The uncertainty matrices

I’m going way back today to when I was working for Datasolve, before the management buyout that merged it with Software Sciences to create DataSciences which was then taken over by IBM as a part of creating IBM Global Services. So that is at least thirty-five years ago. I was then managing a decision support business but also contributing more generally to strategy as a part of the senior management team. One of our most important clients was the Dental Practice Board which was headed by a fairly idiosyncratic Director. There was a general heuristic at Board level that I was a good fit with anyone in the idiosyncratic category so I got sent in. We bonded immediately over an extended discussion of Newton’t role as an alchemist and a general all-around cynicism around management fads. He’d just come out of a presentation which had featured a whole bunch of amateur psychology but one thing he had thought might be useful had been Johari windows. My response was that, in general, I disliked classification systems for individuals and further that Johari tended to the assumption that things were knowable when in practice they are not. This was pure practical wisdom on my part, then complexity science wasn’t even on my radar and I was playing around with Cybernetics, the LSE Pod, Tavistock and Soft Systems. I was then challenged to produce something better.

So on the wipe board of his office in Eastbourne, I drew the first green matrix which contrasted known with knowable in respect of the nature of the system, with known and unknown in respect of the decision maker. That worked well and a few months later I created the second red matrix which, in respect of the nature of the system talked about knowable and unknown, and in respect of the decision maker unknown and unimaginable. It was then a fairly simple effort to overlap the two and create some slides. In those days we were very excited because we had Lotus 1-2-3 and could print onto acetates which had to be protected in cardboard sleeves. I spent a lot of alcohol-fueled time in pubs with Kath and Penny using masking tape to do that before marketing seminars the next day. The constraint was useful as it took effort (and cost) to create slides so you spent time getting them right.

The two uncertainty matrices, here shown in their overlapping form, are the oldest, still in use, frameworks in Snowden’s body of work stretching back over thirty years where the first version was a response against Johari windows.  They have gone through changes over the years but at their heart lies a contrast between the nature of the system in reality on one dimension, and our knowledge/perception of the system on the other.  I’ve used them fairly frequently ever since so they are the oldest framework in my wider work. I famously presented them in the Pentagon one day and a short while after had to put them on the back burner for a bit. To this day I don’t know if the whole unknown unknowns speech was my fault, but in case it was I have always argued there is more significance in the unknown knows. The things that are known more generally in the organisation but are unknown to decision makers.

More recently I have been reworking the two matrices to map against Cynefin domains, the opening picture. But I have also used that to identify the use of Cynefin tools and methods. Some of these overlap a little and a better artist than myself could do a better job, but this is shown to the right.

The other thing I do with that matrix is to draw a diagonal from the bottom left. The known knowns are all about robustness, everything in the red matrix is about resilience. Creating systems and processes in advance of need, and unforecastable need at that. Such systems will need a degree of inefficiency if the system as a whole is to be effective. In hindsight, this should probably have been in the EU Field Guide to provide a context for the methods outlined there. We have added this into the Hexi kit by the way to correct that error. I think both pictures are pretty self-evident. A CoP is a Community of Practice for those not familiar. All other methods and tools are laid out in the EU Field Guide and on the wiki. A lot are described on this blog so a search will find material. It will be a part of the Strategy/Estuarine Field Guide when that is available later in the year.

Experienced practitioners will also recognise elements of the domain models which are undergoing a refresh at the moment; more on that shortly. That includes the positioning of chaos as a point where you may need to make hard choices. So more to come, but that, for the moment is it, other than to note that I am continuing the theme of knots.

Join our Next In-Person Training “Anthro-Complexity on one slide – Advanced Cynefin & Sense-making Intensive, UK 2024″

The banner picture is cropped from an original by David Clode on Unsplash