How are diversity and gender equity in the scientific workforce related to interdisciplinary research collaboration? The research that looks at this question is surprisingly scarce. Initial survey analyses suggest that women on average report more interest in interdisciplinary research than men.1 A few coauthorship studies show some limited support for gender differences in publishing interdisciplinary research, but when controls are added to the models the evidence is not so clear.2 The murky answer from quantitative studies suggests the need to look more qualitatively at the experience and meaning of interdisciplinary collaboration for women and men scientists. We find that women scientists face a particularly poignant catch-22: they are expected to engage in interdisciplinary collaboration, but at the same time are often discouraged from it by well-meaning colleagues.“Nearly all scientists expressed interest in interdisciplinary collaboration, but faced different constraints in practice across the sites.”
In this essay, we draw on results from our National Science Foundation (NSF) funded project on collaboration in the chemical sciences to dive more deeply into the question of gender equity and interdisciplinary collaboration. The project involved collecting ethnographic data over ten to twenty-four months at four sites ranging from a traditional academic lab to an industrial firm. Members of our team spent time at these sites observing lab meetings and conducting 106 interviews. An important methodological contribution of this project is that qualitative data were collected from both academic and industry contexts, providing a unique opportunity for in-depth, cross-sector comparisons of collaboration. One thing we found, for example, was that interdisciplinary collaboration was discussed in coherent ways across academic and industry labs, but was much more ambiguous in implementation.3 Nearly all scientists expressed interest in interdisciplinary collaboration, but faced different constraints in practice across the sites. In academia, tensions arose around practices linked to evaluation, such as publishing, credit, and tenure. In industry, by contrast, time pressures and competition within the firm often strained how collaboration was practiced. These time pressures created tensions between work groups and inhibited open collaboration.
A key finding of our project across all of the settings is that for successful chemical scientists, interdisciplinary collaboration is inseparable from science. A US-born man and principal investigator (PI) of a large interdisciplinary research center, for example, described his next scientific project as a blend of collaborative and scientific considerations:
I went to visit [prospective collaborator] last year; he sent a student down here, et cetera. And believe it or not, the first thing that I look at is, almost simultaneously, the quality of the science being done and how it fits programmatically with our work. And whether or not I would trust this person or not to supervise one of my students. …I’ve learned this over the years. And that generally means that I’m looking (pause) at other values, other than scientific values, um, because I’ve had some disastrous collaborations.
This respondent was one of the most successful and well-resourced chemical scientists with whom we spoke and observed. Both his practices and discussion about his work wove interdisciplinary collaboration and scientific research together seamlessly. He talked about interdisciplinary collaboration and science as being inseparable, whereas some of the less well-resourced scientists we spoke with talked about interdisciplinary collaboration as being something that was a choice, separate from their main scientific research.
While it is a core research practice, very little attention is given to training scientists how to collaborate, and little explicit guidance is available for how to set up scientific organizations to overcome challenges in interdisciplinary collaboration. The result of this unreflective approach to interdisciplinary collaboration is that it then becomes difficult to assess and improve collaborative research, including highly productive collaboration that brings together scientists with diverse backgrounds by discipline, gender, race, and nationality.
The catch-22 of collaboration
This finding—that interdisciplinary collaboration is key to doing chemical sciences research—is gendered. The interview data revealed that junior female scientists sometimes receive bad advice about interdisciplinary collaboration (i.e., to avoid it at all costs), while senior male scientists (like the research center PI quoted above) do not see a bright line between interdisciplinary collaboration and the typical conduct of research. Some more marginal female academics—untenured and without clear signals that promotion will go smoothly—may be responding (and perhaps overreacting) to the idea that women may receive less credit and recognition for collaborative work with men. An untenured faculty member who was a foreign-born woman told us:
I’ve heard that more than once, informally from colleagues or because of how the collaboration was being interpreted, that um, especially junior, junior level, women should be careful about collaborating because they tend to be viewed as less independent anyway and so if they are collaborating and with like a man collaborator, the perception from the outside may often be as to the fact that the woman is the secondary collaborator. Whereas the man is the lead collaborator, especially if it is an older man. But even if it’s a man of the same age, women tend to be viewed as, you know, someone who will not take the lead in collaboration.
The faculty member quoted above had been given the strongest message not to collaborate (perhaps with good intentions); and yet collaboration is related to the highest levels of productivity in science. Interdisciplinary collaboration (especially for foreign-born scientists) is associated with high-impact publications that receive many citations,4 so this respondent seemed to be receiving some not very helpful advice about collaboration.
In general, our data suggest that to be perceived as independent and productive scholars, women need to manage interdisciplinary collaboration carefully. Some women may well benefit from a clear division of labor, as when they are the only expert in their discipline on an interdisciplinary project. This specialization approach seemed to be a strategy for some successful women scientists—e.g., to be acknowledged as the expert in her area and receiving clear-cut credit for her piece of an interdisciplinary project.
The inadmissibility of gender
Another key finding from our study of collaboration in the chemical sciences is related to how diversity is discussed. We found that gender is not an admissible or discussable category for talking about differences in scientific collaboration in our interviews, but disciplinary differences are common and comfortable narratives to talk about the benefits of interdisciplinary collaboration. Take, for example, the quote below from one of our respondents who is a US-born female graduate student. She articulates in a short statement the two trends we saw across our interviews with chemical scientists in all career stages and in all four of our settings: you can’t see or talk about gender, but you can talk about disciplinary diversity. The interviewer from our team asked whether the different backgrounds of collaborators—by discipline or gender—had an effect on research. She replied:
Well, discipline I can definitely see, ‘cause in my lab group, we have some physical chemists and they definitely approach things differently. And the physicists are much better at coming up with observational problems. And the chemists seem to be better at laboratory problems. But, gender-wise? I’ve not seen.
Because of the persistence of ideas about the neutrality of science, identity-based differences such as race or ethnicity, gender and sexuality, and class are unwelcome social distinctions which make discussing diversity difficult. But interdisciplinarity is an acceptable form of “diversity discourse.”
Our research has also examined how the gendered narrative of doing science for societal benefits does not seem to be particularly salient for working women scientists,5 despite its invocation by policymakers as a recruitment ideal for getting girls into STEM fields. Similarly, the perceived desirability of interdisciplinarity and collaboration for women6 may be a gendered cultural expectation that does not match up to the reality of the costs and benefits in practice. The expectation that women are interested in and particularly ”suited” to interdisciplinary collaboration meshes with gendered ideas about cooperation, mentoring, and flexibility, which successful women scientists may not find relevant in their ideas about the desirability of collaboration. The women we interviewed did not especially invoke discourses of doing science for society.
The absence of talking about gender, and the complexity of discussing it without invoking essentialist stereotypes (e.g., making assumptions that all women are the same and that way by nature), will continue to challenge future research designs. Our findings about the invisibility of gender in narratives about collaboration, and the lack of gender difference in approach to interdisciplinary collaboration and claims to be serving society with one’s science, seem to present no easy path toward thinking about policy implications. The findings do, however, speak to the need for further research. We see that gendered organization of science, and in interdisciplinary collaboration, is subtle. It is not discussable in many ways, but nevertheless real in shaping everyday practices, careers, perceptions, and perhaps outcomes. The very inadmissibility of gender is perhaps what makes it more powerful as a social force in scientific collaboration. If inequalities cannot be named, they are more difficult to address.
What can be done
Still, there are ways that some of the gender inequities uncovered in our research might be addressed. We found that untenured women faculty face a catch-22—a real predicament with interdisciplinary collaboration. Women scientists encounter both expectations to collaborate, and, at the same time (through discouraging advice from mentors and peers), significant disincentives to do interdisciplinary and collaborative work.
Eliminating the catch-22 of interdisciplinary collaboration for women faculty—particularly managing the tension between establishing independence and establishing a network of resources and collaborators—requires transparency and explicit discussion of institutional norms and practices. Both independence and connection can enhance scholarly reputation, but they are often understood as mutually exclusive rather than things that can be carefully managed. Collaborations that are unsuccessful can be devastating to career prospects, particularly if one is seen as the disruptive collaborator not fulfilling responsibilities. But having no collaborations at all can also be harmful, as intellectual and material resources may be out of reach and networks that might enhance scholarly reputations underdeveloped. Explicit discussion among collaborators as to responsibilities, credit allocation, and work processes are essential, as are institutional discussions in departments and colleges as to how collaborations are to be valued in career development.
In other words, one policy implication of this finding about gendered perceptions of whether a junior scientist will be rewarded if they engage in interdisciplinary collaboration is that universities should sponsor open conversations about tenure and promotion criteria, and how credit is allocated in interdisciplinary collaboration. These credit practices will vary by discipline, which means special attention (and perhaps extra credit) should be given to interdisciplinary collaboration. The development of a range of explicit criteria may allow for more transparent decision processes around what counts for collaborators.
The authors are grateful for support from National Science Foundation grants 1064121, 1063944, 1413898. All conclusions and opinions are the authors’ and do not necessarily represent those of NSF. We thank Tim Sacco, Angela Stoutenburgh, and Itai Vardi for research assistance on the project. We thank the Items editors and Tim Sacco for helpful comments and editing advice.