The globalization of research, combined with improvements in technology, are resulting in ever larger collaborations among researchers. The average number of authors per paper is now around four (though in disciplines such as physics, the numbers can run into the hundreds or even thousands), but typically only the lead author gets full credit — for example, when it comes to the faculty review and promotion process. Other contributors who have provided essential support – from carrying out the actual research to collecting, analyzing, and managing the data – are at best acknowledged, at worst completely uncredited. A group of funders, publishers, institutions, and others, led by Amy Brand (VP Academic & Research Relations and VP North America for Digital Science) and Liz Allen (Head of Evaluation for the Wellcome Trust) is now working on a possible way to address this inequity through the creation of a contributor role taxonomy. Amy, who was previously Assistant Provost for Faculty Appointments and Information at Harvard University, has kindly agreed to answer a few questions about this new initiative for the Scholarly Kitchen.
What was the initial spur for developing a contributor taxonomy – did you have an ‘aha’ moment or was it an idea that developed over time?
Definitely more of an “aha” moment. I was reviewing a biostatistician’s tenure dossier a few years back and was struck by the different pictures of his scholarship as portrayed in the assessments of his peers, who argued that he was an outstanding scientist and the leading methodologist in his field, as compared with his publication record, which included exclusively multi-authored articles on which he was almost always a middle author (as opposed to first or last author). I had seen this pattern before and knew that this would be a difficult case for the presidential tenure committee to work though. That is, I understood that he was middle author only because he was the methodological lead and not the conceptual lead on the majority of his publications. I knew committee members would understand this as well after reading the peer letters. But I also knew that it would be difficult for them to discount the publication record, because they were trained to expect tenure-worthy professors to be in “leading author” position on most of their journal publications.
So I found myself wishing that there was a way for publishers to capture and display structured information about who contributed what to multi-authored works, instead of, or in addition to, the list of author names. Since I was very involved in the ORCID initiative at the time, and had also worked at CrossRef for many years, it occurred to me that if it was possible to create a controlled vocabulary of contribution tags, then those tags could be included as additional metadata in association with the DOI and, ultimately, with an individual’s ORCID.
What problem or problems do you hope to solve by creating such a taxonomy?
Several. First and foremost, this effort addresses the pressing need for a more fine-grained and transparent system of research and publication credit. If it is successful, there will be far fewer author disputes, and fewer disincentives for sharing data, for example. So it could positively influence both the collaborative culture of the lab, and academic incentive structures more generally.
Because our bibliographic conventions for representing authorship effectively lag behind the semantic capabilities of the web, they persist in obscuring the contributions of those engaged in collaborative research. The ways in which we apportion publication credit today are highly subjective and open to abuse, since name order in author lists has no consistent meaning. Yet who gets credit for research and discovery has a huge impact on people’s lives; it affects career advancement and tenure, and the transparency and integrity of the scholarly record too.
Who have you been working with, and what progress have you made to date?
The contributorship effort has been highly collaborative from the start. We are working closely with the Wellcome Trust evaluation team, and have some financial support from Wellcome as well. Publishers including Nature, Elsevier, PLOS, AAAS, and APS have been involved at various stages in the project. We also had some help early on from Access Innovations. And individuals from several institutions, including Cornell, Harvard, MIT, Columbia, NIH and NSF have been engaged as well. We are currently collaborating with both CASRAI (the Consortia Advancing Standards in Research Administration Information) and NISO in the development of the taxonomy.
What challenges do you still need to overcome?
We now have a core working group being facilitated by CASRAI, focused on fleshing out the taxonomy itself. We had previously tested a taxonomy with 14 roles and it was fairly well received, but we’ve known all along that we would need to refine it in consultation with a broader group of stakeholders, and that’s what we’re doing now.
Once we reach consensus on a set of roles and descriptions that works across the sciences, we will need to develop a selection of implementation models along with best practice guidelines, and then engage several partners in piloting the taxonomy.
I think the biggest challenge, though, will be cultural. Getting senior researchers to change how they think about authorship and credit is no small thing.
What are your goals for the next 12-18 months?
Our main goals for the coming year are to publish the taxonomy and launch a series of pilot implementations. We are planning a workshop in conjunction with the fall 2014 CNI meeting to preview our model and gain pilot commitments from members of the stakeholder community.
What changes would you like to see in how scholarly research is communicated and rewarded if this initiative is successful? Who will be the main beneficiaries?
I am thinking about younger scholars, and those research team members that tend to play non-traditional authorship roles such as coding and data curation.
I am hopeful that, with this initiative, we will ultimately take some of the politics out of authorship and credit because it will be transparently clear who contributed what to a publication. The senior person in the lab who obtained the funding, for example, can be tagged as having obtained the funding for the project without bumping a more junior person who did most of the research and writing – and who needs the publication credit for career advancement — out of prominence on an author list. Similarly, the lead methodologist can be called out as such and not simply buried in the middle of a long author list.
What sort of support are you looking for from the scholarly communications industry – how can we help?
Get involved! We are happy to include interested parties in the reviewer circle for this project, and we’ll be looking for implementation partners soon.
Last but not least, do you have a name for this project yet?
No catchy name or even a website yet, but stay tuned.
15 Thoughts on "An Interview with Amy Brand on a Proposed New Contributor Taxonomy Initiative"
It would be useful if we could see a draft taxonomy to get some idea of how the problem of roles in research is being approached. There are some interesting bibliometrics possibilities if this were actually implemented. The biggest challenge might be how a research team can come to agree on who did what? Lots of room for disagreement there. The complexity of allocation is also a potential problem. Suppose there are five authors and twenty roles. Each author might lay claim to ten roles or more. How is this helpful?
A very preliminary draft of the taxonomy was published in Nature a few months back — http://www.nature.com/news/publishing-credit-where-credit-is-due-1.15033. What is presented there is currently being revised and expanded via a highly consultative process. But we are under no illusions that what we end up with will cover all roles throughout all disciplines. And, yes, some contributors will have multiple roles; some roles will go unassigned; some truly equal collaborations won’t benefit from calling out contributions in this way. But I do believe that there is value in capturing a core set of roles that reflects new types of contribution and collaboration in a structured, machine-readable way. Which doesn’t preclude additional, field-specific free-text contribution statements. Finally, there are many possible ways the additional metadata could be captured and exposed — DOI and ORCID are key in all the scenarios we’ve considered thus far. Prototyping will hopefully get underway with interested partners in early 2015.
Thanks Amy! If I get a chance I will apply my taxonomy of confusions to your taxonomy of roles. http://scholarlykitchen.sspnet.org/2013/02/05/a-taxonomy-of-confusions/.
In terms of the science of science what you are doing is also very interesting. I have worked on taxonomies of science (e.g., http://www.osti.gov/taxonomy/) but not a taxonomy of doing science in this sense. It is quite a challenge. Just as people do not understand all the roles involved in being a publisher they may not realize what being a scientist involves, especially the collaborative aspect.
This could be potentially useful – just the other day, I was discussing the politicking that occurs in the biomedical sciences, where you often get collaborating clinical and non-clinical research labs, and order of authorship is really important. If you could have an official “first author” as well as “first clinical author” (etc.) it could save a lot of arguments.
The comment about the cultural challenge is spot-on, although it’s not so much the academic culture that would need to be changed I think – as soon as something like this becomes accepted at a managerial/administrative level as valid for assessment, it’d be rapidly adopted by academic authors. (Conversely, of course, if your institution’s promotions committee, say, still insists on traditional authorship order for judging contributions, then no-one will care that you can make authorship contributions more ‘granular’.) I also note that it wouldn’t have to be universally useful: I can easily envisage this being important to some fields in the future, and not others.
We already see cases where notation is used to indicate that several authors contributed equally, so there is no first author. This raises the issue of the taxonomy possibly including roles that may not always occur. Research is a complex and subtle case so taxonomizing it usefully will be difficult.
True. I guess I feel that it could be immediately useful in some fields, even if difficult/impossible to design a system that would be appropriate for all research papers.
Contributor indexing is an interesting idea, but not new. In 2001, Davenport and Cronin (http://dx.doi.org/10.1002/asi.1123) proposed such a taxonomy based on XML-markup, but sadly the idea never seemed to take off in spite of the available technology of the time.
Perhaps the problem is not technological but cultural. Contributions are highly contextually-dependent, as you describe in the case of the statistician. In addition, descriptions on one’s contributions can be fuzzy, somewhat ambiguous, and may not lead themselves to strict classification, which is what this contributor taxonomy proposes.
In the end, the letters of faculty support may be the best we have; that, and perhaps some open-text fields at the end of each paper that can allow authors to list their contributions without the confines of limited and fixed categories.
Another overlapping idea is transitive credit, as defined in http://dx.doi.org/10.5334/jors.be and as being explored in http://arxiv.org/abs/1407.5117 This idea requires the assignment of fraction credit to the authors, acknowledgements, and citations that contribute to the development of a paper, software package, data release, or any other scholarly product that is intended to be of use to others and is assigned a DOI of some type.
It’s clear that there are both social and technological issues involved in this. Some of these are being discussed in the context of the WSSSPE theme, with a next workshop occurring in Nov. See http://wssspe.researchcomputing.org.uk for more about this theme.
To me, technology is the easiest issue to address when dealing with credit.
Fractional credit is also an idea that has been discussed for quite some time and it assumes that authors (or a taxonomy imposed upon them) will be able to agree upon a method to divide up incoming credit (e.g. I was P.I., so I get 30% plus 10% for reviewing the paper, plus another 5% for suggesting supportive reviewers…etc.). In addition, it assumes that this kind of fractional counting is constant across fields. A experimental biologist may feel that the statistician should only be given 5% of total credit whereas a computational biologist may value the statistician at 20%.
While an author list may be interpreted many different ways, that may indeed be its strength, not its weakness. Ambiguity is viewed by computer scientists as a problem, but a sociologist may view it as a solution that allows scientists to work (and get credit) together.
This is a great idea, but from a workflow point of view there are numerous questions still to answer: Who has the authority to make the contributorship designations (submitting author, lead author, journal staff, etc.)? There are likely to be divergent, valid opinions about this. Do authors want to deal with this additional submission burden? Is there an expectation that individual contributors somehow confirm the designation(s) that have been assigned to them by somebody else? If they disagree, how is this handled by the journal? How should journals handle post publication disagreements concerning contributorship? How is the contributorship record corrected in primary sources, bibliographic databases and IRs? Who will pay for the extra administration associated with this? Hmmm.
It would be helpful if illustrators and photographers could also be given credit for their scholarly contributions to an article using a taxonomy/metadata schema. With the rise of Creative Commons licensing, figures and data visualizations are often mistakenly credited to the first author rather than the medical / scientific illustrator who created the image and who often owns the copyright. If a picture is worth a thousand words, then the value of that image is worthy of proper attribution.
I suspect a lot of this sort of work is done under “work for hire” terms, which may not require attribution.
Not necessarily. In STM publications, medical and scientific images are often commissioned from independent freelance illustrators and visualization companies who create figures for journal authors. It’s becoming more common for universities / foundations with communication departments to retain and control copyright to creative works (e.g., Mayo Clinic, Cleveland Clinic, Mayfield Clinic) as a means to sustain their services to faculty. Especially to control derivatives.
If we are discussing giving the statistician credit, then the academic honesty and integrity of the photos and figures are equally important. Indeed, the figures are often the most visible expression of an authors work and used / cited extensively in presentations.
Where I have had drawings made for publications, it’s been done either by an independent illustration firm, or by my university’s graphic arts department. In all cases, it was done under a work for hire agreement–the services rendered were paid for, and all intellectual property around the illustration transferred to the person who paid.
If any service wanted to retain copyright, I probably wouldn’t use them–it’s easy enough to find another that doesn’t require this. If it was under a separate copyright, then I need to go through the hassles of obtaining permission when I publish the work, and if I want to publish it under an open access license, I have to go to great pains to make it clear that while the paper is available for reuse, the figure can not be reused by others. Not really worth the hassle when easier alternatives exist.
But you raise an interesting question about how far we must go in offering credit for contributions. I don’t think I ever knew the name of the person who did my illustrations. Should they be included in authorship of the paper? What about technicians in a lab? Different labs have different policies about whether they are eligible for authorship on a paper (treated differently than a graduate student or a postdoc). If a technician does some of the routine busywork that contributed to the paper, what credit should they receive? Must we credit the vet tech who changes the water bottles on the mouse cages? The technician who poured the agar into the petri dishes used? The worker at the chemical supply house who isolated the Sodium Chloride that was used to make the saline solution used? The secretary who did a copyedit on the manuscript?