Doping ID

Over at ID the Future, Paul Nelson has a brief post regarding the doping scandal that will likely deprive cyclist Floyd Landis of his recent Tour de France victory. For those who do not follow cycling, several tests performed after the race showed that Landis had an unusual ratio of testosterone-like hormones in his blood, and that the hormones found contained amounts of specific carbon isotopes not compatible with endogenous origin (for a thorough discussion of the tests and the reasons for Landis’s failure, see this post at Jake Young’s Pure Pedantry blog, as well as links and follow-ups therein). The conclusion from the anti-doping agency was that Landis had (voluntarily or not) taken artificial steroids, and therefore ought to be disqualified.

Nelson extracts his own moral from the story, which is that we can scientifically detect the result of intelligent action without having to exclude every possible natural source of the hormonal imbalance, and, implicitly, that therefore ID is a viable scientific program and - ta-dah! – those evil Darwinists who claim otherwise are just selling smoke. However, Nelson’s attempt at ‘roiding up ID is just as easy to spot as Landis’s.

Since Nelson’s post is short, let me quote it in its entirety:

“…it is almost impossible to be caused by natural events. It’s kind of a downer.”

That’s how Greg LeMond responded to news that doping tests may have implicated 2006 Tour de France winner Floyd Landis [see the final paragraph of the story]. The World Anti-Doping Agency, which supervises the international standards for the licit and illicit use of hormones, defines the presence of “exogenous” (illicit) chemical agents as follows: “Exogenous” refers to a substance which is not ordinarily capable of being produced by the body naturally.

We can expect that Landis will defend himself by trying to find “natural” – non-intelligent – causes for the anomalous ratios discovered (and now confirmed) by testing. We can also expect testing agencies to weigh those proferred explanations in terms of their plausibility.

What we won’t see is anyone saying that intelligent action – in this instance, the deliberate use of intelligently-synthesized steroid compounds to gain a competitive advantage – cannot be detected, in principle, because such inferences involve a universal negative (“natural causes cannot produce x”).

It is possible to catch cheaters. Happens all the time, in fact.

There are so many issues with this argument, it’s hard to know where to begin, but let’s try.

First, in the case of evolution we have a series of known mechanisms that bring about organismal change, and all the end results we observe today are compatible with such mechanisms (though they need not be - every organism could have a completely different genetic code, for instance). This is true also of things like “irreducible complexity” - so much so that not even Behe claims that IC is absolutely impossible to evolve, just highly improbable, in his opinion. On the other hand, in the case of Landis, we also have a series of well-known biological mechanisms (steroid biosynthesis, carbon isotope ratios in biological samples and their origin) and a result that cannot be obtained through such mechanisms (especially the isotope result). In other words, for Landis’s blood data to be natural, we would need to postulate entirely new physiological mechanisms, whereas for, say, the flagellum to have evolved naturally, only known evolutionary mechanisms would have to apply.

Second, it is some ID advocates, most notably Bill Dembski, who claim they have devised systems to reliably prove universal negatives - i.e. that something cannot possibly have been generated naturally - based on statistical considerations and the artificial conflation of “natural” with “by either regularity or chance alone”. Indeed, one of the scientific objections to Dembski’s explanatory filter is that its purported reliability in eliminating natural causes is utter nonsense. (Dembski has waffled on occasion about this, but he has repeatedly said things like: “… whenever the Explanatory Filter attributes design, it does so correctly.”)

Third, of course, is that in Landis’s case, parsimony hugely favors a design conclusion because we have a good idea of who the “designer” could have been (Landis himself, and/or someone on his medical and training staff), and what his methods and motivations were. No need to hypothesize a violation of natural law, supernatural and/or alien interventions, or some other mysterious undescribed entity. This is even more the case in a legal proceeding such as Landis’s doping evaluation, for which the applying standard is simply that of “reasonable doubt”.

How this all comes together becomes rather obvious with a simple thought experiment. Suppose that, instead of finding the unusual hormone features in a professional cyclist’s blood, scientists had found them in a newly discovered, isolated human population deep in the Amazon jungle. Would the same inference of design now be justified? Or would scientists hypothesize and test new hypotheses of natural mechanisms causing the anomalies, before assuming purposeful doping? I don’t think there’s really any doubt what the answer is. (In fact, I would venture that, in such a case, even if a natural mechanism were not identified after extensive research, scientists would still be extremely reluctant to conclude purposeful doping, in the absence of a candidate doper with means and motivations to perform the deed.)

So, Nelson’s last paragraph is right after all: cheating can be detected, and intellectual cheating is no different.

[Note: The original version of the post mentioned carbon “radioisotopes”. As pointed out by a reader in the comments below, this is incorrect: the isotopes in question, C12 and C13, are both stable. The error has been corrected.]