Complexity: Life at the Edge of Chaos [Englisch] [Taschenbuch]

As a second book, narrowly focused on the question of complexity in biology, it is outstanding. Specifically, the question is one of how self-organization (complexity) relates to evolution and what this means for natural selection. Complexity is frequently talked up as the unifier of the sciences. Lewin takes a balanced approach, taking the time to talk to complexity theorists and understand their ideas, then talking to mainstream biologists to see how the ideas relate. His conclusion shows no inherent bias. Where other books on complexity show extreme (perhaps undue) enthusiasm, Roger Lewin's concusion is decidedly "wait and see". I found his insights to be on target and relevant.

I mentioned that this is a good second book. For an introduction to complexity, read John Holland's "Hidden Order". For a history of the Santa Fe Institute and some of the personalities there, read Mitchell Waldrop's "Complexity". Either or both of these would serve as an adequate introduction to this book.

In this revision of his book originally published in 1992, Roger Lewin explains what the science of complexity is all about through interviews with some of its most important practitioners (and critics) organized around some of the central ideas. As such this is both a fine introduction to the subject and an interesting read. Lewin includes 16 pages of photos of the scientists he interviewed captioned with a significant quote from each. He has added an afterword on the application of complexity science to business, and an appendix about John Holland, whom he dubs, "Mr. Emergence."

"Everything works toward an ecology" is an old dictum of mine. I have the sense that I came up with that myself, but I probably read it somewhere years ago. At any rate, what is being said here is that complex systems work toward a state of equilibrium near a transition phase, near "the edge of chaos." This equilibrium can be an ecology (Darwin's "tangled web"); indeed it can be the entire planet, as in the concept of Gaia in which "the Earth's biological and physical systems are tightly coupled in a giant homeostatic system" (quoting Stuart Kauffman on page 109).

A central idea is that "...large, interactive systems-dynamical systems-naturally evolve toward a critical state" (physicist Per Bak, quoted on page 61). These systems include weather, financial markets, piles of sand, and most significantly, ecologies, so that evolution itself is seen as shaped by the dynamics of complexity. Complexity is the "interesting" middle ground between order and the purely random, between the crystalized structure of ice and the Brownian motion of molecules. I had a curious sense of understanding when I compared these three states with positions at chess. First there is the even, static position, perhaps with bishops of opposite color in which no progress can be made, a drawn the inevitable result. Second there is the wildly chaotic position so complex that no one can completely calculate it, say the board after black takes white's queen knight pawn in the "poisoned pawn" variation of the Najdorf Sicilian. In between are the "interesting" positions in which one side might have a small advantage or there might be a dynamic balance of advantages, space versus material, for example, in which a startling combination might be hidden.

These states-"at the edge of chaos"-are seen here as analogous to the phrase transition states of matter, from liquid to gas, for example. The idea is that there is a naturally occurring property of the physical world that forces complex systems into stable, readiness states near the edge of transition. What is exciting is that these states, because they are so "ripe" for change can be influenced or manipulated into change with small resources. Out of complexity comes something that could not be predicted by an analysis of its individual components, an emergent property of the system. I would note that such a natural phenomenon would be attractive to those who believe in punctuated evolution (e.g., Steven Jay Gould) and to those who believe that social and political change typically comes suddenly and with great force.

Central to what complexity science is saying is that reductionism-which is the technique that has driven science to its present position of power and influence-is limited. "...[Y]ou have to look at the interactions as well as the parts," John Holland is quoted as saying on page 220. In other words, you have to take a holistic approach. However, the use of the word "holistic," a New Age shibboleth, is the just sort of thing that makes traditional scientists wince.

Consequently, complexity science is not without its critics who argue that the fundamental mechanism of complexity exists only in a mystical sense and is therefore anathema to mainstream science. Even its practitioners, such as University of Michigan "complexologist" John Holland, admit they are still searching for the fundamental mechanism of this new science. He is quoted on page 214 as saying, "Our present understanding is not much better than the child saying that Jack Frost explains the wondrous colors of autumn."

However most complexity scientists would say that the mechanism isn't mystical at all. It's just not understood yet. I would add that much of what we think we know about the world is based on relationships and phenomenon that we assume we understand, but really we don't. For example physicists say that gravity curves spacetime, but they don't say how it curves spacetime. Presumably gravitons do the trick, but they haven't been discovered yet! So it could be said that gravity is mystical. I like to compare this lack of understanding to the task of watching grass grow. (This also works for evolution.) Every day I look but at no time do I ever see the grass growing, yet after a while I know it has grown. It seems that it always grows when I'm not looking! By the same token we see the results of complexity, but we do not yet see the inner workings of the process. We may never see the process, but through complexity science we may yet understand it.