About
How are representations of violence influenced by the “data agency”, that is, the social practices of data collection and analysis in quantitative studies of armed conflicts?
A large number of causal interpretations of armed conflicts have benefited from empirical validation by quantitative research tools, long before the current debates on Big Data. The first large databases “measuring” conflicts were built after the emergence of behavioralism in the social sciences in the 1960s. Since then, publications using large data sets have helped reinforce and specify new theoretical propositions such as the decline of interstate conflicts since 1990 or the thesis of “democratic peace.” Furthermore, political institutions, NGOs, and the media increasingly rely on the results of quantitative studies to “predict” armed conflicts and develop their analytical and normative positions.
However, within the scientific community, the validity of results produced by positivist quantitative research on conflicts is increasingly being questioned. Criticisms focus on the reliability of collected data, the quality of mathematical models used in statistical analysis, and dissemination practices by major scientific journals.
But how do the results of quantitative research on conflicts actually influence the perceptions of practitioners regarding armed conflicts? Existing literature in political science mainly analyzes how figures are instrumentalized by governments in controlling populations and the digitization of the battlefield. In contrast, in-depth surveys on the reception of scientific production practices, including internal logics of data collection, analysis, and publication, are largely absent in existing research.
The main research question of this project is: how does the “agency” of quantitative data on conflicts, defined here as the set of scientific practices associated with the generation, processing, analysis, and scientific dissemination of large data sets on armed conflicts, influence representations and expectations related to war in the media, political institutions, and NGOs?
DATAWAR will conduct the first systematic study of scientific practices in the field of quantitative conflict studies and their impact on representations of war by practitioners, covering the entire life cycle of conflict data, from collection and analysis to reception and interpretation by journalists, NGO representatives, and officials of three major countries actively involved in managing international crises: France, Germany, and the United Kingdom.
The project explores the hypothesis that scientific production in conflict quantitative studies is less driven by theoretical innovation than by the “data politics,” that is, the availability, reputation, and mathematical manipulability of quantified conflict observations. Consequently, we anticipate that interpretations of conflicts developed by practitioners are subject to erroneous perceptions caused by the nature of available data, the type of mathematical models used for analysis and potentially “predicting” conflicts, and the overly restricted use of available theoretical arguments.
Leaders
Sciences Po
– Frédéric Ramel (project coordinator)
– Louise Beaumais
– Iris Lambert
Ecole Polytechnique (Linx)
– Thomas Lindemann (work package coordinator)
– Grey Anderson
Sciences Po Lille (CERAPS)
– Eric Sangar (work package coordinator)
– Sami Makki
