There was a recent interview with Bill Dembski, leading light o' the ID movement, on Dick Staub. It's mostly the same old garbage, but I thought I'd highlight one part in particular:
Q. Well, let me ask you this. If intelligent design became broadly accepted within the scientific community, how would science have to change? In other words, what about the current scientific process would have to be different? Is intelligent design testable right now within science, you know? There's a strict adherence to certain testability quotients and so forth. How does science change if they accept intelligent design?
A. Well, I just want to speak to the testability business. I would say intelligent design is testable and, in fact, Darwinian evolution is not testable. Darwin said that for a complex organ to form it would have to form according to a series by a numerous successive slight modifications. And then he said, you know, I can't think of anything that couldn't have formed that way. Well of course, I mean, if you don't specify a process any more specifically than numerous successive slight modification, that anything might be the result of that process, such a process. [...]
Claims like this would be irritating if they weren't so hilarious. It's not too uncommon for Dembski to contradict himself, but sometimes the degree is startling. This is the plain fact of the matter: Dembski's method of "detecting design" requires, absolutely, that Darwinian evolution is testable. That's all there is to it. Since he and his fellow ID travelers find "evidence" of design whenever some natural process is supposedly incapable of doing the job, then there's no way he can claim to have evidence of design unless the natural process in question can be definitively ruled out. Dembski has just admitted that his whole design detection methodology is bunk. Nice going.
And that's all, the Darwinists assume no burden of evidence of proof as a consequence. And that holds to this day. Now, with intelligent design there, you can look at certain biological structures. We're arguing that they are intelligently caused.
Dembski's argument that biological structures are "intelligently caused" consists entirely of him saying that Darwinian evolution can't account for them. And this is right after he says that Darwinian evolution can't be tested. In typical fashion, he's trying to have his cake and eat it too. Either Darwinian evolution is testable (and found incapable of doing the job), or it's not testable, and therefore we can't reject this particular natural mechanism, let alone the set of all possible natural mechanisms. (And yes, his ID argument requires that all natural mechanisms, even those we might not have thought of yet, be rejected.)
Many of them are now at the sub-cellular level, these are molecular machines. The most popular one that's been investigated is the bacterial flagellum. It's a little bi-directional motor-driven propeller on the backs of certain bacteria, marvel of nano-engineering, and so we've started to analyze systems like that and argue for their intelligent design.
It's worth mentioning that Dembski has done one (1) analysis of the bacterial flagellum, which appeared in his book No Free Lunch, and that this analysis consisted of nothing more than multiplying each of the amino acids that make up the flagellar proteins times one another in order to get the probability of the flagellum coming about completely at random. No evolutionary process is even considered. Since no one has ever argued that this was how the flagellum came about, it's a pointless waste of space. Calling it an "analysis" is probably a bit too kind.
And as far as anyone can tell, that's it. No other "analysis" has been performed either by Dembski or by any other ID advocate on any subcellular system.
Now, it would be an easy enough thing, in principle, for the Darwinists to come along and say, Hey, this is how sub-systems could have formed. They would have to get a detailed testable step-by-step scenario of how these systems could have formed, according to some Darwinian trajectory or pathway, and if they did that for a number of such systems, I think intelligent design would crumble.
Except for the minor inconvenient fact that this has been done. There are detailed and testable step-by-step scenarios for the evolution of many subcellular systems that Dembski and others have claimed couldn't evolve, including the bacterial flagellum. So how does Dembski respond? He claims that they aren't detailed enough. Exactly how detailed a hypothesis would need to be in order to satisfy Dembski is something he's never clarified, so this criterion is an endlessly moveable goalpost. No matter how much detail one provides, he can always ask for more. And this brings us right back to the original point: Is ID testable? Not if there's no way to rule it out, or at least reject it for being too unlikely. And that's precisely where Dembski has placed it. Neither he nor any other ID advocate has ever given us a way to independently judge its likelihood (it's always just assumed that some supernatural *Poof!* is perfectly likely), and now he's given us no way to rule it out by virtue of parsimony. If it's not detailed enough, then the evolutionary explanation isn't complete, and so therefore ID cannot be ruled out. (Just a quick aside: there would be no way to rule out ID even if an evolutionary account were complete down to the last quark, so Dembski has to rely on Occam's Razor, i.e., the rejection of an unnecessarily complex explanation. By doing so, he implicitly concedes that ID cannot be tested by virtue of empirical evidence.)
Most of the rest of the interview is the same old canards, misleading talking points, ad hominems, undefined terms, self-contradictions, and so forth. So I won't waste a valuable Saturday (at work) going through the whole thing line by line. But anyone else can feel free. :)