Science and its significant other: Representing the humanities in bibliometric scholarship

Résumé : Bibliometrics offers a particular representation of science. Through bibliometric methods a bibliometrician will always highlight particular elements of publications, and through these elements operationalize particular representations of science, while obscuring other possible representations from view. Understanding bibliometrics as representation implies that a bibliometric analysis is always performative: a bibliometric analysis brings a particular representation of science into being that potentially influences the science system itself. In this review we analyze the ways the humanities have been represented throughout the history of bibliometrics, often in comparison to other scientific domains or to a general notion of the sciences. Our review discusses bibliometric scholarship between 1965 and 2016 that studies the humanities empirically. We distinguish between two periods of bibliometric scholarship. The first period, between 1965 and 1989, is characterized by a sociological theoretical framework, the development and use of the Price index, and small samples of journal publications as data sources. The second period, from the mid-1980s up until the present day, is characterized by a new hinterland, that of science policy and research evaluation, in which bibliometric methods become embedded.