All for one or one for all? Authorship and the cross-sectoral valuation of credit in nutrition science

ABSTRACT

The passionate pursuit of authorships is fuelled by the value they represent to scholars and scientists. This article asks how this value differs across scientists and how these different processes of valuation inform authorship articulation, strategies, and publication behavior in general. Drawing from a qualitative analysis of authorship practices among nutrition scientists employed at universities, contract research organizations, and in food industry, I argue that two different modi operandi emerge when it comes to authorship. These different ways of working produce different collaborative approaches, different credit distribution strategies amongst collaborators, and different value placed upon (the pursuit of) authorship. These different valuation processes are neither explicit nor recognizable to those reading (and judging) author lists. As a consequence, in the politics of authorship, the names standing atop a scientific publication in nutrition science represent different types of value to both the individuals and employing organizations.

KEYWORDS:

• Authorship
• value
• credit
• public-private collaboration
• collaboration
• nutrition science

Introduction

Nutrition science journals publish articles by authors from universities, governmental research organizations, food companies large and small, as well as contract research organizations on the same pages without actively discriminating between them. After all, research on nutrients, foodstuffs, foods, and ingredients happen at all of these organizations and dissemination of scientific results in the form of scientific publications is standard practice for all of them. Authors on a single article may even stem from two, three, or all types of organizations, depending on the structure and location of the project that resulted in these publications.

Scientific authorship has drawn attention from sociologists of science, science and technology studies scholars, and bioethicists for decades. The combined body of work ranges from social and ethical questions on the applicability and power of existing moral and legal frameworks and guidelines (Biagioli 1998; Kovacs 2017; Marušić, Bošnjak, and Jerončić 2011; Shaw 2011; Street et al. 2010), through constructivist linkages between authorship and knowledge-making (Collins 1975; Latour 1987). More recently, the role authorships (and the way in which they are counted) play are explicitly taken into account in the context of evaluation practices, as well as how such evaluations discipline specific behaviors in scientists (Hammarfelt, De Rijcke, and Rushforth 2016; Rushforth and de Rijcke 2015). Rather than merely acting as tools of distributing credit, responsibility, or value, contemporary authorship practices in science also present a problem.

Scholarly work on authorship firmly ties together authorship pursuits with credit gathering and distributing strategies to advance careers, or to gain influence (Fochler, Felt, and Müller 2016; Müller 2012; Seeman and House 2010a, 2010b). Normative structures guiding careers coexist with norms that govern authorship attribution and credit weight resulting in conflict and moral pressures (Kovacs 2013; Shaw and Elger 2017; Youtie and Bozeman 2014). The well-known publish-or-perish and grant-or-perish are, of course, part of this more complex array of pressures. Whether these pressures are “real” and how they shape or influence the organisation of science, as well as the epistemic content of science, is increasingly becoming object of scholarly debates. For instance, Fanelli and Larivière (2016) demonstrate quantitatively that researchers have not started writing more over the years, at least when normalized for the rise of authorships per paper, which leads them to question assumptions that publish-or-perish pressures lead to scientists cutting corners. That does not mean, however, that pressure to publish is not experienced in practice. In a mixed-methods approach on the issue, Tijdink et al. (2013, 2014, 2016) argue that such pressure is prevalent in practice, up to the point that it harms the mental health of (bio)medical researchers. Müller and de Rijcke (2017) dive into the epistemic consequences of pressures coproduced by researchers and the evaluation regimes they are subjected to. They argue that assessing worth based on metrics (of authorship) influences and narrows the epistemic trajectories researchers consider viable.

However, those pressures and consequences are situated, as Jabbehdari and Walsh (2017) argue, in the sense that authorship practices, attribution, valuation, and articulation of authorships differ across disciplines. Complementarily, Paul-Hus et al. (2017) demonstrate that authorship practices extend into acknowledgments, and that the boundary between these two crediting practices differs across disciplinary boundaries, meaning that across disciplines and sectors the line between the amount and type of work that gets you authorship and the amount of type of work that gets you mentioned in the acknowledgments is drawn elsewhere. Nonetheless, whether credit serves the same function to all types or researchers in terms of career status, or employment situation and, as a consequence, how credit is valued in the light of diverse scientist biographies (across or even within a discipline) remains largely opaque.

Accordingly, this article studies and reports the valuation of credit distribution through authorships in nutrition science. Drawing from an ongoing series of interviews with junior, mid-level, and senior researchers in nutrition science across Dutch universities, government research facilities, (contract) research institutes, and the food industry, this article touches upon the developmental trajectories of authors and careers in the diverse institutional landscape in Dutch nutrition and the moral evaluations of the self as author. It explores the different articulations active researchers offer regarding the act of authorship and its consequences, its organizational role, and how it is tied in with research evaluation and audit, career development, and local responsible research practices (RRPs). In general, it aims to contribute to our understanding of the social, economic, and organizational context of authorship.

I will first describe the qualitative research approach followed to collect data for this study, as well as the sensitizing concepts informing the analysis. I then present two ways in which authors pursue credit and assign value to credit, followed by a discussion of the salient differences between them, and how both misalign with ideal norms for authorship as forwarded in authorship guidelines.

Approach

This article draws from a series of 16 interviews with professionals who self-identify as nutrition scientists and who are actively publishing or who have participated in academic publishing in their career.

Respondents

All respondents self-identified as nutrition scientists through a paid membership of the Dutch National Academy of Nutrition Science (Nederlandse Academie van Voedingswetenschappen, NAV). Even though many Dutch nutrition scientists list their membership on their resumes, compiling a full list of NAV members is not possible. The NAV does not disclose membership lists for privacy reasons, also because all memberships are on an individual title, not representing employers. Those employers range from Dutch and foreign universities and government agencies, to Dutch and foreign food industries, contract research organizations, and more, including self-employed nutrition scientists. Respondents were recruited by enlisting the help of the NAV administrative office. They contacted random members via email, all with an identical standard summary of the research proposal (provided they fit the desired distribution across organization type and seniority) and invited them to participate. The NAV only put them in contact with me if they agreed to participate (10 of 16) and some reached out to me themselves (6 out of 16). They informed me about the amount of people that declined, but not their identities. The NAV and I recruited respondents in three rounds of invitations, to allow the desired distribution across organization type and seniority to inform preselection for the next series of recruitees. A total of 35 invites resulted in 16 nutrition scientists willing to be interviewed (46% response rate).

All respondents are active nutrition scientists. However, the field of nutrition science is itself not homogenous and is very interdisciplinary. Twelve out of 16 respondents did not commit themselves to a single subdiscipline or identify any subdivision underneath the “nutrition science” label as relevant. Those who did, mostly mentioned nutritional epidemiology (2) or molecular nutrition (2) as a secondary epistemic division. Out of the 16, 11 were female and five were male; six respondents worked at a university, three at a contract research organization, one of which with strong government ties, and seven worked in industry, ranging from big multinationals, to a small consultancy firm. Five self-identified as junior, six as mid-career, and five as senior scientists, distributed across all organization types (see Table 1). All researchers were published authors, with amounts of published peer-reviewed works ranging 1–231 at the time of the interview.

Table 1. Respondents.

Interview Code	Organization Type	Gender	Seniority
IM1	Industry	F	Junior
IW1	Academia	F	Junior
IW2	Academia	F	Junior
IT1	CRO	F	Senior
IV1	Industry	M	Senior
IU1	Industry	M	Senior
IU2	Industry	M	Mid-career
IW3	Industry	F	Mid-career
ID1	CRO*	F	Mid-career
IT2	CRO	F	Junior
IW4	Academia	F	Senior
IS1	Academia	F	Mid-Career
IS2	Academia	F	Senior
IB1	Industry	M	Senior
IB2	Industry	M	Mid-career
IB3	Academia	M	Junior

* This particular CRO works exclusively for the Dutch government and has no industry ties.

Interviews and analysis

Interviews were semistructured around preexisting sensitizing concepts drawn from literature on authorship and research integrity. In the interviews, the interviewer and the respondent reconstructed individual biographies centring on team work, collaboration, authorship, responsible research, and career trajectories, requirements, and desires. All interviews were conducted with the same topic list and questionnaire (55 questions). The sequence of the questions was adjusted to the flow on the interview and regularly questions were skipped because they were answered before being asked. Initial questions were phrased as open questions (e.g., Why do you publish? How did the trajectory towards your last publication look like?) with ad hoc follow-up questions, both open and closed.

These relatively long interviews lasted between 90 and 130 minutes. All interviews took place at a site selected by the respondent. These were their offices (n = 8), their homes (n = 3), a restaurant (n = 2), or the interviewer’s office (n = 1). Two interviews were conducted via Skype, also on the respondents’ request. All interviews were conducted in Dutch and were recorded and transcribed verbatim afterwards. The interviews were conducted between May and December 2016, and respondents were all given the opportunity to check the transcripts afterwards. This check lead to a small number of factual corrections, especially in references made to published works, collaborator or competitor names, and times and dates of events. Only the quotes used in this text were translated into English.

The analysis was informed by a series of sensitized concepts drawn from previous work and existing literature on authorship and research integrity (Bowen 2006). These were (1) collaborative working relationships (Parker, Vermeulen, and Penders 2010; Penders, Vermeulen, and Parker 2015; Perneger et al. 2017; Wuchty, Jones, and Uzzi 2007), (2) (e)valuations of credit (Hammarfelt, De Rijcke, and Rushforth 2016; Rushforth and Sarah 2015), (3) links between authorship(s) and career advancement (Fochler, Felt, and Ruth 2016; Müller 2012), (4) authorship in the context of research integrity (Cutas and Shaw 2015; Penders 2017; Perneger et al. 2017; Shaw 2016), and (5) public-private partnerships in science (Okraku et al. 2017; Penders and Nelis 2011). These sensitizing concepts informed a series of predetermined codes¹ used to pre-order and analyze the interview transcripts. The empirical material subsequently informed an iterative refinement of these codes, after which a number of salient issues emerged. Salience is here understood as the prominence of an analytical issue in the empirical material regardless of the position on the issue (e.g., vivid disagreement counts towards salience as much as consensus). These were the misalignment between ideal and practice, and especially the ways in which this misfit was different across organization types, different valuations of credit across individuals and organizational types, and the moral stress associated with competing normative structures.

During and following the analysis, I organized informal and formal member checks. These were first informal, through continuous conversations with the nutrition science field, even up to the point that I participated as an actor myself in authorship conversations with nutrition scientists honoring the occasion of a joint publication (Penders et al. 2017). Second, I provided a narrative account in the form of a preliminary version of this article to three members of the nutritional research community, two of the respondents (IB1 & IB3) and one member outside the sample outlined above (Cho and Trent 2006; Lincoln and Guba 1985).

Results

Nutrition scientists offer diverse testimonies on the role and value of authorship, or the pursuit of it, in the context of their work and careers. From these testimonies, however, we can identify two dominant trajectories of knowledge making. Within each of these, collaborative forms, career narratives, and assessments of the value of authorship(s), as well as the credit distribution that go with them, shape one another.

Most respondents (14/16) argue that writing scientific publications is part of their job because it is the preferred way to communicate research findings. Researchers refer to their professional responsibility towards research subjects, as well as to a professional responsibility to publish all research, accessible to peers and possibly the general public when they explain why. However, how this process of publication takes place, what it actually means to the participants, and how researchers articulate the value of authorship, as well as the credit that comes along with it, in general or in the context of their careers, is another matter.

One for all

In academia, authorship, careers, credit and value are strictly tied together. Below, I will first show how authorship is considered a shared attribute by junior and senior academics alike. However, the size of the group sharing authorships matter with respect to the value attributed to a particular authorship. Next, I show how in individual biographies, processes of sharing authorship and valuing authorship result in difficult and political negotiations and processes of socialization in which issues like career perspectives and the desire to pursue specific collaborations all factor in.

Ph.D. student² IW1 is about to defend her thesis and advance to a postdoc position. Reflecting on the papers she wrote and published for the benefit of her thesis, she considers authorship to represent too much credit to claim solely for oneself: “Imagine that I were to have set up and conducted something all by myself, even then, I wouldn’t dare be an author all by myself. I do not feel I have enough authority to say things that way (IW1). She prefers collaborative publishing to share responsibility, to share credit, but also to amass more weight behind the claims in the publication, to grant them credibility. Faculty member IS1 agrees. She argues that “If there is only one or a few authors on the paper, one wonders what is going on. I tent to find it suspicious […]. Especially if it is experimental work, I simply do not find that plausible” (IS1).

When it comes to the credibility or authority of published work, a longer author list helps. However, when it comes to those who contributed to the work and have to share credit, preferences may swing the other way. Postdoc IW2, for instance, argues that she prefers papers with fewer authors on it: “I liked it, the feeling that is was more mine; that I earned it more myself … [that] my share was bigger” (IW2).

Authorship per paper is thus not only a matter of quantity. Scientometric research indicates that papers with more authors are cited more (Valderas et al. 2007; Wuchty, Jones, and Uzzi 2007), suggesting more authors on a paper would maximize credit for the work as a whole. Simultaneously, however, researchers are in pursuit of a paper that is “theirs.” Early career researchers in particular, IS1 argues, are startled when they find out that a paper they considered to be theirs, turned out to list way more authors than they expected:

It is their paper and they are just first author. That first authorship [usually] is not up for much discussion. However, they often arrive at this question “we are doing this, but who are these other people on this paper whom I do not know, why do they have to go on it?” […] Sometimes there are people on it who have done data analysis or data preparation, invisible to them. We do encounter that. Those who oppose that, you’ll have to educate them more actively. Some just go along. (IS1)

The education IS1 is referring to in the quote above is how to deal with the mismatches between existing authorship regulations (whether at the level of the organization, at journals, the International Committee of Medical Journal Editors [ICMJE] or at the Committee on Publication Ethics [COPE]) and real-life practices of writing. Education, here, means a process of socialization into the politics of authorship. Most of our respondents working at universities (10/11) did not know the overall content of authorship guidelines, if they even knew they existed. This extends across guidelines issued by their home institutions, the Dutch federation of Universities, as well as, for instance, ICMJE authorship guidelines. Ph.D. student IW1 even argues that she does not know their content, because that content does not matter in her day-to-day practices: “I do not believe that it matters what these regulations say. Stupid or not, who fights the hardest or the smartest get on a paper and in a spot. Yes” (IW1). Faculty member IS1 even argues that such authorship regulations hamper research collaborations: “It is not so easy to set up collaborations if you can only award authorship according to the rules. […] If there is no reward, there is no interest to collaborate. For instance, we usually do not have money to pay for data access. A cheap way is to get the data for free, but allow the other party to be an author” (IS1). Faculty member IS2 confirms this and labels the practice accordingly: “In our discipline authorships are scientific currency. You can buy access or data with them” (IS2).

That does not mean that authorship criteria are of no concern in nutrition science. Industrial researchers are well aware of their content, but even in academic circles, such criteria influence team work. Rather than describing the type of work that results in authorship, in practice the temporal dimension is reversed, and being awarded authorship stimulates people to at least try to live up to the criteria in as much as they can, for instance by engaging with draft texts critically and providing detailed feedback. This, however, seems to be the exception rather than the norm. Researcher IS2 recalls a recent conversation with one of her Ph.D. students as follows:

I find it terrible to have to tell my Ph.D. student, who has just sent his first paper to a set of senior co-authors he doesn’t know and he says that almost no one responded. That I have to tell him that this is normal, I find that severe […]. Apparently, that is normal, if you are high up enough, you don’t have to anymore… (IS2)

To allow authorships to be commodities to be traded for access, and as being up for grabs among the privileged only to deflate their value for junior scholars, sadly resonates with existing literature on authorship and career trajectories in academic research (Biagioli 2000; Biagioli and Galison 2003; Jones 2000). For authorships to be used in data (and other) trade agreements, also requires a steady output of papers. Faculty member IS1 explains how the pursuit of authorships, the genesis of academic papers, and the organization of science overlap:

For our projects we have defined [one or two] primary outcome measurements. I believe we have also defined fifteen secondary outcome measurements. That will become a total of about seventeen papers. […] During the course of the [4 year] project, we can never write seventeen papers. We only get the results at the end. During the course of the project we can write three or four. That we try to distribute a bit, so that everyone’s authorships fit their expertises […].The postdoc or Ph.D. student is finished and there we are, with thirteen papers unwritten at the end of the project. What doesn’t happen is that the PI, that I write them up myself. What happens is that everyone looks around […] for people to analyse that data and write it up […]. If I have a masters’ student lined up for a Ph.D. grant for three years, then okay, they can do it. (IS1)

The distribution of authorships is then informed by the availability of successor students lined up for writing tasks. Seventeen papers allow for 17 series of authors. Who becomes author and goes in which position is highly dependent on who collaborators are, their authorships conventions and the value that is placed on them attaining authorship. As far as authorship conventions go, senior researcher IW4 reveals difficulties in negotiations with international researchers: “The French, they want scientists as first author, not students.” IW4 works in a lot of international collaborations and has informally mapped different conventions and customs around authorships to facilitate negotiations: “In Africa, it is different again, […] because there, eh, it is more about status. Supervisors are on the paper regardless of what they did and students, even PhD students, are not listed because they are simply too low in the hierarchy, you know” (IW4). IS1 works together with China a lot, and she recalls that “With the Chinese there is little to negotiate. The senior is on the paper, even if he doesn’t speak a word of English […]. If we use Chinese data, some of them get on the paper, […] but how they decide upon order over there, I don’t know. We just get names” (IS1).³

Value placed on authorship is also assessed in the context of careers. IS1 for instance argues that “[Authorships] also depend on the ambition level of the people involved. If I have a new Ph.D. student and a postdoc working on a project, but [the postdoc] has already decided not to pursue a scientific career, I will push for the Ph.D. student getting first authorships. The postdoc will then have to settle for second place. My argument will be that, well, you don’t need it anymore” (IS1).

All for one

Authorships are central in the infrastructures of credit in universities, where nearly all staff covet them. In industrial and contract-research organizations, authorships take on different types of value. In what follows, I will show that authorships have market value in industry, primary to the companies and for a lesser degree to the individual scientists. This changes the way in which credit is distributed. Further, industrial authors (have to) subordinate their individual credit to the credibility industries wish to assemble for their claims. This leads to frustration, but also to a different position of authorship as a transitory phase of ownership.

In the context of nutrition science in the Netherlands, professionals can transition relatively easily from one sector into the other and back again. For instance, out of the 10 nutrition scientists interviewed for this study who work in industry or at contract research organisztions, seven had a (brief) career (beyond being a Ph.D. student) in academia before that. I have described authorships as marketing technologies in the context of industrial research before, arguing that allowing corporate researchers to appear as authors on papers strengthens the image of the organization as a science-based employer, able to attract new high-quality research staff, as well as positioning the organization as a serious partner in future research collaborations (Penders and Nelis 2011). A similar logic applies to contract research organizations and the researchers working there, as illustrated by researcher IT2 explaining her pursuit of authorship: “Yes, [it matters] for [a contract research organization] to get work and projects […]. [We] work for other organizations and they do google me, to see who they are hiring” (IT2). Researcher IB1, who is the sole employee of his own nutritional consulting company, asserts that reputation is key to keeping his company afloat: “As a result, until now […] I have been approached by clients. I have not needed to do any active acquisition” (IB1).

The pursuit of authorship among researchers in industry and at contract research organizations, however, manifests itself differently. For instance, not all food industry researchers are actively pursuing authorships. Senior industry researcher IV1 says, “Well, yes, that is not something we are rewarded for at the end of the year” (IV1), or as IW3 summarizes it, “Nice, but no bonus. Next to a lack of financial reward to individual corporate researchers, the value of authorship is also connected to the credibility of the organization, especially in the context of collaborating with universities: “In all cases in which I work with external academics, the academic party always has the final says […] on whether we can or cannot be co-authors” (IV1). Another corporate researcher argues that “if you work with an external [academic] partner, you don’t want two thirds of the author list to be industry” (IW3). Next to being listed as author, author position is also a sensitive topic when it comes to public-private collaborations, as a senior researcher from another company explains: “This has to do with credibility […]. As I said before, we have the rule that the first author is not a [company name] person” (IU1).

In the interest of protecting the credibility of publications, industry agents are limiting the authorships they claim on papers, and limiting authorships to less valuable positions. However, some contract research organizations and companies do use internal infrastructures to compensate for less public credit by boosting internal credit:

Yes, well, when someone publishes something, it is circulated internally, through newsletters of departments. They summarize who managed to contribute to what. Our intranet also lists who is on which publication, so it does make your name more visible inside the organization. (IT2)

Despite this potential career contribution, a lack of public credit is still evident. Especially amongst the more junior researchers in food industry, this can be frustrating. IW3 argues that “at a university, it is arranged a bit differently, that someone who knows something about a topic, that they can get co-authorship quicker” (IW3), to which IU2 adds as follows:

We do not want more than a certain number of people on a publication, regardless of work load […]. The resistance to this is the personal ambitions of researchers who work here. It is madness that you do most of the work, and that John or Pete from an external university, a person who did a lot less work, ends up on a more prominent position. That does not feel fair. Then you tell yourself, okay, I work at a company and we have that policy. Yes, then you’ll have to respect that. (IU2)

Food industry offers more competitive financial rewards,⁴ yet the lack of reward in the form of credit still irks researchers employed there. This shows that authorship has value to scientists working at companies, yet in a different way, they are invested in their products and writings differently. IW3 argues that “my experience is that at the university I published something and then we threw it over a hedge into public space and it disappears. The next step is ignored, you don’t see it. Now, here at the industry, publishing is the first step of a long trajectory” (IW3).

Discussion

The exact content of authorships rules or guidelines is not at the center of attention for nutrition researchers, in line with, for instance Bhopal et al.’s (1997) study of authorship practices in British medical researchers. This is in remarkable contrast to more recent results reported by Nylenna, Fagerbakk, and Kierulf (2014), reporting high awareness by Norwegian medical researchers.⁵ Despite many nutrition researchers not being aware of the exact content of authorship guidelines, they are aware that research practice does not align with these rules (whatever their details are). And indeed, ICMJE guidelines on authorship (which are also used by many nutrition science journals) prescribe significantly different authorship criteria, formalizing minimal required tasks and contributions to qualify.⁶ Strikingly, when asked whether or not nutrition science had any shared norms on authorship and if so, what these might be, all agreed that the first author position is the most important, with second and last place following suit---but across all interviews little agreement existed on anything else, suggesting very local articulations of these norms. This was also recognizable in the researchers’ struggle with the author lists of the papers they read and their difficulty and diversity in understanding what these authorships lists represented in terms of work, credit, and responsibility.

Especially among industry researchers, individual journal and ICMJE guidelines were mentioned, even if respondents were not always fully aware of their content. Industry researchers were very aware that company conventions or policies that suggested minimizing or limiting the number of “their” authors on collective papers may contradict these guidelines. In this labyrinth of guidelines, local conventions, and practices, many experienced difficulties navigating their way. In the context of corporate research, not all contributions are caught through authorships, leaving labor underacknowledged. In the context of academic research, authorships not only represent intellectual or experimental labor, but also overacknowledge the outcomes of political and resource-motivated negotiations, resonating with other analyses reporting on ghost and guest authorship (Gøtzsche et al. 2007; Shaw and Elger 2017; Teixeira Da Silva and Dobránszki 2016).

Across the board, authorships and the credit they are associated with represent value for the individual (career) and for the organization in the form of reputational work (Whitley 2000). However, the distributions between the two differed remarkably. Companies actively (or implicitly) excluding their scientists from a first-author position and limiting the amount of their people on papers limit the capacities of individual employees to amass credit. In practice, it means that companies push forward a limited number of their own scientists to build a reputation through authorship. Taken to its extreme, this means concentrating as much credit as possible in only a few experts---actively catering to the Matthew Effect in science (Merton 1988). Successful individual careers in food industry research do not necessarily require (many) authorships, despite, as indicated above, frustration among some researchers. Companies and contract research organizations, however, do require authorships to establish identities as science-driven partners and employers. Contract research organizations take up a similar approach in their quest for clients. On the contrary, in universities, individual careers do require authorships. Some of the interviewed authors have hundreds of publications on their resumes. Similarly, their institutes and organizational units require authorships to build or maintain their prestige.

Credit distribution and the construction of value, even when collaborating on exactly the same research project, look remarkably different between universities and companies, resulting in two modi operandi pertaining to authorship. Following Hammarfelt et al.’s (2016) argument of the gamification of academic selves, I will opt for the use of two metaphors in competitive sports to highlight the differences between both trajectories.⁷ Competitive sports fits both the narrative of intense competition and that of the goal of continued self-improvement. However, not every athlete competes in the same way, and especially in the context of team sports, collaboration and credit-distribution, and well as the respective value assigned to credit, differ wildly---as do results. In the context of nutrition science, nutrition scientists are participating in two different games on what appear to outsiders to be a shared field. The rules governing credit distribution among team members, as well as the consequences for valuing that credit by team members differ remarkably (Duch, Waitzman, and Amaral 2010).

Authorships and the value they represent for universities and academics are located on both the individual and “team” level. Universities compete with one another through rankings and metrics in which, next to a series of other proxies for research quality, authorships feature prominently (Hammarfelt, De Rijcke, and Wouters 2017). Simultaneously, the individual profiles of researchers are of value to themselves, preparing for transfers to other institutes, promotions, pay raises, and most often, grant applications. In a way, academics are playing professional football, a sport in which simultaneously individual players and teams are scored, compared, and tested against one another. Each individual has the potential to aggregate value for the team and for herself or himself. In parallel, nutritional scientists working for contract research organizations or food industry are also pursuing authorships because they represent value. However, individual profiles are less important and often serve team interests. That is, however, not the case for everyone. Contract research organizations and companies want to employ a few people with good (even star) profiles as researchers. They are competing in the style of a professional cycling team, in which most riders (or domestiques) serve the victory of the team and that of a single (or few) leaders.

The pursuit of credit (through authorship) is intricately tied to evaluation regimes to which laboratories and scientists are subjected. As a consequence, different credit pursuit strategies---different games---will also translate into different epistemic footprints and different evaluations of their career. Müller and Sarah (2017) already expected that evaluation practices contribute to local pressures that might make academic careers less interesting. Industry research IU1 offers clear support for this position: “I have to say, that eh, with my transfer from university to industry, that I heaved a huge sigh that I got rid of that, rid of the fighting and [rid] of that two percent chance to get the right grant” (IU1).

Conclusion

Credit and credit distribution through authorship in nutrition science are not uniform ingredients in a universal process of valuation. As a consequence, in the politics of authorship, the names standing atop a scientific publication in nutrition science represent different types of value to both the individuals and employing organizations. The negotiations, considerations, and struggles that went into authorships differ radically because they are driven by different value pursuits. Contributions, in the intellectual, political, or conceptual sense, to articles differ spectacularly, and the meaning of authorship is very local, yet with a few features that align with organization type. While in universities, contract research organizations, and at companies alike authorship guidelines and the moral boundaries distinguishing acceptable from unacceptable practices are transgressed regularly, the meaning of such transgression is equally local. Overacknowledgement and underacknowledgement of research labor is not always identified as a moral problem, and even if so, ownership of that problem is, in practice, unclear.

Practically, authorship guidelines offer little help to (early career) researchers in authorship dilemmas because of their articulation of science as a practice from which (even if only temporarily) political pressures can be removed to reach decisions (on authorship) free from all values except those associated with ideal science. On the contrary, all values and ensuing articulations of worth, those associated with careers, with organizational goals and aims, and with the distribution of credit between collectives and individuals, are exactly what drives authorship decisions.

Guidelines are not the only things informing research integrity in general and authorship practices in particular. However, they are important and powerful. Policies, practices, or traditions in commercial settings to boast credibility of their work by limiting their association with the work may explicitly fall within such guidelines. Matheson (2011) already demonstrated that, in the context of clinical research, ICMJE authorship guidelines (of that time) allowed a specific interpretation that removed industry authors completely from the author list. Given that valuation of authorship differs significantly across social sectors, we will need additional research, normative and descriptive, into the scientific work that legitimately translates into authorship as described in authorship guidelines, and the ways in which this (dis)aligns with how early career researchers are socialized into authorship practices.

Limitations

This study, like all others, has limitations. Concretely, they point to the size of the study, its focus on one community of scientists, the ways in which recruitment of participants took place, and the analytical focus on value.

First, this study is based on 16 long interviews. While they provide a wealth of information on authorship practices, they do not, nor can they, provide empirically generalizable insights in the practice. Rather, the lessons drawn from their analysis provide a critical vantage point for reassessing authorship practices and the ways in which valuation of credit takes shape. Second, the focus on nutrition science limits the disciplinary wingspan of these lessons. From the work of Sismondo (2007, 2009), we know that authorship practices on the boundaries of medical research and the pharmaceutical industry differ and research practices in which the roles of for-profit knowledge making is more limited may yet experience different distributions of credit valuation. Third, participant selection with the aid of the NAV may have influenced the actual invites sent. However, what is of influence more likely is the self-selection among respondent willing to participate in discussions of sensitive issues, possibly inflating or deflating the observations made. Given that this study does not pursue empirical generalizability, it slightly ameliorates this concern. Fourth, the analytical choice to focus on credit valuation highlights new elements drawn from the empirical material. Other concerns around authorship are, as a result, a lot less prominent in this text. This is not meant to suggest their absence from nutrition science publishing practices.

Ethical review

The Dutch Medical Research Involving Human Subjects Act (WMO, Dutch abbreviation) does not apply to this study, and IRB approval is not required under the WMO.

Acknowledgments

I offer my sincere thanks to all nutrition scientists who participated in this study. I thank Hellen Heutz for assisting in transcribing all interviews. I also thank Guido de Wert and David Shaw for providing critical and constructive feedback that helped improve this article.

Notes

¹ These were, in alphabetical order, as follows: career advancement; career requirements; credit distribution; credit evaluation; instrumental goals for authorship; moral evaluation of authorship (practice); political goals for authorship; team hierarchies; team relationships (9). Codes were split and merged a few times, resulting in a final structure as follows, alphabetically: Authorship ideologies; authorship objectives; career status; credit distribution; credit expectations/wishes; credit valuation; team structure (7).

² It is important to realize that, while I used the label Ph.D. student for reasons of recognizability, in the Netherlands, individuals pursuing a Ph.D. degree are employees of the university and, as a result, enjoy a different status. Co-publication would rarely be framed in a student-faculty dichotomy (Welfare and Sackett 2010).

³ Anderson et al. report, in more detail on the relevance of international collaborations in the context of authorship and research integrity at large (Anderson et al. 2011) and McFarlane argues that, specifically in Chinese contexts, gift economies inform authorship practices (Macfarlane 2017). Similarly, Salita (2010) argues that respect for authority and general courtesy influences allocation of credit and authorship across Asia.

⁴ “Comfortable, according to IB2.

⁵ Next to a potentially relevant difference between nutrition scientists and medical researchers, and the site for data collection, a notable methodological difference between this study and Nylenna et al. is that they collected data via a survey, possibly allowing researchers to provide desirable answers or even consult guidelines while filling out the survey.

⁶ See http://www.icmje.org/recommendations/browse/roles-and-responsibilities/defining-the-role-of-authors-and-contributors.html for ICJME guidelines on authorship.

⁷ It must be noted that Hammarfelt et al. employ gamification as a metaphor for the neoliberal context of the quantification of the self, whereas in this article, I will limit myself to the roles and rewards academics articulate for themselves (Hammarfelt, De Rijcke, and Rushforth 2016; Oravec 2017).