Frontiers Christmas Carol

The Swiss academic publisher Frontiers, owned by the EPFL professor Henry Markram and his wife Kamila (plus some investors, including the publishing giant Holtzbrinck, owner of SpringerNature), is not at all prepared to accept being placed by the US librarian Jeffrey Beall on his list of potential, possible, or probable predatory scholarly open-access publishers. The Lausanne-based publishing house, represented by its Executive Editor Frederick Fenter, battles on many fronts to defend its respectability. As I report below, this includes a personal pre-Christmas visit to Beall at his university library in Denver, as well as some publisher-mediated attempts at ensuring editorial and peer review quality, which occasionally appear somewhat ham-fisted.

Frontiers has joined some respectable organizations like the Open Access societies OASPA and DOAJ, got self-enlisted as following editorial ethics guidelines with ICMJE, and finally became member of the Committee on Publication Ethics (COPE). It the latter case, the double function of the Frontiers journal manager Mirjam Curno as COPE council member and trustee might have been only coincidental with Frontiers joining COPE just after Fenter has sacked almost all of Frontiers medical chief editors for demanding editorial independency.

Beall was highly criticized by the academic community for his decision to place Frontiers on his list of predatory publishers. Many scientists stepped forward to defend Frontiers journals, where they act as editors and reviewers, like Daniël Lakens, psychology professor at Eindhoven University of Technology. Beall’s possible scepticism towards Open Access (OA) publishing was used as argument to utterly disregard all his concerns about the predatory behaviour of some OA publishers and journals.

In the heat of dispute, Björn Brembs, neurobiology professor at the University of Regensburg and former Frontiers editor, described Beall and everyone who takes him seriously as “insane” on Twitter:

Frontiers however took a more reasonable and respectful approach. Shortly before Christmas 2015, Fenter wrote to Beall, asking for a meeting “at a location of your choice”, in order to “discuss in further detail how Frontiers operates, and learn from you in turn how you analyse academic publishers.” In follow-up emails, the meeting was set for December 14^th 2015 at Beall’s university library, while Fenter reassured him that Frontiers has “no intention of taking legal action”. Just hours before the meeting, Fenter surprised Beall with an announcement that his colleague Curno will be joining the meeting as well. Beall was first given “a short presentation about Frontiers” and then had his “specific questions and concerns” addressed by Fenter and Curno. Later on, the two Frontiers executives had lots of time to evaluate their pitch after a huge snowstorm hit the airport and had them stuck in Denver for days. Unlike Ebenezer Scrooge, Beall remained unmoved by his pre-Christmas visitors, despite the charitable atmosphere of the yuletide season: Frontiers is still on his list of potential predatory publishers.

It is worth remembering that Curno is also trustee and board member of COPE. In this regards, Virginia Barbour, Chair of COPE, has written to me:

“We had no knowledge of this visit, nor would I expect to – this an internal issue for the publisher, as would be any other such meeting between a publisher and an outside individual or group”.

In parallel, Frontiers was cleaning its house. After being criticised for negligent attitude to conflicts of interests (COI) and employing academically immature editors and reviewers, Frontiers so-called “editorial office”, which is actually publisher’s office, took sweeping actions. A postdoctoral scientist was sacked as reviewer for not having enough post-PhD years in academia. Natalie Matosin  was re-installed after Lakens and other Frontiers editors expressed their protest over Twitter.

Next, Frontiers editorial office either decided to widen or completely misunderstood the concept of the conflict of interests (COI). Usually, COI is present when manuscript’s authors on one side and its editors or reviewers on the other side are personally or professionally connected to each other. Personal relationships or professional competitions, collaborations or even collegial proximity can severely undermine the impartiality of peer review and editorial decisions. Exemplary for COI is a Frontiers in Medicine publication by the controversial Italian cancer researcher (and Frontiers editor) Alfredo Fusco, which was peer reviewed by his institutional colleague.

Frontiers now apparently decided that a COI is also constituted by an editor and a reviewer working in close proximity to each other, even if none of them has any relevant connection to manuscript’s authors whatsoever.

An experienced Frontiers editor Guillaume Rousselet, senior lecturer in psychology at the University of Glasgow, was asked to revoke a reviewer, whom he assigned to a manuscript submission in Frontiers in Psychology, for being “affiliated with the same institution” as himself. Rousselet was told by Claire Martin, Journal Operations Specialist: “as per our review guidelines this constitutes what could be perceived as a potential conflict of interest”. The email ended with a “QUICK FACT: Frontiers in Psychology ranks third in Impact Factor and is the leading gold OA journal in the field in volume”.

After Rousselet refused to remove this reviewer, he was offered by a Journal Operations Assistant, Kirstin Sonne, that he can keep that reviewer if he provides a public statement: “The reviewer and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review”. Rousselet was further encouraged “to assign a third reviewer to the manuscript, in order to erase any doubt that the review did not meet the highest standards of impartiality and objectivity”. He did no such thing, in fact, Rousselet instead resigned from his positions as Frontiers journals editor altogether.

I approached the journal’s EiC, Axel Cleeremans and the responsible chief specialty editor, Philippe Schyns for a comment. Both of them were rather communicative over phone and email, but categorically declined to be quoted for this blog. Shyns however indicated to me on the phone that he also has resigned from his Frontiers editorial post.

The interference by Frontiers publisher’s office in Rousselet’s case is particularly striking given the fact that Frontiers handling editors occasionally function as reviewers of the same paper. This happens at Frontiers surprisingly often; apparently no-one there saw this contraption as a possible COI. Here as an example: a Frontiers in Computational Neuroscience original research article, (criticised on PubPeer as a re-use of published material), was reviewed by two referees, one of whom was its handling editor and journal’s Editor-in-Chief (EiC), Misha Tsodyks, professor at Weizmann Institute of Science, Israel. It is literally mindboggling why Tsodyks had did not have any COI by acting simultaneously as editor and as reviewer, while supervising himself as EiC doing both tasks, but Rousselet and his institutional colleague did.

Shortly after that incident, Rousselet described to me at length his Frontiers experience as editor and reviewer:

“At the start, I really enjoyed the reviewing forum, which could give, at least in theory, authors, reviewers, and editors, a space to discuss research and reach consensus about a paper. […] In practice, it works in certain situations, and fails miserably in others.

As author & reviewer, my experience is very clear: editors do not engage in the forum to moderate the discussions. I do that as editor, but it is a painful process. You have to scroll through a growing webpage containing comments and replies, and replies to the replies. […]. This is a major complaint I have heard from other reviewers and authors: the forum can turn into a collection of fighting cages (one cage/tab per reviewer), in which certain authors try to wear down reviewers to submission (pressing the finalise review button) by submitting many small incremental updates to their papers. Instead of the 1 to 3 reviewing rounds you would get in other journals, certain authors would easily submit more than 5 different revisions. […] In some situations, authors are trying to avoid making painful changes, so everything is argued and revised versions with minimal changes are submitted. In other cases, authors are willing to make more drastic changes, but they require more help from the reviewers. So it goes back and forth many times. In this last situation, as editor I’ve had many reviewers leaving the forum, because although they think the paper would be publishable eventually, they realise it would take them too much time to bring it to a sufficient level of quality. […]

As editor, it is easy and tempting to do very little, because the forum is all geared up with decision buttons and pre-populated emails. So in practice, I think most editors just let reviewers and authors battle in the forum. Once two reviewers have finalised their reviews, the editor pretty much has to accept the paper. However, the system has some pernicious features: a paper cannot be accepted unless 2 reviewers have finalised their reviews. So if a reviewer is very late, or if as editor you think the paper should be published based on one review, you have to contact the publishing office to make that decision [apparently here, the publisher decides, not the chief editor! –LS]. Also, if reviewers leave the forum (which is how reviewers indicate the paper should be rejected), then the editor can keep inviting new reviewers until two finalise their reviews, and the paper can get published […].

I also heard from editors at other Frontiers journals that they have been explicitly encouraged by their EiC to be lenient, which aggravate the perception of inferior research quality in Frontiers journals. Once reviewers have accepted a paper, as editor I have little choice but to accept it too, even if I have important points yet to be addressed. There is no page in the forum for a discussion between the editor and the reviewers. The editor has to use the reviewers’ tabs to make comments. These tabs are closed once the reviewers finalise their reviews”.

It seems therefore, there is little quality control or editorial supervision happening at the Frontiers journals. Everything stands and falls with the integrity and dedication (or the lack of it) of the individual editors and reviewers. Dennis Eckmeier, neuroscientist at the Champalimaud Foundation in Lisbon, has been Frontiers neuroscience reviewer for three years, and his experience with the reviewer discussion forum was:

“very positive. It is a quick and informal, but recorded, way of interaction. In many cases misunderstandings could be clarified through the forum that would not have been addressed in a traditional review system. These interactions also highlighted unclear parts of the manuscript, which was usually addressed by the authors”.

Eckmeier cautioned though that many submitted manuscripts were of very low quality and gave examples for some cases:

“There were about 2-3 re-submissions, before both reviewers endorsed publication. In one case both reviewers agreed that for publication they needed to add more data. The data were provided and both reviewers endorsed publication.  In one case both reviewers agreed that the work was not publishable. According to the message in the review forum, the authors retracted the submission”.

As author with Frontiers, Eckmeier experienced a reviewer whose “review was basically useless” and who “endorsed publication already after I promised to revise the introduction”. But he also reported that “the second reviewer and the editor were knowledgeable, constructive, and patient”.

Maybe the best way for Frontiers to get removed from Beall’s list would be to sack all incompetent and COI-disregarding editors and reviewers, instead of the dedicated ones who demanded editorial independence, quality control and ethical guidelines. It could also help, if Frontiers chief editors would properly supervise their own journals, instead of leaving this task with the publisher’s junior managers. Many of these field outsiders simply lack relevant research field competence or even, with only with a bachelor degree, the sufficient academic credentials to perform such tasks. However, they are apparently quite busy with doing the jobs of EiC and other senior editors, on behalf of the publisher.

Rousselet and other researchers have been demanding from Frontiers a retraction of a certain article by a retired neuro-oncologist, Ivo P Janecka, titled Sensing risk, fearing uncertainty: systems science approach to change and described by many scientists as utterly nonsensical. As I reported, it was edited (and reviewed!) by a totally field-unrelated medical surgeon Tobias Alecio Mattei, who according to his own employment record, is not and never has been an active researcher, in any field. How he came to be associate editor at Frontiers in Computational Neuroscience, is anyone’s guess. Yet also here, the journal’s EiC Tsodyks apparently chose not to interfere or even participate in the discussion. Instead, Rousselet was told by the Journal Operations Specialist, Martina Haller, that:

“The manuscript has undergone peer review as part of the Research Topic “Application of Nonlinear Analysis to the Study of Complex Systems in Neuroscience and Behavioral Research” and was approved by the Topic Editor, Dr Mattei. Our investigation did not identify plagiarism, reporting of unethical research, established misconduct, data manipulation, data fabrication or breach of third-party legal rights which would substantiate the need for retraction of this publication”.

Good to know that at least the publisher’s office thought the Janecka paper was scientifically sound, and the handling editor Mattei a fully qualified neuroscience researcher. Maybe Fenter and Curno should have discussed this Frontiers paper with Beall, to get some advice on their, as Fenter put it, “common goal: to hold academic publishing to the highest standards possible”.

Update 24.01.2015; 19:30. The figures from the above mentioned paper by Ivo P. Janecka in Front. Comput. Neurosci., 31 March 2014, were later on apparently re-used in a later 2015 publication by same author. Details in my comment below   –LS