Authorship is sometimes described as the currency of the research community. Publication is a competition tool in the fight for tenure. They are required for gaining merit, are important for those applying for research funding, and governs how funds are allocated to departments.
For these reasons alone, it is crucial that the right persons are listed as authors of works to be published. But it is well-known that this is not always the case.
This was shown in a survey aimed at persons who were awarded doctorates from Swedish medical faculties in 2009 and 2016. In 2009, 30 per cent of the respondents stated that their thesis contained articles where persons were included on the author list without having contributed to the research in such a way that they should be included according to normal ethical guidelines for publication. In 2016, 39 per cent responded in the same way.
“This problem persists. It hasn’t got any better, despite more teaching about research ethics, and despite this now often being a mandatory item in third cycle education. My view is that there is quite a lot of cheating about co-authorship going on internationally too, not least in medicine and other fields where research collaboration is common,” says Gert Helgesson, Professor of Medical Ethics at Karolinska Institutet, and co-author of the book Publiceringsetik.
Research is conducted in different ways
The reasons for deviation from the rules vary. They can be about traditions, exercise of power, or pragmatic reasons – such as someone needing one further publication for a docentship or associate professorship.
In some academic fields, issues about co-authorship are more topical than in others. One explanation can be found in how the research is conducted. Research in social sciences and humanities is often conducted and published by individuals or small groups.
In medicine and natural sciences on the other hand, collaboration is standard. Five or six authors is not unusual, and many more may be included. In particle physics, major international collaboration is now common, and there are articles with more than a thousand authors.
“In general, the number of authors of scientific articles has risen over the years. And the more fields of knowledge are linked together, the more this increases.”
Four requirements must be fulfilled
Currently, there are no generally accepted, overarching rules for defining scientific authorship. There are international guidelines, of which the best known are the ‘Vancouver Rules’ (see fact box). They were formulated by the International Committee of Medical Journal Editors (ICMJE), but are today also used in natural sciences, and sometimes in humanities and social sciences.
The Vancouver Rules state four requirements that must all be fulfilled for a person to be counted as an author. The first rule is that the person must be responsible for a substantial contribution to the idea or design, collection of data, or analysis and interpretation of data for the work.
What counts as ‘substantial’ is not defined, but the guidelines still show that certain inputs are not sufficient as grounds for authorship. Examples are the person who only secures funding, or is responsible for general leadership of a research team.
Authorship is also linked to responsibility for the published research, which is touched upon in the fourth requirement of the Vancouver Rules. But this criterion is unclear, according to Gert Helgesson. It can either be read as all authors being held responsible for everything in the article. Or it can be read as all authors being responsible for facilitating any investigation if the the article is suspected of containing irregularities in any part.
The best interpretation
A third possible reading is that the responsibility for investigating any irregularities applies to all, and throughout the work on the study, not just if the published article becomes the subject of investigation.
“I think this is the best interpretation – that if you find out something that might be associated with irregularity, you must bring up this suspicion in some way. But the guidelines should be clarified, so that it everyone knows what applies.”
He thinks that responsibility for the published text should relate to the work that each person has done. Requiring somebody to be held responsible for everything in a study is not realistic, Gert Helgesson considers. Not even the person leading the project.
“It’s not always possible to have that level of insight. After all, one reason for collaboration is often that different groups have differing competences, and sometimes you really don’t know what the others are doing. But certain procedures increase the chances of spotting irregularities, and it could be argued that a research leader should create ways of working that make it more difficult for people to sneak in irregularities.”
Many journals today state that they support the requirements for authorship set by specific guidelines, or have formulated their own rules. But it is difficult for journals to check whether a researcher has actually contributed to the work in the way they claim, and Gert Helgesson says that journals largely leave this task up to the researchers.
Cannot mediate in disputes
This picture is confirmed by the editors of the journal Science:
“It is our responsibility to clearly state the criteria for authorship, and to provide tools for authors to state their contributions [to the work, ed. comment]. We cannot mediate in disputes about authorship. We believe that this is the responsibility of the research institutions, as they have access to lab books and other data that can clarify the contributions and be used to determine whether they match our criteria for authorship,” the editors of Science write in a comment to Curie.
Before an article is accepted by Science, all authors must fill in an author form, and confirm that they fulfil the journal’s authorship criteria. The authors must describe their contributions to the article in the credit system known as ‘contributor roles taxonomy’ that is used by an increasing number of journals. This divides up the work on the origin of an article in different items, and each co-author is credited with the roles that they have contributed.
Special undertakings apply for the corresponding author. A senior author from each team shall also have scrutinised the raw data the team has produced. The editors of Science reviews the author forms to see whether the answers are reasonable.
The entire process is, however, made more difficult, as the definition of authorship can vary between different scientific fields. A contribution that justifies a person being included in the author list in one discipline may only receive an acknowledgement in another discipline, and no mention at all in a third.
“The differences are deeply rooted in the culture and practice of the disciplines. It is therefore very difficult for a broad, multidisciplinary journal, such as Science, to determine across the entire spectrum whether the right authors have been included in every case,” the editors of Science observe.
Encourage common rules
The authorship rules for Science are based on recommendation in an article published in PNAS 2018, which encourages journals to adopt common rules to clarify what is expected of authors.
The article reformulates the authorship criteria of the Vancouver Rules to suit more scientific fields, encourages use of the credit system, and lists the responsibilities that should rest with the corresponding author.
“A list of author responsibilities, in particular those of the corresponding author, has in my opinion been lacking for too long. At a time when authors are keen to share the limelight when an article receives positive notice, but try to avoid it if the article encounters problems in terms of quality of data or analysis, it is important to know who is responsible for which parts of the process,” Marcia McNutt, the primary author of the article and Chair of the National Academy of Sciences in USA, writes in a comment to Curie.
Several journals now follow the recommendations in the article, in part or in full.
Most are written by one author
For the editorial committee of the Nordic Journal of Educational History, auhorship is not a major issue establishes Henrik Åström Elmersjö, Docent of History at Umeå University and responsible for the committee’s work. Almost 80 per cent of the articles in the journal are written by a single author, and they have not seen any increase in the number of articles with several authors.
“In educational history, it is fairly unusual to write articles with more than two authors, and in those cases it is usually immediately obvious why there are several – they are often comparative articles, with authors from two different countries, for example. But we are aware of the general movement towards increased co-authorship, and have formulated rules for authorship to enable us to handle this,” says Henrik Åström Elmersjö.
The journal’s rules are based on the Vancouver Rules, and place particular responsibility on the corresponding author to ensure that the right authors are included.
Challenges of co-authorship
Johannes Westberg is Professor of Pedagogics at Örebro University. His assessment is that co-authorship ahs become more common in his areas – pedagogics, educational sciences, and history – but observes that it is usually small collaborations, and that it is still not unusual to work alone.
Johannes Westberg has himself published research articles together with others, and regards it as an opportunity to conduct more multi-disciplinary research by combining his own competence with that of others. But he also sees challenges:
“Co-authorship should be an engine for increased quality and new collaboration, it must not become an engine in the rush to publish. The reason for co-authorship should not be to add more articles to your CV.”
In his role as reviewer of research ahead of doctoral degree awards and job appointments, Johannes Westberg also sees a problem with assessing the value of co-authored articles.
“Independence and originality are important grounds for assessment. If someone always co-authors, the question arises whether they really can write articles on their own. It is also important, not least for doctoral students, to have the opportunity to formulate and publish their own research,” he adds.