Disagreeing about fiction

Jury:
Stacie FRIEND : Senion Lecturer in Philosophy, Birkbeck, University of London (Présidente)
Françoise LAVOCAT : Professeure de littérature comparée, Université Paris 3-Sorbonne nouvelle (Rapportrice)
Emar MAIER : Assistant Professor Philosophy & Linguistics, University of Groningen (Rapporteur)
Anne REBOUL : Directrice de recherche au CNRS, Institut Marc Jeannerod, UMR 5304 (Examinatrice)
Paul EGRE : Directeur de recherche au CNRS, Institut Jean Nicod, UMR 8129 (Directeur de thèse)
François RECANATI : Professeur au Collège de France (Directeur de thèse)

Abstract :
In this dissertation, I contribute to contemporary debates in analytic philosophy about truth, interpretation and reference in fiction. I defend a version of "functionalism" (originating in Kendall Walton’s work) which says that the key concept for analysing fictions is pretence or make-believe. In the first part, I argue against the modal account of truth in fiction and then introduce "pretence semantics". The modal account says that fictional statements are similar to counterfactual statements, which can be given truth-conditions using possible-world semantics. But the fictional well exceeds the possible, and also the impossible (of hypothetical impossible-world semantics). Moreover, the modal account is incompatible with a causal theories of semantic information which can be argued for independently. As for pretence semantics : it is a formal apparatus delivering fictionality- conditions (instead of truth-conditions) which derive from real features of the props used in games of make- believe and some "principles of generation". In the second part, I fine-tune pretence semantics on a case study. It is a literary debate about what is true in Kafka’s story "the Metamorphosis". Nabokov once argued against critics who say that Gregor Samsa has turned into a monstrous cockroach. Cockroaches do not get stuck on their backs ; Gregor is stuck on his back in the opening scene of the story. So, Nabokov argues, he cannot be a cockroach ; he must be a big beetle. In order to analyze this "great beetle debate", I use the notion of "faultless disagreement" which comes from epistemology. I thus investigate the indeterminacy of fictional events which, I argue, is neither linguistic nor ontological but pragmatic in nature. In the third part, I defend anti-realism about fictional names which says that fictional characters do not exist and that fictional names do not refer. Anti-realism is the same doctrine as functionalism applied to reference. Though intuitive, the view has to meet a powerful counterargument based on "metafictional uses" of names. For instance, one can say truly : "Emma Woodhouse is a fictional character". Given compositionality, it follows that the name "Emma Woodhouse" refers in such contexts. This argument leads to a form of realism : fictional names refer to some kind of "abstract artefact". I show that the best realist theories are inadequate. Then I provide an analysis of the linguistic data introducing a notion of perspective. It enables me to circumscribe the problematic metafictional statements for the anti-realist. Ironically, anti-realists mainly struggle with negative existentials, although these put into words the central tenet of anti-realism, namely that fictional characters do not exist. Pretence semantics (which yields fictionality-conditions) is helpless for giving truth-conditions to them. To account for them, I use a version of positive free logic which combines with pretence semantics. Anti-realism is thus both intuitive and tenable.