Intervening in a complex domain

A good conversation in Moscow this morning with David Anderson around and about the issue of managing a complex system and as is the way with discussion a half formed idea or two coalesced.    I've been teaching how to intervene in a complex system for years, developing and then using the ABIDE perspective question for example. More recently, based on UNDP work we've been mapping out what elements can be scaled.  Then we have three basic questions: (i) what can I change? (ii) Of those things that I can change, where can I monitor the impact of that change? (iii) Of those things where I can monitor the impact, where can I amplify success or dampen failure? All of that is good stuff, but then we come to question of more specifically how do I achieve change.

Now David started to talk about some of the things he does, all of which made sense.   I was comparing it with some of processes we teach, for example the shallow dive into chaos which contrasts with the use of chaos with malice aforethought that characterises deliberate schumpeterian disruption.  I characterised the shallow dive as creating multiple small infections of cowpox to avoid the greater disaster of smallpox.  As he talked I started to see a pattern here which matched some other things we teach on the accreditation courses.  And from that pattern I started to think about how to make it memorable and from that I got to the three Cs.  It's an old trick in methods and nothing wrong with it.

Underneath this of course we have some principles, namely that you cannot determine the outcome in a complex system.  Instead you have to make changes in the present and respond to emergent properties.  So how to make those changes?  Well you can:

1. Constraint management, removing or creating constraints to allow new possibilities to emerge.   Increase the constraint and you may be able to shift into order, which can be desirable.   But if the constraints don't bite think again, pull back and try something else,
2. Coupling can be managed easily.  Who connotes with who and with what?  How tight is the coupling?  Can we allow for novel connections to form virtually or physically which will change the properties of the system?
3. Catalysts introduced can institute new patterns of behaviour, attractors can emerge as a result (see the children's party story) or boundaries dissolved or shifted from being rigid to permeable.

Now of course none of these are categories, they are ways of thinking about interventions from different perspectives.   But that is the whole point of complexity thinking.  Change the perspective and the change the people.