In this edition of the Journal of the Norwegian Medical Association, Midgard et al. retract their article Optikusnevritt - diagnose, behandling og oppfølging (Optic neuritis - diagnosis, treatment and follow-up), which was published in no. 4/2005 (1, 2). This has been done because a considerable part of the article was in fact a direct translation of an article published in 2002 by Hickman et al. in The Lancet (3).
Retracting an article implies that one acknowledges that it should not have been published in its present form. A correction is therefore insufficient. As the article has already been published, it is remain in print printed. One might think it would be possible to remove electronic traces of the article on the Internet. Most medical journals, however, have opted for another approach: They make publicly known - both in the journal in which the article was published and in databases such as Medline - that the article has been retracted and is not to be used as a professional reference. This is done by publishing a letter of retraction, by labelling the article “retracted” and by creating links in the medical databases between the article and the letter of retraction (4).
There may be many reasons why an entire article has to be retracted. At one end of the scale, the authors of an article”s authors may discover they had inadvertently made an error in interpreting the data, while at the other end of the scale the data may have been fabricated or knowingly manipulated. The article in question has been retracted because it is deemed to constitute plagiarism.
What is plagiarism? According to the definition given by Webster”s, to plagiarize is “to steal and pass off the words or ideas of another as one”s own without crediting the source.” Most people would agree that such behaviour is unacceptable. At the same time, our entire scientific culture is based on the notion of passing on the knowledge, findings and ideas of others. For this reason, there is a grey area between plagiarism and what we perceive as the desirable communication of knowledge. In 1993, the Office of Research Integrity in the United States organized a conference on Plagiarism and the theft of ideas to discuss the bounds of this grey area. The definitions of plagiarism, in particular those provided by Dr. Marcel LaFollette are useful. She identifies four elements that must be present for plagiarism to have taken place (5): The use of another”s words, text, ideas, or illustrations; failure to credit the original (“real”) author; the implication statement that the plagiarist is the author; and failure to seek the original author”s consent. If we apply these criteria, Midgard et al.”s article clearly falls within the definition of plagiarism.
The problem of plagiarism appears to be on the increase, particularly within the realm of specialized academic work. In Norway, considerable attention has recently been given to this problem - we need only to name professional authors such as Karsten Alnæs and Jahn Otto Johansen (6). In the case of the major biomedical journals, plagiarism has become a genuine problem, and countermeasures have been discussed (7). One option is to develop software programs that compare texts and count identical characters and words. This approach is not particularly relevant for the Journal of the Norwegian Medical Association, partly because it would not work with translations, but mostly because one precondition for the Journal”s existence is that we trust that authors will not consciously attempt to deceive us. We cannot act as police.
Why then would anyone take the risk of plagiarizing the work of another person? The scope and extent of scientific publications makes it difficult for anyone to acquire a thorough knowledge, even within a limited field - which in turn makes it easier to avoid getting caught. The fantastic access we now have to literature on the Internet makes it easy to find something to plagiarize. The number of articles a scholar has published - and not only the content - counts in hiring decisions fo academic position, grants for research and can also affect his or her financial standing and position. Some people may be tempted to find quick solutions - e.g. by copying rather than writing for themselves. The entire professional system of merit would therefore be undermined if the rules concerning plagiarism were not enforced.
The problem with plagiarism lies not with the content; rather it is its source that is false. This often causes legal and financial problems relating to copyright laws. In the case of scientific publishing, there is another key issue: A scientific article represents a contribution to a pool of knowledge, and one of the roles held by the scientific journals is to ensure the integrity of the scientific record. This applies both to the content and the person or persons who is/are credited as its author. For this reason, the Vancouver Group (The International Committee of Medical Journal Editors - ICMJE) has drawn up principles for good publishing practice and rules which all the leading biomedical journals follow (8). These require that the biomedical journals also react harshly to, correct and, if necessary, retract unethical and incorrect material - not only out of deference to their own journal, but also in order to uphold the integrity of the pool of medical knowledge (9). An important part of these guidelines is who should be listed as author.
Who the author is will play an important role in how we read a text, particularly if the text is a review article. For Norwegian readers, such an article will have particular clout if the upper echelons of the Norwegian professional expertise are listed as authors, thereby endorsing the quality of the article”s content. Readers of the Journal of the Norwegian Medical Association are interested in reading review articles in Norwegian, adapted to Norwegian conditions. Such articles must of course base themselves on international literature. It is equally obvious that one cannot translate what others have written and pass it off as one”s own. In the case of translations too, it is the original author”s name that is listed. Permission to translate the text must be given, and the translator may only be credited as a translator.
The rules governing authorship and plagiarism have not been made in order to complicate matters; but without them the very foundations of professional medicine would be pulled away from underneath us. It is important that readers can trust the journals. Hence when an article has been retracted, it is the retraction that is preserved for posteriority, not the non-original article.