This is a complicated case which involves possible plagiarism, double submission and reviewer misconduct. The timeline is as follows:
- In year n, a paper P1 authored by A1 and A2 was published in the English language journal X. The paper describes a theoretical analysis of a particular phenomenon.
- In year n+6, paper P2 was published in a non-English language outlet by authors A3 et al, which cites P1, but carries essentially the same scientific message.
- In year n+8, A3 et al submitted paper P3 to conference Y without referencing P1 or P2. The main content of the paper was essentially the same as that in P1. This paper was awarded a best student paper prize at the conference and journal X, which has an arrangement with conference Y to fast track “extended” versions of best papers , invited submission of such an extended version. Journal Z, unbeknown to journal X, also invited a paper to be published in journal Z based on P3.
- In year n+8, paper P4 was published in journal Z; the paper did not cite P1 and was only a very minor extension of P2; under journal X’s rules, P4 would have been rejected as not being sufficiently different from P3 since conference Y is regarded as archival in the field. The editors-in-chief of journal Z were two of the reviewers of P3.
- In November year n+8, A3 et al submitted paper P5 to journal X. P5 has the same theoretical content as papers P1–P4, but also has a new experimental section, which does make a new contribution. P5 does reference P1, but only incidentally and does not properly acknowledge that the theoretical content of P5 has previously appeared in P1 (or indeed in P2–P4). The editor-in-chief of journal X was not alerted to the overlap with the previous papers by the two reviewers (who were, in fact, the editors of journal Z). It is not a coincidence that the reviewers of P5 were the same as for P3 since this is part of the journal’s fast tracking process. The editor-in-chief of journal X accepted the paper and it was published in year n+9.
- Around 6 years later, journal X (with two new editors-in-chief) received two independent complaints that P5 contains large sections of material plagiarised from P1, noting that although P1 is referenced, the reference is not sufficient. Journal X starts investigations. Two editorial board members and an independent reviewer confirm the facts as stated above. One of the complaints was submitted in the form of a paper for publication; at present, this has not been sent out for review but is simply being treated as additional evidence/confirmation of plagiarism. (We have recently discovered that this paper has been posted on a web site devoted to plagiarism discussions.)
Other points:
- The co-authorship has changed over the papers P2–P5. A3 is constant (although not always first author) but the “et al” changes. Of particular note is that the authors of P3 are not a subset of P5, despite the fact that content-wise P3 is a subset of P5.
- P5 has become highly cited and A3, although junior at the time of submission of P2–P5, has become well known with many papers and sits on the editorial board of journals. This should of course not affect our action, but it is worth noting that our decision could have a significant impact.
- A3 has admitted in a non-English language web site that he was invited to submit a revised version of P2 to journals X and Y.
- One of the complainants has just pointed us to another publication by A3 et al in a foreign language journal which again appears to have a high degree of overlap, published in the same year as P3 and P4. At the time of submission of this case we have not yet contacted A3 or the referees of P5.
Questions for COPE:
- It is clear that the theoretical part of P5 is effectively plagiarised as the reference to P1 is insufficient. How severe should our response be?
- There is some element of double submission (P4, P5): is this worth pursuing?
- Should we take any action against the reviewers of P5 who have arguably acted unethically, or at least less than ideally?
Other comments are welcome on this complex case.
The Forum discussed this complicated case and agreed that there was some culpability on the part of the editors, given that authors A1 and A2’s work was plagiarised, there was redundant publication and possibly dual submission. The advice was to contact authors A1 and A2 and solicit their opinion. This will give the editor a stronger case against author A3. For multiple papers, the editor should assess the level of overlap and consider retraction of the second paper if the overlap is unacceptable. A3 was a junior author but was also the supervisor on the paper but it may be that publication practices were not correctly understood. The advice from the Forum was to address the plagiarism issue. If there is an acceptable explanation, then the editor should consider the less serious offense of redundant publication. Or the editor may wish to issue a correction, mentioning that papers P1 and P2 should have been cited in the other articles.
We sent a letter to the author A3, setting out our concerns and asking for a response. We also followed the advice from COPE to contact the authors (A1 and A2) of the allegedly plagiarised article. One of these two authors (A1) replied (the other is now emeritus) and said that they had already been contacted by A3 asking for their help in defending the charges. A1 confirmed our view—essentially that the paper did replicate ideas without proper acknowledgement. He was perhaps inclined to be lenient to a junior researcher and regard it as ‘unintentional plagiarism’ but left the decision to us of course. A3 replied, heavily defending their position both on the count of plagiarism and that of double submission. This response was reviewed by the Editors-in-Chief and the two editorial board members who had advised initially as well as the same external referee.
Our conclusion was that the charge of double submission could be dropped (since P3 contains sufficient novel material) but there was definitely inadequate reference to the earlier paper in the theoretical part of the paper. We also decided that there was sufficient novel (and interesting and important) novel material in the paper so that a retraction was inappropriate. We therefore wrote to A3 et al again asking them to sign a short note to be published in the journal acknowledging the inadequate reference to prior work, and apologising for this (we also said that if no response was received we would publish the note in the names of the editor in chiefs. This note also (implicitly) acknowledges that no citation was made to P1 in P3 published in conference Y.
We have recently had a response from A3 et al, agreeing in principle to sign an apology note to be published, but disputing the exact extent of what was to be apologised for. We are currently reviewing these questions within the Editorial Board and will respond to A3 et al shortly.
We also considered the issue of possible unethical behaviour by the editors of journal Z who published paper P4 and who reviewed P3 for journal X. We have decided not to pursue this further owing to lack of hard evidence.
It seems that the substantive issues have now been addressed and the case can probably be regarded as closed (subject to our final editorial board review).