The ‘Nature’ problem – taxonomy and descriptions

This post has been sitting around for a while waiting for the right moment, and today seems to be it. Matt Wedel has just put up a large post on the problems of impact factors and as that is touched on here (albeit from another angle) it seemed prudent to use the opportunity to weigh in on the subject.

If there is one outstanding issue in palaeontology right now it is what I call the ‘Nature’ problem of descriptions. Exciting and above all important specimens get into journals like Nature because they deserve to get a wide audience, but even with the current trend for extensive online appendices, the descriptions of the specimens are often reduced to a few lines of apomorphies as space is made for colour photos and discussion of the evolution and lifestyle of the new taxon (in what is already a tiny paper). Now this is all very well in principle – I would be delighted to get into Nature and if I ever got material exciting enough to warrant submission there, I would do so. But, and this is a big but, these kinds of papers are completely insufficient for basic research.

In many ways, the fundamental unit of palaeontoloy is the taxonomic or specimen description. Unlike biology where if necessary most labs can get their hands on a living or dead member of any given species, we typically have only one to work with (for dinosaurs at least) and that is generally incomplete to boot. They can be extraordinarily delicate and too fragile to move (like an insect) or so vast as to make movement impossible (like a big sauropod). This means it is common for the researcher to visit the specimen and not the other way around, but this is not always realistic. Money is always tight and if for example you want to make a survey of the hindlimb lengths all theropods, you can hardly spend six months circling the globe just to take 200 measurements. Thus we rely on descriptions, drawings and photographs of specimens to provide us with much of the basic information which we require for our research.

However, anyone who has written a full description of even a partial skeleton will tell you just how long it takes. Describing something that is largely complete, with lots of interesting anatomy like sauropod vertebrae, or hominid skulls (which is, of course, exactly the kind of thing that generally makes it into Nature), and in the kind of real detail that is desired by other researchers can take months. My relatively brief description of Fodonyx came in at 20 pages and that was only a skull (with no proper palate or braincase) and one hindlimb, and that could have easily been more than doubled if the skeleton had been complete, and perhaps doubled again if there were multiple specimens, or there were some exceptional features that required detailed work. Making up a 50 or 70 page paper is no easy task, especially when you have all your normal duties – as a PhD student I could work on Fodonyx to the exclusion of virtually all else for weeks at a time, something that is all but impossible now.

Now for the catch – there is enormous pressure on researchers (at every level, but especially when competing for jobs at the lower end of the academic ladder) to publish as much as possible. ‘Publish or perish’ is no idle threat. This approach favours short papers and in highly ranked journals over long paper or those in lowly journals. This leaves researchers in a difficult position – do you complete several short papers quickly on a hot topic that will be well cited over the next few years (when citations count), and that give you some credit on your CV, or spend months working on a huge paper or monograph that will be extremely valuable over say, 50 years, but of little use to anyone tomorrow. You can guess which takes priority, and this goes doubly so for a journal like Nature when papers are so short, publication times so fast, and citation indices so high. Here you need to provide only a tiny description (however brief) and the specimen has been ‘described’, but with so little detail it is not of great use, and the pressure is off the describer to write anything longer now the thing has a name.

The fault here of course lies with assessment systems that pressure researchers in this direction – they want to quantify and measure the unquantifiable. It is not your fault that you work on an area (like say Triassic footprints) that has few other experts, your work can be valuable and of excellent quality, but you will be little cited (few other workers in your field to cite your work) and will only get into mediocre journals (not enough general interest). If your work also takes a long time then you will be further disadvantaged by not publishing many papers (even if they are long ones) compared to say someone who is working on bird origins – lots of short papers on certain morphological features, lots of colleagues and competitors who will cite your papers, and a lot of general interest. You might do the same amount of work as your colleague down the hall, and produce the same number of pages of excellent research, and get in similar amounts of grant money, but only one of you will ‘appear’ to be doing well. Thus, everyone is driven towards getting out more papers (not necessarily more pages) and of greater general interest (rather than greater practical worth). You simply cannot simply quantify such output, especially over the short term – who knows what will turn out to be a marvel of insight, or an essential analysis ten years from now, yet the capriciousness of peer review (especially with the competition to get into the most senior journals ensuring that most are rejected almost on arrival) means that brilliant papers will be overlooked, and average ones published in their stead. Time will uncover the gems and brush over those without lustre, but in the immediacy of measures of scientific worth, the reverse can be true.

If you scout the literature just a little you will see the there are plenty of great and important species that got their first airing in Nature or Science, and yet 10 or even 20 years later, no good description exists. Often, somewhat by definition, these specimens are the most important of their kind (or they would not be in Nature in the first place!) yet no detailed description or set of photographs and drawings exist, just two or three pages with a couple of drawings (that have probably dated in that time too). As I mentioned, the increasing trend for extended appendices online is alleviating some of this issue, and several researcher teams have managed to get the best of both worlds by published the description in chunks – thus getting out a very thorough description, but in several short papers that can be submitted to better journals and relatively quickly. (I should add here that obviously there is a very strict limit on paper lengths in journals like Nature, but even some of the mainstream palaeo journals will baulk at being presented with a 70 or 80 page manuscript and refuse to take it, demand it be cut, or require money to cover the additional printing costs, all things that might not be possible for some researchers). It is not ideal, but it certainly helps and it is reducing the time it takes to get the work into the literature though of course that does not help in clearing out the back-log of those ‘undescribed‘ pieces. Some of them are not that accessible either (aside from their geographical location), describers have the informal ‘right’ to hold onto their material while they are still doing the primary work on it (i.e. making it inaccessible to others for detailed analysis while they do their work), and thus some (but by no means many) pieces are largely unavailable as the ‘owner’ is still working on them. But then with no pressure to complete that work (the name has been published), and in fact pressures to avoid taking months to write long descriptions, this just drags on and on with no description appearing.

I cannot really offer a substantive idea as to how to combat this problem. There is no sign of administrators changing their measures of how we as scientists perform, and it can be just as hard for hiring committees with hundreds of applicants to process for some jobs – they can’t go a read even 1% of the papers those people have collectively published to assess their worth. One can hardly ask Nature or Science to ensure that full formal descriptions appear in the appendices, and part of the attraction of Nature is the ability to publish a short paper quickly and thus remove the problem of spending months writing a long description – you want to get the paper submitted fast and get your important discovery published and recognised, so this would actually act against you in many ways. One cannot either expect researchers to avoid high-impact journals like these because they do not allow for big papers – you are just then actively preventing yourself from getting important work published in a suitable forum (and penalising yourself in terms of missing out on the credit that comes with it). It really is an almost unsolvable position unless researchers take it upon themselves to finish what they started and get a full description published to provide the necessary details for their colleagues once the ‘lite’ version is published. Of course good work deserves to be promoted in the most prestigious publications, but it is actively hampering research and that cannot be considered appropriate.

One last thing, at the Wellnhofer pterosaur meeting we had an open session on pterosaur taxonomy and systematics. The general consensus was that the taxonomy was not good enough, lots of genera and species needed revisions and descriptions were not up to scratch having mostly been done in the dim and distant past and so the photos were inadequate (or non existent before the 1950s) and drawings often unrepresentative, in addition to issues like non-standard anatomical names etc. Then the most obvious point reared it’s ugly head, ‘Ok, so which of us is going to do the work’ at which point everyone agreed that while the work had to be done, no-one could do it. Abandon your traditional research (which it seems by law must be called ‘cutting edge’ in grant applications) to revise century only specimens that have often already been described two or three times already, and you can watch you grants disappear in front of your eyes, career stall, and a huge lack of interest and citations in your work.

Of course we are all chipping away at the problem slowly with the odd revision and redescription, but most of it is left untouched by necessity rather than desire. It is just practically impossible to do the work because it will impact on your career and your ability to do other research. People need to realise the value of basic descriptions and taxonomy (certainly many palaeotologists do) LINK and this goes for funding bodies and hiring committees alike – please realise that just because something gets 50 citations in two years does not make it a good paper, and just because something only gets two in the same time (both by the author citing his own work) does not make it bad. What you think is important and what the actual scientists think is important may not be the same thing.