What Is Complexity?

By Gerrit Van Wyk.

All you need is a simple checklist.

In the US Institute of Medicine report To Err Is Human published in 2000, the authors wrote around 44,000 to 98,000 patients in the US die every year from medical mistakes, which means more people in the US die from mistakes doctors and hospitals make than die in car accidents or from breast cancer. There is disagreement if the numbers are correct and about under what circumstances these mistakes occur and people die, but that’s not the point, the point is the media got hold of this and sensationalized it which created a furor which got the attention of the usual suspects, politicians, planners, consultants, conspiracy theorists, and the like. Something had to be done and into the breach stepped Dr Atul Gawande.

By all accounts Dr Gawande is an American success story; surgeon, academic, CEO, and writer. He is also politically very well connected, and has links to the business world. He published The Checklist Manifesto in which he wrote, all you need to make complex surgery safe is a simple checklist, which tells me he doesn’t really understand complexity.

According to the philosopher Immanuel Kant we cannot know anything without making some basic assumptions. One such assumption of vital importance is what the reality around us looks like, and how it works, because that determines what we believe we know and how we act on that.

Enlightenment thinking and the later Industrial Revolution is the foundation on which our thinking in the West is built. We grow up with it, understand the world in terms of it, act according to that understanding, and most importantly, take it for granted, is unaware we do so, and never question it. If that is your point of departure, what Dr Gawande said makes perfect sense, but what if there are alternative perspectives, that our reality and world is complex for instance?

To all of us in the West, our reality and world looks like a clockwork, which means the parts fit together logically, we can take the clock apart to know how it works, we can make accurate predictions about how the clock works, if the clock breaks down we can replace parts, we can design better clocks, and we are in full control of clocks. But notice “we” look at clocks from the outside in as if we are not part of it, and clocks are there to serve us. That is the basis of science and how scientific laboratories as well as modern management works, and it works well if the thing we are interested in is simple and works like a clock, but if it doesn’t, we get into trouble.

Although most social research is done in a clockwork fashion, we instinctively know our social world doesn’t work like a clock; it is complex. What does that mean? One consequence of a clockwork world is we can classify things into different categories and pigeonhole them, which means from that perspective complexity is a property of the clockwork or thing which we can know by analyses, and that knowledge makes complexity predictable and gives us control over it. But notice how from that perspective our world is still a clockwork reality, in other words our foundational assumption about reality remains the same. That is how Dr Gawande sees reality and complexity and how his simple checklist works.

I see complexity completely differently. Something is not inherently complex like the complex clockwork perspective above; it depends on how we look at it. The more pieces and interactions we add, the more complex that thing becomes and the more we take away, the simpler it appears, in other words looks like a clockwork. When we look at something as if it is a clockwork, it is not because it is like a clockwork, it is because we look at it as if it is a clockwork which is useful for example to build a bicycle, but useless when the thing we look at is complex.

Complex things don’t work like clockworks at all. They have properties emerging from the interacting parts which means if you take them apart to analyze them, they no longer have those properties and you can’t tell anything about them. Like Douglas Adams said, if you take apart a cat to know how it works, the first thing you end up with is a non-working cat.

Because there are many parts and they interact in many possible ways, you’d have to accurately know all the parts and their interactions to make accurate predictions, and leaving out even one part is likely to render the prediction useless. Statisticians use mathematical models as a workaround which predict the likelihood or probability something will happen, but that prediction is only true for the parts and interactions you chose to include, and cannot be extrapolated in general without making big assumptions.

The properties of complex entities depend on the parts and how they interact; if you change the parts or interactions, the properties change, hence all parts play a role, and the properties depend on how the pieces interact. It means small events elsewhere can have unexpected big effects far removed (the so-called butterfly effect), the conditions you start with matters, and very importantly, like in the quantum world, we are part of complexity and what we do influences what happens.

So, one can see here Dr Gawande’s simple checklist cannot simplify complexity, because it starts with the wrong assumption about reality. Once upon a time, like in a fairy tale, I saw the world like Dr Gawande and it didn’t work for me, but today I see the world as the complex entity it is, and it does. If we want to meaningfully change health care, we must shift away from Dr Gawande’s world and acknowledge our reality for what it is, as difficult as it may be.