Still recovering from the TOGAF conference – about which I need to do a blog-report later, ‘cos some major shifts there – and still ridiculously tired from the way-too-early-start, way-too-late-finish days of the conference itself.
But a key point came up yesterday in a conversation with Erik Proper that was nominally about the enterprise architecture notation language ArchiMate. We started by talking about various themes in his ‘meta-blog‘ – “a blog about things I want to blog about” – and the conversation wandered onto Cynefin, which he hadn’t seen before. I described each of the Cynefin domains, at the end of which Erik commented that if the real world is naturally chaotic – naturally anarchic, in the science/physics sense of ‘without rules’ – then all the other modes are ways to guide our understanding of that chaos. The ‘unorder’ segment is where we start the assessment, perhaps, but in practice – at the point of contact with the real world – the ‘chaotic’ domain is where we must always end up, because everything else is just an abstraction for the purpose of ‘making sense’ of what’s going on.
If that’s so, we might note that the analytic segment of Cynefin – the ‘knowable’, complicated domain up in the top right of the frame – is in diametric opposition to the real world. It assumes that there are ‘knowable’ rules – which the real world says there aren’t – and it takes time to do its analysis – a luxury that we don’t have in the real-time interactions of the real world.
So in an enterprise-architecture that would work in the real world, would we need an anarchist counterpart to every analyst? Not just business anarchist, but functional anarchist, process anarchist, security anarchist, data anarchist, storage anarchist? And if so, what would be their work and their role within the enterprise? – keeping things real within that realm, perhaps?
Seems a useful idea to explore, anyway.