Dembski vs. Evo Devo


William Dembski exemplifies the empty void of Intelligent Design creationism in his criticisms of Michael Ruse's review of Sean Carroll's Endless Forms Most Beautiful: The New Science of Evo Devo and the Making of the Animal Kingdom. Ruse's review was positive (as was mine—it's an excellent introduction to the discipline), but he takes a jab at the creationists at the beginning:

A major problem with the critics of science is that they have a problem with problems.

Let me be a little less cryptic. The critics—notably the creationists, and more recently their smoother descendents, the intelligent design theorists—are always whining that science has unfinished or unsolved problems.

This did not sit well with Dembski, who goes on to write a complaint that demonstrates that Ruse was exactly right in every particular, and also demonstrates several other creationist traits, such as an inability to read with understanding and quote mining.

In his review of Endless Forms Most Beautiful, Sean Carroll's new book on evo devo, Michael Ruse faults intelligent design (ID) for harping on evolution's unsolved problems. Moreover, Carroll as well as Ruse suggest that evo devo has now resolved one of the major problems on which design theorists have been harping.

Go ahead—read Ruse's review or Carroll's book. There is no blanket claim of a complete solution of anything, and Dembski is making stuff up. Instead, you will find that what we've got is a productive strategy for addressing evolutionary problems.

Wrong on both counts. Intelligent design does not have a problem with problems. It has a problem with bogus solutions that Darwinists like Ruse and Carroll dress up as real solutions to the problems of biological origins.

Note that while he's claiming he doesn't have a problem with problems, there is one figure in the article of developing squid (and truth be told, that tease was the only reason I read Dembski's article) with the question, "Can evo devo explain squid evolution?". To my disappointment, that's all he says about it—he raises the question, as if that is enough to indict evo-devo, and he certainly does not offer a research program of any kind.

The answer to the question, of course, is yes. Evo devo is not a collection of answers, but a set of approaches and principles that help us tackle difficult problems like the evolution of squid. It says that if we want to understand how organisms evolve, we need to understand the mechanisms by which genetic information is translated into form and function…the process of development. It's awfully hard to disagree with that, but Dembski tries. Or rather, he seems to think that bringing up unsolved problems in squid evolution is enough to show that evo devo has failed.

Look, if we want to understand how modern animals evolved from older forms, we should try to understand:

  • How genes produces morphology.
  • The genetics underlying large scale differences in form, between squid and fruit flies, for instance.
  • The mechanisms that generate striking morphological differences between closely related species, as we see in African cichlids.
  • The genetic basis of morphological variation within populations.

Evo devo does not declare that it has all the answers, it proudly announces that it has the right questions and a good toolkit to address them. Those are all questions that scientists are trying to answer, too; the Intelligent Design creationists don't even bother to ask interesting questions. Generalize his tantalizing opening question to "How can we explain squid evolution?". I can say how scientists would work to explain it. Dembski thinks just asking the question ends the search and does not bother to offer an Intelligent Design creationist's research strategy.

Where Dembski is at his most contemptible, though, is when he mangles the scientific literature to get an answer he wants. Here's an example of creationist quote mining:

But that raises a fundamental problem. Elizabeth Pennisi, in a report about evo devo for the journal Science, dated Nov. 1, 2002, stated the problem this way: "The lists [of conserved genes give] no insight into how, in the end, organisms with the same genes came to be so different."

Wow. That sure sounds like an admission of failure. Those scientists must be stymied.

It's a grossly misleading excerpt, however. This was taken from a report explaining how developmental biologists were excited about the promise of new strategies in evo devo, and listed several examples of successes. All you have to do is read on a few sentences further to see the way the work is going.

The lists gave no insight into how, in the end, organisms with the same genes came to be so different. And given the evolutionary distance between, say, a fruit fly and a shark, "there isn't really an experimental manipulation to let you get at what the genes are actually doing," says Rudolf Raff, an evolutionary developmental biologist at Indiana University, Bloomington (IUB).

The solution, say Jeffery and others, is to focus on genetically based developmental differences between closely related species, or even among individuals of the same species. This is the stuff of microevolutionists, who care most about how individuals vary naturally within a population and how environmental forces affect this variation.

Uh, what's that? They aren't arguing that there is no insight at all, but merely that by working on smaller differences in more closely related species, they have a better handle on how to attack the problem? Hmmm. Further, the next page of the article summarizes recent dramatic successes with this approach, such as in understanding the evolution of cavefish eyes, the nematode vulva, and butterfly eyespots, where examination of differences between closely related species has enabled us to track exactly how genetic changes have led to the differences in form.

Dembski doesn't understand how science in general works, and even after consulting his incompetent colleague, Jonathan Wells, he definitely doesn't understand developmental biology. This passage is painful in its ignorance.

To sum up, developmental geneticists have found that the genes that seem to be most important in development are remarkably similar in many different types of animals, from worms to fruit flies to mammals.

Initially, this was regarded as evidence for genetic programs controlling development. But biologists are now realizing that it actually constitutes a paradox: if genes control development, why do similar genes produce such different animals? Why does a caterpillar turn into a butterfly instead of a barracuda?

This phrase, "biologists are now realizing that it actually constitutes a paradox", is simply false. There is no paradox at all there, and it doesn't trouble us at all. If you observed surveyors at work, and noticed that they marked off two plots of land of identical size, surveyed with similar instruments, and staked out with the same tools, and then bulldozers and carpenters and bricklayers and plumbers and electricians show up at both, but then later discovered that a gas station and convenience store was built on one, while a three-bedroom ranch house was built on the other, would you announce that there was a paradox here? Of course not. You're not an idiot.

What developmental geneticists have discovered is that different organisms use remarkably similar toolkits to assemble their form, and that what matters is how those tools are deployed during development. It's the patterns of regulation and interaction between those genes, which do differ in interesting ways, is what generates differences between species. And that's why Sean Carroll can legitimately argue that evo devo is a worthwhile focus for research.


Ah, but Dembski isn’t interested in science. He’s only interested in planning his next Waterloo, at which he’s very good.

As I read Dembski’s article, all I could say was No, no, no your missing the point. No, no, no your wrong. No, no, no you misunderstood the article, so many “no’s”. It’s frustrating to watch someone twist and mangle an idea and try to use the literature to support their preconceived notion.

This thread on ARN discussed the ludicrousness of the Dembski/Wells position almost 4 years ago. In addition to the difference between scientists and IDeologues when it comes to the posing and pursuit of scientific questions, it’s interesting to see that, in contrast to the field of evo-devo (which has made large, bold strides in these 4 years), ID thought has changed not a single, solitary iota.

Note in the thread the inevitable resort to “you haven’t explained everything, thus evolution NO!” that pops up.

“Of course not. You’re not an idiot.”

Awesome. That’s a great line.

PZ wrote:

This phrase, “biologists are now realizing that it actually constitutes a paradox”, is simply false. There is no paradox at all there, and it doesn’t trouble us at all.

Try this. Google “hox paradox.”

Or look at Sean Carroll’s book Endless Forms, discussed above. Carroll writes (pp. 71-72):

The discovery that the same sets of genes control the formation and pattern of body regions and body parts with similar functions (but very different designs) in insects, vertebrates, and other animals has forced a complete rethinking of animal history, the origins of structures, and the nature of diversity. Comparative and evolutionary biologists had long assumed that different groups of animals, separated by vast amounts of evolutionary time, were constructed and had evolved by entirely different means. The connection between members of some groups — among the vertebrates, for example, or between vertebrates and other animals with a notochord – was well established. But between flies and humans, or flatworms and sea squirts…no way! So prevalent was this view of great evolutionary distance that in the 1960s the evolutionary biologist (and an architect of the Modern Synthesis) Ernst Mayr remarked: “Much that has been learned about gene physiology makes it evident that the search for homologous genes is quite futile except in very close relatives…” This view was entirely incorrect. The late Stephen Jay Gould, in his monumental work The Structure of Evolutionary Theory, saw the discovery of Hox clusters and common body-building genes as overturning a major view of the Modern Synthesis. Gould states, “The central significance of our dawning understanding of the genetics of development lies not in the simple discovery of something utterly unknown…but in the explicitly unexpected character of these findings, and in the revisions and extensions thus required of evolutionary theory.”

Data suggesting that one’s theory (the Modern Synthesis) needs a “complete rethinking” have the flavor of a genuine paradox. Which explains why the phrase “hox paradox” is widespread in the evo-devo literature.

Paul A. Nelson -

Your post does not address the substance of what PZ Meyers said. It merely quibbles, irrelevantly, as to whether anyone has ever used the term “paradox” to describe hox genes. By its nature, it suggests an incapacity on the part of its writer to address the substance of the post.

You do provide a source for your quote box (thank you). But within it, this is unsourced…

“Ernst Mayr remarked: “Much that has been learned about gene physiology makes it evident that the search for homologous genes is quite futile except in very close relatives…” “

If Ernst Mayr remarked this, I strongly suspect that the remark has been taken out of context. It is trivially wrong. Humans and bacteria have many homologous genes for basic metabolic functions, for example. Of course, if Ernst Meyer merely put his foot honestly in his mouth, once, it means little. But if your source is twisting Meyer’s words, that may mean a great deal.

The fact that homologous genes underly basic aspects of development supports common descent. The opposite would not (distorted quotes from Meyer notwithstanding).

Meanwhile, if there were NOT conserved genes underlying development, you would claim that this “supported intelligent design”. But if there ARE, you will merely claim that this, too “supports intelligent design” with the rubric “common design”. (Neither of these would be true, of course, it’s just that the latter supports common descent more strongly than the former.) Thus, intelligent design is worthless for explaining or understanding hox genes, and makes no testable claim with respect to hox genes.

Indeed, if we were to practice “intelligent design”, we wouldn’t even know that hox genes exist.


You are a good sort by all accounts. Do you agree that William A Dembski has been engaged in ‘quote mining’ and that his article is misleading as a consequence?

Admiral Paul Nelson stands on the deck of the sinking HMS ID, holds his teleoscope up to his blind eye and exclaims “I see no evidence!”

Meanwhile, below decks, Swabbie Dembski wearing an Abba costume mans the bilge pump while belting out “Waterloo.”

The existence of the so-called “hox paradox” – the deployment of homologous genes and their protein products in the development of classically non-homologous structures, such as vertebrate and arthropod eyes – has been one of the most widely-discussed topics in evo-devo over the past decade.

PZ knows this, as does anyone who follows or studies evo-devo.

The quote from Ernst Mayr is taken from Sean Carroll’s book, and (as Carroll argues) represents the neo-Darwinian, or Modern Synthesis, understanding of the genetic basis of homology.

Yes, Paul, and if you actually read the articles that you can find when searching for the “hox paradox” in the scientific literature, you’ll find that typically what is said is, “hey, there’s this idea called the ‘hox paradox’, but it actually isn’t paradoxical at all”, and you definitely won’t find them suggesting that this is a serious problem for evolution.

Wagner et al. answer it quite simply in this way in PNAS:

It is now widely accepted that the divergent body plans are based more, but not exclusively, on differences in the regulation of a conserved set of genes rather than different gene complements.

Victoria Prince out your way at the U Chicago has a very nice article on the subject, which uses the ‘paradox’ as a springboard to discuss differences in the Hox genes and their pattern of regulation as an evolutionary explanation.

I would hope you wouldn’t try to claim that Sean Carrol sees Hox genes as an obstacle to our understanding of evolution. You’ve read his book, and he’s asserting something entirely contrary to that.

Do you guys search the scientific literature for words like “paradox”, “problem”, and “controversy” just so you can pluck them out of context and invent imaginary hurdles for evolution to jump?

Mr. Nelson,

Googling “hox paradox” as you suggest, then reading the papers that address this question, confirms precisely what PZ said.

From “Resolving the Hox Paradox,” Science, Vol. 292, Issue 5525, 2256–2257, 22 June 2001:

The first approach to resolving the Hox Paradox was to deny that distantly related animals are really so different after all. The notion of cryptic anatomical similarity became a touchstone for biologists intent on uncovering the conserved genetic underpinnings of animal development. The logical conclusion of this approach was the conceptual reconstruction of Urbilateria, the latest common ancestor of all bilaterians (which is to say nearly all animals). Under the assumption that similar gene expression denotes conserved gene function, Urbilateria became a rather complex beast—with eyes, a heart, appendages, and a segmented body.

But doubts slowly crept in. By the mid-1990s, it was clear that virtually all developmental regulatory genes control several different processes, some of which plainly evolved within insects and mammals. If regulatory genes can acquire new developmental roles, then their domains of expression cannot be taken at face value as indicating anatomical conservation. On reflection, some of Urbilateria’s reconstructed features, such as segmentation, began to look less certain.

This realization has led to a second approach to resolving the Hox Paradox, based on the notion that although developmental regulatory genes are evolutionarily conserved, their interactions are not.

From “Hox cluster duplications and the opportunity for evolutionary novelties,” PNAS 100(25):14603–14606, December 9, 2003:

The first so-called Hox paradox was the discovery that homologous genes ”code” for fundamentally different body plans. It is now widely accepted that the divergent body plans are based more, but not exclusively, on differences in the regulation of a conserved set of genes rather than different gene complements (2–4). This commentary discusses a second Hox paradox: Why is it that in the evolution of chordates (vertebrates) the number of Hox gene clusters has increased several times (Fig. 1), often in association with major radiations (5, 6), whereas no evidence exists for such a trend in invertebrates (7)? It is hard to believe that this difference should be due to differences in the frequency of genome and chromosome duplications between vertebrates and invertebrates. In this commentary, we argue that vertebrate Hox clusters, in the absence of duplication, are structurally less evolvable than their invertebrate counterparts. The constraint on Hox cluster structure may be temporarily lifted after cluster duplication, which may make an association between Hox cluster duplications and adaptive radiations more likely in vertebrates than in invertebrates.

If you intended to imply that a search engine would reveal a different conclusion, please tell us what it is and provide a link.

This is a strikingly illogical statement for a proponent of ID to make…(attributed to Dembski in the above post)…

“But biologists are now realizing that it actually constitutes a paradox: if genes control development, why do similar genes produce such different animals? Why does a caterpillar turn into a butterfly instead of a barracuda?”

It is certainly interesting that highly conserved genes would be active in development, across different lineages. I don’t find it a ‘paradox’. A ‘paradox’ is something contradictory. It is certainly an exciting spur to future research, but again, that isn’t what ‘paradox’ means.

But it also strikes me as rather overwhelming evidence of common descent. Why would an ID advocate trumpet it as a problem for the theory of evolution?

I don’t mean to be sarcastic and don’t wish for this to come across that way but without boring everyone with a long comment, I can only respond

ID The paradigm always stays the same

PZ, would you agree (as Carroll argues) that the neo-Darwinian understanding of the relationship of genes to development and morphological form turned out to be wrong?

Paul Nelson -

You’ve established the point that the misnomer “hox paradox” has been widely used. It has also been demonstrated that users of the term often acknowledge that it is something of a misnomer.

Do you agree that conserved genes active in development, across lineages, supports common descent?

Do you have an alternate, testable, scientific explanation for this observation? What is it and how can we test it?

This is a strikingly illogical statement for a proponent of ID to make…

Harold, I’m shocked that you are able to detect a “strikingly” illogical statement from any other statement made by a proponent of ID.

Have you been tinkering with your Irony Meter?

OK, let me get this straight.

Dembski says that only God an Intelligent designer could produce different creatures from such similar genes, right?

Now, the genes are still similar, right?

So is Dembski arguing that every single reproductive event is the result of direct intervention from God the Designer?

That does seem to be where it is leading, logically.

Back 60 or so years ago, most geologists did not believe in plate techtonics because they saw no way that continents could actually move about. So there was the south america/africa similarity paradox: how could the almost exact geometric fit between the two continents be explained given they were always so far apart. The answer came when the theory of plate techtonics, which showed how continents could move about and further that continents did move about, was developed.

From Paul Nelson’s description of the Hox paradox, it seems that people assumed that growth regulatory genes developed independently and at different times in different evolutionary lines. This assumption seems to have arisen without much examination: since different evolutionary lines separated early, when else would those functions have appeared. (This in contrast to matabolism, which would have to develop early and is similar in most all cellular beings – I think Mayr, who, it seems, is being quoted by Sean Carroll, was likely talking about development genes.)

Now new infomation indicates that development genes also arose early and are widely shared. This doubtless produces a “paradox” (how can these genes work the same in wildly different animals?) but one that I suspect will lead to a resolution that will result in much better understanding on evolution.

Could Paul Nelson do something more quote practicioners of evolutionary science and claim the quotes are against the authors? Could he ask some meaningful and relevant questions that might just allow the rest of us to understand what he believes the relevant issues are?

PS: “paradox” means either puzzle or logical inconsistency. That Africa and SA fit together was a pardox to traditional (continents do not move) geology, which had to be abandoned. If evolutionary biology requires that development genes evolved late in each evolutionary line, then the hox genes are a paradox in that sense. If that is not a requirement, then they are just a puzzle. Dembski seems to be using the term in the logical inconsistent sense. Carroll in the puzzle sense.

Harold, it is true that one can’t just take a crationist’s word for a quote, but you still have to check.

Nelson’s quote, in the box, is correct. Carroll’s source for Mayr is: _Animal Species and Evolution_, Harvard press 1963, page 609. Scientists were just beginning to be able to sequence and compare proteins, much less segments of DNA.

Bill -

Well, of course, proponents of ID make many illogical statements. This one is STRIKINGLY illogical because, while employing the “outraged by a their own straw man” tone of creationists (“why doesn’t a caterpillar turn into a barracuda…”), it actually presents rather straightforward evidence for common descent, and mentions an interesting area of research. In fact, stripped of its tone of exaggeration, it is a logical question for SCIENTIST to ask.

Paul Nelson:

PZ, would you agree (as Carroll argues) that the neo-Darwinian understanding of the relationship of genes to development and morphological form turned out to be wrong?

In what way is this relevant? Many hypotheses have been found to be incorrect or inaccurate.

You still have failed to deal with PZ’s point: that Dembski deliberately mis-quoted, and that the so-called “hox paradox” no longer exists.

Dembski’s behavior is dishonest, yours is only disingenuous.

Attempts to resolve the hox paradox do not agree. One attempt holds that classically non-homologous structures (e.g., the limbs of arthropods and vertebrates, both of which express the gene distalless during development) really are homologous after all. Hence, the paradox is only apparent.

Eric Davidson disagrees:

In response to this paradox [the Hox paradox], an almost automatic response has been that though they may look different these body parts are actually homologous; that there are basic and still hidden pattern formation processes underlying their development, and that these were already present in the bilaterian common ancestor. Conservation must be the reason that the same genes are used in the development of each part in diverse bilaterians, so this argument goes.…But if we sum all the assertions of this sort we produce an impossible and illogical image of the bilaterian common ancestor. It would have been equipped with brain, seeing eyes, moving appendages, beating heart, etc. Something is very wrong with this picture because the way these body parts develop in diverse branches of the Bilateria actually share little in the details of their respective pattern formation processes. If the common ancestor had appendages, for example, it could not easily be ancestral to insect and mouse appendages both, because the structures and processes through which these develop are completely different.

E. Davidson, Genomic Regulatory Systems (Academic Press, 2001, p. 189).

Davidson favors a different solution to the paradox (which I don’t have time to summarize right now). In any case, the paradox is a genuine puzzle. Whether it threatens “evolution” depends on how that infinitely elastic word is defined.

PZ and I have talked some about this, one-on-one, and plainly we disagree. I’ll have to leave it there for now.

David Margolies -

I have never encountered the usage of “paradox” to mean “puzzle” that you describe. I have only heard it to mean “logical inconsistency”. However, I have no reason to doubt that you are correct. You also present a justification for considering the existence of hox genes to be paradoxical, relative to what might have been expected under some prior hypotheses.

And so, with regard the the term “hox paradox”, I think we can say…

1) The term has been used. Paul Nelson is correct in this regard. 2) In this context, the term is subjective. Some situations of pure logic may constitute a paradox in an objective sense, but… 3) To some of us, the existence of hox genes does not seem paradoxical. 4) The issue of whether anyone ever CALLED hox genes a ‘paradox’ is relatively trivial, and does not address the substance of the PZ Meyers post.

Paul Nelson -

If you’re still here…

Scientists believe the basic mechanism of evolution to be genetic variability acted on by natural selection. The RESULT of this may be described, tersely but quite accurately, as a “change in the frequency of alleles”, with the understanding that much else usually changes as well.

This conception is not rigid, but it isn’t especially elastic, either.

While the use of relatively conserved hox genes in development of different morphologic features is fascinating, and may disagree with earlier hypotheses of development, it is also strongly consistent with the theory of evolution from common descent, as described above.

If you disagree, can you explain why, very specifically?

Harold, I was using “puzzle” as an abbreviation for 2 a: in the Merriam-Webster Collegiate: “2 a: a statement that is seemingly contradictory or opposed to common sense and yet is perhaps true.” And I note the Wikipedia uses the word “puzzling” – this from the beginning of the entry in

“A paradox is an apparently true statement or group of statements that seems to lead to a contradiction or to a situation that defies intuition. Typically, either the statements in question do not really imply the contradiction, the puzzling result is not really a contradiction, or the premises themselves are not all really true (or, cannot all be true together).”

David Marolies -

Well, language is a slippery thing. It certainly seems to be a matter of subjective preference whether or one chooses to refer to hox genes as a “paradox”.

That last definition is especially mind-bending - it seems to suggest that a paradox isn’t really contradictory, but merely seems to be so. But then what do we call a real contradiction?

What’s interesting is that science leads to paradoxes all the time, relative to how the human mind processes information. But not at the scale of biology.

It would be quite reasonable to describe much of advanced physics as paradoxical, even though it’s empirically verifiable. Of course, some may choose to subjectively deny that any of it is a paradox.

I hope Paul Nelson answers my question.

Paul Nelson Wrote:

PZ, would you agree (as Carroll argues) that the neo-Darwinian understanding of the relationship of genes to development and morphological form turned out to be wrong?

Note that “neo-Darwinism” has no fixed meaning. Here Paul evidently means the Modern Synthesis. The Synthesis is moderately well defined as the understanding of evolution at a certain time, but it had essentially had no idea how development worked at the genetic level. The term “neo-Darwinism” predates the synthesis by quite a bit; on the other hand it can mean anything up to and including our current understanding of evolution.

Now we know that genes in general are expressed or not based on complex regulatory sequences which are usually next to the gene. This is notably true for genes active in building bodies. Dembski pulls a quote to the effect that a mere list of genes doesn’t tell you how they are employed. Of course not. That’s the point. You have to know at least how they are regulated.

The Ernst Mayr quote that Mr. Nelson takes from Sean Carroll’s book is also used by Gilbert’s Developmental Biology, 6th Ed. It seems to originate in the book Animal Species and Evolution, E. Mayer 1966.[…]section.5475

Further searching yields a 1963 publication of the book.[…]amp;n=507846

My first post :-)

Paul Nelson –

As long as you’re here, I would like to know: do you unequivocally repudiate and disown the HIV/AIDS denial directed at children by the magazine on whose advisory board you sit?

From a moral standpoint, I consider this much more important than a blog spat over who is quotemining whom.

The Discovery Institute has demonstrated itself to be seriously confused about the twin paradox and its resolution—why not the hox paradox and its resolution too?

Paul Nelson says that when someone serves you shoe leather, claiming it’s steak, the important thing is that it’s shoe leather, not what the chef calls it.

Funny, in my world most people would regard the word as being inextricably linked with the deed in this case, and would regard the chef as a liar or outright fraud because of it.

But then, you are talking about people who serve us religion, claiming it’s science, and say that the important thing is that it’s what the chef calls it, not what it is.

Nelson says there’s no scientific theory of ID. Given that, I am still wondering, then, just what the heck ID *is*, if it isn’t science. Is it steak, or is it really just shoe leather.

Alas, Nelson seems to have run away, again, without telling me.

Alas, Nelson seems to have run away, again, without telling me.

But he said he would email you the answers. Surely it can’t be that Paul Nelson is a liar.

But he said he would email you the answers. Surely it can’t be that Paul Nelson is a liar.

Alas, it seems that he is.

And I hope I get the opportunity to ask him about it later. Publicly.

About this Entry

This page contains a single entry by PZ Myers published on July 28, 2005 2:13 PM.

Why isn’t the germ theory a “religion”? was the previous entry in this blog.

Kennewick Man hearings is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.



Author Archives

Powered by Movable Type 4.381

Site Meter