Abstract

Introduction We begin with a statement of what we mean by theory. We do not attempt to provide a formal definition here. By theory, we mean a system of statements. A theory has to be a theory about some reality. Since we consider IR as an empirical science, we have to ask what is the reality in IR. In IR we deal with two kinds of reality. The artificial or formal reality of systems and implementations and the reality of documents, users, relevance, preference, etc. We hesitate to call a mathematical theory about systems an empirical theory because then all models 1 that can be implemented would give us an empirical theory. When we talk about empirical theories in IR, we refer only to theories about the reality that lies outside systems; in other words, about documents, users, relevance, etc. In this sense, we make a distinction between IR-theory and<F31.93

Citations