Steve Matheson, with installment two of three, on the problem with Michael Behe:
But Behe’s reasoning is stunningly dumb. It goes like this.
- In the fifty years since antimalarial drugs were brought to bear on P. falciparum, more than 1020 (that’s 10 to the 20th power, if the superscript isn’t working) of the parasites have been born.
- In that time, the parasite has adapted to the drugs, through selection acting on random genetic variation, but hasn’t developed any completely new proteins or biochemical functions.
- This means that the Darwinian process is unable to generate significant novelty in fewer than 1020 tries.
That’s the argument. And Behe uses it to identify the “edge of evolution.”
Now, I hope it’s already clear to most readers why the argument is spectacularly bad, but here are some comments.
- Most basically, it should be obvious that demonstrating the failure of X to happen in one particular situation is hardly proof that X cannot happen. To extrapolate from a single negative observation (even if it were representative of the scenario in question) to the blanket impossibility of the phenomenon is a foolish mistake.
- More specifically, it should be obvious that we need not expect dramatic new functions to appear during adaptation, since we need not even expect adaptation to occur at all. If functional innovation were as inevitable as adaptation, the dinosaurs would not merely have survived, they would have mastered apparation in additional to intergalactic travel (and world peace). Behe wants you to believe that evolutionary biologists expect dramatic evolutionary innovation to occur, at the level of molecular machinery, whenever selection is applied to a population. That’s nonsense, and I think he must know that.
- It is important to keep in mind that Behe solidly affirms common ancestry, and knows that mutations account for the differences between lineages. This means that he acknowledges the existence of a continuous genetic tree of life, which means that he should be able to formulate scientifically credible approaches to his hypothesis. Pointing to the lack of innovation in one special (parasitic) scenario is hardly a substitute for a direct examination of the myriad exemplars of evolutionary novelty. In other words, the way to determine whether gigantic population sizes are needed for the stepwise generation of novel functions via random mutation is to: a) identify examples of such evolutionary innovations and to work on elucidating the genetic trajectories that could have led to their development; then b) work on determining the population sizes, mutation frequencies, and other parameters that apply to the trajectory. (I like to call this “science.”)
This kind of detailed critique from a scientist who isn’t a ranting atheist–is necessary–and welcome, because too often proponents of Intelligent Design fall back on complaints about the ‘nasty’ tone of their critics as a way to justify not responding to their scientific points. That just won’t fly anymore. (It’s no surprise that Behe’s Amazon blog allows no comments whatsoever.)