The who and how of authorship.
by David M. Post (with input from many!)
Multiple-author papers are a reality of life in many research labs; however, the criteria for determine authorship too often follows unwritten and nebulous rules. This essay on authorship derives from my experiences and discussions as a graduate student at UW-Madison and Cornell University, and as a postdoctoral fellow at the National Center for Ecological Synthesis and Analysis. Many have contributed ideas and criteria to this guide.
Many disputes can be avoided by a clear common understanding of standards for authorship (especially in multi-disciplinary groups). Authorship should be discussed between researchers at an early stage in any project and renegotiated when necessary. Where possible, there should be agreement on which papers will be written jointly (and who will be first author on each paper), and which will be single authored. Early drafts of papers should include authorship to help resolve any future disputes.
I. Who should be an Author?
In general, it is agreed that authorship should be reserved for those who have made significant intellectual contribution to the research; however, what is meant by significant contribution may vary widely. The criteria listed below represents my personal view based on many years of work within and around large multi-disciplinary groups and collaborative labs.
1) If you find yourself saying “I could not have produced this paper without Dr. X or graduate student Y” because they brought to the project a unique and indispensable skill, perspective, or data set then they should be a coauthor. If many other people could have provided the same skill, perspective, or data then they may not necessarily merit being an author. If their role was not unique then see point 2.
2) If a person had a (significant) role in three of the five following criteria they should be a coauthor. These criteria fit an empirical project best, but are generally applicable for theoretical research. (Modified from multiple sources)
a) Formulating the initial idea. This can be very hard to judge so it is best to be generous with these criteria.
b) Planning/facilitating the research. May include figuring out how to collect data, outlining a modeling approach, writing/getting grants to fund the work, providing key equipment, etc.
c) Doing the research. Includes collecting data, coding a model, or working through tricky math.
d) Analyzing the data. May include database manipulations, statistical and graphical analysis, or providing new insights that derive from the results.
e) Writing and publishing the results. This could include writing some section(s) of the manuscript, providing extensive editorial comments, etc.
3) When in doubt, be generous! There are very few examples of careers being destroyed by adding a coauthor. There are, however, many examples of important relationships being destroyed by not including someone as an author. Do not let your ego get in the way.
On the other hand, recognize that adding authors who have not substantially contributed to a paper will devaluate the role of authors who have made substantial contributions. There is no easy decision once a potential author falls under this criterion.
II. Implications of Authorship
Here are just a few thoughts on the implication and connotation of different author structures.
1) Single author – A single authored paper says that you did all the essential work, although others may have contributed by editing, make suggestions on research, providing labor and supplemental data, etc.
2) Two authors – A paper with two authors suggests a shared enterprise. It is often assumed that the second author did almost as much work as the first author, but that the first author did much or most of the writing.
3) Three or more authors – Multiple authored papers are open to many interpretations. Often it is assumed that the second author contributed more than the 3rd through nth authors, but this is not necessarily true. First authors may be project leaders, a collaborator on equal footing with other authors, or students drawing upon the expertise and data of others to answer broad questions or to produce a more complete story. The 3rd through nth authors may have made very large contributions or may have contributed very little; it is always hard to tell.
III. Author order
Authorship order can also become a sticky point, and a point of stress among collaborators. Here are a few thoughts:
1) The person who has made the major contribution to the paper and has taken the lead in writing the paper should be the first author.
2) Decisions about authorship and the order of authors should be made by the first author in consultation with other authors.
3) Individuals who have made a major contribution to analysis or writing (i.e., more than just commenting on successive drafts) should follow the first author immediately. Where there is a clear difference in the size of the contribution, this should be reflected in the order of these authors. However…
4) Often there is no clear difference in the size of contributions, or contributions are sufficiently different in kind that comparisons are impossible. In those cases, consider listing in alphabetical order of their surnames all authors who fulfill the criteria for authorship. Remember, once a paper is in the et al. stage, it really does not matter much where you are in the author list.
Regardless of author order, each coauthor should have a chance to critically review successive drafts of the paper and should approve the final version. Furthermore, each coauthor should be able to defend the paper as a whole, although not necessarily all the technical details.
IV. The student-mentor relationship and authorship
A word of advice to students early in their graduate career - above all, follow any criteria your advisor might have (e.g., some advisors attach their name to all papers derived from their lab group). Otherwise, a good advisor will generally have an important role in at least three of these criteria. They usually help you formulate or refine your ideas, discuss how to test an idea or collect data, provide funds and equipment that allow you to conduct your research, and help you with analysis and publication. Often, the only thing they do not have a role in is the data collection. That means students early in their career should typically include their advisor as a coauthor, while students later in their career may publish a paper or two without their advisor. If your advisor feels he or she has not had significant input or you have been particularly independent, let them make the decision to remove their name from the manuscript.
This also puts the onus on mentors to carefully weigh their role in papers produced by members of their lab. Mentors should remove their name from papers that they have made little contribution to in order to maintain the intellectual integrity of authorship. Mentors should also recognize and acknowledge in an appropriate way the contribution of lab members to synthetic paper they write.