This thesis delineates the boundaries of theories of naming, placing particular emphasis on the demarcation of the semantic function of referring expressions from the pragmatics of referential usage. The need for such emphasis derives from frequent confusions in work on singular reference where pragmatic phenomena are used to refute semantic theories and vice versa. Part One of the thesis examines various conceptions of the semantic function of names. Frege's notion of 'Sinn' is shown to be incoherent as are its reincarnations by modern, neo-Fregean commentators: a more modest role for sense is suggested. Kripke's intuitions about scope-insensitivity in modal contexts are amplified and criticised. An investigation of singular thought yields a discussion of Evans 1 views and the removal of the apparent impasse that substitutevity failure causes for Millian theories. Part Two of the thesis concentrates on the mechanism of namereference, which works by connecting a contemporary use of a name via a causal umbilicus to some (or no) supposed referent. A number of anomalies are examined, from which the notion of a name-using practice and an elucidation of the criteria in virtue of which we judge a term to be a name are developed. In Part Three of the thesis the artificial super imposition of logical structure onto natural language is explored. The argument hinges on a plausible generalization of Donnellan's referentialattributive distinction. From this there emerges a picture of the complex network of intentions that surround the referring process. The role of definite descriptions in grounding the use of names is explained. Both the semantic and pragmatic pictures of name-reference are applied to the perennial puzzles of substitutivity failure and existential statements. The results underline the fact that a great deal of work needs to be done by those who subscribe to the growing orthodoxy of Millian theories, but that this is work in pragmatic areas that are too often ignored by those studying 'pure' semantics.