Tingjiang Kuang

Theories concerning fictional entities and fictional names face the so-called “wrong kind of object” problem. The problem, as outlined by Semejin and Zalta (2021), is that if fictional names denote abstract objects, then some parafictional statements (e.g., Sherlock Holmes is a private detective) would involve the wrong kind of objects (properties only applicable to concrete objects are applied to the abstract objects); if fictional names denote concrete objects in fictional worlds or worlds of pretense, then some metafictional statements (e.g., Sherlock Holmes is a fictional character created by Conan Doyle) would involve the wrong kind of objects (properties only applicable to abstract objects are applied to the concrete objects). This essay addresses the problem from the perspective of possibilism about fictional entities and fictional names. A solution based on two-dimensionalism and counterpart theory will be given. It will be argued that a metafictional sentence is true if considered uttered in a fictional world but evaluated in the actual world. If the sentence is considered uttered in a fictional world, a fictional name picks out a concrete object, and a proposition about the concrete object is determined. This proposition is true if evaluated in the actual world because there is a fictional-person idea (Everett & Schroeder, 2015) in the actual world, which is the counterpart of the denoted concrete object. As a result, the “wrong kind of object” problem is circumvented, as it is the idea that properly receives ascriptions of metafictional properties.

This paper develops Lewis’s (1983) analysis of truth in fiction with conventional semantics to deal with the problem of impossible fiction. Since impossible fictions depict impossible states of affairs, Lewis’s framework, based on possible worlds, struggles to account for their truth conditions. Previous approaches, including quantifying over consistent fragments of fiction (Lewis, 1983) and appealing to impossible worlds (Badura & Berto, 2019), have their limitations. Instead, this paper proposes a novel solution based on conventional semantics, in which convention shifts can alter the meaning of names, predicates, and logical constants. It will be argued that conventional semantics is especially suitable to explain truths in unintended impossible fiction, but the explanation can also be extended to intended impossible fiction. The paper concludes that conventional semantics provides a compelling alternative for analysing truth in impossible fiction.