A peer reviewer discovers plagiarism

In the following, certain details are anonymized to protect the confidentiality of an academic journal's peer review process.

The Story 

Last month, I was asked to serve as a peer reviewer for a manuscript submitted to a certain European academic journal, published by Springer Nature.  I have no prior history with that journal, either as an author, peer-reviewer, or even a reader.  Nonetheless, the abstract of the manuscript was sufficiently intriguing that I agreed. 

I began working on my peer review.  I found that the paper was of appallingly low quality.  I have intermittently served as a peer reviewer for various journals over the last quarter century, but I have never, until now, thought to unequivocally recommend outright rejection of a manuscript.  I began drafting extensive comments to return to the author, with constructive comments on how the paper's quality could be raised to be of publishable value.  One of the points I made was the poor engagement with prior literature, especially with the manuscript's claim that previous investigators achieved little success in predicting certain features of the world.  To challenge the author, I decided to see if I could find papers that contradicted the claim about poor prediction success.  I started to find papers that claimed good prediction success, that were not cited in the manuscript.  I then had the idea of looking up the handful of papers that were cited in the manuscript in Google Scholar, to see who else had cited them.

Lo and behold, I quickly discovered a paper published the previous October in a different journal, published by MDPI, that was nearly identical to the manuscript that I was reviewing.  A few superficial changes were made, but vast portions were word-for-word identical, including the tables and graphs.  I was shocked.  The Springer journal I was reviewing for has, on its journal policies page, a warning that all manuscripts would be checked using CrossRef's similarity checker, using iThenticate, for plagiarism.  And I confirmed that the "victim" paper was indeed in the CrossRef database, so if such a check had been run, the manuscript should have been flagged prior to it being sent out for peer review.  

I next thought about contacting the author of the "victim" paper, but I discovered an online obituary for him, indicating he had passed in December, the month before the plagiarizing manuscript was submitted to the Springer journal!  I submitted my findings to the editor, using the journal's manuscript submission portal.  About a week later I emailed him separately, as I began to worry that other peer reviewers might be wasting time on this plagiarized manuscript.

The editor replied, and indicated he would reject the manuscript.  Indeed, he did so, but in the formal decision letter (for which the reviewers were cc'd), the reports of all the reviewers were included.  Thus I was right that we had all wasted considerable time on the manuscript.  There were a total of 3 of us reviewers, and I was the only one who had discovered the plagiarism.  One of the other reviewers recommended publication, but leveled lengthy criticism on the paper; the other was much more negative, saying the paper was "potentially interesting" but leveled numerous additional criticisms.  They both complained about many of the same things I did (in my draft comments, which were not shared with the journal) though I had also identified other issues not mentioned by them (again, not shared with the journal).

After the rejection, I decided not to let my draft peer review comments go to waste, so I posted them on PubPeer as comments for the published MDPI paper (which I've been calling the "victim" paper).  Unfortunately, the author is no longer with us, and unable to reply to my criticisms.

Reflections on the Incident

Several things went wrong here.

1. The MDPI journal should not have published the original paper.  It likely would not have passed peer review at the Springer journal where it was also submitted.  Moreover, according to Wikipedia, MDPI is a controversial publisher that has been suspected of being a predatory journal.  From now on, I will refuse to submit to, or peer review for, any MDPI journal.
2. The plagiarizer should not have picked such a terrible paper to plagiarize.  Chances of getting it published would be low, except at another predatory journal.  But what was the point of plagiarizing it in the first place?  Why plagiarize?
3. I strongly suspect that the Springer journal did not follow its own policy of running a plagiarism check on the submitted manuscript.
4. The peer reviewers should and did assume that the journal would follow its policies, and none of us attempted to check the manuscript for plagiarism.  Two of the three peer reviews therefore failed to detect the plagiarism, and the third one (me) did so but not deliberately; I stumbled upon the "victim" paper while trying to accomplish an unrelated task.

Peer review is a purely voluntary effort, and journals (including for-profit journals, such as was the case here) rely on the goodwill, skill, and effort of volunteers to perform a thorough manuscript review.  When reviewers are asked to review plagiarized manuscripts, it results in a colossal waste of time and energy.  Journals often have trouble convincing researchers to agree to review manuscripts, and when incidents like this occur, it decreases the reputation of the journal and creates resentment on the part of reviewers whose time was wasted.