Authorship

Some Views on Joint Authorship in Scientific Papers

Richard M. Murray

Division of Engineering and Applied Science
California Institute of Technology
Pasadena, CA 91125

27 September 1992
Revised, 20 November 2022

This note is a description of my views on joint authorship of scientific papers. It lays out my opinions on who should be included as an author on a paper, the order of the authors, and how to resolve conflicts between authors. It is a working document which represents my current feelings on the subject; it is under constant revision as new experiences and information add to my understanding of this complex issue.

Deciding who should be include as an author

The first thing to acknowledge when discussing joint authorship of scientific papers is that there are different views on what is appropriate and that these views vary between disciplines and individuals. The most important aspect of joint authorship is to have a consistent, justifiable philosophy which is within the norms established by the scientific community.

Some resources on authorship:

• My personal thoughts on publishing and differences between science and engineering publishing practicies
• Caltech authorship dispute guidelines
• IEEE authorship guidelines
• AAAS authorship guidelines

There are several ways to acknowledge a person's contribution to a scientific work. Any results which have been previously published by another author and are relevant to the work being presented should be cited appropriately. Specific ideas or suggestions which have not appeared in print might be cited as a "personal communication". This is appropriate if an individual made a very specific suggestion or comment which concerns only a portion of the paper. More substantial suggestions can be noted in acknowledgements included at the end of the paper. Acknowledgements often include individuals who read rough drafts of the paper and provided comments, had discussions on the general subject area which the authors found useful, or pointed out specific results and techniques which were beneficial or instrumental in completing the research.

Any individual who had significant impact on the overall content of the paper should be included as an author. For students, this almost always includes one's advisor, even if that person did not directly carry out calculations, perform simulations or experiments, or write large portions of the paper. Of course, this assumes a student-professor relationship in which the professor takes an active part in guiding the research, providing ideas about specific techniques that are employed, and assisting in generation of the paper (perhaps by providing detailed and critical comments as the paper is drafted). This is above and beyond any nominal direction that an advisor might provide in helping a student to select and choose a research topic.

Other individuals who have made significant contributions to the paper might included students working in your lab, other professors who were consulted as part of the research effort, or colleagues at other institutions. To be added as an author, the contributions of these individuals should be more substantial than a simple discussion about the problem or a reference to some related work. Acknowlegements can serve the purpose of thanking someone who provided information that was crucial to solving the problem without actively working on the problem themselves. In my opinion, being the PI on a grant that supported the work is not, in itself, sufficient to be listed as a co-author, though not all PIs agree with this approach.

I usually use the following guidelines:

• Primary author: In most cases, the bulk of a paper will be written by one individual, the primary author. The primary author should *always* be the first author listed on the publication. It is the first author's responsibility to decide on the remaining authors that will be listed on the paper and the order in which they will be listed. If there is no clear ordering of the authors based on their contributions, alphabetical order should be used.

• Author order: If there are more than two authors, then the order of the remaining authors should either be alphabetical or in the order of the level of contribution (with the last author making the smallest contribution). On papers that I contribute to, I am most often listed toward the end of the list.

• Acknowledgements: Individuals who might have contributed ideas or feedback to the content of the paper but whose contributions were not at the level of serving as a co-author can be acknowledged at the end of the paper. I am usually fairly generous here, listing anyone who has participated in the activities that are being described.

In scientific journals (and increasingly some engineering journals) there are some additional approaches that are often used to define authorship and contributions:

• Corresponding author: many journals require that one (or more authors) be identified as the corresponding author (or sometimes co-corresponding authors). I have always used the approach that the primary author should be the corresponding author, but others often list the senior author (typically the advisor) as the corresponding author and list this person last on the paper. I continue to prefer that the first author serve this role.

• Co-first authors: some journals allow multiple authors to be listed as "co-first" authors, typically by annotating the author list in a way that states that some set of authors "contributed equally" to the material. There is still the issue of who gets listed first, but this at least allows a mechanism for identifying equal contributions. I continue to prefer having a single author listed as first.

• Statement of author contributions: many scientific journals now require that a statement of author contributions be give at the end of the paper and this can be used as a mechanism for describing contributions in a bit more detail. While not common in engineering journals, I believe that a statement of author contributions is a good mechanism for resolving disputes over contributions to a paper. For example, here is a fairly detailed statement trying to make clear who contributed to what:

Summary of Author Contributions: A1, A2, and A3 jointly conceived the conceptual framework. A1 and A2 jointly developed the problem formulation and theoretical approach. A1 worked out the main proofs with input from A2. A2 drafted the manuscript and figures with input from A1. A3 and A4 provided guidance on the overall approach and provided feedback on the final manuscript.

While the final decision on who to list as authors rests with the primary (and/or corresponding) author, the primary/corresponding author also accepts responsibility for insuring that all parties involved are given appropriate credit. It is inherent in the scientific process that new results build on top of old results. Giving proper credit for previous work is the backbone of research.

Before a paper is submitted to a journal or conference, copies of the paper should be sent to people who are listed in the acknowledgements or have otherwise been involved in the paper. This provides those people an opportunity to discuss any problems in properly attributing research results before the paper is accepted for publication. It is also a good way to insure that people have been credited appropriately: if you feel awkward sending someone a copy of a paper, there may be a reason.

Resolving conflicts between authors

For the first 20 years of my career, I had very few issues some up around authorship. In particular, the notion of "co-first" authors was not a concept that was really used and in the first 20 years of my career I believe that I was never part of a paper in which the term "contributed equally" was used in describing authorship. We simply agreed that someone should take the lead on a given paper and left it at that.

This situation changed when I started working in synthetic biology, where publications tended to be in the style of scientific publications rather than engineering publications. In 2013 I was a co-author on a couple of journal and conference papers that used to "equal contribution" approach.

Despite this change, I still believe that for engineering papers, and especially for conference papers, it is preferred for one author to be given the responsibility for leading the writing of the paper and that person should be listed as the first author. Obviously that person should also have contributed to the paper in a manner that is consistent with taking the lead in the writing as well as being listed as the first author. One thing that is clear is that whenever possible discussions and decisions on authorship should be made before the writing starts, and not after the paper is being prepared for final submission. In my experience, the thorniest cases have been those in which the authorship was not clear at the time the paper was being written/drafted and different co-authors had different expectations.

Unfortunately, no two collaborations are alike and no two author lists are alike, so there is generally not a hard and fast rule that one can apply to authorship. It simply has to be worked out individually for each paper and for each group of co-authors. Having said that, my personal preference continues to be that one person serves as the lead author for a paper and the "equal contribution" terminology should not be used unless the work was truly done in a fully collaborative manner in which all "co-equal" authors agree that things are equal.

Some PhD students in CMS and MCE, led by my former student Karena Cai, put together a very nice workshop on research collaborations that included a module on authorship. I highly recommend taking a look at these pages, since they include some great insights and links to further resources.

Additional resources

• https://healthycollab.caltech.edu - Collaboration tips website, put together by CMS and MCE students at Caltech
• IEEE authorship guidelines
• AAAS authorship guidelines