Glenn McGee in The Scientist: Me First!
Glenn's column for September proposes two rules for fixing the system of academic authorship:
This spring, The Scientist received a letter signed by Fertility and Sterility editor Alan DeCherney, asking to retract comments he made three months earlier, in which he accused authors of an F&S paper of plagiarism. The authors of the 2005 paper, on mitochondrial DNA in ovarian failure, are still facing allegations from a scientist who claims they stole his research and left him off the author list.
Authorship disputes are a fairly regular occurrence in science, a natural offshoot of the oppressive demands of a "publish or perish" system. So much can be at stake: If a postdoc or young professor receives that all-important top billing of first author, they are more likely to earn tenure, or a lifetime-guaranteed career. (See "Does Tenure Need to Change?")
There are so many problems with the current system of scientific authorship, it's hard to know where to start looking for solutions. Physicians and professors are offered authorship of virtually or totally completed papers by companies whose medical writers and scientists have completed the bulk, or all, of the work. This past January, a study in PLoS Medicine revealed that two-thirds of industry-initiated randomized trials contained evidence of ghost authorship. (1)
All too frequently, scientists allege that some senior faculty claim senior authorship on every paper that originates in their labs, even if they were only marginally involved. And, authors don't know what each other is doing: Another January study in the Canadian Medical Association Journal (CMAJ) found that more than two-thirds of 919 corresponding authors disagreed with their coauthors over contributions to the paper. (2)
According to the International Committee of Medical Journal Editors (ICMJE), authors must offer "substantial" contributions to the design or data, draft or revise the manuscript, and give final approval of the finished version. Anyone who does not meet these criteria, the ICMJE says, must be listed as a contributor in the acknowledgements section. These rules are simple, straightforward, but not followed: In the CMAJ report, only 3 of 10 noncorresponding authors interviewed met all the ICMJE criteria for authorship.
Clearly, the safeguards we currently have to ensure proper authorship are not working. It's not enough to ask authors to take responsibility for authorship; journal editors and institutions must be accountable, as well. I propose two new rules to help restore fairness in scientific authorship.
Rule #1
Journals must be ready, willing, and able to check whether a submitted paper has been published elsewhere. The technology to do this already exists: In universities across the country, professors routinely plug student papers into programs such as Turnitin, which spit out similarities to other online papers. The program would have to be more sophisticated to detect whether papers were published in different languages, but you can spot many things even without that added level. In the past three years, I've received approximately a dozen calls from scientists claiming a published paper is based on a lecture they gave that one of the study authors attended. Both lecture and paper were in English.Obviously, coauthors should also ensure the work is original before signing off on it, and perhaps peer reviewers should do so as well (although I hesitate to recommend this last point, given how much time many peer reviewers already spend on manuscripts).
Rule #2
Universities need to do a better job of preventing authorship disputes in the first place, and accepting responsibility when these disputes occur. Many institutions provide little or no training in publication ethics, so I propose that universities teach every new research hire about authorship and other aspects of ethical scholarship. In addition, why not send out a monthly e-mail to researchers that discusses a recent case of disputed authorship and lessons they can take from the example? In this way, we can all learn from disputes such as the one over the F&S paper - not just lament them, feel relieved that we're not involved, and move on.References
1. P.C. Gotzsche et al., "Ghost authorship in industry-initiated randomized trials," PLoS Med, 4:e19, 2007.
2. V. Ilakovac et al., "Reliability of disclosure forms of authors' contributions," CMAJ, 176:41, 2007.
contribute a comment
Your contributions to the conversation are very much appreciated. We do have a few simple guidelines, though. Be civil. Stay on topic. We reserve the right to remove comments that violate the aforementioned guidelines. One more thing: comments are moderated, so it may take a little while for your comment to be posted. Thanks.
comments
Disputes in authorship will always occur, as you mention, as long as the "publish or perish" system is valid. And it's getting more extreme.
Oh, and department politics always play a huge role, no?
But I think articles like this are needed to show a reality that many pretend is not there.
Now, if I did all the bench work, spent hours and hours researching and writing the paper... what are those other 11 stupid names doing there?
- by A. Marques on Sep 11, 2007 at 6:02 PM | link
I would have to disagree on the effectiveness of those ideas:
If anyone should have to check for plagiarism I submit it should be the dean, the office of the head of research or someone like that. But in the end these things are not easy to find because the duplication is rarely word for word. Plagiarism is the defining sin of research but actually rare; prevention would be nice but punishment seems more apt. Then the rest of us can get on with what we do without even more delays (I have had a journal go 2 years before responding, the last one took 8 months and then used only one referee).
A monthly email to all researchers would go in the bin with all the other monthly weekly and random emails we receive from on high. I don't go a day without getting something and most are well-meaning but impotent warnings about one thing or another. Researchers are grown up people who *do* in fact know about these issues just like young people know about STDs and that cigarettes cause cancer. Some individuals choose to take the risk and pretty much nothing you *say* to them makes any difference. People do it for the benefits and due to the low chance of effective censure.
I would suggest something simpler. First author should be the person who physically performed most of the work--last author should be the one with whom the buck stops (e.g. grant-holder, most senior, the one charged with oversight). This would also help prevent the most common sort of dishonesty, having the one who did most fo the work appear as a later author or not at all. It would also identify the one who is going to take the staick if it goes bad. Perhaps being aware of that would help focus that person on being a watchdog.
- by emily on Sep 12, 2007 at 9:36 AM | link