New Delhi: An unpublished research manuscript is at the heart of a battle being fought by two Indian scientists to establish that they had done the work for which credit went to a counterpart in a British institution. The rare case also exposes the intense competition in the world of scientific research and the ethical conflicts it can sometimes trigger.
Karmeshu, 59, the scientist who goes by one name, alleges that Demetres Kouvatsos, a professor of information systems at the University of Bradford in the UK, copied ideas from a paper he and his doctoral student, Shachi Sharma, had submitted to Performance Evaluation, a peer-reviewed journal that rejected the manuscript.
Karmeshu, who is a professor in the New Delhi-based Jawaharlal Nehru University’s school of computer and systems sciences, and Sharma claim that Kouvatsos added a few incremental equations to their research and presented a paper—along with his student Salam Adli Assi—at Euro-NGI 2007, a conference held in Norway. The conference was sponsored by the Institute of Electrical and Electronics Engineers Inc. (IEEE), a non-profit organization working among engineers and scientists, with a presence in more than 150 countries and a membership of 35,000.
Crying foul: Karmeshu, a professor at Jawaharlal Nehru University, alleges there are 23 similarities between the two research papers. Photograph: Harikrishna Katragadda / Mint
Kouvatsos, who has denied plagiarism, didn’t respond to emailed queries from Mint on the issue. He is making a formal appeal against a ruling by IEEE in June that he had violated the organization’s principles and would be banned from publishing in its journals for a year.
To be sure, cases of plagiarism in academic and scientific research are not uncommon. As recently as in February, P. Chiranjeevi, a professor at Sri Venkateswara University, Tirupati, was indicted by an internal committee of plagiarizing and falsifying more than 70 research papers. Elsevier BV, the largest scientific publisher in the world, has since retracted 13 of his articles published by group journals. Although Chiranjeevi has denied any wrongdoing, the committee barred him from holding administrative posts or taking on new research students.
K.R. Rao, associate editor of Current Science, an Indian research journal, says he has detected nearly 80 plagiarized submissions to his publication in the past three years. Even so, Prof. Karmeshu’s “case is quite rare and I can’t think of any?recent?precedent”,?he?adds.
The case involves an ethical conflict stemming from Kouvatsos having been one of the three reviewers who read the paper submitted by Prof. Karmeshu and Sharma to Performance Evaluation, which publishes research papers on the performance of computers, computer systems, and telecommunications. The journal rejected the Indian paper “in its present form,” about 14 months after it was submitted in July 2005 for publication.
The paper broadly suggested, for the first time ever, the authors claim, a way to mathematically analyse communication networks, such as Internet traffic, using an approach called Tsallis entropy—traditionally used to analyse a range of phenomena such as the magnitude of earthquakes over a region; the number of species in a genus and even the popularity of websites.
The process of peer review rests on trust and good faith. Reviewers are privy to a new idea, and if a paper is rejected, they are the only ones who would be aware of the contents, besides the authors.
Seeking amends: Shachi Sharma, who co-authored the research paper along with Prof. Karmeshu, and is now a researcher with IBM India. Photograph: Harikrishna Katragadda / Mint
Normally, manuscripts submitted for publication to a journal are reviewed by two or three experts, whose identities are known only to the editors. The reviewers themselves don’t know whose papers they are evaluating. Although the editor has the final say in deciding a manuscript’s fate, reviewers’ suggestions play a vital role.
“Rejection of a paper doesn’t always reflect the quality or originality of the ideas in it,” says Parthasarthee Banerjee, acting director of the National Institute of Science, Technology and Development Studies, a unit of the Council of Scientific and Industrial Research. “Many of Einstein’s papers were rejected, so which journal you submit, your presentation, play a crucial role in getting a work published.”
Prof. Karmeshu and Sharma accepted the rejection and got on with other research. Until they discovered a paper by Kouvatsos and Assi presented at IEEE event almost a year after the rejection, and shot off an email to Werner Bux, then editor-in-chief of Performance Evaluation, requesting him to investigate a case of “suspected plagiarism”.
Mint has viewed the entire email correspondence between Prof. Karmeshu, Kouvatsos and Bux, but has received independent responses from neither Kouvatsos nor Bux. “This paper deals with the same problem, using the same approach as we had presented in our paper to Performance Evaluation. Their paper includes similar mathematical expression at several places; two of the figures are same as in our paper; identical algorithms and essentially same analytical framework have been used,” Prof. Karmeshu wrote in an email to Bux on 17 August 2007.
Prof. Karmeshu is a PhD in statistical physics from the University of Delhi and a 1993 recipient of the Bhatnagar Award, named after CSIR founder-director Shanti Swarup Bhatnagar, given annually by the government in recognition of outstanding scientific achievement. Sharma is now a researcher at IBM India Pvt. Ltd in New Delhi.
“We hadn’t submitted the paper anywhere else and we were suspicious that Kouvatsos had access to these papers by virtue of being a reviewer of our paper,” says Prof. Karmeshu. His instinct proved to be correct. Kouvatsos was indeed one of the three reviewers of the paper—and the only one who had recommended its acceptance, subject to a “revision to explain further the motivation for the work and improve the presentation.” The other two had recommended its rejection.
‘Citations are everything’
Credit for their work is extremely important to those who dabble in scientific research. Fiercely possessive about their ideas and the ownership of work resulting from them, researchers are particularly sensitive about citations in scientific publications that could sometimes hold the key to progress in their careers.
“Citations are everything to a scientist,” says Nandula Raghuram, secretary of the Society for Scientific Values (SSV), an India-based independent watchdog that investigates cases of scientific misconduct.
“Awards, performance incentives, memberships to top academies, research grants, all relate to your publications and the number of times your work is cited,” added Raghuram, who teaches biotechnology at the Guru Gobind Singh Indraprastha University in New Delhi.
In a detailed response to Bux, who initiated an inquiry into the plagiarism charges, Kouvatsos said the Performance Evaluation topic editor who invited him to review the paper in 2005 had been aware that it was similar in scope to his student Assi’s research work. Assi had never had access to the paper, he said.
Bux ruled that Kouvatsos “behaved inappropriately”, and that the journal would not use him as a reviewer. “Kouvatsos was clearly aware of the problematic relationship of the two papers and I cannot view his behaviour as a mere oversight or thoughtlessness,” Bux wrote in an email to Kouvatsos and Prof. Karmeshu on 10 December last year.
He, however, didn’t explicitly say whether Kouvatsos was guilty of plagiarism on grounds that it would be inappropriate for him to do so as Performance Evaluation had published neither paper.
On its part, IEEE wrote to Prof. Karmeshu on 11 June this year that it has recommended that Kouvatsos be prohibited from publishing his works in all IEEE journals for a year, and the article carry a note in its online version saying that it had violated the organization’s principles.
IEEE also says any reader accessing the article would be directed to the “original paper”. Most allegations of plagiarism are settled this way, except in this case, the so-called original paper doesn’t exist in public domain.
“And that’s the hurtful part, because it doesn’t help our case,” says Prof. Karmeshu. “Typically, matters are solved when the original paper is referenced, but in our case, the paper was rejected by Performance Evaluation, and thus doesn’t exist anywhere. Where will IEEE refer this paper to? How will anybody know that the paper is based on our ideas?” The researchers wants IEEE to withdraw Kouvatsos’s paper so that they can rework theirs and submit elsewhere. IEEE has no procedure to do that. The case may drag on.
Ian Palmer, dean of the School of Informatics at the University of Bradford, said Kouvatsos is currently making a formal appeal against IEEE and it would, therefore, be inappropriate to comment until this process is concluded.
Bux is a senior scientist at the IBM Zurich Research Laboratory in Switzerland, a Fellow of IEEE and no longer the editor-in-chief of Performance Evaluation.
“The IEEE hasn’t withdrawn that paper. So even if I revise my work and submit it to another journal, the reviewers will say, “Hey, all this is known and published work,” said Prof. Karmeshu. Anthony Vengraitis, from IEEE’s intellectual properties division, confirmed the society’s decision in an email but declined to answer more questions, citing third-party confidentiality issues. He referred Mint to a manual that sums up IEEE’s approach to plagiarism cases.
Nowhere does the manual discuss conditions under which a paper is withdrawn. Level-3 misconduct, of which IEEE says Kouvatsos has been found guilty, involves “uncredited verbatim copying of individual elements”, and that means cases where paragraphs and individual sentences up to 20% of the original paper have been copied.
Prof. Karmeshu alleges 23 similarities between the two papers. “There was an error in our manuscript which we discovered later. Even that has been reproduced,” he says. “There are very similar paragraphs, equations and graphs in the Kouvatsos-Assi paper, but with different variables (for instance, an alpha instead of a beta).”
Kouvatsos, in his correspondences with Bux and Prof. Karmeshu, denied any wrongdoing. There are only “generic, coincidental and trivial” similarities between the papers, he argued, mainly because Prof. Karmeshu’s approach was a special case, a subset, so to say, of the work already being done by Kouvatsos and his student. He cites two related papers on the subject in 2000 and 2002, but Prof. Karmeshu doesn’t agree. “There’s no relationship between our work,” Prof. Karmeshu said. “If there were, there would be previous citations and that would be reflected in the literature. These citations are irrelevant to the ongoing dispute,” he said.
After Bux’s verdict, Kouvatsos suggested to Prof. Karmeshu in an email that they publish a joint paper that could be incorporated into a book, according to the Indian researcher. “It’s an unacceptable proposition,” Prof. Karmeshu said.
Unlike the US, which has a government-managed Office of Ethics and Research Integrity, or Europe, which has an Ethics Commission to take up plagiarism issues, India doesn’t have an institution to deal with such disputes.
“Had there been such a body, we could have approached that organization,” said Raghuram at the Society of Scientific Values. “It’s nearly two years since this dispute has been on, and it requires considerable individual motivation to fight for your rights.”
Some scientists say such cases are inevitable. “As long as there is science, plagiarism and misconduct will continue,” said P. Balram, director of the Indian Institute of Science, Bangalore. “There are technologies and software that detect plagiarism, and compare similarities between text, but beyond a point you need good old faith,” he said.