Complicated versus complex, tight versus loose coupling

I was inspired to write this post after reading Taryn Davis’ blog post Working with Complex Systems and Complicated or Complex – knowing the difference is important by Will Allen (a link within Taryn Davis’ blog post).

More and more I hear the words ‘complex adaptive systems’ in our ‘industry.’ When I started working in international development in the late 70s, such words were used only by a small group of people, far from common and well understood. The latter is still true I believe. It is imperative that we understand what the relevance is of these words for our work. Given the poor showing of the amount of money that has been poured into health system strengthening when Ebola arrived on the scene, and the difficulty we have of showing results that are commensurate with the amount of money spent, maybe we should take a closer look at this notion of complex adaptive systems.

Complex is raising a child. Complicated is exploring a new planet. For the former, no matter what guides we read, what ‘best practices’ we use in our parenting, we can never predict the outcome of this parenting. Some kids end up well, others not. Growing up is a complex affair. There may be some things we can predict, but there are too many variables and interdependencies that interact to produce the grown up child.

Getting a robot to walk on the surface of the planet Mars, we now know, is technically possible, even though complicated. We know that if we do all the steps right (and the knowledge on how to do these steps right exist somewhere), then we can predict the outcome with a great degree of certainty.
Working with governments is complex. It is complex because governments are run by people, not robots or rockets. Rockets, always built to very precise specs, are the same and one can expect they behave in the same ways, but humans do not as we all know.

Yet in our work with government agencies we often act as if ‘the system’ that we have to interact with is run by robots that will do the right thing; that government systems (agencies) are complicated rather than complex. In complicated systems we will do a (predictable) good job if we clearly define roles, make ‘the right’ decision, structure a clear and unambiguous chain of command, make decisions when they are due and stay focused. Sound familiar? Job descriptions, org charts, decision-making tools, SMART results, indicators, M&E plans.

But we don’t work with robots or rockets. We work with people who have hopes, fears, agendas (hidden or out in the open), complexes, accounts to settle, aspirations, family pressures, etc. This is what makes complexity, well, complex. The leadership role must then be different as well.

Leadership in complex (adaptive) systems requires attention to relationship building, sense making, continuous learning (the act/learn/plan/act loop), noticing what works and building on that and loose coupling. The latter term is used most widely in computing and system design and refers to the degree of direct knowledge that one component has of another. Back in the 70s Karl Weick used the concept to describe, among other things, educational systems. In that context, more relevant to us, loose coupling referred to situations where several means can produce the same result, where there is a lack of coordination, an absence of regulations and highly connected networks with very slow feedback times. Weick recognizes that these qualities appear negative but then explains that, “they actually may help the organization by allowing the organization to temporarily persist through rapid environmental fluctuations, improving the organization’s sensitivity to the environment, allowing local adaptations and creative solutions to develop, allow sub-system breakdown without damaging the entire organization, allow more self-determination by actors.” (Karl Weick, “Educational organizations as loosely coupled systems”, Administrative Science Quarterly, 21 (1976), 1-9 (part). [1]

His reasoning appears to go against the logic that I infer from common discourse about health systems strengthening which often implies tighter rather than looser coupling. Referring back to the Ebola epidemic, I did hear calls for tighter coupling. But think about this consequence: when the connections are too tight, the whole system will break down when one part of it does. Isn’t that exactly what happened in West Africa? I try to visualize what loosely coupled health systems would look like and how to see this in a positive light rather than a situation to be fixed. But I am coming up short, knowing instinctively that there is something of importance for us to explore in this different vision, but not quite sure what.

[1] See also at https://faculty.babson.edu/krollag/org_site/org_theory/Scott_articles/weick_lcs.html, last accessed on 9/9/2015