Sunday, April 03, 2005


Wilkins's (Temporary) Apostasy

John Wikins, possibly depressed at the approach of winter in the Southern Hemisphere, has said kind things about a paper by William Dembski. Dembski's paper was a response to a paper written by Nicholas Matzke, setting forth a hypothesis for the development of the eukaryotic flagellum. Matzke's paper is fairly technical; he does provide some background for the reader.

The flagellum, of course, is of some importance to Intelligent Design. The design is so complex, and complex in such a way, that it is considered highly unlikely that it could have come about by what ID advocates consider to be Darwin's theory. There is thus, in their view, no escape from the conclusion that the flagellum was intelligently designed.

Kenneth Miller has written a refutation of the idea of "irreducuble complexity", which is the basis of the design inference. Matzke's article is more technical, referring to specific proteins.

Dembski's response, which Wilkins describes as "masterful", deals less with biology than with style. Dembski starts by listing the various screen names Matzke has used on different discussion boards. How this is relevant to anything, Dembski doesn't say. It sounds like a culture clash.

Dembski goes on to mention that Matzke, at the time a graduate student in geography at the University of California at Santa Barbara, graduated from Valparaiso University, "a Christian college in Indiana". More specifically, Valpo is a Lutheran college. I attended Lutheran schools in Fort Wayne, which is not far from Valparaiso, so Valpo was always in our minds. I thought about going there, but my mother said that my siblings and I needed to learn "that there are other people in the world besides Lutherans". After a money-exhausting year at a local college I joined the Navy, and found out in a hurry what she was talking about.

Dembski begins his argument by observing that Matzke devotes 13 pages to references and 14 to figures. Ten pages discuss prior research, three are "scene-setting". Only 20 pages are devoted to the actual argument. Dembski doesn't seem to appreciate the careful approach in Matzke's paper. Biologists like lots of references, and discussions of prior research.

In the end, Dembski grants all the biology Matzke mentions. Dembski's major objection is that Matzke's explanation is not sufficiently Darwinian. But what is "Darwinian", and why does it matter? Dembski says that
Darwinism is a theory of process. It says that you can proceed from point A to point B in biological configuration space provided that you can take small enough steps where each step is fitness enhancing (or at least not fitness detracting). The steps need to be small because Darwinism is a theory of gradual incremental change where each step along the way is reasonably probable. As Darwin put it in his Origin, for his theory to succeed it must explain biological complexity in terms of "numerous, successive, slight modifications." Anything else would cause his theory to founder on the rocks of improbability. (emphasis added)
Darwin emphasized small steps in order to stay away from saltationism, the idea that there are abrupt transitions. This was, for him, too much like creation. Darwin worked more than a hundred years ago, though, and the theory of evolution has made some progress since then. The first big jump was the development of the theory of genetics, building on the work of Mendel and Morgan. (More on this in a later post.) This led to neo-darwinism, the merger of genetics and evolution. Later still came Lynn Margulis's theory of endosymbiosis, still later developmental biology, merging genetics and embryology. Darwin didn't know anything about these areas; he based his theory on what he could see.

Dembski says
The bacterial flagellum is a marvel of nano-engineering. ... If a biotech engineering firm were required to draw up blueprints and design specifications for the construction of the bacterial flagellum, it would require thousands of pages (especially if the individual proteins that go into the construction of the flagellum had to be fully specified in terms of their structures, functions, and properties).
Besides begging the question a bit (he needs to prove that the flagellum was designed, he can't assume it), Dembksi misunderstands biology. Biological structures aren't constructed the way human structures are. There is no specification: (the previous incarnation of) John Wilkins observes that DNA is not information. Mary Ann West-Eberhard argues that the metaphor of DNA as blueprint doesn't work either. DNA and enzymes are just molecules which, by reason of their chemical shapes, allow other molecules and atoms to fall into minimum-energy relationships, with some rearrangement of valence electrons along the way. As to probability: The process that Matzke hypothesizes may have taken place over millions of years, involving astronomical numbers of bacteria. The mutations don't have to be sequential, either: Bacteria can exchange genes, in a way that serves some of the functions of sex in more-complex organisms.

Dembski says that Matzkie's model is not testable:
As for Matzke's claim that his model is step-by-step, that's trivially true -- after all, he defined the model as a series of steps. But are those steps reasonably small so that they constitute what Darwin called "successive, slight modifications." My sense is no -- getting from a type III secretion system to a bacterial flagellum in six steps seems on its face to require at least one big leap somewhere. But intuitions aside, given that Matzke's model is not detailed, there's no way to decide whether the steps are small enough to be accommodated by the Darwinian mechanism.
But cooption is in a sense a large step, maybe even qualifying as a saltation. As to detail: A hypothesis of this scope often leaves considerable room for details to be filled in.

Dembski's finale:
Two years ago cell biologist Franklin Harold published a book with Oxford University Press titled The Way of the Cell. In it he explicitly repudiated intelligent design: "We should reject, as a matter of principle, the substitution of intelligent design for the dialogue of chance and necessity." And yet he continued, "But we must concede that there are presently no detailed Darwinian accounts of the evolution of any biochemical or cellular system, only a variety of wishful speculations." (p. 205) Is Harold, a noted cell biologist, right? Or has Matzke, a geography graduate student, disproven Harold? To be sure, stranger things have happened. But Matzke is not in the same league as the mathematician Galois, who at a tender age resolved outstanding mathematical problems that had lain open for millennia. Matzke's model, far from resolving the evolutionary origin of the bacterial flagellum and despite his protestations to the contrary, is yet another exercise in Darwinian storytelling. In this regard I commend him, because he has told the best Darwinian story to date concerning the evolutionary origin of the bacterial flagellum.
I haven't read Harold's book, but defenders of evolution learn to mistrust attackers, because it turns out too often that immediately after a damning quotation from a book on evolution the author says something like "however..." Or the attacker might leave out some important words in the quote. I'm not saying that Dembski did quote Harold out of context, just that he didn't quote enough of the passage to let us be sure.

Maybe the thing that turns scientists off most is Dembski's forgetting that the burden is on him to show that there is a design. Some ID advocates say that ID is only about recognizing design, not about identifying the designer. The most popular method of identifying design seems to be to show that something couldn't have come about by natural means. Maybe this is the only way, absent direct evidence for a designer, or a design process. The problem with this method of argument is that each year we know more about natural processes than we did last year, so saying that there is no possible way, because we can't find it this year, is not very satisfying.

Dembski forgets the difference between did and could have. Matzke gives what seems to be a reasonably possible explanation of how the flagellum could have evolved. Whether it actually did or did not evolve that way may still be open. But Dembski's argument for intelligent design requires that every "could have" argument be false, i.e. that there is no way something could have happened naturally. It is not enough that an individual "did" is false, because there might still be another possible natural way. And to call it "non-Darwinian" is not an argument for a designer. Either we can say that evolution has gone beyond Darwin, just as physics has gone beyond Newton, or we can say that Darwinism has been extended to include genetics and molecular biology. (The title of this blog suggests that I prefer the extension choice.)

Fortunately for the world of philophiles, the next spin of the earth restored John to his senses.