained and avoiding radical innovation.  We may simply have failed to realise that the context has shifted, or it may be a paradigm change, a significant switch in the industry, government etc.   Here we need to transit disorder in order to allow a shallow dive into chaos to open up radically new possibilities.   I’ve used the Longitude problem to illustrate this in the past and that is still a good example. I really need to make some more recordings for the same reason I need a new passport – that video is the current me plus thirty five kilos.

Now there is more to come on disorder. It may make sense to have more than two subdomains (my first two examples open this up). I also need to continue the work I am doing on containers and coupling as a new approach to constraint and also to problem definition. I ran the first ever exercise on that in Milan earlier today. But for the moment its time to move on from disorder.

  • Chris Corrigan

    This is important. I wrote a while ago about the shift towards compliance in the child welfare systems in Canada (http://www.chriscorrigan.com/parkinglot/why-rules-cant-solve-everything/). The issue here is that the more children die, the more rules come into effect to ensure future accountability for future child deaths. And of course the more the rules are followed the more likely it is that future child deaths will be completely anamalous.

    In teaching Cynefin to social workers and leaders and managers in this field, we do indeed initially crack the idea that different onotologies require different responses. Having place complexity at the heart of a leadership program for more than 400 people in this system now in British Columbia we have arrived at the realization that what we are doing is helping people prepare for the inevitable collapse of the rule-based approach to social work. Soon the system will enter a period of chaos (at the moment it keeps edging into disorder and the compliance folks keep pulling it back into order, while the innovators try to tug it towards complexity). We are hoping that a sizable number of the 400 we have worked with will be ready to enter into the mess when it happens and respond with a more sophisticated approach to the problems that need addressing.

    Great post Dave. Thanks.