Monday, August 03, 2020

Chaos

Chaos, James Gleick

Thirty years on my bookshelf and I finally "read" it as an audiobook, and boy do I not regret it.

Of course it's a science book, but it's more a story of discovery and the explorers behind it than the science itself. Quite inspiring actually; it makes you feel like you can do stuff. I started programming the famous (and I guess overdone) Mandelbrot Set visualization because of this book. But it touches on so many other fields and applications, like the weather, epidemic patterns (hey!), animal population trends or heart defibrillator design.

For one thing, where I could see myself doing amateur research is in phase space reconstruction. As far as I understand, although there is a ton of technical jargon that makes it all sound hard to get into, chaos is all about finding a simple formula that visually resembles a complex phenomenon. And phase space reconstruction is a way to reduce complex behaviour to an essential dimension that then makes it easy to find patterns in and match to a formula you might invent.

I'm impressed by this book's journey. Written in the 80's just after chaos threw off its messy and obstructive reputation, it quickly earned a place in public vocabulary through the butterfly effect, its striking graphics and its namesake subject. Yet forty years later, no attempt seems to have been made to re-popularize the subject in a way that would send more research its way. Perhaps it's not necessary? Perhaps we all intrinsically understand chaos? But I doubt that's the case, as many of the ideas here are wrenchingly unintuitive.

It's all about contradiction. Scientists initially thought chaos was either untrue, or useless. How can you use a theory that says "everything is a mess?" But it turns out it's a new kind of mess. Other people wondered what exactly the study of chaos was: neither math, nor science. Observable only as a function of other observations, not really existing on its own. Or as a final example, the book itself. It's about scientists, but much more about their frustrations and dreams than about laws and formulae. Such an enormous subject, one we mention so often, but we talk so little about.

No comments: