For the past couple of years, I’ve been working through arguments about techniques of enumeration and the role of surveys, disease surveillance systems, and other quantitative modes of knowledge production in health, development and relief programs. A recent collaboration with medical anthropologist friends and colleagues even produced a special issue in Medical Anthropology.
I am currently working on a paper about how numerical accounts are derived from surveys – as opposed to narrative ones –and how these accounts are used to justify and legitimize existing programs and public health interventions. While this idea is not new (see any social science literature relying on Ian Hacking’s “Avalanche of Printed Numbers” in the 30 years since that article was written), surprisingly few ethnographic accounts of survey practices exist. Perhaps some folks think it’s boring to study and write about the nitty-gritty of survey research: how survey instruments are developed; how sampling strategies are devised (and interpreted by the surveyed); how research teams gain access to ‘the field’ (delineated by them); how researchers translate seemingly universal survey concepts like “the household” or the “family;” or how to get as few “other” responses as possible…
I think we can learn something by investigating these practices. For example, how do the categories we take for granted come into being?How do they gain authority or legitimacy while others fall to the wayside? How do categories change over time? What kinds of negotiations happen to get us our “clean” data and our “facts”?
That said, in my paper, I reflect upon my experiences working on a gender-based violence prevalence survey amongst Liberian refugees some years ago. I address the assumptions about the nature of ‘gender-based violence’ upon which survey instruments are based: why do we focus solely on women, when we are concerned about gender norms and relations that influence violence—concerns that should impel us to focus on men and women (and the fluid categories along the gender spectrum) ? And what happens when some of our findings question our fundamental assumptions of how rape and various forms of sexual violence affect psychological response?
Surveys convey stories. They are populated by a range of characters (e.g. the adolescent victim of sexual violence, or the militia perpetrator), specific events and time periods of significance to researchers (e.g. before the war, during flight), and settings (e.g. Aleppo, the refugee camp, the household). Studying how these stories are actively produced may help us find new ways forward for addressing pressing problems.