Some debates about persistence lie squarely in the domain of metaphysics, but the central issue at stake between the worm theorist and stage theorist of persistence actually concerns the semantic value of ordinary referring expressions. The worm theorist says that ordinary referents are temporally extended objects, while the stage theorist says they are instantaneous temporal parts of persisting objects. Stage theorists claim to give the best account of our intuitions about the truth conditions of counting sentences such as “there is just one person in my room” when I am in my room alone, even in a fission case when I am about to be split into two people.

I argue against several versions of the stage theory and in favor of an alternative theory of counting sentences. I begin by introducing a problem for the stage theory. This problem naturally suggests an alternative theory of persistence, one that resembles extant theories of reference developed by formal semanticists. I introduce the relevant formal semantics literature on “event-related” readings of counting sentences and develop a perspective theory of persistence, according to which we do not ordinarily refer to only short-lived stages or extended worms, but temporal segments of objects whose length is fixed by the context of utterance. I argue that a modified version of the perspective theory can account for a number of intuitions about counting sentences. I demonstrate that the perspective theory is a natural recourse for any former stage theorist, and then I present some challenges for the perspective theory and defend an error theory of our ordinary language judgments about counting sentences. Finally, I argue against analogs of the perspective theory defended in the formal semantics literature, and I introduce and defend my own theory of judgments discussed in that literature.