With few exceptions, science studies has neglected potential insights that could be gained by studying sociology, a discipline that has played an important role in its own development. This thesis addresses the specificities of sociology from the perspective of constructivist science studies and explores the tensions that arise as a result of this encounter. The theoretical framework is based on actor-network theory (the work of Bruno Latour in particular), and supplemented with the work of Pierre Bourdieu (his theory of sociology in particular) to examine issues specific to sociology. Via a document and interview-based study of sociology in Latvia set against the background of Soviet science, the thesis argues that an open-ended and normatively saturated conception of knowledge hampers the stability of the discipline. I suggest, however, that these qualities matter vitally to the long-term development of sociology as a form of knowledge. Applied to sociology, a science studies understanding of science illustrates what happens when the intimate connection between political representation and scientific representation is not concealed, and hybridity is acknowledged.