Thursday, December 9, 2010

The Curse of the Nonexistent Dinosaurs

You all probably remember Darwinius, that little extinct primate that caused such a huge fuss last year (the image at left is from the original paper). In addition to the massive hyperbole surrounding the specimen (much of which has now been tempered by additional analyses from other researchers), an early issue concerned the validity of the name Darwinius itself.

Scientific names for new animals, living or extinct, are governed by a set of rules from the International Commission on Zoological Nomenclature (ICZN). This body - which has no real authority, other than that granted it by us scientists - is by its nature a rather conservative entity. This isn't necessarily a bad thing, particularly because stability in our naming system is so critically important. When I say Kosmoceratops richardsoni, everyone else should know I'm talking about. But, this conservatism also means that the ICZN has been very, very slow to react to electronic publishing. Right now, names published in electronic-only form are not valid! A paper version, even if it is otherwise identical to the electronic version, must be created (although the ICZN is working to change this).

So back to Darwinius. . .when the name was initially published, it was only in electronic form, and thus not "real" in the eyes of the ICZN. Vigilant eyes noted this, and the problem was rapidly broadcast in the blogosphere. Some wags suggested naming Darwinius out from under the authors, but thankfully cooler heads prevailed. PLoS ONE, the journal in which Darwinius was published, was able to produce a paper copy of the article in order to satisfy the ICZN. It took a few days, but finally Darwinius was "real"!

Thankfully, PLoS ONE learned its lesson. Their publication policies (full disclosure: I am an editor there) have been revised to accommodate the needs of the ICZN, and all is happy now. Or is it?

It turns out that other journals haven't necessarily learned this tough lesson. Time and again, scientific names invalid in the eyes of the ICZN are being published by some pretty major journals. This "zombie nomenclature" walks among us, even today. Here are just two examples.

On September 17, 2009, Science posted a preprint of a paper naming a new small tyrannosaur, Raptorex kriegsteini. As near as I can tell, this name existed only in electronic form at the time, and thus was not yet valid in the eyes of the ICZN. Raptorex was very well publicized (and rightfully so - it's a neat specimen), yet no paper edition appeared until October 16, 2009 - a month later. For four weeks, the name Raptorex was in taxonomic limbo.

Koreaceratops (image by Nobu Tamura)
This early horned dinosaur is hot off the digital presses. As you can probably guess, it hails from South Korea, and the authors hypothesize that it may have been aquatic. Koreaceratops appeared on November 17, 2010, as an online preprint at Naturwissenschaften, and some media attention started popping up earlier this week. Yet. . .the name is not yet valid in the eyes of the ICZN! Until Koreaceratops appears on the printed pages of that journal, the name has no real weight. Hopefully this will get resolved soon!

Cryptovenator (image by Charles R. Knight)
Dinosaurs (and Darwinius) aren't the only organisms afflicted by digital naming issues. On September 16, 2010, the genus Cryptovenator appeared on the website for Acta Palaeontologica Polonica, as a new kind of early synapsid (closely related to the fin-backed Dimetrodon, depicted at right). It hasn't yet appeared in print - so remains "undead" according to the ICZN.

Some Thoughts
The rise of online pre-prints is a Good Thing. It speeds the flow of scientific information in a way that just wasn't possible before. Yet, it also creates a bit of a headache for authors. The window of time between pre-printing and real-printing is a dangerous place for a new scientific name. Although the threat is slight, there is nothing to stop someone from accidentally or intentionally "scooping" the rightful authors. As near as I can tell, the proposed ICZN rules won't do anything to fix the situation, either.

So, what solutions are there?
  1. Don't publish pre-prints of papers with nomenclatural acts. This is the simplest solution in some ways, but also produces the undesirable effect of slowing down scientific communication. Canadian Journal of Earth Sciences has adopted this strategy (hat-tip to Hans-Dieter Sues for the reminder on this one).
  2. Petition the ICZN to give priority to the first appearance of the name in the first widely disseminated pre-print of a paper. This would be a nightmare on multiple levels - an addendum to the Code could take years, and how would one decide which version of a paper containing a name is the authoritative one?
  3. Continue to publish pre-prints, assume that all is going to be well, and worry about it when something bad happens (taxonomic scooping, perhaps?). Highly undesirable, but basically the course of action that we're on now.
  4. Expunge the names from the paper in its pre-print form (suggested by Martin Brazeau at another venue). This does solve the problem, but it requires someone to go through the paper and remove all mentions of the taxon. Despite all of the talk from commercial publishers about "value added", I suspect it's either the unpaid editorial volunteers or unpaid authors who would get stuck doing this. Removing names is not so tough in the text, but figures (such as phylogenies) would be a bear. Also, we're left with multiple versions of the same paper floating around.
What thoughts do you have on the matter? Is it really an issue? Or am I making a mountain out of a molehill? Speak up in the comments section!

Disclaimer: Mention of the "invalid" names above is not, not, not an invitation for paleontological wannabes, ethical miscreants, or rabblerousers to wreak havoc on paleontological nomenclature. Although I hesitate to draw attention to particular instances, I feel that without some concrete examples, my arguments would be less effective. Furthermore, the names are already out there in some very widely read venues! Leave the names for the authors; the situation is not their fault.

Update December 31 2010: Bill Parker over at Chinleana has a great post relevant to this issue.

11 comments:

Nick said...

2 and 3 are the obvious choices. I am not bothered by 3 though. But that could be wishing too well for basic human nature.

Matt Martyniuk said...

Another solution, which is used by the upcoming PhyloCode and is also being worked on (in some form or another) by the ICZN, is to register names in an electronic database at the time of electronic publication. The serial number for each new taxon can also be included in the preprint (I've seen authors in PLOS One already including database registration numbers). That way the name can be locked down even before the slow process of getting an issue to print is finished.

Mike Taylor said...

I would be remiss if I didn't point people at my own recentish paper on this subject, Electronic publication of nomenclatural acts is inevitable, and will be accepted by the taxonomic community with or without the endorsement of the Code. It's v ery easy to find online for anyone who wants to read it, but then the title pretty much says it all. The interesting thing about this, from my perspective, is that it was published in the Bulletin of Zoological Nomenclature (the ICZN's own journal) which I am optimistically interpreting as a sign that they are more open to change than usually thought.

Mike Keesey said...

On Matt's point, it's worth pointing out that the PhyloCode, as currently drafted, does not allow electronic-only publication, so this issue isn't entirely resolved.

Victoria said...

Solution 1 is already being used by the Canadian Journal of Earth Sciences. The title of the article will appear on the issue's webpage, with no link to the PDF and the following note: "This article contains taxonomic descriptions. Therefore, the link to it will not be made available until the print version has been distributed." Once the paper version is available, the PDF goes online.

Jim L. said...

It may be reasonable to strike up a deal with the "electronic only" journals to produce special printed versions for the papers in question, that way this is a mute point. I'm not sure how valid a solution this is though, not knowing that much about the ICZN rules and what they consider a valid printing.

Anonymous said...

I was hoping that there were still a chance that Raptorex could find a better name.

If an specimen turns out to be a synonym, like Nanotyrannus may be (I guess), can its name be "recycled" to an actual species?

Nanotyrannus is so much better than Raptorex.

Andy said...

@Jazinator, this is essentially what PLoS ONE is doing. . .it would be a very messy solution for the preprint problem, though, as the preprint is not really the version of record.

dinogami said...

You left off an option: ignore the ICZN's antiquated rule on electronic publication and do what the (much, much larger) publishing industry does: consider the electronic publication valid. Yes, people have pointed out before that it's silly, capricious, and arbitrary to ignore some rules but follow others from any organization, but when the solution is patently obvious and the "ruling" body stubborn to the point of inaction, then...

Andy said...

Very good point, Jerry. . .

Unknown said...

I like Mike Taylor's comments published by the ICZN, and also think that dinogami has a good recommendation. The ICZN is pretty much past its window of opportunity on this question, and many of us think that the naming game prior to the 1895 creation of the ICZN was at least as good as what we have now. Ultimately, the few specialists working in a specific area will have the final say in what names are acceptable, what names are useless, and what names are to be synonymized. Names are just tags, after all, and history has shown us that many are relatively impermanent as we obtain more information about the animals that we have named. This is a dynamic process. The ICZN, although showing some promise, is now bogged down with implementation of a simple online database (!). If they cannot handle this simple development (about 2 days worth, including testing), then what use?