Organisation by Cynefin domain

I was driving Sonja back to Cape Town airport this morning after a very useful session yesterday with Marietjie and Zhen working through various ways in which we can visual data from SenseMaker®. We took advantage of my being in South Africa for a client project and Zhen being over to teach Stellenbosch University (a new partner) the basics of SenseMaker® projects. We met on Marietjie’s house (with a vineyard and olive grove) just outside Stellenbosch and mixed good conversation with business and excellent wines. More from that as we start to roll out standard offers powered by SenseMaker® over the next few months. There will also be some material in the forthcoming Cynefin and sense-making course in London.

We were talking about various ways in which Cynefin in explained and I mentioned the tetrahedron models that came from Cynthia Kurtz’s first encounter with Cynefin. They were used to illustrate organisational structure in the New Dynamics of Strategy article that we published in an IBM journal. My initial reaction to their value was sceptical, but one needs to respect one’s co-author’s views. After a period of scepticism, I started to use them in courses but after some years my initial nervousness returned. They were leading people to prescribe organisational types by domain and that sort of categorisation always worries me. A hierarchy for example is more than able to handle complexity if it focuses on the right things, but it can never fully constrain or control, something that is possible in the ordered domains. Either way, I stopped using them around five years ago, but their use persists and Sonja challenged me to come up with an alternative.

So after I dropped Sonja off I set off for a four-hour road trip with the camera (restored from its dip into a tidal pool near Sydney) around the coast for a spot of Whale spotting and then back over the mountains. I’ll link to the pictures when I have loaded them at home tomorrow. I was listening to an audiobook of Swallow Dale, which I know by heart anyway so between stops for pictures of waves and penguins I had a chance to think about it. Settled into the lounge at Cape Town I had a chance to play around with a first draft of an alternative. To be clear I don’t think this is it yet and I may go to a graphic artist or cartoonist for something simpler and more usable. But it is a first step.

The opening picture shows where I got to, using different images to make a point. In a sense, this follows my recent post on killing sacred cows in organisational development so I will include aspects of this as I develop it on days one and three of the above-referenced course. So to the domains:

It is far from complete, the cliff needs to be right and disorder is probably a black hole but it is a start. Open for comment and feedback

1. Obvious in the tetrahedrons was shown as a rigid hierarchy with network activity disconnected, top-down direction is all.  I haven’t really changed that as it is the only appropriate model in the domain of best practice.  This domain has rigid coupling
2. Complicated is again largely unchanged. I have shown it as a matrix, the most common form and the most appropriate. In the tetrahedrons, it was shown as both hierarchical and network connections as rigid links.  In this domain, the coupling is looser but still controlling.
3. Complex is where it changes. With the tetrahedrons, we showed the network connected with the centre loosely connected. This was the only I didn’t like and with the benefit of hindsight think is plain wrong. I’ve replaced it here with a fractal image – you need the same representations up and down and strategy cannot be separated from operational action in a linear manner, either way. More on this tomorrow but take this as a holding post for the moment.  Coupling is loose but exists
4. In chaos we show a state of randomness, the tetrahedrons show it as no connectivity. So again no major change.

Now I am just starting here, more work to do. I may keep a variant of the tetrahedrons, but they just don’t work for complex.