Oct 19, 2018


Jennifer Garvey Berger discusses Cynefin on Farnam Street:

This framework comes from Dave Snowden, who is a thought leader in the complexity field. Cynefin itself means those things that come together to make a place a place... This Welsh word is like those forces or pieces that all come together to make a whole...

It differentiates between those things which are predictable, repeatable, ordered systems – they happen the same way each time – and those things that are unpredictable, that just because it’s happened once or even a dozen times this way it doesn’t mean it will happen that way the next time.

Snowden further breaks down the ordered part, the predictable part into those things which he calls simple, or he calls them obvious, depending on when you read them. Those things which are simple or obvious are those things where cause and effect is so tightly related that everybody knows what’s going to happen next, like the parts of organizations where you should be able to push a button and this thing will happen. I should be able to push a button and my paycheck will come out. I should be able to request something and this other thing will happen. The way the organization works is often founded on these obvious or simple processes. Often, what we offer to customers in organizations is also founded on this. A customer should be able to go to an ATM, put in a pin, and get money out. That’s the way it’s supposed to happen. If that starts to happen randomly, then you know there’s something really wrong.

So, then he says, also in this predictable space, he says there are some things where cause and effect is still connected and it’s still going to happen the same way each time, but it’s tricky. You need some expertise to figure it out. You have to analyze exactly which thing is it, and I need some experience or background to be able to know, “Oh, this is one of these and so I should treat it that way.”

This is that kind of tax problem and so I should bring my tax expertise and so you should treat it this way, but this over here, that’s a different tax problem. We should think about it in a different way. We’re still going to get to the same outcome. We’re still going to be paying our taxes to right government in the right way, but the way we go about it will be different depending on the issue it is and the expertise I have. All of those are predictable, but complicated...

Once you cross the midpoint into that unpredictable world, the rules of the game change and you move from being an ordered system, a system that has predictable patterns over time, to being a chaotic system or a complex system. Snowden calls one of those distinctions complex. That’s where there’s a connection between cause and effect, but that connection is not available until afterwards. So after somebody wins the election, we can look back and say, “Oh, these are the reasons why that person won,” but before, it’s anybody’s bet what’s going to happen.


Then his fourth distinction, the fourth area, is this chaotic space. In chaos, there’s no connection between cause and effect, either in the moment of afterwards. We just can’t tell. It’s like stuff is happening and it leads to other stuff that’s happening, but we feel lost and confused. That’s a time—it can be a time of great innovation and it can also be a time of just massive confusion and terror.

He says that for most of us, we spend most of our time in the space he calls disorder, which is a space of not knowing which of these domains we’re in in the first place. There, we tend to act out of our own preference... I’m an academic by training and so there’s a piece of me that just wants to sort and solve stuff. I want to use my expertise and be able to predict and control things in a complicated way and probably, if you just let me at it without any of these frameworks, I’d still be trying to solve these problems this way. So, the framework itself helps me ask a different set of questions, and therefore take different kinds of action.

Also there's this beautiful sketch of the framework (a).