The Loom

Things have not been going so well on the political front for the advocates of intelligent design (a k a the progeny of creationism). This election season their allies on state boards of education in Kansas and Ohio went down to defeat. On the scientific front, things have never really gone well. The Discovery Institute in Seattle claims that it has spent millions on research. They have precious little to show for it. As I wrote last year, a single evolutionary biologist produces more papers in peer-reviewed biology journals than the entire staff of the Discovery Institute. You'd be hard-pressed to find a single paper that actually claims that intelligent design is supported by original evidence. The closest they got to such a minimal standard--a review of the Cambrian explosion--was later retracted by the journal. The Discovery Institute claims that it's got all sorts of stuff in the works, but they aren't ready to share it with the world. Instead, they'd prefer to attack journalists.

In September, Casey Luskin of the Discovery Institute posted a 31-page attack on fellow scienceblogger Chris Mooney. Mooney is the author of the excellent Republican War on Science, which details some of the strategies the Discovery Institute uses to promote Intelligent Design, and the resounding rejection of intelligent design by the courts. As I wrote at the time, Luskin's charges were empty.

Now I'm getting the Mooney treatment.

The occasion is the publication of my article in the current issue National Geographic about the evolution of complex features. In the article I survey several examples of complex features such as flowers and eyes, and I describe the latest research into their evolution.

Yesterday, Casey Luskin took to the Discovery Institute's "Evolution News and Views" site to post part one of an attack on the article. How many parts do have to look forward to? Luskin doesn't offer a clue. But in part one, Luskin has already offered a wealth of misrepresentations and faulty reasoning.

The headline of Luskin's piece is "National Geographic Evolution Article Discusses Evidence that Supports Intelligent Design." Luskin claims that I am selectively and misleadingly presenting information to make the case for evolution. If I presented the whole picture, the case for Intelligent Design (the idea that life is so complex it must have been created by an Intelligent Designer) would become clear.

To explain why this is wrong, I need to take a minute to explain how most science writing is typically done. You become familiar with a subject by reading recent papers in the leading peer-reviewed journals. You call the authors of those papers. You talk to other experts who have a similar track record. In some cases they agree on the big picture. In other cases, they differ sharply. You then try to craft all this information into an article that can introduce the subject to readers who are not familiar with it. Do you quote all the papers you've read, verbatim? Of course not. You choose the best examples of the sort of work that has led scientists to their general conclusions. But to make sure that you've chosen the right examples and described them accurately, fact-checkers take a close look at your work to find errors.

In other words, the simple absence of some piece of information from an article does not mean the reporter is misrepresenting the subject. Luskin would like us to believe otherwise. What's more, he claims that the "extra" information missing from my article brings evolution into question. It does not.

Case 1: Embryos. On page 115 of the article, there's an illustration of three embryos: a fish, a chicken, and a human. The caption explains that the early embryos of these animals look much the same, but the genes in corresponding parts of the embryo guide development by different paths. It adds that evolution often reshaped organisms by altering the genes that control development.

Luskin responds by showing a picture of embryo development from egg to adult. In the gastrulation stage, vertebrate embryos look more different from one another (some are ball-shaped, for example, and some are flattened). Eventually vertebrate embryos pass through a "phylotypic" stage in which they look a lot like other vertebrate embryos, before taking on their distinctive identities.

"These facts don't so fit neatly with Zimmer's claim that '[e]volution often reshapes organisms by tinkering with the genes that control development,'" Luskin declares, "because the hourglass pattern of development shows that transitions from fishlike development ultimately into other forms of development would require radical restructuring (not 'tinkering') from the earliest stages of development."

This is wrong in many ways. Let me break it down into three parts: the earliest differences in embryos, the phylotype, and the adult form of vertebrates.

As fellow scienceblogger PZ Myers has clearly explained, the differences in the earliest stages are superficial. In many cases, the difference is simply whether an embryos develops as a mass of cells, or a mass of cells sitting atop a yolk. Whatever the differences in how the earliest embryos look, they undergo the same core steps of development, known as gastrulation. And the same genes control that process. In other words, what we see in the earliest stages of vertebrate embryos today are variations on an ancestral theme. Sounds like the standard evolutionary process to me.

As for the phylotypic stage, scientists are still debating why it remains so similar among vertebrates. There may be constraints that prevent natural selection from favoring mutations that push an embryo away from this shape. And once embryos move past the phylotypic stage, they begin to develop distinctive organs. Luskin completely ignores the particular organs that the National Geographic illustration actually illustrates: the limb bud.

In fish, chickens, and humans, tiny swellings of cells along the side of embryos emerge, and the cells express almost identical networks of genes. But over the next few days or weeks, the limb bud develops into very different shapes--a fin, a wing, or a hand. Scientists have charted the genes that switch on in cells in the limb buds as they take different forms. (None of these scientists, it should be pointed out, work at the Discovery Institute.) It turns out that many of the same genes are at work in fins, hands, and wings. The differences emerge thanks to the differences in when and where in the limb bud the genes produce their proteins. This is precisely the sort of evolution scientists are talking about when they refer to tinkering with genes that control development.

Case 2: The bacterial flagellum. Many species of bacteria such as E. coli use spinning tails to swim. These flagella are made up of dozens of proteins that work together. Some behave like a motor, some like a flexible hook, and others like the coiling tail. Still other proteins help build the flagellum, such as a needle that injects proteins into the growing shaft.

I am not surprised that the Discovery Institute would try to attack this example. In my work on my new book on E. coli, I've been exploring the long-running fascination creationists have had with the bacterial flagellum. In 1994 the Creation Research Quarterly dedicated an article to its wonders. "To evolutionists," the article claimed, "the system presents an enigma; to creationists, it offers clear and compelling evidence of purposeful intelligent design." In 2005, the Dover intelligent design case came to be known as the bacterial flagellum trial. But both at the trial, and in peer-reviewed journals, scientists explained why the notion that the flagellum was intelligently designed has no traction in the scientific community.

In Luskin's attack, he misrepresents both my article and the science that it describes. He claims that the only evidence I provide is the needle. Some strains of E. coli and other species of bacteria use a practically identical form of this needle to inject toxins into other cells. Luskin then quotes William Dembski, also of the Discovery Institute, who claims that this needle is just one bit of evidence for the evolution of the flagellum. "What's needed is a complete evolutionary path and not merely a possible oasis along the way," Dembski informs us.

Luskin apparently does not understand the meaning of the word phrase "for example." I chose to describe one structure from of many in the flagellum that can be found in microbes serving other functions. I based this part of the article on a number of papers that identify evolutionary links between proteins in the flagella and in other structures. But it would have been absurd for me to catalog them all. Rather than actually address all of that evidence in scientific papers, Luskin prefers to attack an article in a general-interest magazine for not reading like a scientific paper.

Luskin also would have us believe that the only way to find support for evolution is to provide a mutation-by-mutation account of evolutionary change spanning billions of years. The way science is actually done is quite different. Scientists do not go into a time machine and recreate history step by step. They assemble evidence and judge whether they support a hypothesis or not. If flagella evolved from earlier structures, the discovery of related forms of those structures would come as no surprise. I also explain in the article how flagella themselves provide evidence for ongoing evolution--including even cases in which flagella have been lost, with only the disabled genes for them surviving as molecular vestiges. These are other lines of evidence that support the hypothesis that flagella evolved.

All of these evolutionary links have allowed researchers to begin to put together detailed hypotheses about how bacterial flagella evolved. I describe one of these hypotheses, which was published by Mark J. Pallen and Nicholas J. Matzke in "From The Origin of Species to the origin of bacterial flagella," in October 2006 in Nature Reviews Microbiology. It begins with a simple structure for injecting molecules, and then becomes increasingly elaborate. All of the structures that get combined in the process can be found in living microbes. In some cases, combinations of these structures have also been discovered. And scientists have barely begun to explore the microbial realm, so scientists can test Pallen and Matske's hypothesis in years to come. If you rely solely on Luskin, however, you'd think I sat down and made up a story.

Luskin claims at the beginning of his post that I'm writing about evidence that supports intelligent design, but nowhere in the post do we find out why. Luskin instead tries to poke holes in evolutionary biology, and fails. It's remarkable that he calls for an absurdly detailed reconstruction of history as evidence for evolution, while expecting nothing of the sort from advocates of intelligent design. Apparently the rules are different at the Discovery Institute. There you need only make vague references to a designer, claim some supposed shortfalls of evolution, and you're done. "With its irreducibly complex nature and machinelike properties," Luskin concludes, "perhaps the simplest explanation for the origin of the flagellum is intelligent design." The term "irreducible complexity" is meaningless when it comes to flagella, and relying on a "machinelike" appearance of something is hardly a compelling argument that something was designed rather than evolved. I am reminded of the wise advice of Albert Einstein: Make everything as simple as possible, but not too simple.

(Note: If Luskin gets around to Part Two of his attack, I'll update this post as needed. I suspect, however, that it will be more of the same.)

Update, 11/16: PART DEUX

In his second post, Luskin announces that my article "ironically discussed much evidence which ID-proponents often contend supports intelligent design."

Where exactly is that irony? Imagine my article had been about the formation of the Grand Canyon--about the strata that formed its walls over millions of years, about the discontinuities where the layers of rock were heaved and eroded, only to be covered up again by more layers of rock and then finally cut through by the Colorado River. If Casey Luskin was a Young-Earth creationist, he might have written that the article "ironically" discussed much of the evidence which Young-Earth creationists often contend supports the idea that the Grand Canyon was formed by a worldwide flood a few thousand years ago. Just because they talk about the same evidence doesn't mean that their conclusions are valid. The same goes for intelligent design.

This irrelevance doesn't stop Luskin from going ahead and offering up bits and pieces from the article and claiming them as evidence of design. He rehashes various examples I offered of different groups of species using ancestral sets of genes to carry out different functions. For example, the axis of a developing fly's body is controlled by genes from the same family of genes that control our own.

As I explain in the article, scientists have done a lot of research into how such genes evolved over time. An ancestral set of body-axis genes (called Hox genes), for example, was passed down to different lineages of animals. In each lineage, new copies were made, some old copies were lost, and the genes gradually changed their genetic sequence thanks to mutations that were favored by natural selection. Plenty still needs to be worked out about this evolution, of course, but all the work so far supports the theory that these genes diverged by the same sorts of natural processes we can see today.

For example, natural selection can leave its fingerprints on genes. Some mutations alter proteins, while others have no effect, and when genes take on new adaptations, the protein-altering mutations are unusually common. So if the evolution of body-axis genes was a process of adaptation by natural selection, you'd expect to see the fingerprints on these genes. Some scientists at Yale decided to investigate, and earlier this month, they reported that they'd found the fingerprints. Remarkably, they could even pinpoint the parts of the genes that had undergone strong natural selection: the parts that encode sections of proteins that grab other genes to switch them on. That's exactly where you'd expect to find the fingerprints.

Luskin does not bother to address this scientific research, or any other for that matter. Instead, he suggests that these genes are signs of "common design." Apparently, "designers often re-use parts that work in different designs."

So, on the one hand, we have scientists analyzing 155 different genes using the latest statistical techniques for detecting natural selection, which they followed up with studies on the structure of the evolved proteins themselves. And then on the other hand, we have Luskin saying these genes...well, they just look like they're re-used by a designer.

What is truly ironic is the way intelligent design advocates now try to use shared sets of genes as evidence for "common design." After all, it was not all that long ago that they claimed that a biological structure such as the bacterial flagellum were intelligently designed, and for their evidence they claimed that the flagellum was encoded by a unique set of genes. The only problem was that scientists are getting better and better at finding the relatives of "unique" genes, often doing other functions. Now Luskin and company offer a designer which does anything it (He?) has to do to mimic whatever changes evolutionary biologists document.

In the end, Luskin searches for support for intelligent design in--I kid you not--the article's analogies. In a couple places in my article, I quote scientists likening some evolutionary processes like variations on a theme, or changing elements while remodeling a house. Luskin claims that in both cases, the analogies hint at "intelligent agents."

Before you shout, "Whoa! Dude!", consider another analogy in the article. Sean Carroll likens the development of an embryo to the construction of a building:

"If you walked past a construction site at 6 p.m. every day, you'd say, Wow, it's a miracle--the building is building itself. But if you sat there all day and saw the workers and the tools, you'd understand how it was put together. We can now see the workers and the machinery."

By Luskin's logic, an embryo must be filled with real construction workers--or at least "intelligent agents." But lo and behold, the construction workers are but mere molecules, strings of carbon, oxygen, hydrogen, and other elements with no intelligence in our sense of the word. And yet these genes and proteins are quite capable of building an embryo.

I use analogies a lot in my articles, because they help make alien realms of nature more familiar to us. I assume that readers will recognize that embryos are not filled with guys with hardhats. But now there's at least one reader about whom I have some doubts.

We're almost done. Luskin promises one more part. Any guesses where he will go next? Word counts?

Update, 11/18: Part Three:

"Was the Ford Pinto, with all its imperfections revealed in crash tests, not designed?"

This is not from a satire of intelligent design in the Onion. These words actually appear in the final part of the Discovery Institute's lengthy response to my article. I am tempted to just let those words speak for their own absurd selves and end this post right here. But it's worth taking this trip to the very end. Having confused a magazine article for a scientific paper, having searched for evidence of design in figures of speech, Casey Luskin now sets the rules for what anyone can and cannot discuss when it comes to intelligent design. The only problem is, he can't seem to abide by them himself.

For his final target, Luskin chooses my description of the evolution of the eye. In recent years, scientists have unveiled hidden connections between vertebrate and invertebrate eyes. Those clues are helping scientists understand how light-sensitive cells in early animals evolved into the variety of eyes seen today. Luskin skirts quickly past this research, which he calls "reminiscent of common design."

Reminiscent is a strangely unenthusiastic word for an intelligent design advocate to use. You can't tell if he really thinks these ancestral genes are evidence of design, or if they just vaguely remind him of design. In any case, claiming these conserved genes are "common design" is a gross simplification of the full evolutionary process scientists are documenting. Evolutionary biologists do not just identify ancestral genes involved in building structures such as eyes. They also identify how those genes duplicated and diverged in different lineages of animals, how other genes were co-opted later to build new kinds of eyes. Our eyes and fly eyes are both filled with a light-refracting substance called crystallin. But the genes for each kind of crystallin are different, and can be traced to different ancestral genes that played other functions in the body. Luskin does not offer any alternative explanation for this process, nor does he point us to a peer-reviewed scientific paper. He just reminisces.

Luskin perks up, however, at the suggestion that in some respects the vertebrate eye doesn't seem very well designed. As I point out, the retina is delicately attached to the back of the eye and can become detached; its photoreceptors point backwards; and its optic nerve has to create a blind spot in order to exit the eye.

These observations take up the bulk of Luskin's response, which climaxes with him asking--in italics--"Why has National Geographic become a mouthpiece for a view of theology that states that a designer must design things to withstand certain types of physical attacks?"

Theology? How'd we get here?

Well, apparently intelligent design only requires "the detection of specified complexity." If something shows specified complexity, then it must be intelligently designed. "Specified complexity" is a term used by intelligent design advocates for patterns that are so improbable that they could only have been created by design. Supposedly, one can use a mathematical procedure to detect specified complexity. Luskin doesn't actually mention whether anyone has detected specified complexity in the eye. I'm not familiar with anything alone those lines, and a search of databases of scientific papers yields nothing.

And I mean nothing. I searched PubMed, a database of over 33,000 scientific journals maintained by the National Institutes of Health, for the phrase "specified complexity. " I got this response: "Quoted Phrase Not Found." That's because specified complexity has been roundly rejected by the scientific community.

To dare utter a single word about the design beyond specified complexity, Luskin warns, is to speculate about the "moral purpose of the designer." In other words, it's theology. Theology, Webster's informs us, means "the study of God and of God's relation to the world." I made no mention of God in the article. It is, after all, an article about science, ie, the study of the natural world limited to natural explanations. I simply pointed out--as others have before me--that the flaws in the vertebrate eye make it hard hard to envision it as the work of a designer, who--by definition--could design something before building it.

Yet Luskin declares that I have crossed the line by discussing intelligent design beyond specified complexity. There's just one wee problem here. Let me indulge in my own italics: When Luskin himself writes about intelligent design, he also goes beyond specified complexity. Luskin bolsters his case for intelligent design by telling us that designers re-use designs. He tells us that designers make things that fail catastrophically, like the Ford Pinto.

Luskin rattles the rooftops when others break his rule, but breaks it himself. He is not content to simply tell us that the eye has specified complexity. Perhaps that's because he can't. His fellow Discovery Institute colleagues break the rule as well, by discussing the purpose of intelligent design--Michael Behe testified at the Dover trial about the evidence of intelligent design in "the purposeful arrangement of parts." The Discovery Institute sets rules for others that it doesn't have to follow.

Yet after accusing me of theology, Luskin is still not content to leave the eye's flaws alone. After telling us that design flaws don't matter, he immediately turns around and claims that the vertebrate eye is not "inefficient." He quotes an article in a creationist magazine that declares the retina must be upside down in order make contact with a layer of tissue called the retinal pigment epithelium, on which the retina's survival depends.

This is an old creationist line, and recently Ian Musgrave demolished it anew. In brief, the retinal pigment epithelium and other features of the eye compensate for its structural problems. For example, thanks to the vertebrate eye's backwards arrangement, the back of the eye overheats and needs blood vessels to carry away excess heat. Octopus and other cephalopods have their retinas arranged forward, which allows them to do without such extra features. They need less blood flow to cool and feed their eyes, which are very powerful (and blind-spot free).

And that's it. That's the sum total of the Discovery Institute's week-long assault. They claim that the illustrations for my article are misleading, when they're not. They claim that I offer almost no evidence for the evolution of various traits, but ignore much of the evidence that I do present (along with the evidence in scientific papers). They claim that I am pushing theology by discussing the notion of design beyond the discredited concept of specified complexity, when they plow beyond it themselves.

I've spent a lot of time on this rebuttal, in part because I don't want accusations against me to go unchallenged, and in part because it's been interesting to explore the sort of arguments used by intelligent design advocates. We encounter lots of stray pieces of information about various species arrayed to give the impression that evolution has no evidence and that intelligent design has a lot. But nowhere in these attacks do we find any actual interest in the organisms themselves. Why do you and an octopus have eyes that are so different and yet so similar? It's a fascinating question, but one that never appears in Luskin's posts. Design is always the answer, like a brick wall to curious minds.