Some remarks on referentiality

The main focus of this paper is on two issues: which, if any, NPs can be used to refer, and which, if any, NPs can be (truthfully) said to refer all by themselves. It is evident just from this statement that we will be dealing with two distinct reference relations – a pragmatic relation holding among speakers, expressions, and entities out in the world (so to speak); and a semantic relation holding among expressions and entities. It should also be made clear that only singular reference will be at issue here, of either the semantic or pragmatic type. However I’ll assume that plural entities and masses are possible singular referents (as is standard these days – cf. e.g. Link 1983, Partee 1986). A couple of other background issues deserve sections of their own. In §2 we look at object-dependent propositions. I will sketch an argument for a view on which such propositions do not contain entities from the outside world, but in their stead contain constant individual concepts. §3 contains another argument for incorporating individual concepts into our semantics, based on the interpretation of indefinite descriptions. These sections will turn out to be relevant to the discussions which follow. In §4 and §5 we will take up in order the issues of (singular) pragmatic and semantic reference, closing in §6