How might we experiment and learn about the processes and teaching of experimenting and learning?

In the work we do with leaders dealing with complexity we put a lot of emphasis on experimenting and learning.  We do this well, up to what we’ve learned to this point.  Here are some first thoughts about what we might do next.

What are we doing well?  We do a good job making the case for experimenting.  Leaders tell us they get real insights and value from:

• Understanding a situation is complex and the best that can be done is to try things out,
• Loosening up, experimenting more widely and in more creative ways, and
• Becoming more active and reflective learners, focusing more closely on the present system and how it seems to be changing.

We also have some fast and fun ways we use to introduce leaders to the ideas of safe-to-fail experiments and to help them generate novel ideas for experiments for each other.  We have them help each other look to the edges of issues, make them small, cheap, and pragmatic, and suggest ways they might each have more fun doing it.

So far, so good.  But then some leaders also ask, “Is that it?  Is that all there is?”  And, “How do I apply this in my work?”  A few go on to point out, because they have to lead big changes, they doubt starting small and experimental will make a big enough difference quickly enough.  They have to “go big or go home.”  And, in going big, they have to get the right answer.

Some of the challenge in this may just be the challenge of complexity for leaders.  It is hard.  There are just too many variables and it is impossible to predict how the situation will evolve.  There is no right answer to be found.  You can go big and go fast, but you have no real way of knowing you are going “right” unless you have been trying stuff out in a small way and trying to learn from that.

And some of the challenge is for us.  We support the development of leaders.  What is our next level of support?

These questions came up again as I was reading a book by Yuen Yuen Ang where she adroitly applies complexity thinking to the question of how China escaped the poverty trap.  Ang uses an adaptation framework described by Axelrod and Cohen in their book Harnessing Complexity: Organizational implications of a scientific frontier.

Then I started to wonder, could an adaptation framework help our work and our leaders?

Axelrod and Cohen’s framework is built around three elements.  They observe how complex adaptive systems evolve by trying out lots of different ways of doing things (variation), enabling connections between many agents (interaction), and then applying a process where some variations survive and others do not (selection).

Axelrod and Cohen propose a core question for each of these three elements:

• Variation: What is the right balance between variety and uniformity? (My take on that question might look more like, “what is a useful balance between variety and uniformity?” Maybe that’s splitting complex hairs…)
• Interactions: What should interact with what, and when?
• Selection: Which agents or strategies should be copied and which should be destroyed?

I am looking for two different things here.  Firstly, I am looking for a framework that takes me all the way through the process of experimenting and learning – a frame that gives what we already do with experimenting and learning a context.  Something that provides guidance on how this fits with other activities to help tackle bigger challenges.  This frame then gives me a sense for how relationships and interactions might impact my experiment—which ones might make a difference (at the core or at the edges)?  Finally, it can also help to think about how experiments will be amplified or dampened and the criteria we would use.  Experimentation, interaction, selection – we can see how the pieces fit together or what we might need to do to get a better fit.  This is win number one.

The second thing I am looking for is ways to teach these adaptation concepts and practices in simple and practical ways that are directly relevant to leaders who then feel inspired to try them out and learn from them.  We already know how to take the idea of experimentation itself from theory to practice.  We also already know how to talk about and work with approaches to interaction (systems maps, meeting in complexity, hero perspectives) and selection (selecting direction, stories about what you’d hear more of and less of).  What we so far haven’t done as successfully is to connect these ideas into experimentation.

This seems to me to be the place we can get more wins.  Focusing on “interaction,” we could expand the work we already do on understanding the existing system and mapping the key relationships and attractors.  We could tweak the system exploration questions we ask and link them more closely to how leaders think about experimenting with these relationships.  The aim would be to produce a practical experimentation tool for leaders—look first at “who IS interacting with whom and what?” and then ask a version of Axelrod and Cohen’s question: “What or whom SHOULD interact with what or whom, and when?”

Axelrod and Cohen further offer a set of questions and strategies about interaction.  They concentrate on whether an interaction is internal or external to the organisation.  They explore whether it’s physical, with people meeting either in time or in space, or conceptual, where we might interact about our different identities.  Each of these interactions has the potential to advance or retard adaptation.  We could almost certainly develop “interaction practices” that would further incorporate these theories about interaction into the ways we help leaders develop and run experiments.

Focusing on “selection,” at Cultivating Leadership, we also have the strong bones of a practical selection tool that leaders could apply to their experiments.  We focus leaders on having a direction they are working toward, rather than a known destination they are aiming to achieve.  We have them define success by stories they would hear “more of or less of.”  Following the “selection” theory of Axelrod and Cohen, we could take such next steps as developing more practical ways to focus on success or learning criteria or tactical strategies for amplifying and dampening experiments.

From these ideas, I have crafted a to-do list:

• Develop a more intentional approach to interaction
• Provide more specific steps for selection
• Pull all this together with experimentation (variation) into an explicit framework for adaptation.

Three wins!  We’ll be winning so much we will get sick of winning! It’s the zeitgeist.

These experiments will mean I will have more to say here as we figure out what works and when and where.  We would love to hear your experiences in linking these ideas and tools.