Vigilant Science Journal Editors Fight Redundancy

They strive to ward off authors who rehash findings in order to rack up lengthy lists of publishing credits In today's competitive job market, some scientists may be tempted to add heft to their c.v.'s by stretching their research a little, engaging in practices referred to by journal editors as redundant publication. Redundancy--attempts to get two or more articles out of the research for one--has always existed in some form or other, editors say. But, spurred by speculation that the pressure to publish is causing the problem to escalate, they are becoming more adept at spotting redundancy, and they are calling investigators on the carpet when they uncover it. Yet what constitutes a transgression is not always clear, so one editor's redundancy may be another's acceptable article. This suggests to some authors that redundancy is in the eye of the beholder, which makes it tempting to submit suspect papers for publication, in the hope that they may be deemed acceptable by the editor in question. Biomedical editors appear to have taken more steps to combat redundancy, are more vociferous in their denunciations of it, and, not surprisingly, seem to write more about it than their counterparts in other fields. So is there something different about biomedical research that contributes to the problem of redundancy? Max Hecht, editor of Evolutionary Biology, thinks so. He says there is a lot more money involved in biomedicine than in biology and it's "an industry, not a science." He feels it's more competitive than biology and that researchers can't get funded by the National Institutes of Health "unless they are out there writing." This increases redundancy, he says. The impressions of Robert Malina, editor of the American Journal of Biology, and P.W. Hodge, editor of the Astronomical Journal, support Hecht's. They view their own fields as more relaxed, less money-driven, and having fewer publishing demands than biomedical research. Malina points to a cycle in which biomedical scientists continually scramble for larger amounts of money and publish more papers than do their counterparts in other fields. They have to publish to get funded, to report on their funded research, and to get future funding as they go from one grant to the next, Malina says. Hodge adds that the amount of money in astronomy is small in comparison to biomedicine and "the pressure to publish is less onerous." Another ingredient in the mix is financial gain, says Jiri Janata, associate director of the Molecular Science Research Center at the Battelle Pacific Northwest Laboratories in Richland, Wash. Speaking from the perspective of his specialty, chemical sensors, Janata argues that redundancy is linked to the degree to which an area of research has direct application. He thinks much of biomedical research has this monetary incentive, in which procedures and products can go quickly to the marketplace, unlike "the more basic sciences." Chemist Sharon Boot of the Council of Biology Editors (CBE) says that in a survey done by CBE in 1991, "redundant publication was identified most often as a problem editors faced." But she says that at the American Chemical Society (ACS), where she worked for four years, "although they had a comprehensive set of guidelines, they felt that redundancy was not as much of a burning issue in the physical sciences as in the biomedical arena." Charlie Bertsch, head of the journals department at ACS, says that the society is indeed cognizant of the problem: "Our editors would be concerned if they were aware of it and wouldn't publish [a redundant paper]." Boot feels that since most ACS editors don't write editorials, they don't have an opportunity to air the problem on a regular basis in their own publication. She adds, however, that it does come up at CBE meetings, but when CBE opened up its membership to virtually all scientific disciplines in the 1980s, few editors from the physical sciences joined, thus isolating themselves from such discussions. As for only biomedicine being an industry, Boot says that when she was finishing her Ph.D. in chemistry at Stanford University, the No. 1 goal was getting research grants to support students. "I don't see [chemistry] as any different from biomedicine," says Boot. Coming to the defense of biomedicine is David Benditt of the University of Minnesota Medical School. Although he agrees biomedicine is more of an industry than some disciplines, he's never felt pressured to publish and says that the people he knows "just like to publish." They enjoy being on the cutting edge, he says, and contributing to the field. If biomedical scientists stand out in any way, speculates Benditt, it may be that the health sciences attract people who are more outgoing, "so getting things out there is more akin to our personalities." --P.M. "Anyone who is a serious scientist complains about this problem," says Jiri Janata, associate director of the Molecular Science Research Center at the Battelle Pacific Northwest Laboratories in Richland, Wash. Journal editors say redundancy comes in various forms. The most blatant is duplicate publication, sometimes called self- plagiarism, which is word-for-word publishing of an article for the second time. Slight variations on that theme include changing the text but not the tables; changing the first author and title but not the remainder of the paper; and, in the view of some editors, publishing findings in a letter to the editor prior to publishing essentially the same findings in a paper. Then there are scientists who want to publish the same paper twice but in journals aimed at different disciplines, arguing that the disparate audiences need the information, but don't read across specialties. Most editors frown on this practice. There is also a phenomenon that editors call "salami science." This refers to cases in which scientists slice and dice the same data for multiple publications. In such instances, judgment calls as to what's acceptable occur on both the editors' and researchers' sides of the editorial desk. Editors find themselves asking: Is this a contribution to the literature, or should two or more slices be combined into a more satisfying piece? For example, Jerold Lucey, editor of Pediatrics, describes how a researcher might try to turn one paper into three. "You report on how many babies survived, then how many survived one month, and then you write up the long-term results," he says. Another judgment call is whether a large enough piece of salami has appeared here that further publication would result in overlapping papers. But even if the slice is not redundant, it often carries a negative connotation in editors' minds, because it implies that the investigator has merely done just enough work to warrant publication. The same can often be said of the least publishable unit (LPU), the smallest piece of research a scientist can prevail upon an editor to accept. Such a paper may or may not be considered redundant or a piece of salami. The size of an LPU is a fuzzy concept--it's certainly smaller than a monograph, but larger than whatever an editor thinks is too small. Its size varies from field to field and journal to journal, so that editorial acceptability will depend on both tradition and the extent to which it is viewed as a contribution. "There are people who have made their careers writing two-paragraph papers," says Robert Malina, editor of the American Journal of Biology. With that said, no one knows how much redundancy is out there. Some editors feel there is a lot, and they editorialize about it periodically. On the other hand, editors in other disciplines say they don't see that much, and, judging from the results of computer searches of the literature, they don't write about it, either. Even so, ask any editor about redundancy and he or she will tell you it's a bad thing. Why? Editor P.W. Hodge at the Astronomical Journal says that with the explosion of scientific knowledge has come a tremendous financial pressure on professional publications. More papers are submitted, he says, at the same time as the monetary investment in reviewers, editors, and production has sky- rocketed. "We just can't afford the resources, the time, and the cost of redundancy," says Hodge. At the Journal of the American Medical Association (JAMA), deputy editor Drummond Rennie notes that scientists sign over their copyrights to journals when they publish, so if they publish twice in two different journals, there's a legal problem. Redundant work also displaces original work from other researchers, which Rennie says irks him. What's more, redundancy distorts the literature and the reward system of science, essentially allowing a researcher to claim authorship of multiple papers when only one set of work has been done, says Marcia Angell, associate editor at the New England Journal of Medicine (NEJM). Redundancy not only erodes the significance of a long publication list, says Angell, but also "can lead a reader to undervalue work if it's fragmented into many small pieces." The extent of the problem hasn't been studied in depth. But Battelle's Janata has come up with some findings that disturb him. He first stumbled on redundancy several years ago when writing a review of the literature in his field, chemical sensors, and says it's becoming a more frequent occurrence. A number of research groups, says Janata, "would publish the same article five times in different journals in only a slightly modified form." Of the 15 papers NEJM considers for publication every Thursday, usually two are redundant, Angell says. And chemist Sharon Boot, chairwoman of the editorial policy committee at the Council of Biology Editors in Chicago, says that when she worked at the American Chemical Society the problem came up often. Physics isn't immune, either. At the Journal of Applied Physics, editor Steven Rothman says, "A lot more gets by than we catch." On the other hand, Bernard Shiffman at the American Journal of Mathematics considers it a minor problem and says that's probably because a mathematician "can publish one paper a year and be the top person in his field." Nels Lersten, editor of the American Journal of Botany, doesn't see much redundancy, either, while Hodge at the Astronomical Journal says, "We occasionally have cases." Biomedical editors appear to see more redundancy and give it more attention (see story on page 7). The International Committee of Medical Journal Editors, for example, which represents more than 400 journals and has its secretariat at the Annals of Internal Medicine in Philadelphia, has publicly opposed redundancy. And since 1990 the National Library of Medicine has used a separate subject heading, "duplicate publication," to index articles that discuss redundancy, while in 1991 seven pediatric journal editors formed a consortium to trade information about it. Editors' means of identifying and rectifying the problem vary. At Analytical Chemistry, editor Royce Murray is sensitive to the issue, and relies mainly on his reviewers to spot it. "The system strives to be perfect," he says, "but is humanly imperfect." Lersten of the American Journal of Botany says, "To be on top of it you have to be following the work of a specific investigator." Henry Cowell, who edits the Journal of Bone and Joint Surgery (JBJS), has some ideas about prevention. Redundancy is an important concern for him and his 400 peer reviewers. So each of the 1,000 or so manuscripts that come in annually is first scanned by an editorial assistant, who looks for similarities with anything already published by JBJS. Then peer reviewers are told to comment on whether the paper "is even close to something that has been published before," says Cowell. Finally, during the editorial process, the manuscript is checked against various reference sources. Still, "we don't catch it [redundancy] 100 percent of the time," says Cowell. Even when they do catch it, editors deal with it differently. Malina, at the American Journal of Biology, just went through two months of haggling with lawyers, publishers, and authors to settle an ordeal in which "virtually the same paper" published by his journal was published at Maturitas. Although the authors claimed oversight, they "had to write a letter to the other journal retracting the article," says Malina. At the Journal of Applied Physics, on the other hand, Rothman says the strongest thing he does is write a letter saying, "Don't do this anymore." When a duplicate recently surfaced in an electronics letters journal, he asked for an apology from the author, "and we published [the apology]," says Rothman. Miscreants can expect some fatherly advice at Analytical Chemistry. Murray will write "a special letter," but he tends to view the problem more as part of the learning process for young academics and will tell them that redundancy enhances neither their own reputation nor the journal's, "and I hope I never see this again," he says. More hard-nosed editors will refuse future submissions from past transgressors, or at least look on them with a jaundiced eye. Cowell at JBJS won't spurn submissions, but in the case of self- plagiarism will tell the researcher to discuss it with his or her department head and have the head contact him. "I'll write him [the department head], if he doesn't call me," says Cowell. Angell favors exploring all the circumstances. She will call the scientist and check with the other journal's editor. If NEJM has intentionally been burnt, "we want to make our readers aware of it," she says. She will write a tough editorial and permit the researchers to give their side in the "Letters" column. And they do, because redundancy often has two sides. Take the LPU. Many scientists argue that slicing and dicing of a database is just science. William Benditt, a cardiac surgeon at the University of Minnesota Medical School, would go even further. He says that the LPU is a nebulous term and if editors don't want them submitted, they will have to better define what they do want. Benditt says that it's easier to define an LPU retrospectively than prospectively and that most researchers just publish at the point when they have found something interesting. He can look back over five years, he says, and say a particular paper shouldn't have been published, but he can't see that at the time. "There are no sign posts in research," says Benditt, "to tell you how far you've come." Psychologist Roberto Refinetti of the Williamsburg, Va.-based College of William and Mary, writing in the FASEB Journal, says he is also unsure about how the LPU is defined. He praises it because it provides rapid dissemination of findings, aids in data management, stimulates integration of findings, and furnishes a measure of a scientist's productivity. Even so, he acknowledges that he's never thought about how long or short his ideal LPU should be. "I go entirely by tradition," he says. Although editorial hackles usually rise when discussing the LPU, unlikely support for it comes from Analytical Chemistry's Murray. His own metaphor, born of editorial license, defines it as "sort of intellectual bite sizes that should go into papers," he says. As fuzzy as the LPU may be, so is the cause of redundancy itself. There is no shortage of editors who point to the publish-or- perish dictum, but there are other factors, as well. Battelle's Janata indicts publishers themselves. They put on conferences, he says, inviting people who have published important papers. The same papers are then published again as proceedings or in special journal issues, he says. Max Hecht, editor of Evolutionary Biology, also blames the publishers. As the number of journals increases, he argues, researchers are pushed toward redundancy. "They want to be noticed," he says, "and there is so much out there, the only way is to repeat it." At JAMA, Rennie is less charitable. "I think people are just slack. They want to inflate their bibliographies," he says. The proposed remedies are as varied as the causes. Janata has suggested that a computer subject-and-author search should be done for every manuscript. When violators are discovered, their names should be publicized. NEJM's Angell, along with several other editors, supports limiting the number of publications a scientist can use for promotion, tenure, and funding. This guideline exists at Harvard University, but so far has no teeth in it, she says. Janata suggests that publishers stop issuing redundant proceedings from conferences, or attach the disclaimer "that they have not been refereed." Hecht, however, thinks the situation will resolve itself. Much redundancy, he argues, is due to the proliferation of journals, but university libraries, their main subscribers, can no longer afford them and are canceling their subscriptions. He expects many journals to fold, which will eliminate much of the temptation to publish redundantly. "The market should solve the problem," says Hecht, "in three to four years." Paul McCarthy is a freelance writer based in Honolulu, Hawaii.

Vigilant Science Journal Editors Fight Redundancy

Interested in reading more?

Become a Member of