Dangerous Science Fiction

Sunday, January 8 2006 - 0 Comments

I keep re-reading the essay by Michael Chrichton that came up on del.icio.us/popular last week called Fear, Complexity, Environmental Management in the 21st Century (which I recommend reading if you're gonna sit through this). Chrichton wrote that science-fiction book State of Fear in 2004 which portrayed eco-terrorists and environmentalists in a similar light and created a bit of a stir in the science world just because so many people were citing the book in their reasoned arguments for or (more often) against the threat of global warming. The essay above is kindof an encapsulation of the logic of the book boiled down to a talk form.

The thesis of the essay is that complex systems abound in nature and our ability to understand them is inadequate. The conclusion that he seems to draw from this is that complex systems (the environment, in particular) should be left to do their complex system thing -- that things will take care of themselves in ways that we cannot forsee. What he's implying is that complex systems have some sort of self-equilibrating mechanism, and I think he's right to a degree. Over certain time scales something like climate can be considered an observable measure that implies equilibrium. But instead of talking about equilibrium and the conditions that lead to and support that equilibrium, Chrichton talks about complexity in a way that makes it sound like the universe has no order and that complexity is really a non-science for all of it's predictive power.

One of the tenants of chaotic systems (or complex systems -- they're the same thing) is that they have sensitive dependence on initial conditions (i.e. small changes in input can yield large changes in output) and Chrichton admirably notes this in the essay. Weather is a complex system where this logic should be applied. Weather is extremely sensitive and can appear almost random at times. Climate (defined as a long term average of weather) is more stable precisely because it averages out the seemingly random elements of weather.

Something that Chrichton gets wrong is the ability to control chaotic systems. It is absolutely true that chaos and complexity can be controlled, though maybe for Chrichton's approval I'll use the term managed (I'd like to say control here actually, because in a scientific and engineering sense it can exactly mean manage). It can't be controlled in the way that you can control an airplane or a rocket, you can't guide it wherever you want to (except when you can), but you can keep some systems from spiraling away toward some undesirable extreme. It's not always possible and it's certainly not always easy, however, complexity does not preclude control of complex systems. Chrichton mentions the stock market as an example of a complex system that can not be controlled, and he's wrong. The stock market could be controlled if we had access to a gigantic pile of money. We could manipulate the price of a stock to move it any direction we wanted when we wanted. Of course that's not the way to make money in the stock market. The stock market is chaotic in reality because we do not know what the millions of inputs (individuals with smaller sums of money betting on certain stocks) are into the system with exact (or even approximate) specificity. Therefore the system behaves like a complex system because any attempts to model it have large errors in input and as we know an error in input can lead to large errors in output. The stock market can't be controlled for practical reasons not theoretical ones (though the SEC exists in part to make sure that fraudulent attempts to control the market at a micro-cap level are not allowed).

For our purposes here complexity means that we can't control weather, but we can control climate (and conversely, we can fuck up climate). The key, going back to the sensitive dependence line, is to keep the system's inputs within the boundaries of conditions that support the desired behavior. Meaning, quite simply, don't change your inputs too significantly or you will overtax the equilibrium mechanism that handles these (relatively) small deviations. That's the key to controlling complex systems, correct the inputs while the deviations are small, before they become large and unfixable. This is what ecosystems do. This is why they are important. They are resilient to change, but not infinitely so. There are many inputs for climate which need to stay within certain bounds, but there is at least one input in particular that humans are perturbing to ranges outside of modern historical norms. The question is how far we can push it before we exceed the system's equilibrating capacity. The corollary question is what new behavior results from leaving equilibrium. This is really where the complexity question comes into play because we really don't have a basis for knowing what might happen if we leave our current equilibrium, but we know that large change is possible. This is, to me, the case for managing the problem. Let's not push ourselves out of a comfortable equilibrium into an unknown system.

While I agree with Chrichton's disapproval of using fear as a motivator and see the observation about civilization naturally moving toward de-carbonization very optimistically (though I question the speed at which we are doing so and the economic penalty we pay for moving too slowly), I take issue with the essay's science by analogy approach to persuasion and his heavy use of media distilled science to support his conclusions. But in the end, Chrichton's main fault is that he misapplies elements of complexity theory to justify a why bother attitude and to nearly demonize the scientific process. To think that scientists are working with linear frameworks for the behavior of complex systems and only science fiction writers are privy to the non-linear way of things is truly ridiculous.

Posted by Ben at January 8, 2006 09:19 PM

Comments

Post a comment