Worst paper of the year?

September 4, 2009 • 6:00 am

A new paper in the Proceedings of the National Academy of Sciencee makes the bizarre and completely unsupported claim that the two stages of the butterfly life cycle: caterpillar and volant adult, result from a hybridization event, with the caterpillar resulting from a butterfly mistakingly mating with an onycophoran (velvet worm). Instead of a single lineage evolving two different life stages, then, each stage reflects a completely independent evolutionary event.,

An article on this paper in Scientific American first explicates the theory:

In the current issue of Proceedings of the National Academy of Sciences (PNAS), Donald Williamson, a wheelchair-bound 87-year-old zoologist from the University of Liverpool in England, suggests that the ancestors of modern butterflies mistakenly fertilized their eggs with sperm from velvet worms, also known as onychophorans. “People have been trying to find one solution that covers all of metamorphosis,” Williamson says. “I say it’s a change in taxon during development.”

Velvet worms, which fall between worms and insects on the tree of life, have soft-bodies and superficially resemble caterpillars, particularly the larvae of an early butterfly relative known as Micropterix. Velvet worms have evolved a variety of elaborate fertilization procedures. Males are known to place sperm packets not on the female’s genital opening, but rather on skin tissue, which the sperm penetrates before migrating to the ovaries.

Williamson believes that an ancient insect accidentally picked up that sperm, and butterflies now contain two developmental programs so they live half their life as velvet worms and half their life as winged butterflies.

“Animals have been able to hybridize since they invented sex,” Williamson says. “With external fertilization, there’s always the possibility that some sperm will fertilize the wrong egg.”

and then gathers the reaction of other scientists, which is uniformly negative:

However, scientists asked to comment on Williamson’s theory were taken back by it and surprised it made it into such a prestigious journal. For example, from insect paleontologist Conrad Labandeira of the Smithsonian National Museum of Natural History in Washington, D.C.: “You’re kidding!”

After looking over the paper, Labandeira pointed out more substantial criticisms. Hybridization between closely related species sometimes occurs in the animal kingdom, but it is highly unlikely that the sperm of a velvet worm could fertilize a distantly related insect egg and produce a viable embryo. He also raises the question of where the genetic program controlling metamorphosis would come from.

“If I was reviewing [this paper] I would probably opt to reject it,” he says, “but I’m not saying it’s a bad thing that this is published. What it may do is broaden the discussion on how metamorphosis works and…[on]…the origin of these very radical life cycles.”

Insect developmental biologist Fred Nijhout of Duke University took a less diplomatic view of the article saying it would be better suited for the “National Enquirer than the National Academy.

Besides being inherently implausible, the theory can be easily refuted by DNA analysis, which should show, in butterflies, two completely disparate genomes, one resembling that of onycophorans.

Now who communicated this bizarre, unsupportable paper to one of the world’s most prestigious journals? Think about it: who has made their careeer asserting that all of evolution results from lateral gene transfer and hybridization?

Lynn Margulis.

Margulis had some great ideas early in her carreer, most famously the bacterial endosymbiont idea of the origin of mitochondria. She’s deservedly famous for pushing those ideas in the face of serious doubt. But lately she’s been spreading bizarre ideas about hybridogenesis being the main driver of evolution.(I was particulary distressed by the book on speciation she wrote with her son Dorian Sagan — also son of Carl Sagan, to whom Margulis was married — Acquiring Genomes: A Theory of the Origins of Species [2002], which argued that hybridization is a pre-eminent driver of speciation. Fortunately, the book has had zero impact on the field of speciation.)

Margulis is showing symptoms of what I call the Big Idea Syndrome, to wit: “Whatever I study is the most important thing in evolution — in fact, the driving force of evolution.” We’ve seen it before with developmental plasticity, epigenesis,evo devo, and other such buzz-fields. It’s all too tempting to think that one’s pet idea applies widely, or even universally.

PNAS of course, permits elected members of the Academy to promote papers they like by “communicating” them and adjudicating the reviews, which were often solicited by friends of the member or those known to be favorable to a paper’s results. This has resulted in the journal’s publication of some egregious work on the past; Linus Pauling’s papers on vitamin C are the most famous example.

Fortunately, this situation is changing, and soon PNAS manuscripts will receive the same rigorous review as regular journals offer — and bigwigs will no longer be able to push substandard work into publication.


Williamson, D. I. 2009. Caterpillars evolved from onycophorans by hybridogenesis. Proc. Nat. Acad. Sci. Early Edition (online)

61 thoughts on “Worst paper of the year?

  1. “Fortunately, this situation is changing, and soon PNAS manuscripts will receive the same rigorous review as regular journals offer”
    I hope that change hasn’t happened yet – I am just about to submit my groundbreaking paper about the evolution of frogs – the accidental offspring of a trout and a kangaroo!

    1. Hmmm – should I apply for a research grant to go to Bangkok to investigate the origins of mermaids?

      Don’t act so shocked. You’ve all seen those pictures.

  2. Thanks for the context. The Scientific American piece was unusually blunt in summarizing the reaction. Do you think they did that well? They certainly presented it as a wildly improbable, but they didn’t say how it could be tested. Or would it have been better to let this sink like a stone?

  3. Margulis is negative on “neo-Darwinism,” and has stated her belief that all speciation is attributable to genome acquisition. We now have complete genomes for Homo sapiens and Pan troglodytes, which are separated by just a few speciation “events.” Show me the genome acquisition.

    1. HA! – We acquired most of the chimp genome from our ggggggggrandparents – from whom chimps also acquired it… SEE – genome acquisition!

      Seriously, though – Margulis and endosymbiosis have always loomed larger than life in my mind and imagination. I hate to see her so far off course these days.

      On the other hand – people were pretty skeptical of her 40 years ago, and she was right.. I suppose if I don’t think about it too hard I might be able to hold out hope that she could still surprise us on this one…. or not…

      1. “On the other hand – people were pretty skeptical of her 40 years ago, and she was right.. ” I know that you didn’t mean this, but this is one step away from every crank’s defense.

  4. It should also be noted that Prof. Margulis is a devotee of Prof. Peter Duesberg and is an HIV/AIDS denier. She, despite being of Jewish origin, has also associated herself with Holocaust deniers.

      1. Did someone coin a term for this recently? i.e. if someone believes in one bonkers idea, they probably believe in other (theoretically unrelated) bonkerdoms?

    1. Margulis really has gotten bugfuck insane recently. Her opposition to natural selection seems to be based more on her political commitments than any scientific evidence, and it shows. She’ll endorse just about any idea, no matter how crazy, if it can be construed as somehow contradicting the notion that natural selection can lead to speciation. For instance, it seems like her endorsement of AIDS denial has a lot to do with the fact that HIV is a pretty astounding example of evolution by natural selection.

  5. Perhaps Mr. Williamson is not so far off. My mother liked to swim where fish are known to spawn and I have webbed toes.
    Also, I saw Waterworld so I know that I’m not the only one.

  6. I’m glad this happened. I think most scientists know there’s something broken with the PNAS process, while at the same time dreaming about being published there.

    1. I’m an author of a paper that’s just come out in PNAS and I can assure you we were NOT allowed an easy ride by the reviewers.
      This monstrosity of a paper, on the other hand should be nowhere near a journal like PNAS as it dilutes the scientific value of the profession of science as a whole.
      I downloaded and read the paper and what it is is pure speculation that is easily testable but not done.
      Why was this allowed to be published without the most basic of molecular genetic analysis being applied?
      You can’t put all the blame on Margulis or even the reviewers for this one (although you can certainly assign a LOT of blame on them). What about the editors?

      1. An irony-free congratulations, Sigmund!

        I was speaking to this point by Coyne.

        “PNAS of course, permits elected members of the Academy to promote papers they like by “communicating” them and adjudicating the reviews, which were often solicited by friends of the member or those known to be favorable to a paper’s results.”

        Obviously not all or even most manuscripts submitted to PNAS are the result of the cronyism or bias of the Academy. However, I believe PNAS has an accountability problem when it publishes biased speculation masquerading as science, which is a direct result of it’s unique vetting process.

        What’s the title of your paper, so I can look out for it?

      2. PNAS has had a “Track II” system in place for direct author submissions without member communication for probably 10+ years, in which the manuscript is handled as in typical journals. However, there is certainly incentive to have it communicated if possible.

      3. “What’s the title of your paper, so I can look out for it?”
        The p53 target Wig-1 regulates p53 mRNA stability through an AU-rich element
        I’m not first author but I’m there! (in my non internet pseudonym persona.)

  7. I was grateful to see the last paragraph of the article. I had raised a question about peer reviews in another thread and I was uncomfortable reading the article – until that last paragraph.

    Turn the ship into that huge wave and we will make it to the next port!

    – a paraphrase from a PBS documentary I saw the other night about Krakatoa. It just came to me when thinking of the PNAS review process.

    1. It should be pointed out that there are different types of review process involved with PNAS papers. Most papers submitted to them DO go through a tough peer review process. One category of papers “contributed” by National Academy members does not (or rather does not if they chose to pick only favorable reviewers that are likely to approve the manuscript.)

      1. Sigmund is right. And just to flesh out this distinction, there are 3 avenues. You can always figure out which is which by checking what describes the editor:

        1) “Contributed by Bob Smith”;
        2) “Communicated by Bob Smith”;
        3) “Edited by Bob Smith”.

        All editors are Academy member, iirc.

        The first requires that the editor be an author. The second requires that the authors contact a member and ask them to act as their communicating editor. Both of these get pro forma reviews, but can be turned around rapidly and are entirely at the discretion of the editor (ie NAS member).

        The final track is more or less like any other journal. And I think it is also usually of higher quality.

  8. This reminds me of how I envision the physicist arxiv system. (Disclaimer: I haven’t used it.)

    I believe the new system is that they try to weed out the bulk of the crackpots by having their _CVs_ “peer reviewed”. (As it is a communication pipeline/data base, there is little review of individual papers as of yet.) Aside from those who crack later, or fool their sponsors, I’m certain there are problems with this type of solicitation.

    In any case, there are papers like this, that in part (well, at least it’s testable) or whole are “not even wrong”, out there.

  9. Didn’t Margulis also claim, incorrectly, that every organelle in the cell is an endosymbiont? And that cilia are as well?

    My general thought is that she gets too much credit for what she got right. You don’t give a marksman’s badge to the guy with the shotgun, after all, just because he hit the bulls-eye a couple times.

    1. She doesn’t have a shotgun, she has a Gatling gun and she is spinning in a circle spraying bullets all around. One hit the target.

  10. How does it rank vs. last year’s paper briefly in (before it was removed by) Proteomics that asserted that the mitochondrion was the seat of the soul?

    A lesser journal, to be sure, and not the fault so much of the review process since the lunacy was apparently slipped in during manuscript revision, and with tons of plagiarism to boot.

    Search: mitochondria soul warda if you missed that one.

    Otherwise, since it sounds like the PNAS author is a Fossil, will some now point to the paper as fossil evidence of genome capture 🙂 ?

  11. There’s an article out in the New Scientist with a similar theme. It talks about sea squirts and starfish being hybrid organisms. Is there any more evidence for that than butterflies?

    1. This New Scientist piece is apparently referring to an earlier paper by the same guy that appeared in American Scientist (95:509) in the November-December 2007 issue (I’m not sure why New Scientist is digging up a two year old article for comment). In that article, no evidence is presented that sea squirts are inter-phylum hybrids. Someone actually tried to cross starfish and sea squirts, but it didn’t work (some starfish eggs developed, but molecular genetic analysis showed they were just starfish: there may have been sperm contamination, or parthenogenesis) Quite the opposite of hybrid origin, the take-home message from the sea squirt genome project is that sea squirts are, as had long been known on morphological grounds, chordates; see Science 298:2157, 13 December 2002. In fact, recent molecular systematic work indicates that sea squirts are even closer to vertebrates than lancelets (Amphioxus; the traditional closest relatives of vertebrates) are.


    2. There was also a bit about this in New Scientist’s infamous “Darwin Was Wrong” issue in Jan 2009.

      Interesting speculation, but apparently nothing more to it.

  12. The issue is not limited to biology though. Fred Hoyle had some great ideas early in his career but later on he became known as the “guy with the crazy ideas”. Some people made a sport of showing how Hoyle’s ideas were wrong and some ideas were just so far-fetched that people put them in the “obviously wrong” category and wouldn’t even think about it.

    Now the problem with Hoyle (and many others) was that he made statements but did not put forth any framework for testing his claims; he essentially made claims and left others to think about the implications of his claims and how to test them. I wouldn’t mind people putting out crazy ideas if they also give people some indication on how to progress; that would at least show that some thought was put into the crazy idea.

    As a crazy idea let’s take the “ether” (except that many people didn’t think it was crazy back then). A.A. Michelson devised a scheme for proving the existence of the ether (he was an ether believer) but demonstrated that the ether could not be detected. Believing that he just wasn’t doing a good enough job of detecting what was not there, he improved his experimental setup. Ironically, his experiments produced excellent data which helped bury the notion of the ether. Michelson developed instruments which were incredibly simple and yet so sensitive and so useful for many fundamental investigations. So crazy ideas can lead to good work if people can state the ideas in such a way that they can be tested. Not all crazy ideas are wrong, but as far as crazy ideas go, as Carl Sagan put it: “Extraordinary claims require extraordinary evidence.”

  13. “PNAS of course, permits elected members of the Academy to promote papers they like by “communicating” them and adjudicating the reviews…
    Fortunately, this situation is changing”

    I don’t know if this is a good thing. Sure, the current system means that papers like this one occasionally get published. But what’s the harm done? Everyone knows it’s crazy.

    Given that the Academians are all (I hope) very accomplished scientists, I think it’s worth listening to what they want to communicate, even if it does mean that dodgy papers will come out.

    So long as they’re read in the right frame of mind (i.e. with a pinch of salt), I say, let it continue. It’s like with Medical Hypotheses, many of the papers in which are absolutely ridiculous, but it provides a useful service.

    1. Sure, the current system means that papers like this one occasionally get published. But what’s the harm done? Everyone knows it’s crazy.

      Sure, you and I and everyone who is sane, educated, and does not cut themselves of from reality knows, but what about the general populace? And the children? Won’t somebody think of the children?

      Homeopathy believers still cite the 1988 Benveniste paper that got into Nature.

      1. True but that’s Nature, which is meant to have a very rigorous peer review process.

        Whereas an article communicated to PNAS by a member of the NAS is, and everyone knows it is, not.

        So being published in PNAS is not the same kind of endorsement as published in Nature.

  14. I was going to mention the American Scientist review in which the insect/onychophoran idea was just one of a truly bizarre parade of similar howlers. ALL metamorphosis is to be explained by hybridization. Very odd stuff.

  15. Thanks for writing this article. I’m going to show it to my conspiracy theory friends that there are no authorities in science and that even the biggest names can get their butts kicked (i.e. unsupported claims successfully challenged) by any scientist anywhere at anytime without consequences. One friend even believes that the Royal Society suppresses science. Yeah, sure, like they could (and would want to) pull that off. If RS members did do that, they’d have to answer to their friends, grad students and thousands of their peers in a very short period of time. Unlike religion, politics and the corporate world.

  16. Not being an insect or onychophoran specialist, my biggest gripe with the paper is the disconnect between the title and the content. For such an assertive title, the paper offers nothing in the way of convincing data as support, just some pictures and conjecture.

    And where are the phylogenetic hypotheses? If his hybridogenesis hypothesis is correct, then we would expect a random sample of markers from a holometabolous genome to produce one of two strongly incongruent phylogenies consistent with the positions of the two potential parents. A simple trawl through the literature for such conflict would have strengthened his case. Why are such papers not cited? Could no such evidence be found?

    And any paper that begins with “I reject the Darwinian assumption that larvae and their adults evolved from a single common ancestor.” is begging for a creationist quote-mine…

  17. For years, ICR’s Duane Gish included in his “debate” spiel a challenge to evolutionists to explain how the “miraculous” metamorphosis of the butterfly could have evolved. ICR even published a book about the Monarch butterfly by a creationist (a retired engineer) demonstrating how it proved the wisdom and goodness of the Creator. It will be interesting to see how (and if) ICR handles this flap. Ordinarily it seizes eagerly on anything that can be made to reflect badly on “evolutionists”. Here we have a putative answer to its challenge, which while seemingly daffy is both evolutionary and testable (if anyone cares to test it genomically).

  18. Not to disagree with your post overall, but the sentence “Besides being inherently implausible, the theory can be easily refuted by DNA analysis, which should show, in butterflies, two completely disparate genomes, one resembling that of onycophorans.” does not make much sense. Unless I’m missing something, currently not enough DNA of onycophorans has been sequenced to be able to do this analysis. So all this is saying is that you expect the results of such an analysis in the future (when the data become available) to falsify this prediction. You could say that about any scientific hypothesis you don’t believe (provided it’s specific enough to make falsifiable predictions, i.e., doesn’t fall into the “not even wrong” category), so this statement does not really add anything to the critique.

    1. ok, I’m sure enough about this hypothesis being wrong that I’ll bet you $100 that when the data come out, it’ll show me right. You on?

  19. I like to think of Lynn Margulis and Kary Mullis as the trickster goddess and god of biology. They’re extremely creative thinkers who are occasionally spectacularly right and more often spectacularly wrong, but love stirring up a fight more than they enjoy actually doing research.

    Seriously, this hypothesis is testable, but I’d expect the author to have done some of the basic background work. Insect sequences are extremely well represented in DNA databases, and a quick search of the GenBank nucleotide database pulled up 29 pages of entries for onychophoran sequences. (Maybe nowhere near enough to clinch the case, but it’s a start.) “I’ve got an idea — now someone else go test it” is okay for a freshman term paper, and passable for a theoretical paper in a field where experimental data is sparse, but in this case, it’s a sign of vanity or laziness or both.

    1. I came across this late, so I’m sure you’ve all moved on somewhere else, but for posterity, a couple of observations on this blog:

      1) Williamson had a long career of careful study of larval marine organisms, published in respectable marine science journals and filling in gaps in knowledge, without rocking any boats. His more recent and radical ideas came from decades of careful natural history observation, most of it conducted long before we had the scientific tools to study genomes. The larval transfer hypothesis was put forward to explain a paradox he observed in marine larval morphology -why some bi-radially symetrical adults had penta-radially symetrical larvae, and vice-versa. There wasn’t an explanation available, so he proposed one. It seems strange that some of you have taken the time to dismiss the scientist, as well as the science, without at least taking a cursory look at either.

      2) Williamson was 87 years old at the time of publication of the PNAS article. If you were asked to referee a grant application from him to test these ideas, would you accept it? Would any of you risk your careers or reputations by lending him your time or your lab facilities? No? Why do you think anyone else would then? Given this level of skepticism and its consequences, how was he supposed to get the resources needed to provide the evidence? In such circumstances, the best you can do is theorize and hope your ideas will spark an interest and someone will think it worthwhile to test them. Which I think was the point of this and his other publications. Mainstream science works by constructing an argument based on existing evidence, identifying a research frontier, and then seeking the resources needed to test your hypothesis. Margulis saw merit in this hypothesis and championed it. She may well be wrong. I’m not an evolutioanry biologist or geneticist, so I don’t know if they have merit or not. If not, ignore them (its what most people have done). If they do, test them.

      Seems simple enough. So why all the anger?
      Accusing the author of a paper you disagree with of madness, laziness, arrogance, and even of the crime of being old (fossil comments) is not warrranted.

      By the way, I agree evolution is true, and PNAS Track 1 refereeing sucks – or sucked, so I agree with everything you say here – I just don’t like to see ad hominem attacks in a science debate.

      1. Eddie,

        I support your points: In particular, not only is Williamson about 90 by now, but he is also confined to a wheelchair by illness or accident, so it’s even harder for him to explore his ideas in a very hands-on way.

        Another point: It is true that the whole notion of a critter acquiring a phase of metamorphosis as a part of its development process is wildly unlikely (at best). However, the acquiring of a phase of metamorphosis by stepwise evolution is also very unlikely: It’s an archetypal example of the “hopeful monster” concept by Gould, et al., because a half-evolved metamorphic process is a non-starter. So even from a more standard perspective, the evolution of a metamorphic phase is, quantitatively, a “then a miracle occurs” moment. In a sense, Williamson has simply proposed a particular form for this “miracle”.

        I do not know nearly enough about gene expression and development to evaluate whether Williamson’s concept is totally impossible or just wildly improbable. I will note that I have heard that genomic information from extremely different species is very much the same, which would be a pebble on the side of “maybe this isn’t completely impossible.”

  20. PNAS is NOT a prestigious journal.
    It exhibits the worst nepotism of any journal. Most papers published there are garbage published by friends of the editors.
    Not only do I not submit papers there, I do not even cite papers published there. 50% of the papers there have major scientific issues.

Leave a Reply