The work of Brenda Zimmerman and her colleagues, Curt Lindberg and Paul Plsek is published in 1998. They introduced leaders to Edgeware: Insights from Complexity Science for Healthcare Leaders and it is this work where I will take a deeper dive.
Edgeware (1998) was published at a time of significant flux in the organizational life of hospitals in Canada. Conservative governments were promoting significant reorganization in what was being positioned as “right-sizing”. There was merit in the work; health care costs, which are a significant part of provincial budgets in Canada, had been driving upwards for many years and experts indicated there was potentially some slack in the system and room for new ways of doing things. As is common is such situations, there was a great deal of angst among those charged with “changing” and there was expected job loss as a result of new expectations. Edgeware with its practical tools and ways of progressively understanding systems was a welcome addition to my reading list and to the vocabulary I was trying to master.
At that time, I was retained to support the change efforts of the nursing department of a large complex pediatric academic hospital. It was challenging work in a highly complex adaptive system and Edgeware offered new insights, useful mental models and tools to consider. At the time the prevalent thinking was not CAS but industrial era, mechanically-driven bureaucracy and organizational change was laden with anxiety and fear about the future. At the same time very ill children and their families needed to be cared for with high levels of nursing skill, collaboration and compassion. Stradling the two worlds – industrial era modeling and complexity thinking – was a constant challenge especially given that my own earlier education was limited to industrial era models with all their rules, boundaries and machine metaphors. It was becoming more and more obvious that complexity thinking was what was needed in so many current situations.
The first Edgeware principle I took on immediately was to view your system through the lens of complexity, the idea that there are many ways to view, examine, and understand a situation. We are creatures of habit and upbringing, how we learned to think in the past didn’t necessarily mean there weren’t other ways to view a situation. I learned, not for the first time, that there are many lenses available to us and we can cultivate ones that are useful in different situations.
I liked a good enough vision right away, the idea that a vision doesn’t have to be perfect to “get it out there” and start using it. Iterations and improvements come along as they are discovered and/or required. I began to put out ideas that constituted a “good enough vision” of where we needed to go and why, and to work with people who could help to put more viewpoints and ideas forward. This in itself was a great way to get people involved in discovery processes, to foster new thinking, new ideas, and engaging discussions. We also learned together to utilize the concept of minimum specifications –to avoid too many rules and pre-determined habits and let people explore, practice, do things their own way, and let the results emerge. Trust the process is an essential mantra to consider for life experiences such as these and it was important in this situation.
Chunking was and is very useful. When you feel overwhelmed, it’s important to break the work into manageable pieces and just get started – almost anywhere. As you work at it, it begins to come together. Trusting the process is very useful here too.
Complexity requires the use of both Clockware and Swarmware. When I was first introduced to this idea, I needed time to understand what it meant and even what it was. Management process terms, coined by Kevin Kelly, founding editor of Wired magazine, clockware refers to rational, measured, standardized mechanisms where control is paramount. These are the time-honoured management tools in industrial era, hierarchical and bureaucratic factory models. Swarmware, on the other hand, are tools and processes that are often experimental, compatible with self-organization, and untested approaches, where learning models and the use of multiple action in a trial-and-error approach are useful.