París, Francia
This article argues that, within French sociology, there exists a little known and rarely demonstrated tradition in the use of statistics. This tradition is original in as much as it attempts to escape from the usual alternative of unconditional defense of statistics or its systematic criticism. This French tradition considers statistics to be a potentially valuable, powerful and sometimes irreplaceable tool, provided that their use is accompanied by a reflection on their limits and scope. In this sense, this tradition advocates a reflexive use of statistics in sociology, i.e., a use that integrates sociological reflection on statistics, on their limits, and on what it means and implies to use them in sociology. This approach is not shared by all French sociologists, though it tends to bring together some of their leading figures. The article focuses on Auguste Comte, Durkheimian sociologists (mainly François Simiand and Maurice Halbwachs) and Pierre Bourdieu. Despite their differences, due particularly to the fact that they worked at different times, one can identify a continuity in their thinking on statistics.