A short reminder of Cynefin and the difference between domains:

• Obvious situations can be categorized and have fixed constraints.
• Complicated situations can be analyzed and have governing constraints.
• Complex situations can be probed (trying something out that’s safe-to-fail) and have enabling constraints.
• Chaotic situations require you to act, and have no constraints.

The fixed constraints of the obvious are easy to understand. Think of putting a plug into the wall. There’s really only one way to do it, and it’s very hard to get it wrong!

Chaos is caused by a lack of constraints; meeting constraints will cause it to dissipate. Think of fire burning until it runs out of fuel or oxygen. This is what makes Chaos transient and short-lived; it will rapidly grow until it meets constraints, at which point the situation resolves (but not necessarily in your favour).

Robust constraints

In complicated situations, the governing constraints provide limits to what can be done. In terms of our policies and processes, these are hard-and-fast rules. They are context-free, which means they apply to everything, regardless of context.

In one of his later articles on constraints Dave Snowden refers to governing and fixed as robust constraints, which survive by becoming stronger and which, when they fail, do so catastrophically. Think of things which snap, shatter, or result in someone being immediately fired.

Resilient constraints

Enabling constraints, though, are contextual. That means that they can be adapted or escaped for context. In policy and process terms, they’re guidelines or heuristics. Dave uses the analogy of a salt marsh, as opposed to a sea-wall; the water is contained, but can gradually permeate the marsh.

Enabling constraints are resilient. That means that when they break, you usually get a bit of warning, or the impact is small to start with.

Coupling (connecting) and Containing constraints

Resilient and robust constraints can both be either connecting or containing. A connecting constraint is one which holds things in place without limiting them; a tether or an attractor. A containing constraint keeps things in one place without centering them; limiting or guiding them.

Disorder and Constraints

Disorder often occurs accidentally when we’re using the wrong kind of constraints for the situation.

For instance, think of the difficulty of getting innovative work done when you’re at a highly regulated big bank with all of its processes, its governance and change control!

Or alternatively, think of what happens when you fail to obey the safety notices or read the instructions, treating something as if it’s safe-to-fail when it isn’t.

Using the wrong kind of constraints – treating something complex as if it’s complicated, or vice-versa – is a form of disorder, and when disorder persists, it usually results in chaos.

Creating Probes around Constraints

Wicked, complex problems are often hard to solve directly, and they’re usually made up of combinations of human behaviour, and the constraints within and around which we act, whether complex or complicated.

By looking at the ecosystem of constraints – all the processes in place, and all the rules and guidelines and legacy that help us or hinder us – we can consider whether relaxing or tightening constraints might produce a beneficial outcome. That relaxing or tightening can become a probe.

In complex ecosystems, especially human ones, be careful! Dark constraints – things which cannot be seen, but the effects of which can be felt – are often cultural, and changing things can cause unexpected side-effects. Remember to check that it really is safe-to-fail!