Paper short abstract:

In a global health program, local knowledge systems inform how village-level enumerators participate in monitoring activities. Exploring the collisions between local knowledges and universalistic project imperatives raises crucial implications for understanding intervention 'success'.

Paper long abstract:

Contemporary global health projects are increasingly expected to conduct 'evidence-based' activities. As such, monitoring and evaluation procedures tend to rely on the collection of quantitative data guided by predefined, fixed indicators of 'success'. This approach often neglects the messier and more complex realities of the implementation process that shape intervention outcomes. Drawing upon eight months of fieldwork in a Muskoka Initiative-funded maternal newborn and child health project, I demonstrate how ethnographic methodologies, when used alongside more conventional project monitoring and tracking systems, can reveal the complex and integral role that 'local knowledges' play in the everyday workings of interventions. Expanding on Geertz's concept of 'local knowledge', I argue that local knowledges enacted during the Muskoka project are traceable to a history of intervention projects in the region. As people participate in a multitude of development and research projects over extended periods of time, universalistic knowledge unfolds in locales to become grounded epistemologies. As new projects are introduced, individuals draw on these grounded knowledges to make sense of project goals; it guides their participation in project activities, including monitoring and evaluation procedures. In particular, by analysing the jottings that village-level enumerators mark down on survey instruments—displaying highly relevant understandings that do not fit pre-determined survey response options—I show how this knowledge unsettles the expertise of the scientists who designed the survey and also questions the assumed the authority of the supervisors who reinforce universalistic scientific logics.