What happens if you try to disperse the Impermeable Fog

Why do bad papers get published in proper journals? Well, mainly because the Impermeable Fog, known as review confidentiality, obscures the entire decision process and ensures no-one can be made responsible. Like in the case of the recently published (pseudo-)phylogeographic study by Zhang et al. (2017) in Tree Genetics & Genomes.

The corpus delicti

A citation alert pointed my attention to the study of Zhang et al. on Zelkova. We published a paper on the evolution of this small elm genus more than a decade ago, there were a few papers on genetic differentiation in the western Eurasian species and Korea/Japan, but that's pretty much it. It is too small to attract a lot of attention by molecular phylogeneticists. [Explanation for non-scientists: these are people who try to infer the evolutionary history of a group using gene sequences or similar data.] Extra-tropical tree genera, large or small, can nevertheless be very challenging, hence, interesting. And it looked like somebody put the effort and further studied the complex mix of ITS signals and plastids including new Chinese material.

I asked a colleague to send me the PDF.
[Info for non-scientists, in particular tax-payers: although most scientists that publish non-profit research are state-employees, and we do most of the work when it comes to publishing, acting on a voluntary (unpaid) basis as authors, peers, editors, and not rarely proof-readers of what the “professional” proof-setters produce, our libraries and institutes pay for the access to most journals including our own research. Unless the authors also pay for free (open) access. Tree Genetics & Genomes is a journal published by Springer-NPG, a private company charging 3000 $ for open access.]

What I got was a paper with quite a lot of strange phrases, even strange to a German like me (and we Germans have a large arsenal of odd English phrases). And even weirder content, and based on an extremely poor data basis.

Charging the windmills, Don Grimmxote is on the ride!

From a recent experience, I knew the journal has a stringent review process in place (two rounds with two different sets of reviewers, and an editor checking the 2nd revision himself); so how could they fail so badly? Not worried about burning bridges, I wrote a couple of unusual (in our business) mails to the corresponding authors and Editor-in-Chief (EiC), and I also confronted the communicating editor [Info to non-scientists: This is the one who makes the call]. I asked him, on which basis he decided to publish this study. His answer was as swift as evasive. He cannot comment, because of peer review confidentiality.

Meanwhile, I learned that the poor English relates to the fact that the only native speaker in the author list, a long retired but quite well merited U.S. American population geneticist named Stewart C. Sanderson (apparently a devout Mormon and high priest of his church; fun-fact: Mormon religion advocates creationism) died June 2016, seven months before the paper was submitted. An information the authors were too shy to give. Zhang and the late Sanderson published three joint papers this year, and about two dozen since 2011 published in an illustrious set of (pseudo-)journals; the machine was running well till the very end. In their response mail, the authors thanked me for my comments and confirmed that a certain James I. Cohen checked the English instead, as stated in the acknowledgements. Either the authors lie or he gave up at some point, because the English of the submission, and partly the 1st revision, was quite bad and one reviewer spend some time on correcting the worst. They did incorporate some of the reviewer's edits and information, but not all; and mostly ignored the to-the-point critiques. I, and you, should not know this at all, because of peer review confidentiality.

Finally, the EiC informed me that the review system is not perfect but he stands by the decision to publish (obviously not having read the paper or my mails), but I’d be welcomed to write an “opinion piece”, which then would be reviewed and eventually published. I answered that I’d be happy to do this, but only if the entire process is made transparent. Let’s disperse the Impermeable Fog for a brighter future! A month is gone and I didn’t get a reply (see here for the correspondence so far), so I'll just post the main critiques (the paper is a really nice and educative example of how to not make a biogeographic study).

Laudable and other reasons for publishing a bad paper

Given the circumstances that this is probably the last collaborative work of a dead (and long buried) man, it could be viewed as a kind of scientific homage. Pär's decision may have been a compassionate one, and the poor finish is simply due that he didn't find anyone who was good-willing enough to review this unsavable paper, again. His hands are bound: Advising the authors to find living cooperation partners that can help them to improve the study/text, is considered unethical. Teaming up with a constructive reviewer who already put a lot of effort in saving the paper is naturally a no-go, too (although there have been occasions, where relatively famous reviewers sneaked into co-authorship for much less…but thanks to the Fog, we know nothing). As a well-meaning reviewer, I’m not even allowed contacting the authors to explain them in detail where they went wrong and discuss with them, how to escape the dead end. Prohibited by peer review confidentiality!

Swedish transparency laws vs. peer review practise, and only one way out

I find it quite odd, that any journalist could ask for Pär's or my (when I was still living in Sweden) tax reports and what we did with our VR-research money, but not his decision letters and the reviews they are based on. Although, in this case, two of the involved persons (the editor and one peer bound by confidentiality) are paid by the Swedish tax-payer, and Springer-NPG earns good money by selling their journals also to Swedish institutions and libraries. How can we ask for transparency in all parts of live, but not the peer review deciding on what research gets published and what not?

Once a bad paper is published, it stands. You cannot hold anyone responsible for publishing the paper or having violated good practise: the report(s) on which a decision is (are) based is (are) kept confidential; also a frequent measure outside science to cover up dubious decisions!
And like bad decisions done by politicians and a-like, having no proof of misconduct, one can't force the retreat of a study once published. Or to think forcing the journal to issue an erratum (hard to get any statement out of them, I tried this once during my active career). Hip-Hip-Hurray for peer review confidentiality.

The only way to set things straight is to put (a lot of) effort into drafting a formal response (‘opinion paper’ in case of TGG), which then is going through the same nebulous review process than the original paper. Maybe it will be treated fairly, maybe not; thanks to peer review confidentiality no-one will know. So much for the ideal of scientific discourse. Instead, it’s more the 1916–1918 situation in north-eastern France, with some alien camouflaged spaceships hoovering above you and you have no idea on which side they are (probably the other). Plus, the effort you have to put into a formal response can easily be higher than the effort the authors put in their published paper; and there is little gain in it. So, most bad studies with provably wrong results/conclusions stand for eternity. Many of them will just remain uncited and quickly forgotten (like many good studies), but others (like this one by a famous palaeontologist, a study one of my professors recommended as a good read for fame always beating peer-review or proper data) can have substantial impact before their faults become too obvious and widely known. Some may even haunt you when you submit your next paper, because you need to explain to the editor (and peers) why you’re not citing the crab. [Info for non-scientists: Of course you cannot put it this way, you have to produce an “objective” reason for not doing it, e.g. citing papers with opposing results.]

Publication of scientific results can be a very dirty business. All we see in print is filtered by an Impermeable Fog that no-one can (afford to) disperse. But, it would not need to be this way. All could be avoided by just making the review process transparent! The readers, but also the public paying all the bills, should have a right to know based on which assessments a study was judged publishable. Or at least should be able to inquire, when doubt has raised.

A call to arms!

As a natural-born realist, I know that I will probably not change the system. Still, it can be fun to try. In case you agree that the Impermeable Fog needs to be blown away, or don't like that your taxes are spend on keeping up an intransparent system, I started a petition on; feel free to sign up for the fight against the windmills (a hundred signatures are needed to get it public).

No comments:

Post a Comment

Enter your comment ...