Plagiarism is an issue that unfortunately affects all publishers at one time or another. Recognizing it as a growing concern particularly in the viral age, the Committee on Publication Ethics (COPE) chose it as the theme of its annual London seminar, held on 26 March 2010.

Plagiarism in the Electronic Age

Harold Sox, of Dartmouth Medical School and the Dartmouth Institute and former editor of the Annals of Internal Medicine, began the day with a presentation on plagiarism in the electronic age. He admitted that he found very little in the scientific literature when researching the topic, but he discussed four real-life cases of plagiarism that he came across. After each case, Sox explained the “moral of the story” and how authors can avoid falling into these frequent traps. Outwardly simple actions, including checking proofs for similar texts and assuming personal responsibility for attaining permissions where necessary, were mentioned. The outcomes of each case were also given. Inevitably, they involved investigations held by the relevant institutions and retractions. However, Sox described a case from his own journal in which the author plagiarized a rival’s work; after an apology and a correction were issued, the article remained published with the approval of the original author— demonstrating that cooperation even with rivals sometimes has a place in publishing. Nevertheless, Sox did say that journal editors must retract fraudulent work and that academic leaders should bring more shame to those found to have committed fraud.

Looking at the frequency of plagiarized published material in the New England Journal of Medicine, Sox found that under the leadership of Editor-in-Chief Jeffrey Drazen, there has been only one case out of 70,000 submissions so far in which an article was published that proved to be plagiarized. That is impressive, but how many such articles go undetected? A survey of researchers found that anywhere between 0.02% and 0.2% of published articles could be fraudulent, on the basis of a sample of 21 surveys of scientists.¹ With manuscripts being published every day and more articles perhaps being found to be plagiarized, those numbers may soon be obsolete, and similar, more recent studies will be needed.

Sox reiterated throughout that the scientific record is “sacrosanct and continuously being built upon”. Therefore, fixing errors is the responsibility of the entire scientific community. But given high numbers of submissions, how can publishers prevent plagiarized articles from being published? Inculcating high morals into scientists so that they do not consider plagiarism as an option is one method. So is detecting misconduct and performing investigations. But should routine screening of manuscripts be introduced by publishers as a matter of course? To that end, and perhaps controversially, Sox described how every article could be considered a possible case of plagiarism until proved otherwise. In the following discussion, many felt that to be a step too far; authors might become too apprehensive to submit their research anywhere for fear of being accused of plagiarism. It would also require a lot of work by editors and publishers. Another speaker helpfully discussed a tool that could aid in plagiarism detection.

An Update on CrossCheck

At the 2008 COPE seminar, a new tool was announced that would detect plagiarized material by using a database of content published online.² Kirsty Meddings, of CrossRef (a nonprofit association³), updated attendees on how the service has progressed since its inception. More than 40 million records are available on CrossRef,⁴ and more than 3000 publishers and societies have signed up as participants. CrossRef identified plagiarism as a growing concern in 2006, and CrossCheck was launched in June 2008. Using an article’s unique DOI, users can compare a submitted paper with the full text of published articles to see whether there are overlapping sections.

The tool seemed familiar to most delegates. The use of that and similar services not only deters would-be plagiarists but protects the reputations of publishers and enhances the value of the submission and publishing process. However, despite the accuracy of the tool, Meddings emphasized that it is just a tool. Editors must check the report that CrossCheck generates for plagiarism and use their judgment and experience where necessary.

Cultural Differences in Plagiarism

The next speaker, Vasiliy Vlassov, of the Moscow Medical Academy in Russia, described plagiarism and scientific fraud as global problems beset by cultural differences, inasmuch as attitudes toward intellectual property differ among countries. Using his native Russia as an example, Vlassov said that no Russian state-funding body has a facility for managing cases of research misconduct. Indeed, only in 2009 did Moscow State University introduce a code of ethics that included plagiarism in its remit. Perhaps treading in political waters, Vlassov stated that if text ownership is a new, Western concept, other cultures will differ from the West.

Vlassov noted that in 2006, the chairman of the Russian Higher Attestation Commission stated that up to one-third of dissertations were bought, and plagiarism was common.⁵ In an entertaining aside, noting that no quarter seemed untouched by plagiarism, he described how former President Vladimir Putin had been accused of plagiarizing part of his thesis as a student.⁶ Reiterating Sox’s point, educating authors is important, especially if they have a poor grasp of English. Therefore, institutions and educators have their part to play in ensuring that correct ethics are instilled in future authors when they begin to publish their research.

Panel Discussion

The COPE organizers introduced a panel discussion to end the proceedings. They invited attendees to submit questions to presenters and members of the committee, and this provided the day’s greatest debate on publishers’ and editors’ experiences with plagiarism and on what to do when confronted with it. Soon clear were the many gray areas of plagiarism, the forms it can take, and the actions that editors should pursue in these situations. Self-plagiarism in particular was a major sticking point for delegates. Can authors plagiarize their own work? Ultimately, this comes down to novelty rather than ethical behavior. There may be copyright issues, but authors reduce the originality of their work if they submit the same results in different articles.

On the whole, attendees listened to and accepted the advice given by the panel. However, there was one cause of disagreement: cases in which authors submit to several publishers and succeed in having an article published in more than one source. COPE members felt that although such authors had clearly been unethical, the article itself should not be completely removed from the literature, because the importance of the work itself is paramount. Instead, the first journal found to have published the article should issue a correction and not remove it, and other affected journals should retract it. It was emphasized that retraction should not be used to punish authors.⁷ Attendees disagreed with this course of action: if authors lie, all copies of the article should be removed, because this behavior cannot be tolerated. The point was left open, with most delegates on one side and COPE apparently on the other.

It was clear that the onus falls on both authors and editors to prevent the publication of plagiarized material. Authors must be educated to ensure that they do not “accidentally” duplicate published text, and they should have misgivings about doing so from a moral standpoint. For editors, all the tools that are available cannot absolve them from responsibility for not only taking steps to prevent publishing plagiarized material but enforcing correct and appropriate action when such articles “slip through the net”. Through seminars like COPE’s, publishers will be able to take a collective stand against unethical behavior and send the same clear message to all authors, in whatever discipline or country.

Acknowledgments

I thank Alexander Kroll and Frances Mulvany for their assistance in editing this article.

References

1. Fanelli D. How many scientists fabricate and falsify research? A systematic review and meta-analysis of survey data. PLoS ONE. 2009;4(5):e5738.
2. Hrynaszkiewicz I. Privacy, plagiarism, and publication: report from the COPE seminar 2008. Sci Ed. 2008;31(5):159-161.
3. CrossRef: History/mission. [Internet] [cited 4 April 2010]. Available from: www.crossref.org/01company/02history.html#.
4. CrossRef: News release: 40 million CrossRef DOIs preserve the record of scholarship. [Internet] [cited 4 April 2010]. Available from: www.crossref.org/01company/pr/news020210.html.
5. Kommersant. [Internet] [cited 5 April 2010]. Available from: www.kommersant.com/pda/doc.asp?id=662935.
6. Times Online: Putin accused of plagiarising his PhD thesis. [Internet] [cited 28 March 2010]. Available from: www.timesonline.co.uk/tol/news/world/article695235.ece.
7. Wager E, Barbour V, Yentis S, Kleinert S on behalf of COPE Council. Retractions: guidance from the Committee on Publication Ethics (COPE). [Internet] [cited 4 April 2010]. Available from: publicationethics.org/files/u661/Retractions_COPE_gline_final_3_Sept_09__2_.pdf.