Patterns & pragmatism 3 of 3

January 14, 2024

Observant readers will have noticed that I am using pragmatism in two respects: yesterday in reference to abduction and the Pragmatists today in respect to the more normal day-to-day use. In the opening post, I talked about our natural ability to manage patterns in our daily lives and how language carries highly contextual meaning in family and social groups. I should have referenced Deacon’s Symbolic Species there for those who hung up on theories of language on which AI depends, but I will focus more on that in future posts. I referred to the role of art in meaning-making and introduced the idea of requisite ambiguity as essential for sense-making. In a direct criticism of models seeking to identify qualities and competencies, I pointed out that those are emergent properties with no causal effect. If you change how people interact, the qualities and competencies will emerge in parallel. I also challenged their static nature. In effect, we are dealing with Deluzian Assemblages: an ensemble of heterogeneous elements which compose a territory. The English translation of the French agencement doesn’t help here as it implies an assembly of components. To quote Anna and Ellie in a book on SenseMaker®, we are about to publish, it means “to arrange, to play out or to piece together“, and it is “not a unified whole, but more a heterogeneous co-existence”.

There is a social context to any human capability, which is not just about the individual but also about a body of practices that have emerged over time and are highly path-dependent. Tradition and ritual have a value function, in the main. For example, this video shows a modern craftsman creating a medieval book. Considerable individual skill is demonstrated, but that skill comes from a long tradition of experiments, errors, and consequential learning. I found that on Guy Gabriel-Kay’s Twitter feed. Readers may remember he featured in my Christmas blog series. My second post was all about abductive reasoning and its role in human decision-making. I also introduced the role of intuition in that post and the role of surprise or anomaly detection in getting us to see things differently. I also brought into place a previous set of posts on issues of granularity, abstraction and coherence, which are essential here. Our work on abduction, with SenseMaker® and otherwise, is focused on abstraction and is complementary to both the Pragmatists and the work of Bateson (father and daughter) on metaphor.

One good place to start is the idea of pattern languages, a term initially coined by Christopher Alexander in 1977. His work was to break down various architectural forms into design patterns that ordinary people could reassemble to solve complex problems. Patrick Hoverstadt and Lucy Loh have taken that approach in their book Patterns of Strategy, which I recommend. I’m curious if Patrick is happy with me doing this, and that may relate to the qualification I will make in a minute. By its nature, language needs vocabulary, and the eighty patterns in that book provide it. There are elements here of object-orientated design, which was important in my early work on methods over thirty years ago – seeing software and people as objects or patterns allows for more adaptability and resilience in system design. In any assembly or agencement, the interactions also matter, and not everything can be joined to everything else, and there may be dependencies in the process anyway. But the underlying principle of ensuring that your ‘components‘, ‘objects‘, or ‘patterns‘ are coherent and make sense as such but require combination with others to have practical application is essential. Although it arises before we see the adoption of complexity science in organisations, it works well with scaling by decomposition and recombination rather than imitation, but not with avoiding aggregation.

Our approach with Hexi links to this. We started it last year and got a lot of uptake, and we also had a lot of learning! The 2024 version is finally available for sale. The new Estuarine pack will be announced soon, as will the opening up of the approach to participation by other methods and tool providers. We’ve been working with the Flow System Playbook via Nigel and John to understand what is needed here. What is critical here is to know when an assembly becomes an assemblage. That also includes when it needs to make the change, which is only sometimes the case, and how to make it happen. In relatively crude terms, an assembly handles complication, but an assemblage is needed for complexity. An assembly creates a homogeneous whole, while assemblage is a heterogeneous evolutionary process which is never fully stable and is requisitely diverse.

Now, there are several ways in which that switch can happen, and here, I want to focus on two: firstly, when the acts of assembling and their practice become ritualised within a community and secondly, where we change the interactions and enable fractal engagement through multiple interactions. I’ll go through those now, but on the second, I am also building to a significant post and product launch midweek (I hope) so there is more detail to come.

So, to the first. I’ve previously argued that the process by which Black Cab drivers acquire The Knowledge over several years also, through the interactions, creates the abstract qualities of trust and professionalism that you would not get by simply assembling drivers controlled by apps. Both have utility, but they are very different. The social processes engender emergent properties that make the assemblage more than the sum of its parts. On the downside, while that can introduce energy efficiency, it can also stifle innovation and change. The conflict between generations of craft results in innovation in part because of the constraints of the prior generation. Much innovation in the arts arose because of the need to find ways to work around censorship, for example. At a certain level, those constraints encourage innovation, but when they become stifling, we need to engage in radical disruption. These options are all in the various types of Cynefin Dynamics and the new framework Estuarine Mapping, which de-territorialises a negative assemblage by changing the granularity.

Which leads me to the second. As a general principle, the finer the grains, the less we disagree about what they mean, and they can combine more quickly than course graining. Estuarine mapping breaks a situation down to the level of actors, constraints and constructors understood in the context of what energy and time is required for change; from that point, we can assemble something novel that achieves agreement by emergence over time and if it stabiles heterogeneity then it retains the capacity to evolve and is an assemblage. Suppose it stabilises in a homogenous manner as best or good practices. In that case, it still has high utility in the context of its creation but less adaptability when the context shifts. So, one of the things I would like to do with the 80 patterns in Hoverstadt and Loh is to create several hundred, nay thousands, sub-patterns and diverse perspectives on the original 80. As they stand, they are categories. Fragmentation and fractal search would make them something that is constantly being as becoming. Years ago, I did a similar project with an Architectural school, which allowed citizen stories to combine and recombine at a finely trained level through abstract interpretation to suggest novel and different opportunities in brownfield development. That same process applies or organisational design, strategy and many other fields. To quote myself: reduce the granularity, distribute the analysis, and disintermediate the decision makers.

Nowhere is this more important than in the field of mapping attitudes. We have done that work in the context of safety and culture and are now starting to look at ethics in general and in the context of AI. Attitudes are lead indicators from which we can spot anomalies or weak signals early enough to say something about them. And, critically, if our interventions take place in diverse contexts, we have a heterogeneous process. In effect, one-size-fits-all change and education programmes in companies try to homogenise the system based on a complete perception of what is required. That is what we need to change, and we will announce a set of offerings that are fast, low cost and easy to adopt, without the need for any special expertise, shortly. In these three blogs, I have tried to lay down some of the foundations of the approach,

The banner picture is cropped from an original, licensed under the Unsplash+ License, meaning I paid for it. There is a theme in the banner pictures of this series, starting with the formalism of a highly structured field system to the potential of a boat anchored in the mist to this more purposeful one. My opening picture in the first blog has a sense of mystery or discovery, and I then used an AI-generated image yesterday. I’m doing the same today, and this time, the instruction was to “create an image that creates patterns of meaning in a complex environment“. This one is interesting for the selection and positioning of the elements but note the regularity. I will return to that tomorrow when I start a theme on AI.

Recent Posts

About the Cynefin Company

The Cynefin Company (formerly known as Cognitive Edge) was founded in 2005 by Dave Snowden. We believe in praxis and focus on building methods, tools and capability that apply the wisdom from Complex Adaptive Systems theory and other scientific disciplines in social systems. We are the world leader in developing management approaches (in society, government and industry) that empower organisations to absorb uncertainty, detect weak signals to enable sense-making in complex systems, act on the rich data, create resilience and, ultimately, thrive in a complex world.

Cognitive Edge Ltd. & Cognitive Edge Pte. trading as The Cynefin Company and The Cynefin Centre.


< Prev

Patterns & pragmatism 2 of 3

Before Thursday's trip to Liverpool, I talked about the pattern basis of human decision-making and ...

More posts

Next >

Algorithmic Induction

This is the opening post in a short series looking at AI, which, while most ...

More posts

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram