Title of session: Complexity Session
Initiator(s): Peter Hobby
Participants: Leia Isanhart-Balima, Johannes Schunter, Rachel Cardone, Lucie Lamoureux, Nancy White
Trying to understand how complexity might be more acceptable/palatable to help us in KM. What is happening, what could be done. Rachel and Nancy have used the Cynefin framework (Kurtz and Snowden). Complex Adaptive Systems.
Rachel: we know that a lot of spaces in development are complex and people say “this is a system” but it only gets you so far. You need something tangible to move forward (fund, design, implement). Can we actually identify what space we are operating in and design programs accordingly. How to APPROACH a complex/ systems problem. The right side of the Cynefin framework is fact based, left side is pattern based. So I’m going to design a program for fisheries. Some things are simple. How do we behave in a simple (not simplistic) environment. Sense, categorise, respond. May be the checklist or a beurocratic process. Lessons learning a lot of time, organizations will try and simplify things and pu tin simple that are actually complicated or complex. An apple becoming a fish. Simple is fact based management. There is a “right” answer. I know to delete spam. After 10 RCT you don’t need an 11th if it is repeatable. Standard protocol. Complicated is still fact based, but you need a level of expertise to diagnose, predict outcome. It is discoverable but not immediately apparent. Might be more than one right answer. So you need more of a process to identify the right answer(s). Known unknown. Some complicated, as they become more known, chunked into smaller bits, might move into simple.
Most management and leadership is based on (assumption) of fact based systems. That works in simple or complicated.
A reorg creates a complex system. So is climate change, but different degrees of complexity. In complex spaces different agents are at play and interacting in different (and unpredictable ways.) The pattern is proble, sense, respond. Can have rigor or s simple probe/test the water. In Singapore when SARS came out, hugely unpredictable and emergent. But because Hong Kong hospitals could coordinate different responses, they could test a number of options. Cycles are shorter than in a typical complicated space. Emergent, instructive patterns. Unknown unknown, many competing ideas, a need for creative, innovative approaches. Pattern based leadership. That
Chaotic – high turbulence, no right answers, high tension, many decisions to make and no time. Act, sense, respond. Disorder – things from chaotic and simple, when you look top to bottom, can go in and out of disorder quickly. Something upsets the underlying issues of something simple. Might lead to another domain. Things aren’t necessarily static. Helps us understand situations and know how to act.
Rapid prototyping, making safe fail experiments, ways to use failure productively in the complex domain. Failure in the simple side is not pushing the envelope. So looking at one’s KM strategy, what sits in which domain and are the interventions appropriate for that domain. Pushing everything into databases (simple) may not address a complex context. Massive failure in the complicated domain is a potential disaster. It is a learning space in the complex.
How do you do this? What are the KM tools? No KM tools but diagnosis tools.
Diagnosis tools? How do you do it? Different ways, for e.g.:
- Butterfly stamping - using post-its
- Future backwards – work to past state. We tend to make assumptions based on past evnts. Can’t predict future but examine assumptions about future.
KM – complex but has been approached on the right side of the framework. Development is also complex. Do we understand causality? We can’t measure KM – this framework could make KM more measurable Explore tensions – get the hell out of the complicated
When we talk about including everyone – we need to discern WHEN to do this. As a Seattle utility customer, they don’t have to ask me how to run the dam, but when they want me to conserve, I need to be engaged. There is a similar question about KM, and making a better business case and ability to evaluate/ROI. Leia’s experience in Rwanda. 10 Country HIV grant. Mandate was transition to local partners. In Rwanda, ministriy of health. It had never been done. Everyone said it could not be done. We did it. Had a great minister. Thinking back through the KM piece, primarily thinking about the lessons learned paper we did. What we expected, what happened, how the partnerships worked. Learned a great deal and shared it. The process of how we measured how well we were meeting our transition benchmarks. One was we needed to involve government, patients, providers and ask the mhow they felt we were doing meeting the benchmarks. Their perceptions. Their sense of the indicators. M&E? KM? Where you are going to get the pieces of information to tell you, what processes to elicit. How to triangulate if you need to redirect your efforts to meet your benchmarks. And community internally, to governments, to patients and providers. Massive M&E frameworks, then repositories that had the supporting documentation. As the evidence was compiled.
PEPFAR – Presidents Emergeny Plan for AIDS relief. Was it largly supply chain? (Supported by MSH and JSI) but also community and health systems strengthening. Seeking patterns. Farmer to Farmer video story. Are there times when you identify things that are complex and you must get it to complicated. Bounding things on purpose to force them ofer. (Nancy talked about Sensemaker, using narrative fragments and other data for pattern recognition…)
Relating this to KIM. How do you apply this to learning? (Future backwards, butterfly stamping, and other methods). How do you make the knowledge available and people can learn? This gives you a vocabularly to place the issue as grounding and level setting. Solutions Exchange example – working well in complicated and simple. How can CoPs help us in complex. You have to absorb the discussion and body of knowledge and make something with it, try it out, bring it back to community and discuss it further. One aspect of how to work in the complex. Challenge is people want to give you “the” answer. Sometimes that reductionist model is not helpful. Need to be aware of the complex domain. Recent post by Dave Snowden on 8 practices for working in the complex space (Nancy reblogged it at http://www.fullcirc.com) and Chris Corrigan (“Parking Lot” blog) has been translating some of that work into practical approaches.
Org change story. What are the implications of the change? Why aren’t we doing AARs all along the way. Who is going to do the work of the deleted jobs? AAR rapid, frequent during work vs at the end.