Dave Snowden

The complicated domain

RSS Feed

Part of the new training programme is a day two focus on Cynefin and the sub-domain models.   I've previously posted on these, but they have been in a state of evolution that will continue for a bit.   As is Cynefin itself although that is fairly stable now.  Today in Melbourne I taught them all and finally sorted out a problem with the axes on the complicated one and completed (after the training itself) the shift of all of them away from ambiguous circles to a more structured use of the three by three.   The complex domain model has already been through that process and was published here and here; although the form there related to Agile and Scrum as a part of Agile rather than the generic version.  For the moment in three posts I am going to update the other three (and I am still debating if there is one for disorder. 

I'll blog next week on some of the reasons for creating a more familiar form (the three by three) at this stage; its all a part of understanding life cycle management.  However for the moment the essence of a domain model (although there are variations) is as follows:

  • They have an XY axis for which I have created versions for each, but where I anticipate others using the form to represent different perspectives.   Don't assume any such variant has my endorsement unless I say so though!
  • The bottom left represents a transition to a less ordered domain (or inauthentic disorder for Chaos)
  • The top right represents a transition to a more ordered domain, bottom left to a more disordered one
  • The diagonal from bottom left to top right is the line of coherence.  This (in homage to Ashby and Boisot's use) represents a sensible pathway for transition.  In complex and complicated its two way over the whole range but that is more restricted for simple and chaotic.
  • Deviations from this path are either colour coded green for useful and can do something with or red for deal with urgently
  • Blue represents management action
  • Domain models can overlap each other, as the transitionary states in the top left and bottom right are boundary areas for both.   I will need to show this in diagrammatic form at some stage
  • Corners on the opposite corners to the line of coherence represent potential transitions to disorder

In general they are designed to be a cooks handbook, not a recipe.  A if you are do something to achieve this effect type of instruction or reminder.  Obviously there are links to SenseMaker® as a mapping and monitoring tool but I will talk more about those in the future.

Now the previous complicated domain model was a bit messy and I was not happy with the axes but I had a flash of insight during the training course today and produced the new version shown above.  The two axes are:

  • Strategic importance ranging from critical, through significant to trivial; in the last case it may have been significant once but it is no longer.  Note that something can be important but not strategically significant.  Also note that here I mean strategic importance for the leadership of an organisation in respect of their active engagement.  This is an important qualification.
  • Involvement or possibly engagement ranging from contrary, through relevant to everyone; contrary means that small groups of people with different views need to be engaged, everyone is similar to synchrony on the complex domain and relevant implies a 'need to know' element.

So the ideal transition then goes through three stages:

  1. Concepts emerge from the complex domain and the top right box of that overlaps completely the bottom left here (working on the representation of that at the moment) but the idea is that out of a series of safe-to fail experiments something emerges that can shift from exploration to exploitation.   In the early days it is important to check that the exploration is complete (and it probably is not) to you have small groups of people with different views working on the issue.  Dissent here is your friend.
  2. As views start to converge you engagement more people in the process of formalisation checking that it is stable enough for exploitation.   A lot of projects will stay here, most in fact and then pass back into complex on the explore-exploit cycle that is so important to Cynefin overall.  I may need to put a double end/start point to the coherence line to show this, thinking about that.
  3. Finally some projects or ideas are stable enough to shift to formalisation and rigid process control shifting to simple. This should only happen where there is no longer any strategic advantage in shifting between exploration and exploitation.  However some very important things can end up here.  Think of hospital operating theatre health and safety which may require leadership attention to get things fixed, but should then move to purely operational work.

We then get two extremes positions with appropriate management actions:

  • Synchrony around key strategic issues means you have a strong danger of the Longitude problem.  An idea or approach which has moved to mass adoption ti quickly and now has powerful politics associated with its adoption.   Here if you are entering the space (the importance of monitoring) from bottom left aim to disrupt the process.   if you are in it, put power at risk so attitudes shift.
  • At the other extreme we have the time wasters who are continuing to argue minor points of doctrine when strategic issues are resolved.   If you are entering this space from the bottom left then you break up the cabals; if you are already there then cut the crap and enforce standards.  There is a time for conformity.

So that's it for now, more to do and I already thought of some changes writing this post.  But evolution requires some stability so I throw it open for comment/use.

 

Larger version below

Top