The correspondence between attention states and the form of kind-referring NPs - general explanations for seemingly ad hoc facts

Two seemingly ad hoc properties of kind-referring NPs have often been mentioned in the literature on generic nominals. One is that definite singular NPs that refer to kinds (in examples like The lion is not yet in danger of extinction) denote a so-called "well-established" kind. This does not hold for all types of kind-referring NPs, though. A second issue which has often been mentioned is that indefinite singular NPs that refer to kinds (in examples like One lion (namely the Berber lion ) is almost extinct) always refer to a subkind of the most general kind of thing their descriptive content is compatible with, which does not hold for all types of kind-referring NPs either. In this paper, I show how these, and other, discrepancies between different types of kind-referring NPs can be given a perfectly general explanation if one assumes with Gundel, Hedberg, and Zacharski (1993) that an NP form signals a certain cognitive status for its associated discourse referent, regardless of what type of entity this discourse referent is.