Abstract

Summary One of the central questions concerning theories of reference has been the problem of how the reference of scientific terms gets fixed. Descriptive causal theories of reference, as discussed in this paper, have re-introduced the role of theoretical beliefs and conceptualisations in term introductions and reference-fixing. The present paper argues that the idea of reference-fixing as a dot-like event (baptism) is wrong: a number of episodes from the history of science are discussed to support the claim that reference-fixing is a historical, drawn-out process. This, however, does not stand in the way of successful reference. The two processes are simply separated. A criterion is suggested to determine successful reference. From this approach two further ideas follow: not all scientific terms actually have the power of referring and even those that do will always retain a residual indeterminacy