Chaos Theory

A Book Review of ‘Chaos’ By James Gleick

I’ve always been intrigued by chaos theory but never dived very much into it. This book seemed incredibly popular, so I figured it would be the perfect opportunity. It’s rather old, being published in 1987, back when chaos was a relatively new field. This has the downside of making chaos theory sound a lot more exciting than it is, but it doesn’t feel like a fatal flaw.

I really enjoyed the historical aspect of chaos theory. How it came to be, the problems it solved, and the problems it encountered. It’s a very fitting example of a paradigm shift in some sub-fields, and it was delightful to read. I also found it very interesting to see how it was aided by the development of computers, and their power both helped and limited what could be known.

How personal Gleick made some of the advancements, focused on the individual scientists’ discoveries and how they got to that particular point, made it super interesting, and it brilliantly captures the beauty of scientific discovery.

I deepened my knowledge of chaos theory many folds. And some of the sections were really well-written and easy to understand. The fundamental insight is that chaos theory tells us that we don’t live in a deterministic Laplacian fantasy where we can predict everything in advance. A good example that is covered in the book to understand this is weather prediction. As we increasingly became more knowledgeable about the weather, models were built to predict it, but it was surprisingly difficult, especially in the long term. Any error in measurement multiples and makes the whole process unbearably complex. The issue with chaos theory isn’t that it breaks cause and effect, by rather that you can’t know all the causes by principle alone.

For example, in the weather, you can’t have sensors in every city. And if you have it in every city, you can’t have it in every street. And if you have it in every street, you can’t have it in every meter, and so forth, to infinity. But let’s say you have infinite sensors, which will always be impossible. Then we have another problem, the sensors will always round. You can’t get infinite decimals. But those decimals, at some point or another, will start to matter. They will make larger and larger variations, and as the whole process complexifies, your whole model breaks down.

A similar example is given with tracking coastlines. We can say that coastlines are infinitely long because whatever method you choose to measure it, it won’t capture all the small twists and turns. And if you use a more precise tracking method, then the length of the coastline will increase. But you can always be more precise, up to a point where we simply cannot measure it, and the deeper you go, the longer the coastline is.

The entire field is incredibly fascinating. However, it did have a big drawback. It’s fascinating but hard. The concept as a whole, as I just described isn’t very hard, but there are many nuances and other examples which aren’t very easy to grasp. As the book progressed, I found it harder and hard to keep up, and it started to have more and more math, something I almost have a phobia for. About halfway I got so tired of it that I decided to stop reading. The topic is really interesting, but some of the technical descriptions got a bit too dense for me, at least given what I wanted to get out of it.

It’s not a bad book, and I really enjoyed most of it, but the math ruined it for me. I think for most people, especially with a good hard science background (which isn’t me, I’ve always been more interested in meta-science than science), it will be a good book. If the math doesn’t scare you and if you’re interested in math, then you’ll likely enjoy Gleick’s work. If you’re not familiar with chaos theory at all, then it will be incredibly fascinating.

If you want something lighter, I’d recommend Chaos: A Very Short Introduction by Leonard Smith.

Thanks for reading! If you like non-fiction book reviews, feel free to follow me on Medium. If you don’t use Medium, you can subscribe to my Substack.