originally posted at www.designinference.com
On October 11, 2003, the Talk Reason website posted an article by Nicholas Matzke titled "Evolution in (Brownian) Space: A Model for the Origin of the Bacterial Flagellum" (http://www.talkreason.org/articles/flagellum.cfm). Talk Reason advertises itself as a website that "presents a collection of articles which aim to defend genuine science from numerous attempts by the new crop of creationists to replace it with theistic pseudo-science under various disguises and names." The most obvious target here is intelligent design. Indeed, Matzke's article attempts to rebut one of the main challenges that intelligent design has raised against Darwinian evolution, namely, how to explain the emergence of irreducibly complex biochemical machines like the bacterial flagellum.
Before reviewing and critiquing Matzke's article, I want to offer a few remarks about Matzke himself and my past interactions with him. Matzke's day job is as a geography graduate student at the University of California at Santa Barbara. Nonetheless, he is also one of the most active participants in online discussions concerning evolution and intelligent design (the sheer volume of text that he is able to generate is remarkable). In such forums, he tends to go by various pseudonyms. His main one until a year or two ago, when he blew his cover by publicly attacking Jonathan Wells at UCSD, was "Nic Tamzek." On the ARN bulletin board (www.arn.org) he has used "Niiicholas." On the ISCID bulletin board (www.iscid.org), through which I know him best, he goes by "Yersinia." He uses still other pseudonyms in other forums (as in the Talk Origins newsgroup).
Matzke's interest in intelligent design and evolution goes back at least to his undergraduate days at Valparaiso University, a Christian school in Indiana. As far back as 1996 (and perhaps earlier) he was posting online in the American Scientific Affiliation's evolution discussion forum. All these discussions are archived, and it appears that at the time Matzke was still keeping his options open about where he would come down in the debate over biological origins. In the last four years or so, however, his views have ossified so that his defense of Darwinian evolution and his attacks on intelligent design have become unswerving if not predictable.
I learned of Matzke's latest article through a New Zealand biochemist named Robert Mann. Mann maintains an Internet mailing list critical of intelligent design. Like many, Mann worries that intelligent design, in claiming to show that biological systems exhibit signs of intelligence that lie beyond the reach of the Darwinian mechanism, is committing an argument from ignorance. Thus, in his most recent email to me, he wrote, "Admittedly, the submicroscopic details of the flagellum, or a fortiori the chloroplast, remain unexplained by neo-Darwinian theory. The logical gist of IDT is a 'designer of the gaps' inference from lack (pro tem) of scientific knowledge. This is open to getting filled in, as a neophyte has recently suggested at http://talkreason.org/articles/flagellum.cfm." The neophyte Mann cites is Nicholas Matzke.
So has Matzke in fact filled in the gaps that intelligent design claims are insurmountable for the Darwinian selection mechanism? In particular, has he provided a detailed, testable, step-by-step Darwinian model for the origin of the bacterial flagellum? Matzke claims that he has. Thus we read at the beginning of his article: "This article will propose a detailed model for the evolutionary origin of the bacterial flagellum." And at the end of his article we read: "Finally, in light of the organized complexity and apparent 'design' of the flagellum, the very fact that a step-by-step Darwinian model can be constructed that is plausible and testable significantly weakens the suggestion that extraordinary explanations [read intelligent design] might be required."
In fact, such claims by Matzke about what his article is supposed to have demonstrated are highly misleading. Matzke at one point in his article refers to the bacterial flagellum as an "icon of intelligent design." Certainly it's understandable (and even commendable) that as a Darwinian he should want to knock this icon down. But to do so he must make good on his claim to provide a detailed, testable, step-by-step Darwinian model of how the bacterial flagellum could have originated. Unfortunately for him, that claim is false under any reasonable construal of the terms "detailed," "testable," and "step-by-step." The further claim that he has significantly undercut intelligent design is therefore false as well.
To see that Matzke's proposed evolutionary model for the origin of the bacterial flagellum is deeply flawed, I want to grant Matzke most of the actual biology he cites and focus instead on the logic by which he arrives at his conclusions. Matzke, as is evident from his Internet postings as well as from the article under consideration here, has a tendency to overwhelm with citations to the biological literature. Indeed, one of my colleagues in the ID community refers to him as a "PubMed junkie." Yet when it comes to putting arguments in his own words and rigorously following through a train of thought, Matzke is decidedly less in his element. Let's therefore focus on the logic and structure of his argument.
For starters, let's do some simple bookkeeping. My print-out of Matzke's essay weighs in at 58 pages single-spaced. Of these, 13 pages are devoted to references. Another 14 pages are devoted to figures. That leaves 32 pages for his actual argument. Of these, 3 pages are devoted to concluding remarks reviewing and plugging his model for the origin of the bacterial flagellum. In addition, the first 10 pages of the essay are stage-setting, describing past research that attempts to get a handle on the flagellum and its origin. Thus only 20 pages of the article are in fact devoted to Matzke's actual model for the origin of the bacterial flagellum.
Why are these page numbers significant? They are significant as a reality check. The bacterial flagellum is a marvel of nano-engineering. As Matzke himself admits, thousands of research articles have been written about it, many of them trying simply to discover the role and function of its various components. Howard Berg describes the bacterial flagellum as "the most efficient machine in the universe." 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). And yet, somehow, with the Darwinian mechanism in hand, all that design work can be passed over. A "detailed, testable, step-by-step" engineering approach to the construction of the bacterial flagellum would require thousands of pages, and yet a "detailed, testable, step-by-step" Darwinian approach to the construction of the bacterial flagellum requires only 20 pages. On its face, there's something funny going on here.
This is a reality check, not an argument. Let's now see why this reality check in fact points to deep problems with Matzke's model for the origin of the bacterial flagellum. For Darwinists like Matzke, the reason 20 pages suffices for a "detailed, testable, step-by-step" Darwinian model for the evolutionary origin of the bacterial flagellum is that the Darwinian mechanism of natural selection and random variation is supposed to constitute a divide-and-conquer strategy for explaining complex biological systems in terms of simpler systems. Provided the bacterial flagellum can be explained in Darwinian fashion in terms of several simpler systems, the presumption is that these simpler systems can in turn be explained in Darwinian fashion in terms of still simpler systems, and so on until we reach a level of simplicity that does not require a Darwinian explanation at all. In consequence, it is enough for Matzke to show that the origin of the bacterial flagellum can be explained in terms of an evolving sequence of simpler systems.
If the bacterial flagellum did indeed evolve, then a bacterium, call it B, with a bacterial flagellum evolved from a bacterium, call it A, lacking not only a flagellum but also all the genes for the flagellum (including any genes homologous to the genes for the flagellum). For a Darwinist to be convinced that an account of the evolution of B from A is plausible, it is enough to find a few intermediates between A and B that could plausibly have evolved from A through those intermediates into B. Matzke follows this strategy. Thus he proposes the following intermediates: (1) He starts off with a bacterium that has what he calls "an ancestral type III secretion system." (Note that Matzke requires 7 pages to justify such a system as a precursor to the bacterial flagellum; that leaves 13 pages for the remainder of his "detailed, testable, step-by-step" evolutionary model.) (2) Next, this bacterium evolves a pilus or hair-like filament that extrudes through the type III secretion system; this pilus will later become the "propeller" that drives the fully evolved flagellum. (3) Next, this pilus experiences, as Matzke puts it, "rapid improvements ... under selection for increased strength, minimizing breakage, increased speed of assembly, etc." (4) Although initially involved in adhesion, the pilus evolves motility, which initially is quite crude, being nondirectional and simply for "random dispersal." (5) Next, this "crudely functioning protoflagellum" gets a chemotaxis and switching system tacked on so that motility becomes directional and interactive with the environment. (6) Finally, this entire system gets refined through natural selection, which evolves a hook and additional axial components and thereby forms a modern flagellum.
Question: In what sense is this model detailed, testable, and step-by-step? To answer this question, we need to consider more closely how Matzke justifies this model. The key elements of Darwin's theory that underwrite Matzke's six-step model for the evolutionary origin of the bacterial flagellum are cooption, change of function, and natural selection. Given an ancestral type III secretion system, additional elements get co-opted and incorporated into the evolving system. As they get co-opted, the system changes its function. What's more, between times of co-option and change of function, natural selection enhances the system by honing the existing function.
Given this picture of how Darwinian evolution moves from step (1) to step (6) in his model, Matzke needs to do two things to justify it. First, he needs to show that systems being co-opted were in fact available to be co-opted (in particular, he needs to justify making the type III secretion system the starting point in the evolution of the bacterial flagellum). Second, he needs to show that each evolutionary transition from one step to the next represents an increase in fitness and could with reasonable likelihood have occurred. Now, in Matzke's mind, he has satisfied both these requirements. But has he?
To justify the availability of the components that later get folded into the ancestral type III secretion system as well as the availability of this system itself as a starting point for his model, Matzke looks to evidence from molecular homology. Perhaps the biggest stretch is getting the ancestral type III secretion system itself to start the ball rolling since, as Matzke himself admits, the best evidence points to this system arising only after metazoans have been on the scene, and thus many hundreds of millions of years after the origin of the flagellum. Thus Matzke makes an unconvincing argument for homologies between the type III system and an ATP synthetase system. As Matzke admits, "sequence similarity searches do not turn up significant hits," so the putative homology is based on some unspecified similarity metric and is therefore less than rigorous. The other components that subsequently need to get folded into this starting system are less of a stretch but still highly dubious. For a number of the proteins that make up the flagellum, there simply are no known homologues. And for others, there is no reason to think that they were available for co-option in the actual ancestral line leading from a bacterium with no flagellum (and no genes for the flagellum as well as no homologues for these genes either) to one with a modern flagellum. Matzke just helps himself to whatever biochemical products are available in whatever living forms are available subject only to a very loose conception of homology. Anything in this ballpark, and even purely hypothesized biochemical products, constitute fair game for co-option in Matzke's model.
Notwithstanding, I'm happy to grant Matzke all the homologues he wants as well as the ancestral type III secretion system to start his model rolling. That's not where the problem lies for the Darwinian evolution of the bacterial flagellum. Indeed, there are forms of intelligent design that require the same concession. Matzke at one point in his article refers to the "antievolutionary 'Intelligent Design' movement." Conjoining the term "antievolution" with the term "Intelligent Design" has become a useful rhetorical ploy by Darwinists for discrediting intelligent design but in fact is quite misleading. Michael Behe, the best known proponent of intelligent design, holds to universal common descent. He is as much an evolutionist as Matzke. Where they differ is on how evolution brought about biological complexity. For Matzke and the majority of biologists, the Darwinian mechanism is all that's required. For Behe, some form of intelligent guidance is additionally required. But clearly, if the bacterial flagellum is evolving under intelligent guidance, then existing designed structures are fair game for co-option into newly designed structures. Insofar as intelligent design is a theory of evolution, it is a theory of technological evolution, and technologies evolve by taking advantage of existing technologies. Thus co-option will play as important a role in any intelligent design models for the evolution of the bacterial flagellum as in Matzke's Darwinian model.
That leaves the other thing Matzke must do to justify his model, namely, to show that each evolutionary transition from one step to the next represents an increase in fitness and is reasonably likely to have occurred. Although Matzke devotes considerable attention to justifying that certain transitions in his model represent increases in fitness, it's hard not to see these justifications as post hoc rationalizations. Fitness is a relation between an organism and its environment. What evolutionary biologists trying to explain the origin of the bacterial flagellum have an empirical handle on is the modern flagellum, the modern type III secretion system, and various homologues of structures embedded in the flagellum that appear in extant organisms. What they don't have an empirical handle on is Matzke's "ancestral" type III secretion system, the environment it inhabited, what sort of environment lacking metazoans this structure might have had a selective advantage in, the intermediates in Matzke's model leading from ancestral type III secretion to the modern flagellum, and the environments in which those intermediate systems (if they existed at all) might have been functioning. Matzke's most sustained argument for selective advantage relevant to his model concerns the minimum functional requirements for a pilus/filament to serve as a motility structure capable of overcoming Brownian motion (hence the title of his article "Evolution in (Brownian) Space"). For a bacterium with an ancestral type III secretion system that has sprouted a pilus, does motility for the pilus (and thus for the bacterium) constitute a selective advantage that overcomes the cost of evolving motility? Maybe. Maybe not. Who's to say? Given our minimal knowledge of the ancestral environment leading up to the bacterial flagellum (and we're talking an environment at least two billion years old when it comes to the evolutionary origin of the bacterial flagellum) and given only hypothesized rather than actual evolutionary intermediates leading up to it, there is no way to decide what does and does not constitute a selective advantage. Just about anything could constitute a selective advantage or selective disadvantage.
Notwithstanding, let me grant Matzke all the selective advantage he needs in moving from one step to the next in his model. The biggest obstacle justifying his model still remains to be addressed, namely, showing that each step in the model is reasonably likely to follow from the previous one. 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.
Are the transitions from one step to the next in Matzke's model reasonably probable? Does each step in his model constitute only a "slight modification" (sensu Darwin)? There's no way to tell because the model is not sufficiently detailed. As I pointed out earlier, all we have in hand is the modern type III secretion system, the modern bacterial flagellum, and various homologous biochemical structures embedded in the flagellum present in extant organisms. We don't have the intermediates that Matzke posits nor the ancestral type III secretion system. We don't know what they look like. We don't have their precise biochemical specification. We don't know if the intermediate systems that Matzke's model hypothesizes would work. We have no way of determining how easy or hard it is for the Darwinian mechanism to bridge the steps in Matzke's model. Matzke, throughout his article, invokes gene duplications and mutations at key points where the Darwinian mechanism is supposed to effect transitions that are reasonably probable. But what gene exactly is being duplicated? And what locus on a gene is being mutated? Matzke never says. Indeed, his model is so unspecific that he cannot answer these questions. But unless we know the answer to such questions, there's no way to know whether the transitions Matzke describes are reasonably probable and therefore of the type required by Darwin's theory.
Matzke claimed that his model for the Darwinian evolution of the bacterial flagellum is detailed, step-by-step, and testable. Let consider these in turn. Is Matzke's model detailed? Hardly. Lynn Margulis's criticism of neo-Darwinism applies all too well to Matzke's model: "Like a sugary snack that temporarily satisfies our appetite but deprives us of more nutritious foods, neo-Darwinism sates intellectual curiosity with abstractions bereft of actual details -- whether metabolic, biochemical, ecological, or of natural history." (Quoted from Acquiring Genomes, p. 103.) Matzke gestures at the types of systems that the Darwinian mechanism would need to produce if it is to evolve the bacterial flagellum. But these systems as well as the changes they must undergo are left unspecified, and we have at this time no way to determine to what degree, if at all, they could be instantiated in biological reality. 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.
That leaves testability. What are we to make of the repeated claim throughout Matzke's article that his model is testable? In claiming that his model is testable, Matzke confuses a precondition for his model with actual evidence for it. Fully two thirds of the 20 pages Matzke devotes to his model are concerned with establishing homologues for structures embedded in the bacterial flagellum. In several cases the homologues are completely absent. Thus, according to Matzke, his model is testable because if the homologues are found, they would confirm his model. But this argument evinces a deep confusion. If the homologues never existed in nature, then the systems that needed to be co-opted for the evolution of the flagellum never existed either, and so the flagellum could not have evolved in Darwinian fashion. In other words, if the homologues never existed, Matzke's model is dead in the water. Matzke's model therefore presupposes the existence of these homologues. They are a precondition for the model, and any evaluation of the model's adequacy proceeds on the assumption of those homologues. That's why I was willing earlier to grant their existence. To actually test Matzke's model requires being able to evaluate the likelihood of transitioning from one step in the model to the next. Because the systems at these various steps are in fact unspecified (they are hypothetical; they do not, as far as we know, currently exist in nature; they are not available in any laboratory; and researchers for now have no experimental procedures for generating them in the laboratory), there is no way to carry out this evaluation. It follows that Matzke's model is in fact untestable.
Toward the end of his paper Matzke takes up the charge by design theorists that "the construction of evolutionary models amounts to nothing more than the telling of 'just-so stories'." Matzke claims that through his model he has effectively overcome this charge in the case of the bacterial flagellum. But he hasn't. 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. If there were time and opportunity, I would substitute his article for the one by Kenneth Miller -- "The Flagellum Unspun" -- in my forthcoming co-edited collection for Cambridge University Press with Michael Ruse. Matzke at least glimpses that irreducible complexity raises a problem for evolutionary biology. Miller just blithely invokes co-option and is done with it. Alas, my book with Ruse is already in production, so I must decline to offer Matzke an invitation to contribute to that volume.
Copyright 2003 William A. Dembski. All rights reserved. International
File Date: 11.12.03