This is a rework of the notes I took in an Open Space session with the title: What does ‘results’, ‘knowing’ and ‘sensing’ mean in the complex domain? – implications for harvesting (July 2015 at Beyond the Basics in Leicester, UK).
Since quite some time – years really – I have been following Chris Corrigan’s journey into the teachings of Dave Snowden and his Cynefin framework. If you haven’t looked into it, please do; as all people who I have explained it to, even with my limited understanding, are enthused as the model really tells something important: it tells in grand detail the differences between obvious, complicated and complex domains (and of course more, but I am not going into that for now).
In the AoH network – and in mainstream society at large - we tend to be confused, or not having good distinctions, about the nature of the system we are working in (ontology) and what kind of knowing is applicable in any of these systems (epistemology). The Cynefin framework can really help with that! Knowing if you are in a simple, or a complicated or a complex situation is really important as it has many, many implications – especially for what we can know, how we can know it and what kind of decision we can and need to make.
We call a situation complex when there is no causal relationship between cause and effect. If you do action A, sometimes you will get a result B, at other times you will get result Z. When people are involved, it is always complex; think about how you deal with your child or your partner, sometimes they respond totally different to your same action! Even speaking of results is not appropriate in this context, as we need to understand that in the complex domain there is no (one) solution. Chris used the example of getting hunger out of the world, or getting climate change sorted… there is no solution possible, it is instead an ongoing work of sensing what is working and sensing what is less working. So then, what does harvesting means in this regard?
Chris mentioned a few principles from this framework, on how to work with DATA in complexity:
- Use fine-grained data objects – instead of letting people scribble on the flipchart paper in a World Café, give participants little post-it notes where they can note elements of importance for the issue at hand. In the language of the Cynefin framework they are called micro-narratives, or anecdotes. The point is to come up with data that can be used as bricks, not too small so that it is like sand and you cannot build with it, or not too big like boulders so that you can’t move them easily… the points is to be able to construct categories or slice the data in different ways; as there is no one good way to do that!
- Distribute cognition: instead of having the hosts, or if you are lucky the harvesting team, make sense of the data have the group collectively do it. This means: build in (way more) harvesting processes in the design – in other words: host the harvesting. Let people do it in many small groups to prevent some kind of group think where you loose the details and the voices from the edges. In Cynefin language: prevent premature convergence.
- Disintermediated sensemaking: instead of clustering and categorizing the data in a way that the terms become too abstract or too general and don’t have any meaning anymore, give the data and the many different ways of sense-making and categorization to the decision makers or the ones who have to act.
In the domain of complexity the roles of the actors, the decision makers and the evaluators come closer and closer together. In the domain of the obvious and the complicated, what are called the ‘ordered’ domains, these roles belong to different people. This means that when we deal with complex issues that everyone needs to learn to become a learner, an actor, an evaluator and sense-maker.
Please remember that the data in a complex system don’t give you the answers – as there is no solution! Data are only there to help the sense making process; which is essentially in service of the emergence of many ‘probes’ – experiments – prototypes. Doing the actual experiments is, again, in service of understanding the complexity in the system better and better.
This all means that we can learn a lot from the rapid prototyping, the social labs, the agile way of working and probably many other methods that are emerging. We need a new culture of working-learning, which will eventually cut out many layers of management in traditional organisations:
- After the collective sense-making, I think there is value in using your intuition to choose out the probes you will actually do. If there are some weak signals that are noticed by a few, that might be the best place to start. Where everyone sees the signals it is not really a space of experimentation anymore; and when you are the only one seeing the weak signal you might not have enough traction.
- Do many small probes or experiments, not just one! Most of the time they are so small that no official permission is needed! Think big, act small.
- There is incredible power in quick feedback-loops! Always think iterative and adaptive. Less time spend, less money wasted, less fear to fail and learning happens more quickly!
- It is not about ‘best practices’ that can scale up (because ‘best’ practices only exists in the obvious domain and not in the complex domain). From the different experiments you do, you can maybe deduct principles and processes that can be applied in other contexts; but there is no copy-paste possible of actual practices.
- What if the Key Performance Index becomes about measuring failures and learn from them??? (cfr. Failure conference, fuckup evenings) Knowing that 80% of systemic change processes fail, it seems that a lot of learning is not captured at all.
- For the ones who know Systemic constellation work, I guess it could be used to check the effect of possible probes and in that way speed up the feedback loop even more!
Still a lot to learn!!!