Abstract
Scientific theories are analyzed in terms of the role that they play in science rather than in terms of their logical structure. It is maintained that theories: provide descriptions of the fundamental features of their domains; on the basis of 1, explain non-fundamental features of their domains; provide a guide for further research in their domains. Any set of propositions that carries out these functions with respect to some domain counts as a theory. This view of theories is developed and defended, and provides the basis for reconsidering a number of issues in the philisophy of science. It is argued that theories need not be unrestrictedly universal with respect to space and time; that the distinction between observable and theoretical entities fails because observables do function as theoretical entities in scientific theories, that there is no genuine philosophical problem of reduction; that the notion of "levels" can be replaced by the notion of "domains"; and that theories are the basic unit of scientific knowledge