Over time, various thematic classifications have been put forward to organize science into a coherent system of specialized areas of research. From an analysis of the historical evolution of the criteria used to distinguish the sciences from one another, I propose in this paper a quadripartite typology for the different thematic classification systems propounded by scholars throughout the centuries. Basically, I argue that the criteria used to differentiate the sciences have been alternately drawn from their respective subject matters, kinds of (...) knowledge, methods and aims. Then, I show that several reclassifications occurred in the thematic structure of science. Finally, I argue that such changes in the structure of learning displaced the modalities of contact between the objects, knowledge, methods and aims of the various branches of science, with the result of outlining reshaped intellectual territories conducive to the emergence of new areas of research. (shrink)