‘The sincerest form of flattery’: How a math professor discovered his work had been plagiarized

Not long ago, it came to my attention that a 2016 paper by my students and me, “Measuring Semantic Similarity Of Words Using Concept Networks,”  had been plagiarized, verbatim. The offenders had added two words to the title, which now read: “A Novel Methodology For Measuring Semantic Similarity Of Words Using Concept Networks.” Their article was published in the journal Webology, which has been delisted from Scopus, Elsevier’s abstract and citation database. My first impulse was to ignore the transgression, but I asked the question what to do on a closed mailing list read by former colleagues:

I know that imitation is the sincerest form of flattery, and I’m sufficiently flattered, and I know that even a Harvard President was let go for lesser forms of plagiarism, but Integral University of Lucknow is not exactly Harvard. We may already live in a post-truth world (if Trump gets reelected it’s proof positive that we do) and I don’t quite have it in me to destroy the futures of some random students (or perhaps faculty?) in India. The online journal where it appeared is published in Teheran, and does not appear on Beall’s list of predatory journals. What to do?

The responses ran 10-0 in favor of doing something. Here is a typical one: 

you should alert the journal where it appeared. I’m not familiar with Webology, but if it’s a reputable, peer-reviewed journal, authors would be required to attest that the material is original (or sufficiently sourced with references). My assumption is that the editors will want to know that this article failed their pre-publication review process and should be retracted. Bad behavior and laziness should not be rewarded. You have every right to point out that careful, deliberative, accepted and established processes were not followed in this case. Sometimes we need to be the “check” in checks and balances.

The final point I would make is that it’s better to respond to this situation quickly (sooner rather than later). Otherwise, a long & growing chain of references could develop pointing to the plagiarized article. It can be difficult to retract and correct mistakes, but the problem only gets worse (more entrenched) over time. good luck,

I followed up with the journal, but have received no response so far. I have not followed up with Integral University of Lucknow, whose website simply claims theirs is the “best university in India”. The head of the Department of Computer Application, where both plagiarist authors work, is none other than Mohammad Faisal, second “author.” I could complain directly to him, or perhaps to the first author, Farooq Ahmad, who seems to be in charge of quality assurance for “Teaching Learning & Education” at the department.

So far I have not done so (some people on the mailing list suggested complaining to their superiors, which I haven’t done either). But again, I am not interested in destroying their careers. If they continue to serve at the Best University in India so be it. Someone on the mailing list wrote:

Flattery is one thing, plagiarism is something else again.  From what you describe, this is far beyond flattery.  The author had to knowingly do this and from what you say provided nothing of his own. I consider plagiarism worse than theft. Stolen physical objects can be recovered or replaced. Stealing ideas can’t be.

But I strongly disagree: ideas reproduce at zero cost; a stolen idea is an idea that somehow gets broader circulation. What is stolen really is the credit for the idea (more on this later). Also, consider this reply on the mailing list:

Back in the 1980s and 1990s I found myself asking large groups of computer science undergraduates to write simple essays. We often found that Indian students (in particular) were “plagiarising”. When we confronted the students we discovered that this was truly a symptom of a culture clash. The norm at their home schools was to do research, often as a collaborative group, draw their own conclusions but include supporting evidence directly from published work elsewhere without attribution. They had never been told that this was unacceptable, and in their previous experience it was completely normal. So we very simply gave them a written statement of what was considered appropriate in a US setting, including attribution of included materials and acknowledgement of the work of collaborators. There were no complaints and the problem stopped.

The fact that the journal claims to check for plagiarism but didn’t catch this seems to be a serious problem.  I can’t tell from your description whether they have a statement for submitters to sign that asserts they haven’t plagiarized — some journals do and some just take it for granted. Based on my experience, the “take for granted” approach isn’t wise in an international context.

What makes the whole affair interesting is that it sheds some light on the seamy side of academic publishing. Since citations are the coin of the realm, the tracking of citations is a life and death matter both for clueless university administrators, the Vogons who increasingly rule academia, and for the professors caught in the rat race. Back in the day it was “publish or perish,” and one side effect was the proliferation of second- and third-rate journals that helped desperate authors improve their publication count. My own university is guilty of this practice, publishing such a journal at least since the 1970s, so who is to cast the first stone on Webology? 

By the standard eigenvector method (nowadays better known as Google’s PageRank), scientometry is perfectly capable of detecting the lower- and higher-quality journals, and this is what impact factor should all be about. But as soon as IF became important (e.g. in evaluating “research output” by universities and grant agencies), it became subject to Goodhart’s Law, and Retraction Watch already discussed very significant efforts to game it. Back in the day, this method was monopolized by Thomson Reuters whose annual Journal Citation Report was largely in line with the researchers’ own assessment of journal quality, except for weird effects that were stemming mostly from misclassification (what constitutes a comparable range), from weak control for the size of the field in question, and from selection bias (it was a lot of work to start indexing a new journal). 

This last effect, which hurt journals published in foreign/mixed languages particularly, is by now greatly ameliorated by publication shifting online, which enables better language detection, better citation parsing and, at least in principle, better duplicate detection. Today there is an active effort by Clarivate to combat the gaming of IF (a similar arms race is taking place between Google and the SEOs) and I, for one, expect impact rankings to become better over time.

This is something that we should all welcome, especially as 2nd- and 3rd-tier publications play an important role in the research ecosystem (besides letting Vogons pretend they have genuine academic credentials). Peer review is far from perfect, and sometimes genuine jewels get lost in the rush to the highest IF journals. But this is rare (and there is no suggestion that my plagiarized paper was such) and the system is largely self-correcting by the well-known mechanism of rejected authors systematically working toward getting their paper published in the next-highest tier (and walking down the hierarchy until it gets accepted). Séminaire Delange-Pisot-Poitou is no Annals of Mathematics, but it publishes solid work that needs to see the light of the day. Just as in industry, there is a genuine structural need for the “Mittelstand,” and even for “SOHO,” journals.

This brings me to the real issue: How did I know about the plagiarism? The paper that got plagiarized is a genuine 2nd rater – it got published in the proceedings of a workshop sponsored by what was at the time the premier professional society of my field, the Association for Computational Linguistics. ACL proceedings are nowadays overshadowed by NeurIPS, ICLR, and ICML proceedings, but at the time they had better exposure, and our paper garnered 35 citations according to Semantic Scholar, 49 according to Google Scholar. A truly first-rate paper is much less likely to have been stolen – the risk would be too high. The problem is not so much stealing Le Déjeuner des Canotiers as trying to resell it — no fence in their right mind would touch it. I don’t read Webology, nor do I think I have ever come across a paper that was (originally?) published there. So how did I find out? As I explained on the mailing list:

I’m again working on a project I started 15 years ago, and this required looking at an old paper of mine “Eliminating ditransitives” […] – let’s call this paper “A”. One of the reasons why I haven’t continued with the project was that it had no pickup whatsoever, of the 20 citations (Google Scholar) it gathered in the meantime 19 were by my students, who of course knew it perfectly well that the subject is dear to my heart. Now that Google Scholar was showing one independent citation I was curious to check what the citing paper “B“, had to say. Finally, finally, someone is interested in my brilliant contribution to the hard problem that “A” attacked!

Sadly, no such luck. It was just that the paper in Webology was a verbatim copy of “B”, which cited “A” – the one I really care about –  and in the process professors Ahmad and Faisal lifted the citation together with the content. 

Perhaps the moral is that we should be grateful to the ever-growing citation databases. Making these public would be a lot better than the current state of the art, just as it would be better to have a lively IF competition rather than the monopolistic Clarivate, but that’s another story. Given the current “H-index or bust” mentality, I should be actually grateful to Webology for having lifted what was a self-citation (which my university, quite correctly, does not count towards H-index) to the more valuable “independent citation” category. For better or worse, I am flattered. 

Andras Kornai is a professor at the Budapest Institute of Technology and Senior Scientific Advisor at the SZTAKI Computer and Automation Research Institute.

Like Retraction Watch? You can make a tax-deductible contribution to support our work, subscribe to our free daily digest or paid weekly update, follow us on Twitter, like us on Facebook, or add us to your RSS reader. If you find a retraction that’s not in The Retraction Watch Database, you can let us know here. For comments or feedback, email us at [email protected].