Even though automated information systems have been used in work life for almost three decades, the academic discipline of information systems development is still in a 'preparadigmatic phase'. There is no central corpus of a well understood and accepted theory of how these artifacts should be understood and designed. What we see is a set of scattered methods and theories, with influences from a wide variety of other disciplines, such as logic, linguistics, philosophy, cognitive psychology, organizational theory, ethnography, etc. There is a practical need for creating an overview and a deeper understanding of how different theories and methods are related to each other, what their relative strengths and weaknesses are and when they are applicable.In this paper we will present such an overview, based on a discussion about different theories of meaning. A framework is presented that divides the usage context of an information system into three areas of interest: referential aspects, individual aspects, and social aspects. Different theories elaborate on these aspects in different ways and to different degrees. There are strong analogies between these theories and philosophical theories of meaning that elaborate on corresponding aspects. To bring out these analogies we discuss the strengths and weaknesses of a set of approaches: (1) the traditional mainstream focus on referential aspects, (2) the adoption of cognitive theories in the information systems field, (3) structured theories about the social and communicative usage of information systems, e.g., theories based on the speech act theory, and (4) 'usage holism' and the criticisms of the role of abstract, theoretical modeling and analysis during information systems development. The paper does not present a final solution to the problems we address. Rather, it should be read as a contribution to what we believe must be an ongoing debate and a long term reflection on the theoretical foundation of information systems-research.