COPE, the publishers’ Trojan horse, calls to abolish retractions

Four private scientists without any agenda whatsoever published a research result preprint on the portal BioRxiv. The “new results” reported in the article are actually new ideas which are just as good as any research results, because they are supposed to bring the field of scholarly communication forward. The question is, where to, and why should anyone go there. Because the idea is to abolish the only tool science now has at hand to punish research misconduct: retractions. Fraudulent papers are to receive instead an amendment, which will notify those particularly inclined readers that research data or ethics approval (for clinical studies) might have been falsified or missing. Those proposing to remove the only punitive measure available in scholarly publishing are in fact the very people who are supposed to be overseeing the editorial integrity. The goats whom science welcomed as gardeners now dropped the pretence and declared their true vision for the garden.

The preprint is titled “Amending Published Articles: Time To Rethink Retractions And Corrections?” and is authored by Virginia Barbour (1), Theodora Bloom (2), Jennifer Lin (3) and Elizabeth Moylan (4), their respective affiliations are: 1 Queensland University of Technology (QUT); 2 BMA House; 3 the Digital Object Identifier Registration Agency  Crossref and 4 the publishing house BioMed Central (BMC), which is part of Springer Nature. Only a footnote said that Barbour’s main work is not in her academic capacity with QUT, but as the Chair of the Committee on Publication Ethics (COPE), a publisher-financed advisory board on editorial processes (see my earlier report here). The publisher executive Moylan is in fact also since last year a Council member at COPE. Bloom’s day-job is that of Executive Editor with the medical journal The BMJ, the journal which co-founded COPE in 1997.

However, Bloom declared to me in an email:

“As indicated prominently on the preprint, the views expressed are those of the authors and not our organisations”.

Of course they are. As Bloom pointed out, there is even a note: this document does not necessarily represent the views of the organizations listed here”. Why would Bloom, COPE president Barbour, or the BMC research integrity official Moylan let their private views on how publishing should deal with misconduct influence their professional day-to-day work, which just coincidentally is mostly about decisions on research integrity and editorial ethics? Such supposition is just plain silly. That the preprint constantly refers to COPE, is just a coincidence.

This is what the COPE president Barbour et al have on offer in their preprint:

“We propose removing the term retraction altogether and instead using the term “amendment” to describe all forms of post-publication change to an articles. The term “amendment” carries a neutral tone and is generic enough to apply to a wide array of cases, including the smallest instances such as a typographical error all the way to large retractions. By employing a uniform term, we hope to remove any associated stigma in the context of scholarly literature. When readers encounter each amendment, they read the notice for details on each case, and can judge the article and its revisions as appropriate”.

This above might as well have been written by the notorious cheater scientists Olivier Voinnet or Paolo Macchiarini, who were so far quite successful to avoid a number of retractions. Voinnet succeeded to keep his record down to 8 retractions so far (see here), by issuing almost 20 shady corrections where he replaced manipulated data with something new and often just as fishy, in a process I like to call ”Voinnetting”. Especially Macchiarini was relatively in luck regarding retractions (recent case here), since his main papers in the journal The Lancet are still standing, despite obvious lies about ethics permits and the patient’s true state in them. This is because the Editor-in-Chief (EiC) of The Lancet, Richard Horton has a strong personal antipathy to the concept of retractions as such (e.g., it took him 11 years to retract the toxic and infamous Wakefield paper on alleged dangers of vaccines, see my report here). Fun fact is: Horton is one of the founding fathers of COPE of 1997. So there is indeed a big customer interest to avoid retraction, certainly among the many cheater scientists, but also among the journal editors who just want to avoid the hustle.

Retractions namely are a pain in the neck. Especially with the rise of the internet portal PubPeer, editors are now inundated with suspicions of data manipulations. Dealing with all this is cumbersome; many journals do nothing at all until someone contacts them in person with evidence. Even then, anonymous complaints are sometimes disregarded straight away. To investigate the misconduct evidence further, journals must invest lots of time and money, when they often can spare neither. Thus, many journals simply sit tight and wait for the results of institutional investigations, if those ever take place. But even when an institution demands a retraction after having investigated and found misconduct, journals still haggle with authors for many months on end and sometimes even drop the issue altogether if the authors oppose the retraction too much.

In fact, some authors even threaten to sue the publisher if a retraction is threatened. The western-blot-breeding specialist Mario Saad did this, but he meanwhile lost a number of manipulated papers to retractions. Some journals operate the policy of never retracting anything unless all authors agree. The EiC of the journal Human Molecular Genetics refused to do anything at all about a manipulated paper despite the retraction request of the corresponding author Philippe Froguel, because the other corresponding author and the obviously guilty party, Kathrin Maedler, did not agree to the retraction (see my report here). Other journals are apparently staff-overstretched, clueless or simply can’t be bothered dealing with their authors’ misconduct. The best example of the latter is the COPE-member Cell: to its EiC Emilie Marcus, data manipulation or absence of original data is no reason at all to retract a “cool” paper, as she recently demonstrated for both Voinnet and Maria Pia Cosma (see this report).  Retraction is a total measure; a retracted paper cannot be cited or referenced anymore. There is no such thing as a partial retraction, though EMBO Press did establish such a weird concept, with the result that they simply applied it to save the most disreputable papers by Cosma from a proper retraction (see this report).

Bottom-line: too many journal editors would sure prefer to hang on some amendment note to a paper and be done with it.

Liz Wager, former COPE chair, who “offered some comments” during the preprint’s writing, explained in the comment section of the preprint:

“I can see many benefits from separating the process of investigating misconduct and alerting readers to unreliable work. However, one important recommendation of this new proposal is that “If misconduct or fraud has occurred, this should be reported on, but such reporting should be considered as distinct from the process of correcting the literature.” But we know that universities in many countries are often unwilling to share information about misconduct investigations, so I hope the authors will provide some more detail (and research institutions will also comment on) how this might work in practice”.

However, the authors who actually invited readers to comment on their preprint, chose not to engage in a critical discussion, but only respond to praise from other publishing executives, as Barbour did here.

I approached Bloom, who then wrote back to me:

“We also try to say emphatically that misconduct needs to be investigated and reported. Our concern is simply that waiting for a full investigation before providing any formalised correction of the literature means that there are long periods when some people know that there are clear problems with an article but most readers do not. Once investigations are complete we would absolutely want that linked to the article, but it is frequently the case that this would be many years after the problems with a paper are identified”.

I then asked Bloom straightforward if the idea was to abolish retractions altogether and replace them with amendments. I received this reply:

“In our view, there is a need to quickly and unmistakably flag papers that are problematic, inaccurate or erroneous. That need is not well met by the current system that requires a full investigation to be completed, behind closed doors, and the full reasons behind any problems determined, before a retraction notice is issued. In our view, in such circumstances it makes more sense to quickly flag the issue – even while people are disagreeing about what might have caused it – and subsequently to add a narrative about the causes. This separates the two aspects – correcting the literature on the one hand, and investigating misconduct on the other – both in time and in terminology. (I wrote something that reflects my own thinking about this issue a few years ago, at http://blogs.plos.org/biologue/2013/11/11/so-tell-me-what-would-you-do/)”.

However, in their provocative preprint Bloom and her co-authors do not propose to flag papers until retraction is sufficiently justified (and conveniently forget that such tool already exists, aptly named Editorial Expression of Concern). They propose to abolish retractions altogether. In the very worst cases, where fraud or patient abuse are officially established, and even the authors gave up defending their opus, the preprint’s authors advice preservation of the paper in full, but decorated with explanatory amendments:

“Wholesale or complete: the article as a whole is considered unreliable in its current form. There may be elements that remain “correct” but large proportions are not. Instead of “retract and replace” as currently practiced by some publishers, we would recommend “retract and republish” with a new DOI that lands on the newer version and makes plain the chain of events. In the case where authors and/or journal may wish to dissociate themselves from it completely, this is fully noted with a full description in the associated narrative and no attempt to insert new text or other content”.

The publisher executives insidiously use as a smokescreen the good and honourable concept of a “living” paper, as opposed to the traditional “static” printed version. There, the authors can add in time their new data or new discussion through amendments, or versioning, to their already published papers. A great idea, and certainly worth pursuing. But it should not be polluted with Voinnetting. Authors caught with manipulated or absent research data should under no circumstances be invited to replace it with something else, through an amendment. If a doctor who lied about ethics approvals or the outcome of the published medical intervention is permitted to “fix” all this with an amendment, actual human lives of real patients are at stake.

Thus, if COPE and the publishers it represent get their will, there will be no retractions in the future. The most outrageously fake papers will get a cautionary amendment, but can still be quoted, referenced, and used for funding and job applications. Barbour and her COPE colleagues sure understand the implications of their proposal to pull the teeth of all research integrity watchdogs. Should we nod in agreement, lulled by the big talk of a “living” paper? Whose ally is COPE anyway? Their guidelines on editorial integrity (which are already full of loopholes) are in fact not worth the paper they are printed on. When called for advice, COPE showed that they are incapable of advising retraction even in clear cases of self-plagiarism (e.g., see here) or even plagiarism (personal communication with a reader). No real revolution can be imposed from above, and the simple fact is: COPE represents the interests of the scholarly publishing industry. Not those of honest scientists.