Abstract
With the increasing spread of not just North–South development interventions, but also research collaborations, the terms of these relationships in terms of power, hierarchy and equity have come under increased scrutiny, especially amongst researchers in the area termed the Global South. On the basis of multiple years of fieldwork alongside four different research projects in Malawi, Cooking Data are an ethnographic examination of how human and non-human actors mutually co-constitute what is popularly known as data. The book examines the political, social and cultural underpinnings of what the numbers that constitute data signify, and how they come to be made in the context of HIV/AIDS-focused demographic research in rural Malawi. The author, Crystal Biruk, conducted ‘research on research’, the focus of her enquiry being how research and the data that it looks to generate are co-produced and negotiated by multiple actors involved, from academics in the Global North, to their less privileged counterparts in Malawi, fieldworkers, respondents and those excluded from survey samples. Not only does the book shed light on how data are imagined, created, circulated and accepted, but also makes evident the racial and neocolonial underpinnings of quantitative, demographic HIV/AIDS research as well as North–South research collaborations.
Cooking data comprise five main chapters, each of which speaks to different elements of data visualization, production and dissemination. The first chapter focuses on how the various imaginations undergird the preparation of a survey format. Pre-imagined forms of data shape the practice of its creation, with choreography of practice aimed at standardization attempting to eliminate its messiness. In doing so, academics attempt to replace ostensible subjectivity with objectivity. However, survey instruments that do not account for the rhythms of everyday life, people, spatially contextual lifeworlds and more prove inadequate to capture a phenomenon or construct imagined to be inherently stable, with data points believed to represent them instead displaying a consequent capriciousness and instability. Part of this Biruk attributes to what she calls the spatial politics of data production, directing attention to the racial hierarchies within partnership relationships. With donors often being located in the Global North, and local researchers dependent on these funds for research activities, the latter find the nature of their engagement with the projects circumscribed with implications for how a project is conceptualized.
The second chapter demonstrates how knowledge workers construct and broker what comes to be known as ‘local knowledge’. Although project heads and external researchers may operate with the assumption that local knowledge is fixed and knowable, Biruk demonstrates how local knowledge and the experience it is built on is ‘embodied, relational’ (p. 71) and constructed, a negotiation and dialectic between different levels of researchers and the researched. Although recruiters/academics may imagine that fieldworkers from the country in which research is conducted are experts, local knowledge itself is a brokered container, with credibility an on-going process, fashioned through the interaction of fieldworkers with research methods, tools and protocols. These interstitial spaces between different actors are created during the process of recruitment and training itself and unfold over time, through practices of boundary-making between superiors, fieldworkers and respondents.
These boundaries are complicated by an often contentious question in research—the compensation of respondents with gifts for their time. Through an outlining of how the same ‘clean’ gift—soap—was received by different participants, we see in the third chapter how seemingly innocuous gifts that function as tokens of gratitude complicate researcher–respondent relationships. Through the promise and deployment of this gift, subject positions are (re)shaped and challenged, with respondent understandings of the value of the work they put into answering survey questions contaminating ideas of soap as being appropriate compensation. Instead, Biruk suggests that the values attributed to compensatory gifts are instead informed by history, and participants’ growing sense over time of their value as research respondents in areas that are heavily researched.
This relationality extends to data production as well, and the fourth chapter highlights how data are seen as a function of a temporally unfolding relationship between the respondent and researcher, challenging the fiction of ‘clean’ raw data. Attempts to ‘see like a research project’ (Biruk, 2012) circumscribe the imagination of what is possible, and draw boundaries around perceived realities. Consequent endeavours to smoothen out errors by aiming to control the data collection process through probes and callbacks result in a stilting of the realities of the contexts they wish to capture, and may not account for the role of bodies, affect and practices in data collection processes. This is best exemplified in Biruk’s description of the use of data collection processes such as bean counting, which were viewed by some as being patronising and were consequently difficult to implement, highlighting mismatches between researcher constructions of respondents’ epistemological processes. This goes to demonstrate how data collection is a negotiation and a performance, located in situated contexts.
The fifth chapter builds on previous chapters to closely examine how data are cooked, deployed and translated in policymaking and academia. The interpretation of seemingly immutable statistics is shaped by the predispositions of interpreters and the audience. Through a description of how statistics about the transmission of HIV/AIDS through men who have sex with men) was presented and received by ideologically segregated groups of people at different conferences, we see that the interpretation and credibility of statistics may hinge on non-data-related considerations including the social relations within which ostensibly data-driven knowledge is (re)interpreted. In the end, some numbers are more useful than others.
This book is a significant contribution to scholarship in anthropology more generally, and ethnography, epistemology and knowledge-making practices in particular. Biruk masterfully brings together insights from across the research process, with implications for how we may reimagine what constitutes data, and how our own disciplinary predilections and traditions—whether as demographers or anthropologists—shape our understandings of what are good data. Given Biruk’s sometimes overstated anthropological affiliation, at times the reader is left wanting for more, especially since at times it appears that the analysis precedes the ‘data’. Some sections also appear slightly repetitive and therefore overstated, the chapter on gifts being one in case. Most troubling is Biruk’s conflation of demography—as an almost caricatured, monolithic discipline—with all quantitative research in general, given that different disciplines that employ quantitative methods are driven by different logics and practices. These considerations notwithstanding, cooking data are a significant contribution to the scholarship on research and knowledge-making practices.