02 December 2007

Do ID theorists generate data?

When I was in grad school, some of us figured out a basic rule of thumb regarding the titles of scientific research papers: if the title is in the form of a question, the answer is almost always "no." Here, I'll review a journal article with a question for a title, and its answer appears to be the same as the answer to the question that forms the title of this blog entry.

Jonathan Wells, a former developmental biologist and current senior fellow at the Discovery Institute, set ID hearts aflutter in 2005 with his publication of a peer-reviewed article in the scientific literature. And it's not just any old paper; in it, Wells "uses the theory of intelligent design (ID) to formulate a testable hypothesis," enthuses the Discovery Institute press release. The article is featured prominently in the Discovery Institute's exhibit wherein they refute claims that ID theorists don't do science. (Their site actually convinces me of the opposite.)

The paper is titled "Do Centrioles Generate a Polar Ejection Force?" Now, there are good reasons to ignore this little article. First, the journal, Rivista di Biologia, is utterly insignificant, and is prone to publishing articles that are clearly on the edge of scientific respectability. Its editor is (reportedly) a creationist and is affiliated with the Discovery Institute. Second, the paper is not a primary research report. It outlines a hypothesis, accompanied by a literature review, but describes no new experiments and reports no new findings.

Indeed, when the article's appearance was first announced in 2005, many ID critics made note of both of those facts. But neither is relevant to a scientific review of the work.

Rivista is, in fact, a joke, and it is indeed notable that Wells sent his paper there. But that's not a reason to conclude anything important regarding the article's contents. After all, every journal is edited by a group of humans who are biased in various ways and fallible in every way. And every scientific author considers those facts when deciding where to submit a manuscript. Rivista was a friendly venue for Wells. Science seems to be a friendly venue for Michael Shermer. So?

And what of the fact that the article contains no data? Well, there are whole journals devoted solely to the proposing of hypotheses. Major journals like Nature will often run articles which do nothing other than outline an hypothesis. In the case of Nature, such articles are clearly indicated as hypothesis-only features. Wells' paper isn't marked 'hypothesis', but at least as it appears in Rivista, it is not presented as anything more than the description of a new model for centriole function. There's nothing unscientific or inappropriate about such a scientific paper, and I see nothing inappropriate about its inclusion in Rivista.

The issue, then, is whether the paper is any good. Are the ideas coherently based on what is already known? Is the hypothesis testable? Is it plausible? Does it have any explanatory value?

And overall, I think there are three questions worth considering:

  1. Is the article a valid contribution to the scientific literature, based on the criteria just mentioned?
  2. How has the hypothesis fared in the two years since it was published?
  3. What does the article reveal about the utility of "intelligent design theory" with regard to the development of hypotheses in cell biology?
In considering question 1, I have prepared a review, proceeding as though I am reviewing the manuscript for publication in a reputable journal. My review considers only what was known in 2005, and addresses some of the technical details of the paper. I've included my "summary & recommendation" section below. My thoughts on the other two questions are below, where I discuss why I think Wells' hypothesis should be rejected, and offer some comments on the relationship between the article and "intelligent design theory."

1. Review of manuscript #QOD-01, "Do Centrioles Generate a Polar Ejection Force?" by J. Wells.

Summary & recommendation. In this manuscript, the author reviews current knowledge regarding the structure and function of centrioles in animal cells, then proposes a novel role for centrioles in the generation of the polar ejection force. The author's proposal is plausible and testable, but suffers from two main weaknesses: 1) the hypothesis is somewhat extravagant, while seeking to fill a knowledge gap that is not similarly dramatic; and 2) the hypothesis is highly speculative and poorly supported by current data. Moreover, the author misrepresents current uncertainty regarding the cause of the polar ejection force, and fails to convincingly establish a strong correlation between presence of polar ejection forces and presence of centrioles. Given the author's stated antipathy toward "neo-Darwinism" it is perhaps not surprising that he omitted the more detailed taxonomic survey that is needed to strengthen the claim of correlation. And the author glosses over the known roles of centrioles in the formation of cilia, structures known to be absent in land plants. Overall, the manuscript is generally clear and well-written, but is weakened by ad hoc swipes at evolutionary theory and some concluding speculation (on cancer biology) that is unacceptably vague and simplistic. Attention to these matters should improve the manuscript such that it will be suitable for publication in Rivista di Biologia, and I recommend that it be published after revision.

2. What is the status of Wells' hypothesis now?

First here's an overview of the hypothesis and the phenomenon it seeks to explain.

Centrioles are tiny structures found in animal cells. Most are cylindrical in shape, composed of longitudinally-oriented "blades" of filaments called microtubules. (My own research is focused on control of the cytoskeleton, and microtubules figure prominently in many of our ideas and experiments.) Molecular Biology of the Cell is a good place to learn more, as are Ian Musgrave's articles on The Panda's Thumb, and here's a simple diagram from a small review in Current Biology a few years ago:

The picture above shows one centriole, but they always come in pairs, tethered together and oriented at right angles to one another, like a pair of numchuks spread out. Centrioles have always fascinated cell biologists, because they have some bizarre properties. First, their shape. Second, their highly regular behavior: each cell has exactly two centrioles, which duplicate during cell division to form the odd L-shaped pairs. And during cell division, the centrioles sit inside a structure called the centrosome, which is centrally involved in the movement of genetic material during the all-important process of mitosis.

At least two major questions concerning centrioles have bugged cell biologists for decades. First, what do they do? And second, how do they duplicate? Wells focused on the first of those questions, seeking to draw a connection between the existence of centrioles and the occurrence of a peculiar phenomenon during cell division. This phenomenon, the "polar ejection force," resembles a kind of wind blowing from the ends of a cell (the poles) during mitosis. If you remember your college biology, you should be able to conjure an image of chromosomes moving away from the middle of the cell (during anaphase of mitosis), toward the two opposite poles. (This is the means by which each daughter cell gets a full complement of the genetic material that had been copied a few hours earlier.) Every picture of that stage of cell division looks like this:

In this picture, the chromosomes are those boomerang-shaped things, and they're being dragged to the right and left by the yellow threads, which represent microtubules. The chromosomes aren't usually shaped like boomerangs; they look like that here because they're being dragged by the red knobs, and the floppy arms are lagging behind. Why do the arms lag? It looks like there's a force pushing on them, driving them back from their destination -- it looks like there's a wind blowing. That's the polar ejection force.

Cell biologists have long postulated that the polar ejection force is due to the influence of motors attached to the chromosomes themselves. These motors, called chromokinesins, are known to control various other aspects of chromosome movement, and even in 2005 when Wells published his paper, the polar ejection force was known to be at least partly due to chromokinesins. Wells noted correctly that the story was incomplete in 2005: candidate motors had been identified, but questions remained about their ability to provide the needed force.

The gap in our knowledge, however, was unremarkable. Motors that were capable of moving chromosomes, and that were present in the right places at the right times, had been described. Wells' first big error, in my view, came when he looked at the gap and concluded that "something more must be involved." Whoa. Might something more be involved? Sure, but that's true of nearly every aspect of cell biology. Must something more be involved? Even in 2005, I would have said no, and if I had reviewed Wells' manuscript, I would have urged him to tone it down. His proposal was extravagant in any case, but pitched as a needed plug for a significant hole in biological understanding, it was flat goofy.

What was Wells' idea? Well, he thinks that centrioles look like turbines. So he thinks maybe they are turbines. And so maybe they rotate around each other, just at the right moment during cell division, and create an oscillation in all those microtubules, thereby generating a vortex that drives the polar ejection force. And maybe that rotation is regulated by calcium. That's the Rivista paper in a nutshell.

The ideas are testable, and plausible to the extent that they don't invoke functions or phenomena from way far out. But they aren't based on any data. There are no observations that even suggest that centrioles rotate (and Wells postulates rotation rates of up to 10,000/second), nor have microtubules been seen to exhibit the vortex-like oscillations that Wells' hypothesis predicts. And most importantly, the polar ejection force is not known to affect anything other than chromosomes. But Wells' hypothesis predicts a "wind" blowing into the center of the cell, a wind that would exert force on every particle between the two poles. In other words, even when the paper was published, its ideas came out of left field.

I doubt that the paper has ever been cited by another cell biologist. No one has published any observations to suggest that centrioles rotate or that a vortex is induced in a dividing animal cell. A proof-of-principle experiment would have been technically challenging but perfectly feasible, if a little expensive. (Here's one idea: use video microscopy to look for movement of inert particles, perhaps fluorescent beads, inside the postulated vortex of the mitotic spindle.) Whether Wells ever tested his hypothesis (or whether he meant to), I don't know.

But now, in 2007, there's no need. Two different lines of evidence make Wells' hypothesis unworthy of further consideration.
Either one of those results would have killed the idea.

3. What does this episode tell us about "intelligent design theory?"

Wells' paper described an unsupported hypothesis that was proposed to fill a minuscule gap in cell biological knowledge, and the hypothesis has since been shown to be completely without merit. For me, it's a struggle to find words that adequately communicate the failure of this article to achieve any scientific respect. The Discovery Institute calls it a "peer-reviewed scientific publication...supporting the theory of intelligent design." Maybe that says it all.

But here are some comments on how Wells went about presenting his idea. First, to reiterate: he overstated the problem he was trying to solve. It seems to me that his enthusiasm for the centriole-as-turbine idea caused him to seek a place for the idea no matter what. In other words, I surmise that he floated the centriole-as-turbine concept because he wanted centrioles to be turbines, and not because he sought an explanation for the polar ejection force. This is a dramatic failure for one of the leading (former) scientists of a movement that seeks to remake the very basis of scientific explanation. It seems to me that the last thing the ID people should seek is a wild-eyed, unsupported speculation, based on no observational evidence and shown to be completely mistaken within months of publication, as an exemplar of their approach to scientific explanation. Worse, this is Wells' area of expertise, and I suspect that this paper was his last and best contribution.

Second, Wells takes some idiotic swipes at evolutionary science in his paper, and I see significant irony in some of them. In the abstract, he disclaims "molecular reductionism and neo-Darwinism." What is his paper about? A laughably incorrect speculation about tiny turbines, in which their properties are wholly accounted for by...molecular reductionism!

Finally, what about this paper could possibly be distinctive as the work of an ID theorist? Only this: that he considered the possibility that tubular structures could be turbines. Now, is that really an ID distinctive? Has Wells noticed how cell biologists refer to motors, cargo, tracks, export, import, zip codes, sorting? Cell biologists regularly identify patterns in biology that make sense through reference to structure-function relationships, and cell biologists are happy to refer to those functions using the vocabulary of design, both literally and metaphorically. If that's all that ID thinkers have to offer, then maybe they should just get back in the lab with the rest of us and try to figure out how all this stuff works.

blog comments powered by Disqus