[In search of politics in knowledge production. A plea for a historical-political epistemology].

Knowledge production has an intrinsic political dimension. Starting from this presupposition, it is argued that the systematic integration of and reflection on the political dimension is necessary for an adequate understanding of historical processes of knowledge production in the sciences. The consecutive plea for a historical-political epistemology proceeds in two steps: First, it is illustrated that in a number of recent historical science study cases, the political dimension is frequently marginal, or even absent. After a short discussion of previous theoretical concepts to describe the impact of politics for the production of scientific knowledge, an approach is sketched which builds on Hans-Jörg Rheinberger's historical epistemology and Bruno Latour's symmetrical anthropology. It is argued that in addition to Rheinberger's program to describe epistemic systems, the political dimension is intrinsic to three stages of the process of data production: First to an initial phase which consists in the arrangement socio-technical configurations to produce new evidence. Here, factors such as the culturally shaped perception and evaluation of "relevant" problems, as well as the perception of career resources have to be taken into account. Second, the political dimension is relevant in view of the continuous re-adjustments of the configuration of epistemic systems, e.g., towards newly available financial, technical, or intellectual resources and "relevant" challenges from outside the system. Thirdly, the data produced and represented by epistemic systems--"evidence"--are yet in need of interpretation. This process is in itself imbued with continuously shifting mechanisms of selecting and creating hierarchies amongst the pool of available data.