ABSTRACT
The usual distinction between 'definite' and 'indefinite' articles is commonly explained as incorporating a semantic distinction between 'specific' and 'non-specific' reference. The traditional and taxonomic accounts cited so far have been superseded in recent years by 'locational' accounts of determiner interpretation, notably in Lyons and Hawkins. Anaphorics turn out to be noticeably more complex than non-anaphorics, at least by one line of semantic representation, and the author claim that this is because they are actually founded upon non-anaphorics, in that they are subsequent references to initial non-anaphoric uses. In natural language, though, descriptive generics are statements of typical or characteristic states of affairs, and not universals; they are normality statements, and not probabilities or certainties. Finally, Dahl takes Lawler's distinction between Universal and Existential generics and translates it into a possible world/possible times account. There is an ample philosophical literature, originating with Russell, which discusses the problem of 'definite descriptions'.
