Login
Increasingly, scholars across the life and social sciences recognize the necessity of multi-method, interdisciplinary research for its ability to adequately understand the world’s complex problems.[1] However, the process of designing and executing these projects can be challenging. Interdisciplinary endeavors often risk privileging one discipline/methodological paradigm with others incorporated in a more consultative manner (i.e. quantitative versus qualitative), or, they run in-parallel without integrating epistemologies and methodologies (Lewis 2021). Examples of symmetric and integrative projects which unsettle disciplinary boundaries to afford new kinds of knowledge remain few and far between.
In the following piece, we (ZB and CB), as members and ethnographers of interdisciplinary teams, reflect on several “Mexican Exposures” (MEXPOS) projects which bring together researchers in anthropology, epidemiology, biostatistics, engineering, and health economics to make better knowledge and “better numbers” about health and inequality in Mexico (Roberts 2021). MEXPOS projects collaborate with long-standing epidemiological birth-cohort studies (ELEMENT and PROGRESS) which are based within Mexico’s National Institute of Public Health (INSP). Through performing and observing the laborious process of integrating the often-disparate methodologies, epistemologies, and analytical aims that each expert brings to the team, we have identified that some team members act as mediators, performing a critical role in making these interdisciplinary collaborations work. Our contribution to understanding how interdisciplinary knowledge is made (Lin et al. 2007) is a focus on the interpersonal aspects of knowledge production through exploring how these teams make better data by destabilizing disciplinary boundaries. By doing so, we hope to elucidate the challenges and opportunities of this kind of collaboration by exploring what is made possible when doing this work together.
ZB has a bachelor’s degree in Anthropology and has spent a year managing MEXPOS projects as she prepares to begin a joint MD-PhD Anthropology training program; as an undergraduate, she also spent three semesters working in the MEXPOS ethnographic coding lab. CB has a background in physiotherapy and medical anthropology, and works with MEXPOS as part of her doctoral research and the Biosocial Birth Cohort Network, which included shadowing the MEXPOS team for ten days in April 2023 and meeting field workers in Mexico City. We developed this essay from our shared experience and observations of interdisciplinary knowledge practices within MEXPOS team meetings.
MEXPOS projects perform bioethnographic work. Bioethnography is a research method which combines methodologies from the social and biological sciences to understand environment-body interactions as relational and situated processes (Roberts and Sanz 2018). The premise of bioethnographic teams is to generate new knowledge by transcending disciplinary boundaries to tackle the complexity of the topics of study. Bioethnographic methods differ from other examples of collaboration between the life and social sciences that break down because of the lack of a shared intent or question (Lewis 2021) and encourages critical implosions between “nature/culture” (Roberts 2021). This methodology calls for the unsettling of traditional epistemological boundaries between the disciplines involved to critically engage with the questions and objects of research at hand in new and innovative ways.
This interdisciplinary research model hopes to create knowledge that truly apprehends pressing problems and questions. In practice, however, we have found that this process can feel awkward, clunky, and falter as members navigate the integration and destabilization it requires, working through the tensions of epistemic purity and interdisciplinary compromise while creating new modes and subject positions towards these blended methods. In our experience, the element which often relieves these tensions and moves teams toward achieving their collaborative aims has been the presence of mediators and the labor they perform.
Mediators are key for facilitating conversations between disciplines that bring underlying “taken-for-granted” assumptions to the surface, enabling these teams to progress past disciplinary limits. Most MEXPOS teams consist of a core of senior academics that are anchored within respective disciplines and act as knowledge-keepers alongside a variety of research assistants and management staff. Within MEXPOS, we have found that mediation is typically performed by the project manager and several graduate research assistants, including ZB. These are junior scholars with varying degrees of training in ethnographic methods as well as survey methodology, epidemiology, biology/life sciences, statistics, and other quantitative analyses. Through their training and background, mediators are well positioned to steer, generate, and develop bioethnographic questions while fostering a group dynamic that advances the team’s goals. The two examples we present here demonstrate mediation-in-action which allowed these teams to move forward amid, and possibly because of, disciplinary friction (Tsing 2011), which was harnessed by the mediators and transformed into something generative.
In the Spring of 2023, a team within MEXPOS worked on a collaborative paper based on the insights of the Household Chemical Assessment Project, a pilot study of two working-class households in Mexico City. This project, involving anthropologists, epidemiologists, exposure scientists, and metabolomics researchers, documented household and personal care products along with their use/meaning and generated a master list of chemical ingredients and insights about household exposure. During these meetings, the team debated how to situate this project and its outputs within an existing paradigm of exposure research, “the exposome” (Wild 2005). The team was stuck; the epidemiologists were aiming for epistemological clarity and a fixed structure to proceed, while the anthropologists were looping back and questioning the paradigm itself by posing alternative questions. The mediator registered that the two camps were talking past one another due to differences in their underlying notions of what “exposure” entailed on an ontological level, and pulled together readings that spanned both sets of disciplines to be discussed as a group at the next meeting. This effectively moved the team forward by 1) developing a new starting point with a shared knowledge base and vocabulary, and 2) opening a window into each discipline’s mode of inquiry in a way which allowed for more nuanced discussion about their respective stakes and assumptions. In this way, mediators can act as disciplinary polyglots thanks to their ability to understand the languages of the different disciplines, recognize and iron out misunderstandings, and summarize the conversations held by senior academics from different camps. This practice of mediation enabled the team to theorize beyond disciplinary limits and pioneer a new orientation towards exposure inquiry and intervention that enmeshes social and life scientists within a framework of shared understanding.
The second example involves another MEXPOS team, comprised of anthropologists, health economists, biostatisticians, and epidemiologists, that leveraged insights from a previous project (NESTSMX) about household water infrastructure. The team created a module of survey questions for the Mexican National Health and Nutrition Survey (ENSANUT) in order to investigate the impact of an intermittent water supply on health, gender, and household finances. One meeting about question revisions for the following year’s survey got stalled when differences in disciplinary aims and timelines surfaced. The anthropologists, who predominated, wanted to ameliorate their own apprehensions around survey methodology by tinkering with existing questions and discussing potential new ones to keep fidelity to the complex ethnographic insights. The biostaticians seemed frustrated by this, as they pointed out the looming due date and advocated for straightforward and generalizable questions to produce data that could be meaningfully compared to the previous year. The mediators suggested narrowing the discussion only to the ethnographic data that could be directly operationalized into the specific module questions that the biostaticians agreed would be worth modifying because they describe experience instead of measuring prevalence. As such, the mediators helped to reconcile qualitative richness and quantitative concreteness in translating ethnographic insights into questions that produce 0s and 1s, modulating between the sometimes-disparate aims and scales of ethnographic and statistical research processes that make integrating them so difficult.
Sketches from CB’s notebook while observing interdisciplinary knowledge practice, integrated and expanded by ZB.
Contemporary academic training calls for more interdisciplinary models, which could produce more mediators for multidisciplinary teams. These mediators do not always squash or quell conflict, but rather harness the productive role of the disciplinary unsettling that bioethnography facilitates among established academics, including the resulting misunderstandings and moments of uncertainty. These moments of generative friction offer critical points of reflection and surprise, and can reveal disciplinary assumptions and blind spots – which might be what is most valuable in bioethnography. This generative friction is a function of the unexpected: to borrow from studies of cognition, it is when a habit, in this case a disciplinary way of thinking, is contradicted and calls for a new way of understanding (Clark 2018), as well as epistemic humility. Here, these new ways of understanding are the interdisciplinary insights that are made possible through practices of bioethnographic integration. The mediators make friction productive by “paying attention to the diverse concerns of different disciplines and incorporating responsive negotiation of their collaborative possibilities and the tensions between them” (Mol and Hardon 2020). Here, mediators provide the crucial vector required to propel the group forward.
The making of truly interdisciplinary knowledge often requires overcoming epistemological paradigms through disciplinary destabilization. Mediators both manage interdisciplinary tensions and foster the generative friction that emerges, allowing for new kinds of knowledge to be produced together. Mediators can recognize, hold, and harness the discomfort of competing objectives and respond accordingly with the symmetry of the meta research process continually in mind. While disciplines will transform as opportunities for interdisciplinary training continue to increase, our experience with these bioethnographic collaborations underscores the importance of maintaining spaces for generative frictions that mediators can render into positive momentum.
[1] See https://www.ucl.ac.uk/soc-b-biosocial-doctoral-training/soc-b-centre-doctoral-training-biosocial-research; https://new.nsf.gov/funding/learn/research-types/learn-about-interdisciplinary-research
Clark, Andy. 2018. “A Nice Surprise? Predictive Processing and the Active Pursuit of Novelty.” Phenomenology and the Cognitive Sciences 17 (3): 521–34. https://doi.org/10.1007/s11097-017-9525-z.
Lewis, Ashley. 2021. “Questioning the Promise of Interdisciplinarity: An Ethnography of an Interdisciplinary Research Project.” University of Nottingham.
Lin, Wei, Rob Procter, Peter Halfpenny, Alex Voss, and Kenny Baird. 2007. “An Action-Oriented Ethnography of Interdisciplinary Social Scientific Work.”
Mol, Annemarie, and Anita Hardon. 2020. “What COVID-19 May Teach Us about Interdisciplinarity.” BMJ Global Health 5 (12): e004375. https://doi.org/10.1136/bmjgh-2020-004375.
Roberts, Elizabeth F. S., and Camilo Sanz. 2018. “Bioethnography: A How-To Guide for the Twenty-First Century.” In The Palgrave Handbook of Biology and Society, edited by Maurizio Meloni, John Cromby, Des Fitzgerald, and Stephanie Lloyd, 749–75. London: Palgrave Macmillan UK. https://doi.org/10.1057/978-1-137-52879-7_32.
Roberts, Elizabeth F.S. 2021. “Making Better Numbers through Bioethnographic Collaboration.” American Anthropologist 123 (2): 355–69. https://doi.org/10.1111/aman.13560.
Tsing, Anna Lowenhaupt. 2011. Friction: An Ethnography of Global Connection. Princeton University Press. https://doi.org/10.2307/j.ctt7s1xk.
Wild, Christopher Paul. 2005. “Complementing the Genome with an ‘Exposome’: The Outstanding Challenge of Environmental Exposure Measurement in Molecular Epidemiology.” Cancer Epidemiology, Biomarkers & Prevention 14 (8): 1847–50. https://doi.org/10.1158/1055-9965.EPI-05-0456.
UPDATE: As a commenter helpfully pointed out, the person whose tweet I’m responding to was a political science Professor, not a historian. This kind of messes with the framing of this post but rather than stealth re-write it I’ll leave it as is and let you interpret my Freudian slip as you like
When I was in grad school, my Department’s grad student organization made shirts that read, “Political Science: Four sub-fields, no discipline.” Behind this joke is a common observation about political science, that it is defined by its focus rather than a formal set of methods or theories. Not everyone agrees with this characterization, and there have been some efforts to craft political science-specific tools. But generally political science is a field that draws on insights and tools from other areas to study politics. This is most pronounced in international relations. IR looks not just to other fields but also other sub-fields of political science to study the world.
Many present IR’s lack of discipline as a critique. They view IR scholars as a group of raiders, pillaging ideas and methods from other disciplines then returning to our barren homeland. Two recent Twitter kerfuffles, however, demonstrate that this aspect of IR is actually our greatest strength.
I don’t spend much time on Twitter anymore, but I still seem to discover the latest controversy. Two academic ones were related in their attacks on political science.
First, a historian responded angrily to a new article in the American Political Science Review by Anna Grzymala-Busse on European state formation. The historian suggested she used overly-simplistic methods to make a point that “real” experts on early modern Europe already knew (I’ve anonymized the tweet as I don’t like engaging in Twitter attacks).
This is a common complaint I’ve heard from historians studying international issues. IR and CP either take history’s insights and repackage them as our own, or don’t realize historians have already said this. Several academic institutions I’ve been a part of have included fierce and rather petty attacks by historians on political scientists.
As some respondents to this tweet noted, however, this historian isn’t really being fair. The role of religion in state formation is hardly settled ground–I took an entire class on debates over the role of religion in nationalism in grad school. Also, isn’t it a good thing to test and confirm certain arguments using a different set of data and methods? And when pressed, he couldn’t point to what historical works the author overlooked.
Critiques of IR and political science take the place of addressing real issues within other areas of study
In a follow-up tweet, the historian also makes an ironic call for interdisciplinarity. Ironic because this is an interdisciplinary work! Grzymala-Busse combined insights from comparative politics and history to generate new knowledge; this is in line with her other work, which involves careful attention to historical detail. Those calling for interdisciplinary engagement should cheer this, unless “interdisciplinary” just means listening to historians…
The second Twitter incident involved a data scientist. A data science grad student tweeted a broadside against the replication crisis in psychology, followed by attacks on political science and the social sciences in general. Another data scientist responded by pointing out that social scientists don’t conduct our own statistical analysis and instead get “real statisticians” to do it.
Again, people took issue with this. Some noted the data science grad student hadn’t really characterized the replication crisis accurately. Others asked for specific examples (which weren’t forthcoming). I’d also point out that it is true data science does have a real impact on our lives, but it’s hardly a positive one; one data science course I took focused on things like getting around CAPTCHA tests and tricking spam filters. And in practice “interdisciplinary” for data science often means using Python to study political or social issues without engaging actual subject matter experts.
These are very different controversies, and I’m sure these two people wouldn’t agree on much if forced to have a conversation. But both involve the perennial attack on political science (and IR by extension): we don’t come up with our own insights or methods, we just steal the former and implement the latter badly.
I thought of this debate recently while visiting the excellent Jorvik Viking Centre in York, England. The popular view of the Norse raiders known as vikings is of pillaging hordes, and that was certainly the case initially. But as often happens, they settled down. And in the case of Jorvik, they created a thriving cosmopolitan society enabled through their wide-ranging travels.
Maybe I’m pushing this metaphor a bit here, but I think of IR as Jorvik.
Yes, we got our start by combining economic models and humanistic insights. Yes, our research tends to include references from disparate traditions. Yes, our data is messier than other fields or even sub-fields within political science.
These issues all became strengths, however.
Because of IR’s broad roots we have to be conversant in different disciplines. When engaging with people from other disciplines, I often get the sense they’ve never really read anything from my field; their critiques are often caricatures. By contrast, many IR scholars are well-read in other fields.
Additionally, we recognize the difficulty of drawing on and testing different disciplines. That’s why you can find IR and political science discussions about combining methodologies or triangulating among competing schools of historiography.
Finally, the challenge of dealing with incredibly messy data has created problems for IR but also led to fertile debates. For example, the problem of selection effects in conflict onset has led to a useful back-and-forth.
Interdisciplinary means each side listens to and learns from the other, not that one asserts superiority and territoriality
Beyond that, these critiques of political science and IR take the place of addressing real issues within other areas of study. In C.P. Cavafy’s poem, “Waiting for the Barbarians” (which I referenced above) the inhabitants of a classical city sit and wait for the barbarians to arrive instead of dealing with the problems in their civilization. We can sense this in some of the attacks I’m discussing.
Historians are rightly frustrated at the lack of support for and interest in the humanities from universities and the general public. They often, however, see political science as the problem, such as Grzymala-Busse making a splash by engaging in historical debates. I am also reminded of a seminar my grad school put on to help students prepare and turn their dissertations into books; we went around the room discussing our topics and one history student made a crack about how “relevant” mine would be in DC. Rather than finding ways to demonstrate the value of a humanistic and historical approach to contemporary issues, some historians seem to blame political science and IR for sucking up all the attention (and student interest).
Likewise, data scientists are rightly tired of inadequate statistical models and badly interpreted findings. But what many of them seem to miss is that this is not a problem of stupidity: it is over-confidence, something data science tends to exhibit. I also sense a bit of frustration that political scientists are still seen as the expert on…politics despite our lack of cutting edge programming skills. This could be solved by closer collaborations between data scientists and subject matter experts, something that is often lacking.
It’s almost like political science and IR have become the Other to our critics, alleviating the need for any deeper reflection. As Kavafy ended his poem: “those people were a kind of solution.”
So what should be done?
Well, I am just finishing a fellowship at Edinburgh University’s Institute for Advanced Studies in the Humanities (IASH), which was funded by the Centre for the Study of Islam in the Contemporary World. It includes fellows from across the humanities, as well as the social and natural sciences. Part of the fellowship is a “work in progress” talk, which I gave last week. The empirical subject was my new work using social network analysis to study international religious politics, but the broader theme was my ongoing effort to test concepts from the humanities using quantitative social science methods.
I was unsure about the reaction. I didn’t know if the crowd of humanities scholars would react hostilely to me as an interloper. Instead, it was an incredibly fruitful discussion. There were tough questions and critiques, but they were in the spirit of collaboration and community. They recognized that I valued their disciplines, and they did not see the fact that I drew on theirs and mixed it with others (i.e., my lack of discipline) as a problem.
In this context, interdisciplinary meant each side listened to and learned from the other, rather than one asserting superiority and territoriality. It’d be nice if that attitude spread outside IASH.
FreshRSS 