Author! Author! Do All Scientific Papers Really Need To Identify So Many Of Them?

Researchers hold mixed opinions, with some advocating a streamlined approach that credits only germane contributors Geophysicist Marcia McNutt routinely reads a stack of journals ranging from the Journal of Geophysical research to Science and Nature. When she reads a paper in any of these publications, she usually feels safe making a few assumptions about the authors listed on the paper. "I usually assume the first author is the person who actually did the research," says McNutt, a professor of earth and planetary sciences at the Massachusetts Institute of Technology. "I assum if there is a senior person who is last author, whether that's second or thid, that that is the person who probably wrote the proposal to fund the research and had the idea to begin with, but generally did not personally carry out the research. "And then, if there is an intermdeiate author, I usually assume that is a person who had some sort of significant input--such as providing a key data set or perhaps working on some key subset of the problem--but was not the person who physically wrote the paper and did more than 50 percent of the research." McNutt's assumptions are typical. She reads the unspoken messages of the order of authorship the way most scientists do. But as common as these assumptions are, are they valid? A survey of secitnists and journal editors from a variety of disciplines reveals that the assumptions are sometimes valid, and sometimes not. Contributors on a paper may choose to list themselves in alphabetical order, in reverse alphabetical order, or in some other sequence whose significance is known only to the authors. In addition, stories abound of `honorary' authors--researchers who have not really contributed to the research--hitching a ride on papers. Who should be listed as an author and who should not, and just how the authors should be ordered, typically are questions left up to the researchers themselves. "We expect the authors themselves to know about the relative contributions of the participants in the study," says Daniel E. Koshland, Jr., editor of Science. "It would be very difficult for a journal to really go into that and ask people how things were actually done. And convention varies greatly from discipline to discipline." Traditionally, for instance, alphabetical ordering has been more common in physics and mathematics than in the life sciences, according to Columbia University sociologist Harriet Zuckerman. While leaving order up to authors satisfies scientists' generally strong streak of independence, it has left them with few guidelines and standards. As a result, says George Lundberg, editor of the Journal of the American Medical Association (JAMA), readers can be--and are--misled. And this, Lundberg believes, is a problem that needs a solution. "Readers, in particular academic personnel committees, do give weight to order of authorship based on their perception of its meaning," Lundberg says. "But since authors and journals have never determined the meaning, the messages sent by the current order of authorship may be spurious, confusing, or even opposite the interpretations that are taken." More than 20 years ago, Zuckerman wrote, in a classic paper on authorship order (American Journal of Sociology, 74:276, 1968): "As the social organization of scientific work has become more complex--more often collaborative and sometimes intricately organized--the visibility of individual role performance is reduced. The changed organization is reflected in the rapid increase in multi-authored papers." At that time, Zuckerman estimated that about 40 percent of the papers in biological sciences, 60 percent of those in physics, and 80 percent of those in chemistry were multi-authored. Today, it is rare to see a single-author paper in biology and chemistry. And in some disciplines, such as high-energy physics, papers can have an author list that includes more than 100 names. "In my area of physics [condensed-matter physics], until very recently, more than four or five authors on a paper was rare," says Daniel Fisher, a professor of physics at Harvard University. Now, he says, with the interaction of various fields, including materials science, "there are more papers with many authors--up to a dozen or 15, of which several might have been involved with growing the material and making the materials." This trend adds to the problems posed by authorship order. "The larger the number of authors, the more squabbling there's going to be," says Lundberg. "If you have papers with only one or two or three authors, there's unlikely to be much trouble. It becomes a big deal if you're at a university in which order matters a lot in terms of whether you get promoted or not." Some academic institutions, such as the University of California, make it a policy during hiring and promotion to establish what a scientist's role was in a paper and not rely simply on authorship order. But that practice can be tedious and isn't followed by every institution. Moreover, Lundberg says, "if you're dealing with journals that limit the number of authors, [authorship order is] a big deal because if you happen to be number nine and the journal will allow only six, you're not an author." Even if an author makes the list for the journal, that researcher still may not be cited as an author of the paper in indexes and references. "Only the first three or four authors would be listed on a multiple-author paper on Medline or in Index Medicus or in references," with the rest represented only as "et al.," Lundberg notes. "So if a person had a great deal to do with the paper and was the sixth or seventh author, it's quite likely that person's name would rarely show up anywhere as having authorship at all." Just how much conflict authorship order generates is impossible to measure. But scientists interviewed for this story indicate that authorship order had not posed major problems in their careers, although most could recall at least one time when they were dissatisfied with the order in which their name appeared on a paper. The dissatisfaction was not always rooted in the predictable case of not getting enough credit for work done. "I'm sure there's squabbling over it, but usually, in the papers I've been involved with, the squabbling is over trying to get someone to be first when they feel that they're not sure they should be," recalls Fisher. For instance, Fisher's name appeared first on a mathematical physics paper to which he was not the primary contributor. He protested to the other authors. But in that discipline, alphabetical listing is common, and the other authors wanted to stick with that convention. Usually, junior scientists, who are just beginning to collect journal credits, are more anxious about name order than senior scientists who already have their careers in gear. But even if a junior's name is first, the value of that placement can be decreased in the young scientist's eyes if there are more than four or five other names on the paper. "The more names that are on there, it looks like more people were involved and it was less your work and more the work of the whole group," says Timothy O'Toole, a molecular biologist who was just promoted from postdoc to senior research associate at Scripps Research Institute in La Jolla, Calif. In other cases, however, young researchers are anxious to have their more-famous mentors add their names to papers. Scientists and editors interviewed complain of two practices in particular that can increase the number of names on an authorship list. One is to include as an author someone who did not actually participate in the research but contributed material or minor technical assistance. Clinical researchers, especially, have a reputation--justified or not--for being particularly aggressive about demanding authorship for supplying such contributions as patients or tissues taken from patients. Scientists interviewed for this article, including clinical researchers, say that in most cases, simply supplying material warrants an acknowledgment at the end of the paper, not authorship. In the October 1990 issue of JAMA, Lundberg and his colleague Don Riesenberg publicly grappled with the authorship order issue by proposing a five-point set of guidelines. It was not the first time someone had tried to solve the authorship-order problem. "There is no shortage of suggested solutions in the literature," Lundberg and Riesenberg wrote. "These have included mathematical formulas; multiple categories of authorship, denoted by coded credits; and a worksheet assigning weight to various aspects of the investigation." Lundberg and Riesenberg's guidelines propose that authors be listed in order of their contribution to the work, including the writing, with the first author having contributed the most. All authors, the guidelines state, must actually participate in the work and the writing and assume public responsibility for the paper's conclusions. There is no room for honorary authorship, according to the guidelines, and simply heading a lab is not sufficient reason to assume a place on an authorship list. Authors should determine as early as possible in the research what the order will be, "although relative contributions may need to be reassessed later by group consensus." Finally, Riesenberg and Lund-berg propose that any disagreements about order should "be resolved by the principals, not by the editor" of a journal. The guidelines are far from exotic. In fact, they match the personal standards the scientists interviewed for this story say they use for determining authorship order. Nevertheless, says MIT's McNutt, "if you call 100 scientists, you won't get the same answer from all of them as to what the legitimate role [of an author] is." Consider, for instance, the case of the university professor who gave his graduate students a problem to solve on a take-home exam. One bright student came up with a particularly creative solution that the professor recognized as a potentially valuable addition to the scientific literature in the field. He encouraged the student to write a paper based on the test answer and submit it to a journal for publication. The student did, and the paper was published. The student was the first author on the paper. Should the professor have been listed as an author as well? "Absolutely not," says McNutt. "This was an answer to an exam question. If there is any case where something has to be a single author, an answer to an exam question has got to be it." However, the professor, a friend of McNutt's who after its publication shared with her how the paper evolved, opted to be listed as second author. Since the student was working in his lab and receiving financial support and encouragement from the professor, and since the professor's question sparked the student's ideas, the senior scientist figured he deserved to be an author. While some scientists may be horrified by the professor's decision to include his name on the paper, others are more horrified by the idea of imposing any kind of standard upon scientists for determining authorship order. "I think it would be a hindrance," says physicist Fisher. "Because then there would always be arguing over exactly what the rules meant and how to interpret them." There's also the problem of who would set the standards and who would enforce them. Some journals have made a stab at encouraging standards, but journal editors themselves are split about what their role should be. After several years discussing the issue, the editorial board and staff of JAMA couldn't reach a consensus on authorship order standards. Thus, Lundberg and Riesenberg's guidelines are only that--guidelines, and not official JAMA policy. "I think it's a tricky issue, and for my money, journals that publish papers should not tell people what to do," says Benjamin Lewin, editor of Cell. "We only tell people what to do if we get stuck with something that we regard as transgressing the normal bounds. If we get a paper with an author who is also on a competitive paper being published here, we feel entitled to say, `You can't publish the same thing twice.' Otherwise, if you present a paper and it has some authors on it, it's really your business to decide who should be an author and who shouldn't." Kathryn Phillips is a freelance science writer based in Ventura, Calif.

Author! Author! Do All Scientific Papers Really Need To Identify So Many Of Them?

Interested in reading more?

Become a Member of