Sunday, July 11, 2004

Gee on ID Don't miss this short article from Nature magazine about yet another problem with ID theory. It's author is paleontologist Henry Gee. Here's an excerpt:

The fact that the work seems so surprising exposes another, more dangerous conceit that scientists are prone to. Dangerous, because it leaves science wide open to the temptations of so-called 'Intelligent Design'. Advocates of this view object to evolution by invoking what Richard Dawkins has called the 'Argument from Incredulity' – that is, if I don't believe that something is possible, it cannot happen. Philosopher William Paley in his Natural Theology famously used this argument when he compared the delicate designs of nature with a pocket watch. Pocket watches are not made spontaneously, so if the existence of a functioning, integrated watch implies a watchmaker, then the same must surely apply to a living creature.

More than a century later, proponents of Intelligent Design use the same reasoning when they marvel at the intricate design of, say, a bacterial flagellar motor. How can one ever give credence to the view that the sophisticated mechanism of the flagellar motor could have evolved to such precision without a guiding hand, when the tiniest of changes to its apparently irreducible complexity might render it useless?

I was discussing the problem with two polymathic friends of mine, reproductive biologist Jack Cohen and mathematician Ian Stewart, co-authors of Figments of Reality, Evolving the Alien and The Science of Discworld. They are working on their next book, Appearance of Design (to be published by Penguin next year) and Cohen sent me a draft chapter containing a devastating response to the challenge of Intelligent Design. It arrived on my desk at about the time that the work on the lamprey immune system was making waves in the Nature office, so it struck a particular chord with me.

Cohen argues that the fallacy in the Intelligent-Design argument about the flagellar motor (or any other system), is that proponents present the motor we see as The Motor, the exemplar, the only one possible, and, what's more the best possible, surely optimized by a Designing Hand. But when Cohen searched the literature, he found that a wide variety of flagellar motors have been described, each arranged in its own way, each its own solution to effective rotary motion in the microworld. There is no such thing as The Motor, no Platonic perfection enforced on bacteria by Divine fiat. Instead we see ad hoc solutions that are not perfect, but idiosyncratic and eclectic – just what you would expect if evolution were working on its own, without a Designer.

ID proponents, most notably William Dembski, tell us that it is the combination of complexity and specification that prove that something was designed. To 'specify' the biomolecular systems they focus on, they rely on the functionality of those systems. Gee is pointing out that there are invariably many ways to design a biomolecular system to perform a given function. Any probability calculation purporting to measure the likelihood of obtaining a given system in the course of evolution must take this into consideration.

Dembski wants you to believe that evolving the particular bacterial flagella used by the bacterium E. coli (his favorite example) is an event that is comparable to flipping a fair coin one thousand times and getting heads every time. What the biological literature actually shows is that evolving this particular flagellum is more like tossing the arbitrary collection of heads and tails we expect to get when we flip a coin one thousand times. The particular sequence we get is doubtless highly improbable, but since something had to happen we don't regard it as suspicious.