The cybernetic roots of (some important bits of) ecology

I recently read Dan Davies’ new book The Unaccountability Machine (a brief accessible summary of which is here.) Among other things, it’s a potted history of, and attempt to revive interest in, management cybernetics. Briefly, cybernetics, as developed by Stafford Beer and others in the decades after WW II, was a set of ideas about how to manage complex systems–things like large companies, universities, and national economies. Those systems are “complex” because their internal workings are far too complicated to be fully described and understood in every detail. That complexity makes them difficult to manage, in part because it creates the potential for unexpected and catastrophic behavior.

But of course, it’s not only human organizations and economies that are “complex” in this sense, and it wasn’t only cyberneticians who were interested in complex systems. For me, one of the pleasures of reading Davies’ book was encountering some cyberneticians whose ideas were foundational to some important bits of ecology.

For instance, one of the other founders of cybernetics was child prodigy, legendary MIT prof, mathematician, computer scientist, and philosopher Norbert Wiener. As an ecologist, you should recognize his name. Except that you may not, because ecologists often either misspell it as “Weiner,” or miscite it as “Weaver.” Norbert Wiener is the “Wiener” in “Shannon-Wiener diversity index.” Wiener’s ideas were a big influence on information theorist Claude Shannon. And so the information metric that ecologists call the “Shannon-Wiener diversity index” is named for both of them.

Norbert Wiener. Source: https://monoskop.org/File:Norbert_Wiener.jpg

Another leading cybernetician was psychiatrist W. Ross Ashby. Ashby was interested in how feedback mechanisms enable adaptive behavior by, and homeostatic regulation of, complex systems such as the human brain. His book An Introduction to Cybernetics popularized the term “cybernetics.” Ashby also wrote the very influential Design For A Brain. I’m guessing this is the first time many of our more junior readers have heard of Ashby, but our more experienced readers definitely know who I’m talking about. He’s the “Ashby” in Gardner & Ashby 1970 Nature, “Connectance of Large Dynamic (Cybernetic) Systems: Critical Values for Stability.” In that paper, Gardner and Ashby suggested that any large complex system could be approximated as a matrix of random numbers, with those numbers describing the effect of a change in each part of the system on each other connected part. Gardner and Ashby used a computer to generate a bunch of matrices of various sizes and degrees of connection, and asked if those matrices possessed a stable equilibrium or not.* Their simulations revealed all the results relating “stability” to “complexity” that physicist-turned-ecologist Bob May famously derived analytically two years later (May 1972 Nature). In particular, the result that a sufficiently species-rich and/or highly connected ecosystem has vanishingly small odds of possessing a stable equilibrium. Implying either that real ecosystems don’t have stable equilibria, or else that they have stable equilibria only by virtue of possessing some sort of highly non-random structure, in terms of which species interact with which others in what way.

In summary, you could argue that the two biggest and most important ideas in ecology in the 1970s both came from cybernetics. I don’t know that you’d be right to argue that. But I don’t know that you’d be wrong either!

But that was the ’70s, man. These days, ecology isn’t nearly so into abstract ideas, and it’s definitely no longer into borrowing abstract ideas from other fields (borrowing statistical methods, yes; borrowing abstract ideas, no).

*Yes, I’m being very imprecise in summarizing Ashby’s work, but it doesn’t matter for purposes of this post.

Hot Topics

Related Articles