In today’s’ post we introduce the Hexagon Sensemaking Canvas. This canvas has been in the making for some time. It is both an output of our practice and an input to our practice. We wanted to gain some experience in using the canvas before publishing it and the canvas has changed as a result of this. So here we go ….
Models and tools
Models and tools come in a wide variety and have an equally wide set of purposes. We all know the scale model which is a smaller version of something real, a plane, a car, a city, etc. And all of use now the weather forecast, which is the output of a computer model that attempts to predict the weather based on recent and current weather conditions. I’m very familiar with that kind of modelling as my PhD thesis work is on Fractal aggregation and transformation processes in silica gels. in an attempt to explain experimental observations by modelling the physico-chemical system and comparing both forms of data. The important word here is attempt, as both the weather and the aggregation and transformation process have complex aspects. This means that they can’t be predicted precisely as we all experience almost every day.
A special class of models are used by humans to make sense of what is going on, events that happened, explore the diversity of perspectives people have, diagnose organisational pathologies, make strategies of what can be changed, etc. Some sensemaking tools that are in use for thousands of years and there are recent ones. The first recent one I encountered around 1996 was Stafford Beer‘s Viable Systems Model. The model is homologous to the human nervous system and used a.o. to diagnose organisations. I didn’t track since then, but I guess that any year or two at most I ran into another one in biology, strategy, management literature, robotics and many other domains and places.
On of the things that intrigued me for a long time was the presence or absence of boundaries in those model. As a chemist we have a concept called a molecule and that system doesn’t have a sharp boundary, while living systems – from one-celled bacteria to the largest mammals have boundaries, often some form of skin. This creates an inside/outside of the system. Ofcourse in systems the boundary of a system is always a choice (f.e. is a human a human w/o relations with other human beings?), but still the System 3 (inside and now) and System 4 (outside and then) of the VSM has influenced how I assess models ever since I found it in 1996.
Confluence and Cynefin
In 2004 a colleague pointed me at the Cynefin model by Cynthia Kurtz, Dave Snowden and – as it turned out later – a group of IBM colleagues working on Knowledge Management. That one resonated with me as it is – like the VSM and my PhD work – partly based on complexity science. I was particularly glad to find people that used Stuart Kauffman‘s work on the emergence of order from chaos (where order also means patterned order like the pattern of seasons) in the social domain. I had done some work myself around 1998 to understand the travelling of information (knowledge and rumours) in human networks using simulated fractal networks so I appreciated their efforts.
As I was working on Services Innovation at the time, the group I worked in at the time start to use the narrative sensemaking approach to aid the development of new services that were emerging in the business. Many services – unlike most products – follow an outside-in path. They first arise in the business, some grow and become an integral part of the business and it is only then that head office and central R&D discover them and efforts may start to manage their further development (f.e. marketing support, creating better tools and infrastructure, delivery process improvement, etc). The outside-in v.s. inside-out perspective was present in the VSM (System 3 and 4), but not in Cynefin.
During projects it turned out that that was often a confusing factor when making sense of narrative data and especially during the design of StoryForms (signifier sets). When Cynthia Kurtz published her Confluence Sensemaking Framework in 2010 it gradually dawned upon me that Cynefin and Confluence had been separate models first (1999-2001) that became integrated from 2001 onwards as is visible in the 2003 IBM Systems journal article referenced above. Over the years I kept track of the development of both models which leads to the following overview:
In general, the CSF has been basically stable from 2001 onwards (but please note that the CSF submodels are not represented here, the full CSF is here.) while there have been minor changes in the Cynefin model:
- Cynefin started as a cultural sensemaking tool, basically a 2x2 matrix.
- In 2001 is had taken its neural/biological shape.
- In 2003 (Complex acts of knowing) terms like known, knowable, complex and chaos are present and the “Seeing eyes from the CSF have been integrated.
- In 2007 the eyes are gone and known becomes simple, knowable becomes complicated and chaos becomes chaotic.
- By 2015, chaotic reverts to chaos and simple becomes obvious.
Towards the Hexagon Canvas
So most of the changes occurred on the ordered side of the model in the known/obvious and knowable/complicated domains. Thinking about that I realised that the splitting of order into known and knowable (at 2.30) in fact introduces an inward and outward perspective:
- Known is the perspective of the agent. It is known to me, it’s internalized (note that it is assumed others know too). And what’s known seems simple to the agent in focus.
- Knowable is about the outside/other/external system. Something can be known about what’s out there by further investigation/analysis). And what’s not known may turn out to be complicated system of relations of the system in focus.
So known/simple says something about the agent (its an inward statement) while in complicated the agent says something about something outside itself (its an statement with an outward perspective). This made we wonder why only the ordered systems were split in the creation of the Cynefin model from Kauffman’s types of sytems in nature (ordered, complex, chaotic). Why not split complex and chaotic too. And why not split disorder too? After all, my thesis work and studying the VSM has shown that in nature there are two types of complexity:
- The inward / autopoietic type from biology (Maturana c.s) featuring terms like closure of organisation, self-organisation, self-maintenance, syn-reference and language development.
- The outward / networked type we find in daily life: the inability to fully control/manage the process around us even to the point that chaotic events occur.
I’ve been crafting drafts based on this idea for years, certainly since 2007, discarded many sketches and sheets. Paraphrasing Stafford Beer, many models went into the grinder. I’ve been posting earlier versions of what I’ve named the Hexagon Sensemaking Canvas on twitter over the last months and recently designed three “forms”. Here they are:
These images may see daunting without a proper story of their development and background materials. That is why we have created a Slideshare presentation (also available in PDF) shown below that presents the most important ingredients that went into the HSC on pages 2 to 9. This way we also want to make clear that the HSC is not our sole achievement, but as so often we stand on the shoulders of giants that came before us and are still with us. Next, on pages 11 to 17 the step that lead to the Hexagon Sensemaking Canvas are presented. The final pages 19 to 23 contain highres versions of the HSC and some remarks on applications:
We hope these sheets were helpful for you to understand where the HSC came from and how it can be used in practice.
The HSC is a work in progress. We publish it with a Creative Commons Attribution-ShareAlike (CC BY-SA) license so that anyone can use it freely and improve it. We encourage anyone to leave comments and suggestions on this page/site so that it becomes a repository for anyone that seeks the latest insights and practices. We intend to update the HSC and will publish new versions using the same [year].[version] format. The current version is 2015.8.