“Clean data—well-collected raw numbers—contain within them thousands of stories,” Crystal Biruk writes in Cooking Data: Culture and Politics in an African Research World (Duke, 2018), an excellent new ethnography of quantitative data production in Malawi. Whether it is a matter of guidelines, questionnaires, or statistics, producing “clean data” is partly a process of scrubbing away all the subtleties of social existence—all those stories—and reducing them to figures and formulae for comparison. Yet despite the great interest in metrics and policy-relevant research, very little ethnography of the actual production of this quantitative evidence exists. Cooking Data is a valuable and welcome addition. Biruk provides an innovative and collaborative view of quantitative data-production—a social project (or assemblage) in which heterogeneous objects and a range of people (fieldworkers, supervisors, respondents, foreign demographers, scholars and policy-makers) play essential roles in the production and evaluation of data. Based on 20 months working on survey-based household research projects, in 2005 and 2007-2008, Cooking Data is an engaging analysis of how AIDS-related numbers are made and how AIDS knowledge is produced, sorted, circulated, valued, evaluated, and seen as trustworthy.
The book is arranged along the main sites of what Biruk calls the “life course” of data to stress how numbers are socialized and transformed and all of the care that they require. While numbers may seem abstract and unscathed by the hazards of the material world, they do not exist abstractly or without our collective care and attention: our writing them down, adding them together and tabulating them into charts, presenting them at conferences, in PowerPoint presentations or articles. But Cooking Data is not mainly a book about how data are socially constructed, so much as how they are “stabilized”—or, how the production of data involves the construction of whole worlds of support around them. “[R]esearch projects,” Biruk writes, “do not so much intervene, treat, or change the contexts they enter into as they co-constitute them” (p.208). Each chapter approaches this idea from a slightly different perspective, mapping out features of a research world and how it is knitted together through the production of data and new subjectivities and socialities, characteristic practices and ways of moving around and acting, exclusions and inclusions, hierarchies and relations of expertise, and ways of valuing knowledge.
Biruk analyzes the labor processes of data production in the ‘sanitized’ space of the office (chapter 1), ‘messy’ interactions in ‘the field’ (chapters 2-4), and in the conferences, presentations of findings, meetings, and policy debates where data is turned into evidence (chapter 5). Biruk shows how health knowledge moves within and between these sites and how it is evaluated and validated at each step along the way: not only in formal presentations and articles (chapter 5), but also during its production (chapter 4) and in informal spaces like discussions about the value of research between researchers and respondents (chapter 3) or about the value of expertise between Malawian and foreign researchers (chapter 1-2).
By describing how worlds are created around survey-based research projects and how they are always entangled with the lives and relations of those who produce them, Cooking Data provides an important lesson about the idea that “cooked data” are flawed. Good numbers are produced by fieldworkers’ creative and skillful negotiations between the standards and plans of demographers and unexpected encounters. By showing how, Biruk extends and brings a new set of perspectives to studies of the use of quantitative metrics in global health (Adams 2016).
The book’s first ethnographic chapter, “The Office in the Field,” begins with a discussion of the politics of space and how the desire to produce findings from survey-based research that can be exactly replicated elsewhere encourages demographers to imagine a kind of mechanical “assembly line for data” staffed by mindless, interchangeable researchers. The closer Biruk looks at the “social life of data,” the more every step of the process can be seen to involve messy negotiations and skillful and imaginative improvisations. Data are made through researchers’ encounters with unanticipated obstacles, divergent perspectives, and different valuations of their work; after their collection data are further transformed in offices and policy discussions. Yet they must emerge “clean,” as if it were untouched by these encounters. High-quality data are never clean; we can have numbers untouched by negotiations only when we have made them up entirely.
Much of the book focuses on mediators of survey research worlds and the work-lives of Malawian researchers. Chapter 2, “Living Project to Project,” focuses on the skilled knowledge brokering done by research supervisors and fieldworkers, how supervisors and fieldworkers perform a kind of local expertise that aligns with foreign demographers’ expectations, and how producing fields for research involves marking off the boundaries that define ‘the field’ and detach it in time and space from ‘the office.’ Biruk describes how fieldworkers enact differences between themselves and rural Malawians: researchers and supervisors specially dress for the field, tell “silly villager stories,” and discuss how best to ask about witchcraft and other apparently irrational beliefs, constructing “rural research participants nostalgically, as symbols of a nation of peasant farmers, bearers of tradition, and masters of cultural knowledge, as foils to their more modern selves” (p.78). At the same time, Malawian researchers draw attention to their differences from foreign researchers, holding difference out as evidence of the local knowledge on which their livelihoods depend. Constructing all these divisions, Biruk shows, is an important part of the production of data.
Chapter 3, “Clean Data, Messy Gifts,” considers the value of data for different participants through an examination of encounters between fieldworkers and their rural research subjects. In 2008, the Longitudinal Study of AIDS in Malawi project distributed two bars of soap (Lifebuoy and Sunlight) as a token of thanks to thousands of participants in a panel survey for their responses to a 25-page survey. Small, standardized, inexpensive, and easily transported: ethics boards, demographers, and researchers saw soap as a manageable and appropriate gift. Unlike money, soap was not felt to risk coercing the participation (or invalidating the consent) of impoverished people in survey research. For many respondents, soap was a welcome and useful gift, even if they would have preferred cash; for others, it was a token of the uneven distribution of project benefits. Some were prompted by the gift to re-consider the value of answering questionnaires or providing blood samples for HIV tests, saying that soap was an inadequate “payment” for their investment of time, energy, and blood. Still, others accused researchers of being “bloodsuckers.” Data are not only evaluated by demographers, they are assessed and reckoned at every stage, by researchers and by respondents. Respondents held out soap as evidence in their critiques of the impersonal relations characteristic of survey-based research that expected poor people to volunteer their time answering questions. Biruk shows that these differing perceptions did not stem from rural Malawians’ misunderstandings of researchers’ aims (or from their embeddedness in an exotic “gift society”), but rather from their recognition that gifts of soap fasten together participants in research worlds that demographers, ethics boards, researchers, and rural Malawians all take part in but whose benefits are unequally distributed.
Contrasts of clean and dirty, raw and cooked, skilled knowledge work and unskilled labor provide a series of counterpoints that Biruk weaves between the perspectives of people occupying different positions within research worlds. Respondents’ critical and reasoned assessments of the value of soap and data complicate familiar narratives of well-intentioned global health projects encountering stubborn culture.
Chapter 4, “Materializing Clean Data in the Field,” describes how standardizing practices of enumeration in demographic survey-research come to be embodied by researchers while also generating new social categories, identities, and practices that sometimes rub awkwardly against values of standardization. The chapter opens in the back county of central Malawi, where, during the winter, routes are impassable to any but the sturdiest Land Cruiser. With the start of rains, fieldworkers walk or hire canoes to navigate “crocodile-infested waters” to reach sample households (p.130). Dedicated (and, evidently, often damp) researchers carry questionnaires and consent forms, notebooks and writing implements, hand-drawn maps made in previous years, photos of panel respondents, bars of soap for gifts, and handheld GPS devices for mapping households. They frequently ask for directions. It is mostly improvisation and doggedness that count when you find that “good data do not lie passively in wait to be collected by fieldworkers” (p.164).
People’s lives in rural Malawi, like people’s lives everywhere, are always complicated, disorderly and contradictory. What survey-based projects do to these lives is to shape them into “clean, high-quality data,” patterns of numbers that are households and people and their histories. Starting in the office with the design of a survey, researchers start to limit their panoramic and confusing view with the epistemic commitments, practices, and instruments (sampling strategies, probing, and so forth) of enumeration. The most nuanced view is not their aim; it is the clearest vision of a particular bit of reality that demographers are after (and generally get) and this may involve screening off from view much of what ethnographers like Biruk construct entire books from. Biruk shows how turning lives into numbers involves researchers’ embodying the epistemological commitments of demographers. These permit “seeing” but also create blind-spots. Of course, there’s nothing inherently wrong with measuring some things and not others. Like subway maps, public health metrics are useful because they abstract from the real world many details that would get in the way. Abstractions themselves do not have any effect on the real world. Our engagements with abstractions do. However, Biruk’s aim is not to show what research projects miss, or to debunk numbers “as simply false, socially constructed, or inaccurate” (p.26), but rather to ask what research projects are in fact doing. What larger social worlds cohere around research projects? How does producing numbers afford certain representation and not others? What representations are valued or discarded?
Chapter 5, “When Numbers Travel,” begins with an examination of how specific numbers become evidence as they circulate after researchers have returned from the field and their work has been typed up and converted into statistical data. Biruk cautions against being swayed by the authority of numbers, noting that even critical accounts risk giving over a power to numbers that they would not otherwise have; “numbers do not stand alone but require cultural, social, and other scaffolding and negotiation to be propelled through the world” (p.168). Numbers seem to move easily, crossing difference without effort; they can be detached, rearranged and compared, tabulated, coded, graphed, and statistically manipulated, converted into dollars and inserted into equations. They grip our imaginations. But sometimes good numbers fail to “travel” or to convince their audiences of their validity (pp.184-191); and, sometimes, ideas based on no evidence whatsoever circulate widely, such as the idea that HIV is spread in Malawi by “harmful cultural practices” (pp.173-184). Whether or not data become evidence is not only a matter of epistemological rigor; their credibility also depends on the confluence of people and institutions with diverse perspectives and interests.
Cooking Data is a readable and engaging book. Biruk builds a sophisticated theoretical argument through ethnographically grounded storytelling. The narrative moves beyond familiar oppositions between the rich complexity of people’s lives and the uncaring simplicity of numbers that characterize much critical writing about the processes that turn human beings and communities, with their complicated life stories and histories, into quantitative data. Ethnographers commonly use this contrast to explore how a focus on what is quantifiable can obscure and devalue what matters most. But by doing so, such accounts risk confusing the interpretive labor of simplifying and schematizing for the bureaucratic violence characteristic of the state (Biruk 2016). Biruk opens up new analytical space for caring for numbers, showing how they can be useful, and not only (mis)represent real worlds and values, but also create new ones (p.212). Cooking Data will be of interest to anthropologists, global health practitioners, researchers and others, and would make an excellent contribution to syllabi on knowledge production, medical anthropology, science studies, global health, and critical development studies.
References cited
Biruk, Crystal. 2016. “Studying up in critical NGO studies today: reflections on critique and the distribution of interpretive labour”. Critical African Studies 8(3):291-305.
Adams, Vincanne, ed. 2016. Metrics: What Counts in Global Health. Durham, N.C.: Duke University Press.
Brendan Tuttle lives and works in South Sudan, where he is Research Manager for Forcier South Sudan’s Monitoring, Evaluation & Learning Unit and for the Evidence and Learning Facility in South Sudan (ELFSS) of DFID's Humanitarian and Resilience Programme in South Sudan (HARISS). He holds a doctorate in anthropology from Temple University and is a Research Associate with the Children’s Environments Research Group (CERG) at the Graduate Center, City University of New York (CUNY).
