Paper short abstract:

This paper examines conflict early warning systems as a hierarchical way of generating knowledge about local contexts in a way that is refracted through postcolonial insecurities and donor relations, with Liberia as a case study.

Paper long abstract:

This paper examines conflict early warning systems from the two perspectives of the political economy of conflict prevention and peacebuilding work in West Africa, one one hand, and the socio-technical context of data collection and interpretation on the other hand. The focus is on national early warning systems, with Liberia as a case study. Briefly, an early warning system of this sort consists of local-level data collection on events that are deemed indicative of a brewing conflict, a central body (a government agency, university, or CSO) that aggregates, validates, and analyses the data, and transmits a recommendation for action. Despite frequent invocations of local ownership in such systems, their work is in practice overwhelmingly steered (and financed) by the vicissitudes and preoccupations of key donors. At the same time, the local-level data collection is subject to a whole range of constraints, ranging from economic and technical (i.e. do local rapporteurs have reliable access to mobile telephony), to the epistemological (e.g., do all rapporteurs understand the categories used in the system, e.g. "gender-based violence", the same way), to the political (e.g., do some rapporteurs face threats when trying to report certain incidents). This paper, based on a research collaboration between the Folke Bernadotte Academy and the Monrovia-based Ducor Institute, examines these and other various issues to suggest that early warning must not just be understood as a relatively straightforward conflict prevention mechanism but as an arena for the (re)definition and silencing of disparate grievances and concerns.