Posted by Nick Matzke on November 28, 2005 09:14 PM

2,4-dinitrotoluene (or 2,4-DNT, a chemical relative of the explosive TNT)Dembski has posted a contest on his blog, seeking the best case of gradual evolution by cooption (hmm, sounds like something evolutionary biologists have been talking about for decades) in the case of human-designed technology.

I’m talking about an actual history of invention in which an initial technology does A, and then a small change allows it to do B, after which a further small change allows it to do C, after which co-opting an existing system (without extensive modification) allows it to do D, etc. The evolution of a motorcycle from a motor and a bicycle is not a good example in this regard because the motor and bicycle require extensive design-work to adapt them to each other.

I hereby nominate the gradual evolution of “intelligent design” during the descent with modification of the manuscripts of Of Pandas and People, reviewed in this series of posts.

I expect my $100 by week’s end. Seriously, though, speaking of cooption: Dembski’s last-ditch, backup-backup argument against cooption as an explanation of irreducible complexity – that we had no well-documented examples of natural cooption – was destroyed by Scott Minnich during the Kitzmiller v. Dover trial.


I think Dembski may now be exploring gradual, intelligence-guided cooption because cooption nuked Behe’s irreducible complexity argument. Unfortunately, Dembski’s specified complexity argument depended entirely on Behe’s irreducible complexity argument in order to succeed in biology. Dembski and Behe first said that irreducible complexity couldn’t evolve because any precursor missing a part was by definition nonfunctional. As a backup argument, they acknowledged that cooption could perhaps produce such systems, but they only addressed a fake, strawman version of cooption, namely all-the-parts-come-together-at-once, tornado-in-a-junkyard-type assembly. When that backup argument didn’t hold water, because of the commonality of functional subsystems (with some but not all of the parts of the “irreducible” system of interest), Dembski switched to a backup-backup argument: “The only well-documented examples that we have of successful co-option come from human engineering.” (“Unintelligent Evolution“, December 2004 – various versions of this same statement are found in several places in Dembski’s writing).

That was from the start a silly thing to assert, but unfortunately for Dembski, his colleague Scott Minnich torpedoed whatever vestiges of credibility Dembski’s backup-backup argument might have had, when he explained in his sworn testimony in Kitzmiller v. Dover that a degradation pathway for a toxic compound (2,4-DNT, a derivative of the explosive TNT), a pathway which requires a half-dozen different proteins, had in fact naturally evolved in microbes on a U.S. Air Force base within the last few decades.

The relevant Minnich testimony is here, and the exhibit they are discussing is this paper: Glenn R. Johnson and Jim C. Spain (2003). “Evolution of catabolic pathways for synthetic compounds: bacterial pathways for degradation of 2,4-dinitrotoluene and nitrobenzene.” Applied Microbiology and Biotechnology, 62(2-3), pp. 110-123. [PubMed] [DOI] [Journal]

Johnson and Spain (2003) is a review paper, based in part on this research paper: Glenn R. Johnson, Rakesh K. Jain, and Jim C. Spain (2003). “Origins of the 2,4-Dinitrotoluene Pathway.” Journal of Bacteriology, 184(15), pp. 4219-4232. [PubMed Central (free)] [DOI] [Journal]

As if Minnich’s testimony weren’t clear enough, the fact of natural cooption in the origin of this system was underscored by DI spokesperson Casey Luskin. In his review of Minnich’s testimony, Luskin wrote,

Minnich, who had previously read the paper, explained that to evolve this pathway required the modification of maybe 2 or 3 preexisting enzymes. There was really nothing new here, and certainly nothing approaching an irreducibly complex biomolecular machine. Minnich called this microevolution.

There you go, a multiple-proteins-required system evolving in an evolutionary eyeblink by a natural process of cooption, something which leading ID expert William Dembski said required intelligence, and we get the “it’s just microevolution” line!

But it’s just a metabolic pathway!

Clever ID advocates will pull out the old emergency defense, used only when convenient by Michael Behe, that “metabolic pathways are not irreducibly complex” (italics original). Unfortunately for this defense, at trial and elsewhere Michael Behe, William Dembski, and Casey Luskin have all made it very clear that metabolic pathways with multiple required parts are, in fact, irreducibly complex:

1. Dembski,Three Frequently Asked Questions about Intelligent Design (PDF)“, September 2003:

Strong Irreducible Complexity of Molecular Machines and Metabolic Pathways. For certain enzymes (which are themselves highly complicated molecular structures) and metabolic pathways (i.e., systems of enzymes where one enzyme passes off its product to the next, as in a production line), simplification leads not to different functions but to the complete absence of all function. (italics original)

2. Luskin favorably quotes this whole section of Dembski’s article on the IDEA center website.

3. Behe, discussing the lac operon (the lactose metabolism pathway of E. coli) Kitzmiller Trial Testimony Day 10, afternoon:

Q. Sir, have some scientists argued that there is experimental evidence that complex biochemical systems can arise by Darwinian processes?

A. Yes, there have been a total of two such arguments which I regard to be very important, because these were claims that there had been experimental demonstrations, not just speculations, not just stories, but experimental demonstrations that either irreducible complexity was incorrect or that complex systems could be built by Darwinian processes.

Q. And one of those claims was raised by Dr. Miller, is that correct?

A. That’s correct. I think on the next slide we see that he wrote in his book Finding Darwin’s God,which was published in 1998, he said, “A true acid test used the tools of molecular genetics to wipe out an existing multipart system and then see if evolution can come to the rescue with a system to replace it.”

So here he was making the point well, here one test of this claim of irreducible complexity and the ability of Darwinian processes to make complex systems, well, is to find a complex system in a cell, destroy it, and then see if random mutation and natural selection can come back and replace it. And I have to say I agree that’s an excellent test of that claim. However, I disagree with Professor Miller’s further comments and conclusions.

Q. What was the particular system that he was looking at?

A. Well, he was referring to what is shown in a little cartoon version on the next slide. This is a figure again taken from that biochemistry textbook by Voet and Voet discussing a system called the lac operon. Now, an operon is a little segment of DN A in a bacteria which codes for a couple of genes, and genes code for proteins, and the proteins usually have related functions or function as a group, and one of them is called the lac operon which is used to, the proteins of which are necessary for the bacterium Escherichia coli to metabolize a sugar called lactose, which is a milk sugar.

And it consists of a number of parts.


Q. So this was the system that Dr. Miller was talking about in –

A. Yes, I’m afraid this is the background for the system that he started to discuss in his book.

Q. Which he sees it as experimental evidence to refute the irreducible complexity claim?

A. Yes, that’s right, and if you look on the next slide you’ll see the part of his book where he discusses that. He says of the system, he says, “Think for a moment. If we were to happen upon the interlocking biochemical complexity of the re-evolved lactose system, wouldn’t we be impressed by the intelligence of its design. Lactose triggers a regulatory sequence that switches on the synthesis of an enzyme that then metabolizes lactose itself.

“The products of that successful lactose metabolism then activate the gene for the lac permease, which ensures a steady supply of lactose entering the cell. Irreducible complexity, what good would the permease be without the galactosidase? No good of course.” And he continues that same discussion on the next slide, he continues, “By the very same logic applied by Michael Behe to other systems, therefore, we can conclude that this system had been designed, except we know that it was not designed. “We know it evolved, because we watched it happen right in the laboratory. No doubt about it, the evolution of biochemical systems, even complex multipart ones, is explicable in terms of evolution. Behe is wrong.”

Q. Is Dr. Miller right?

A. No. Dr. Miller is wrong. Now, Professor Miller is always enthusiastic and he always writes and speaks with great excitement, but I say that when you examine his arguments closely, under close inspection they simply don’t hold up and this is enormously exaggerated, and the results of researcher Barry Hall that he is describing here I would happily have included as an example of irreducible complexity in Darwin’s Black Box.

(bolds added)

So, three prominent spokespeople for ID agree: metabolic pathways with multiple required parts actually are irreducibly complex. And yet Dr. Minnich admitted at trial that cooption had produced just such a system in a matter of decades. He has furthermore said that his colleagues at the University of Idaho have gotten similar results on other degradation-resistant, human manufactured chemicals, like TNT. And, in fact, the scientific literature is full of such examples.

So, to sum up, as a result of Kitzmiller v. Dover, we’ve not only established that ID really is creationism, and that the ID movement has been scandalously hiding this inconvenient fact – we have also established that both molecular cooption, and the natural evolution of irreducible complexity, have in fact occurred within a few decades. And not just once, but many times.