The distinction between complexity and chaos theory can be problematic depending on who you are talking with, or what you are reading. There is no full agreement in the literature as to how the terms are used. There are also some mildly worrying differences emerging between its use in the social and natural sciences; an especial concern as complexity provides the basis for a realist and pragmatic account of social systems and a truly trans-disciplinary approach to human understanding.

While the distinction is important it’s full nature is still seen as through a glass darkly. I’m still working on how I understand the distinction, and as importantly how I explain it. For me explanation and understanding go together; its why I have my best insights on my feet in front of an audience. So this post is contingent. I reserve the right to have the underlying but as yet not fully articulated concepts change over time. Consider it a sort of manifesto; a plea to move away from dichotomies and dilemmas to dialectic and paradox, to understand that the nature of emergence is in part reflective and that we can influence at at times direct the flow.

Note: the passion of said manifesto (which some may consider to approach hyperbole is the next section so feel free to skip it!

Why is this important? with passion aforethought …

The distinction has implications for how we manage or govern social systems and how we deal with the Scylla and Charybdis of state planning and neoliberal free markets. This is of importance in general politics but is a matter of current concern in the Development sector where it manifests either in the inauthentic casualty of double blind trials or the despair of hoping that somehow or another the market will provide a magic bullet. The realisation that we are a part of the world, determined in part by our interactions with the ability to reflect, to enact and to judge, can also remove the false dichotomy between truth and falsehood perpetuated by the crude scientism of the likes of Dawkins on the one hand and the the evil banalities of the religious right. Both interestingly anthropomorphise the idea of God and separate subject from object, both are deterministic in their own petty ways. We need to see ourselves as part of a flow of meaning over time, determined and determining, being and becoming, mediated and mediating. With that we may better understand dignity, virtue and the other qualities that make us human. There is a clear role for religion in creating that, but it is not the religion of absolutism or single points of revelation. In the Catholic tradition we saw glimpses of what that might be in the work of Rahner, of de Chardin and Gutierrez before truth and justice were sacrificed on the alter of sexual (and sexist) orthodoxy. Religion in various forms is fundamental to what we are and absolutist rejection simply creates an equal and opposite reaction.

Having said all of that, the distinction and overlaps between the bodies of theory associated with both words is still not clear even within the field of those who study them let alone more general awareness. Worse, people are absconding with the language of complexity without making any change to the way they think, simply putting their old, frequently stale wine into new wineskins. So we have to work on making that separation clearer. Cynefin does it to a degree but needs further development. One of the strengths of Cynefin is that it has always been a tale that has grown in the telling and it will continue to develop over time. We need to create a language of distinctions that does not depend on static categories or esoteric language. One that allows people to sensitise themselves to clusters in the tails of distributions, to the emergent possibilities of ambiguity while planning. We need people to understand that resilience is not about recovering like some elastic ball, but more the continuity of coherent identity over time. It means a language that goes beyond the crudity of fail fast fail early to something more subtle that avoids the linearity implied by that statement. We may need a new word that encompasses the sense of failure as continuous learning without the pejorative overtones.

What is the distinction? a starting point

Chaotic systems are highly sensitive to initial starting conditions, but are also deterministic at agent level, and the agents are discrete. It is a world determined by rules in which initial conditions cannot be defined but in which the system is susceptible to models. In popular thinking Gleick’s Chaos: Making a new science got a lot of the thinking started and nuanced distinctions however important are difficult to establish once popular momentum is achieved. A complex world is not random nor is it arbitrary, we can manage or direct interactions to allow direction. We have free will (which is not the same thing as license), we can reflect of the nature of the interactions and assemblages of which we are aware, we can make choices for which we can be held accountable. Subject, object and interaction are fluid, merged and merging. Abstract constructs such as human language (itself an abstraction) allow us to create meaning with intent as to direction but not as to goal. Such intent is a product of intelligence and also of fluid and adaptive identity structures. I sometimes reference these as the 3Is of cognitive or human complexity. In a human complex systems we can create order by deliberate intent, or through habituation over time in a more fuzzy way. Seeing assemblages as strange attractors for example may allow us to gain new insight and understanding into what is the natural unit of governance within a society.

The role of constraints and the 3Is mean that we can experiment with channelling or directing the flow of emergence within a social system. But we can only gain understanding by interaction with the system as a whole in multiple ways. hence the emphasis in our methods on safe-to-fail experiments and also mapping the evolutionary potential of the system through micro-narrative landscapes. These are not the same as fitness landscapes per se but derivative of them. The net result of this is that we have to be very very careful with models. We already know that swarm robots with their physical interactions create more variety that computational models. In a human system you can add intelligent, intentional identities and the environment, tools and the like to that physicality. That means that the model of the system as whole is not just the narrative landscape, but it is also the awareness of and self-reflection on the interventions along with the choices made in the process. And complex systems are about process by the way, emergence within enabling constraints not randomness or blind deterministic compliance with rules

In governance, it means that knowing what the contextually appropriate identities are to which authority is delegated or assumed along with the nature of degree of control of their interactions. We have to understand flow to avoid unnecessary turbulence and we also have to understand what level coupling is needed to bring those identities to gather to meet common threat. Above all we have to find a way to educate (and I mean that in the widest sense) actors of varying passivities as to the nature of emergence, resilience and riding inherent and necessary certainty. I know I won’t live to see that through, but I want to provide some small momentum to in it 2015 with the Small Countries and other projects. It also means, I think, that I am now ready to get that book written …

< Prev

Between the ideal and real?

An interesting extended weekend in an Agile Alliance think tank I joined last year.  Then ...


Further Posts

Next >

Creating an architecture for change

Yesterday’s post was partly inspired by my reading David Chandler’s 2014 book Resilience: The Governance ...


Further Posts