With academic jobs depending on a record of accomplishment, authorship issues are of vital importance to postdoctoral fellows and graduate students. And because of its critical importance to the careers of scientists, authorship is shaped by social customs. Like other customs, authorship is usually not determined by explicit criteria and not "discussed in polite company." Coupled with parallel trends toward greater numbers of authors being included on manuscripts and increased collaboration across disciplines, young scientists these days often find themselves involved in complicated negotiations with other academics over authorship.

There are several phenomena that are likely to increase the complexity of these negotiations. These include the use of co-authorship or several first authors, honorary authorship, and, most troubling to academic freedom and accountability, ghost authorship. Fortunately, however, there are a growing number of resources to which postdocs, graduate students, and even faculty can turn for bringing authorship issues into the light and to provide objective criteria for determining who is an author.

For more discussion on the ethics of authorship, see the Next Wave feature How Should Authorship Be Decided?, which comes complete with case studies and links to other Next Wave articles and resources.

The rise in the number of authors on biomedical research papers is striking. In 1930, the average number of authors on biomedical research studies was 1.3; by 1989 it was 6.0. This trend reflects both the growing collaborative quality of research and a dilution of authorship. An extreme example of this trend is illustrated by a 976-author paper by the GUSTO Investigators. This paper received the 1993 Ig Nobel Prize for Literature, for having "one hundred times as many authors as pages."

With every additional claim for authorship status, the issues become more and more complicated. And as a number of recent very public retractions have shown, it is more difficult than ever to know who is accountable for what in papers with large numbers of co-authors.

Disturbingly, the intellectual contribution of these growing lists of authors is unclear and may be diminishing. A 1994 study found that 21% of authors of basic science biology papers and 30% of authors of clinical studies had no involvement in conception or design of the project, design of the study, analysis and interpretation of data, and writing or revisions. These honorary authors may have only provided the laboratory space, funding, or a small proportion of samples for the research. This study is all the more telling because the data is self-reported--the numbers reflect only what researchers were willing to acknowledge.

What may pose even more of a threat to accountability in scientific publishing is the growing phenomenon of ghost authorship. This typically occurs when a medical writer is hired to write up the study, or an individual made contributions that merited authorship but is not included in the list of authors. Scientific dialog with the authors is threatened if the ghostwriter, or even worse, the named authors, haven't understood the underlying experiments and may be unable to explain or defend them to other scientists. Additionally, because the process of reducing the work to writing forces the investigator to more closely examine the study and often gives rise to refinements, much can be lost by farming out the writing, including reducing the academic training opportunities for the student or postdoc authors.

On the other hand, amidst all the clamoring for inclusion as an author, "it is well to remember that having one's name on a paper is not necessarily beneficial to one's scientific reputation" (Wilson, 1952). Consider the case where a paper is later discredited and retracted. All the authors' names will be forever linked with the fraudulent work regardless of their role. It is vitally important to trust those with whom you collaborate and to dissociate yourself from those whose work you question.

Despite all of the above concerns, is it possible that honorary and ghost authorship are victimless crimes? This is a dangerous assumption, because by diluting the responsibility for the research, there is less accountability for the data and its interpretation. This also increases the likelihood and the number of those victimized by fraudulent work. The victims are often postdoctoral fellows and graduate students.

The International Committee of Medical Journal Editors (ICMJE) Uniform Requirements for Manuscripts Submitted to Biomedical Journals are quite explicit. Every author listed must have:

  • made substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data;

  • been involved in drafting the article or revising it critically for important intellectual content; and

  • provided final approval of the version to be published.

  • Acquiring funding, collecting data, or supervising the research group, by themselves, do not justify authorship. The ICMJE advises that decisions on authorship should be made by a joint decision of the co-authors, and that, in multicenter trials, all authors must meet the ICMJE criteria for authorship.

    Another troubling area of authorship disputes is the order of authorship. For graduate students and even more so for postdocs, the order of authorship can seem to make or break an academic career. Furthermore, many researchers are not aware that different academic fields assign different meanings to the order of authorship. For example, in some fields, the last author spot is customarily reserved for the principal investigator or head of the lab. In others, the department chair holds that spot. In yet others, placement as last author means that that person was the one who had the smallest contribution to the work.

    Because authorship placement is often an unspoken tradition, and can vary even from department to department in the same field, it is best not to make assumptions. Researchers should initiate discussions about authorship in the earliest stages of, and to continue to discuss authorship throughout, the development and completion of a project. Researchers especially need to realize that an individual's authorship may change over time depending on the way experiments turn out, the work actually conducted by each author (in contrast to what was predicted), and other factors.

    Written specification and confirmation of the authors' roles can prevent surprises and dashed expectations. Sometimes, faculty set aside oral agreements on authorship much to the dismay of a postdoc who has relied on the oral promise of first authorship. An agreement in writing might not be as easily dismissed and will certainly prompt an open discussion between collaborators. Again, all should remember that written agreements do not constitute promises that are fixed in stone. Situations change, with people joining and leaving projects in mid-stream, and with experiments taking unexpected turns that may lead to restructuring of publications. Thus, any predetermined authorship agreements should explicitly acknowledge that authorship status is contingent on certain situational factors and may be renegotiated if those factors change. However, prior discussion of authorship provides the research group the opportunity to determine two things: first, what constitutes authorship and the order of authorship, and second, what factors would change authorship.

    Ombudsmen are frequently asked to help resolve disputes over authorship, particularly disputes over the order of authorship. Ombuds are a confidential, neutral, and independent resource for trainees and faculty. An ombudsperson may assist the parties in understanding how a dispute arose--often the result of reliance on promises not kept--and how best to consider resolving it. Further, a neutral party can help balance the power disparity between the parties and, if so desired, attempt to mediate the dispute (see sidebar for additional resources that may help avoid conflict).

    Helping Hands

    There are guides available to researchers for use in preventing and mediating authorship disputes. A comprehensive and practical guide to all aspects of publishing has been developed by the International Committee of Medical Journal Editors (ICMJE). In addition to the ICMJE, specific journals, professional societies, and universities often have policies on authorship. See, for example, The Society for Neuroscience, the Journal of the American Medical Association ( JAMA), and Stanford University.

    Newer trends in authorship that could alleviate some authorship dispute include journals requesting written confirmation or specification of authors' roles and contributions. For example, the ICMJE authorship guidelines state that, "Authors should provide a description of what each contributed, and editors should publish that information. All others who contributed to the work who are not authors should be named in the Acknowledgments, and what they did should be described." This way of delineating authorship would improve both the accountability of and trust in researchers and would make it easier to establish how credit should be given. Indeed, JAMA has begun to require that authors specify their roles when submitting manuscripts for publication.

    Another strategy for lessening the pressures of authorship is to reduce reliance on curriculum vitae to describe one's scientific contributions. For example, during the job application or promotion process, explanations of precise roles in published research could be included in letters or addenda to the CV. Some universities have also begun to limit the number of publications considered in the promotion review process, and promotion committees are asking faculty to describe in detail their roles in published work as well as the significance of the work. These trends and guidelines should help trainees and their mentors avoid the potential pitfalls of authorship while encouraging integrity and trust in research.

    *Mildred Cho is a Senior Research Scholar at the Stanford University Center for Biomedical Ethics and Martha McKee is Ombudsperson in the Stanford University School of Medicine.

    Reference

    E. B. Wilson, An Introduction to Scientific Research (McGraw Hill, New York, 1952)