Scientific authorship is important for a number of reasons. New knowledge must be published, to make it known both to the public authorities (who have paid for the research) and to the research community (who are given the opportunity to test the results). The authors are entitled to due recognition of their contribution, which will provide incentive for further research and serve as the basis for new research funds and continued career development. Conversely, undue recognition of authorship makes academic communication more difficult and gives credit to those who do not deserve it. This can in turn lead to research resources being lost, while also weakening the motivation to pursue high-quality research.
Most scientific societies and journals have therefore established ethical guidelines for the publication of scientific research. These guidelines were formerly relatively general but have become stricter in the past couple of decades because of various breaches of ethics in the conduct of research. Scientific studies of publication practice and authorship have also contributed to guidelines being made more specific.
Fundamental requirements for scientific authorship
Any discussion of the problems surrounding co-authorship must start from the fundamental requirements for scientific authorship. The guidelines for scientific publication under the International Union for Pure and Applied Physics (IUPAP 2005) state:
"Individual researchers should understand and scrupulously uphold high standards for ethical behavior in the conduct of research, particularly in relation to the verification and truthful reporting of data, the granting of proper credit, and referencing of the work of others in publication. Plagiarism of another's work is a form of theft and constitutes serious misconduct."
Where co-authorship is concerned, the guidelines are limited to stating that co-authors must take full public responsibility for the work (i.e. in accordance with the guidelines quoted above) and must have made a significant contribution to the work:
"In publication it is essential that each co-author contributed significantly to the research reported and openly accepts joint responsibility for the work. If these conditions cannot be met, the person should not be included as an author."
What significant contribution means is not elaborated further. It could be understood to mean that it is up to the individual scientific disciplines and journals to define this further. In the following we shall look at some examples of how attempts have been made to solve this problem in different scientific fields.
Criteria for co-authorship
In the sphere of physics, it is useful to consult the American Physical Society (APS 2002), which publishes a considerable number of scientific papers. Their guidelines state:
"Authorship should be limited to those who have made a significant contribution to the concept, design, execution or interpretation of the research study. All those who have made significant contributions should be offered the opportunity to be listed as authors. Other individuals who have contributed to the study should be acknowledged, but not identified as authors. The sources of financial support for the project should be disclosed."
While there is still no qualitative definition offered of what "significant contribution" means, contribution is exemplified as the "concept, design, execution or interpretation" of the scientific work. This refers to the different stages and elements of research work and can be used of both experimental and theoretical studies. The American Mathematics Society and the American Statistical Association (AMS 2005 and ASA 1999 respectively) as a main rule demand significant intellectual contributions from all co-authors. Exceptions are possible, but must be justified. The differences in the wording of the physicists' and mathematicians' guidelines are scarcely of general importance, but rather an expression of differences in working methods and publishing tradition. In chemistry (represented by the American Chemical Society, ACS), the criteria for co-authorship are quite similar to those of the physicists (ACS 2011).
If we go to biology (for example, the American Institute of Biological Sciences, AIBS), we find guidelines that are more detailed. These are very similar to the guidelines prepared for biomedical publishing (the so-called Vancouver Protocol, adopted by the International Committee of Medical Journal Editors, ICMJE). The Vancouver rules start by defining an author as someone who has made "substantive intellectual contributions" to the published work. Regarding co-authorship, the guidelines state:
"Authorship credit should be based on 1) substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data; 2) drafting the article or revising it critically for important intellectual content; and 3) final approval of the version to be published. Authors should meet conditions 1, 2, and 3." (ICMJE 2010, chapter on Authorship)
The rules emphasise further that administrative and financial responsibility alone does not qualify an individual for a place in the list of authors:
"Acquisition of funding, collection of data, or general supervision of the research group alone does not constitute authorship."
These and other contributors who do not satisfy the authorship requirements should instead be thanked in a separate paragraph or footnote (acknowledgement).
Some problem areas
There is insufficient space here to explore all the issues associated with the organisation and administration of author collaboration. We shall limit ourselves to considering three problem areas: order of authors, author lists for major collaborations, and co-authorship and "number of authors".
Order of authors
We shall not attempt to provide an overview of practice, since practice varies from field to field, and within individual fields, from subdiscipline to subdiscipline. There have been many attempts to base authorship order on the principle that it should reflect the significance of the author's contribution, as APS observes, for example:
"Standards for determining the order of authorship vary widely from field to field. In most fields, having the lead author position is considered most desirable and likewise would be indicative of having made the greatest contribution to the paper. The lead author is often, but not always, the individual who took responsibility for writing the first draft. However, traditions vary from field to field and collaboration-to-collaboration, so it is difficult to generalize. If position on an authorship list is of concern to you, it may be wise to explore this issue early on in collaboration."
Author lists for major collaborations
The major collaborations that have taken place in modern science have sparked renewed debate on the subject of co-authorship. For example, several of the collaborations in the particle physics experiments at the new Large Hadron Collider (LHC) at CERN involve close on 2,000 participants from about 150 different institutions around the world. This gives rise to a number of questions: What meaning is there in a list of authors containing 2,000 names? Is it reasonable to assume that all these individuals have made such significant contributions to the scientific studies that they deserve to be named as authors? How is it possible to identify the leading forces behind the studies? How can publication lists like these be used to assess young research scientists for academic posts and research grants? These are not new questions for scientific research, and there has been well-considered and well-founded practice for a number of years. Nevertheless, the trend towards ever larger collaborations and, not least, questions and criticisms from peers in other fields, has led to these problems being considered afresh.
Two groups at the Commission on Particles and Fields (C11) of the International Union of Pure and Applied Physics (IUPAP) have been working to address these problems. One group has examined the issue of co-authorship, the other the question of how best to credit the efforts of young researchers. The first group points to the fact that the established practice of including all the members of the collaboration on the list of authors has arisen because the planning, structuring, execution and analysis of the experiments have taken place over a long period of time (10-20 years) and have required multidisciplinary expertise, so all the links in the chain must be intact for it to function. The group has also made a comparative study of publishing practice in major collaborations in other fields, and proposes supplementary publishing channels that can give due and justified recognition to a smaller, select number of researchers. It is however emphasised that it is the right and the responsibility of every collaboration to establish proper and appropriate publishing practice, and that this should be agreed and known among all those involved in the collaboration.
Co-authorship and number of authors
Multiple authorship has a direct influence on the "number of authors" system used by the Ministry of Education and Research to allocate its research funding, whereby the funding is divided among the number of authors involved. Thus, sole authorship counts more (for the individual) than multiple authorship. If all the authors are from the same institution, however, the institution gets the full funding. It is debatable whether the "number of authors" system does anything to encourage international research collaboration. If authors from several countries are involved, the weighting for authorship credit is correspondingly reduced. (In the case of CERN publications, it would mean dividing by 2,000; neither the number of Norwegian nor the total number of CERN publications can compensate for a reduction of this magnitude.) Also if a Norwegian author lists a foreign institution in addition to their own institution, the weighting for authorship credit is correspondingly reduced.
Aside from the "number of authors problem", there may be cause to question whether sole authorship grants higher professional status than co-authorship from a purely reputational point of view. Is it justified, and is it a good thing? Most professional assessments in the sciences point to the fact that the research environments are too small and that many are simply "one-man groups", thus greater collaboration is advised. Current systems and attitudes appear to be opposed to these recommendations.
As mentioned initially, there has been a tendency to introduce ever greater levels of detail into ethical guidelines for scientific publication, inter alia to prevent breaches of ethics in the conduct of research. Above, we saw how the degree of detail varies in the rules that govern different fields. This could be taken as an expression of the need for greater clarification in those fields. There is no real basis for saying that the variation in detail is owing to differences in awareness of research ethics, even though that has been claimed. Should such differences exist, they are apparently being erased over time, since the various guidelines all seem to be converging towards the Vancouver rules.
In any case, opinion will be divided as to how ethical guidelines should be formulated in detail. They must contain clear normative elements while also being capable of gaining broad acceptance among the research community. It is not enough for them to be based on "qualified guesswork". It is therefore important to start from scientific studies of scientific co-authorship, based on bibliometrics and surveys. There is little existing literature in this field, but it is a research field in development. We shall refer to some of this literature and discuss the significance it may have for formulating guidelines for scientific authorship.
E. Tarnow (2002) has studied the extent of "unjustified" authorship in relation to the guidelines issued by different journals. In a survey of over 4,000 members of APS, Tarnow asked them to state how many of their co-authors could be regarded as unjustifiably listed according to three different sets of guidelines (APS, ICMJE, "direct contribution"). In all three instances, the percentage of unjustified co-authors increased with the number of co-authors and reached a saturation point at slightly fewer than one hundred co-authors, at 23%, 67% and 59% respectively for APS, ICMJE, and "direct contribution". A large percentage of those surveyed (90%) responded that the available guidelines had not been used when co-authorship was determined. A comparable survey Tarnow carried out among pathologists produced similar results.
There can scarcely be any doubt that surveys of the type referred to above will be of importance both for raising awareness of existing guidelines and for raising awareness of the need to develop the guidelines further. They will also help prevent rules from being framed so bureaucratically that they fail to win acceptance among the research environments.
Despite the apparent tendency towards convergence in the understanding of the need for detail and for making the guidelines for co-authorship more strict, it is unlikely to lead to uniform practice for all disciplines. Their working methods are too different to make that possible. Therefore, it is important that the guidelines for the various fields should be known and understood among actors and users.
This article has been translated from Norwegian by Lesley Cawley, Akasie språktjenester AS.