Retour vers liste

Détail de la contribution

Auteur: Emar MAIER

Quotation and Unquotation in Free Indirect Discourse

Abstract/Résumé: Free Indirect Discourse (FID) is form of reporting what someone said or thought. It shares linguistic features with both direct discourse (DD) and indirect discourse (ID). If we rely on the interpretation of pronouns to determine whether a given report construction is direct or indirect, FID would be a kind of indirect discourse (Sharvit'08). The obvious differences would then presumably follow from the syntax, i.e. FID is main clause, ID is subordinate clause. However, a quick glance beyond personal pronouns and tenses shows that ID and FID diverge drastically. In FID, indexicals like today, this, here, but also expressives and particles, are always interpreted from the perspective of the character whose thought/utterance is reported. But the crucial contrast between ID and FID is that FID allows the writer to slip into the language of the protagonist, thus shifting the idiolect, dialect, or register: She was angry. Oh, how they misunderestimated her! I conclude that analyzing FID as ID is a non-starter. A popular alternative formal semantic analysis of FID involves tinkering with the semantic context parameter. Schlenker and Eckardt propose that interpretation may depend on two distinct context parameters simultaneously: the context of the narrator (C), and that of a protagonist (c). In FID a split in context dependence manifests itself: pronouns and tenses always get their reference from C, other indexicals from c. This analysis builds the distinct behavior of indexicals and pronouns/tenses into the lexicon. In Schlenker's account in particular, it follows that all and only pronouns and tenses are narrator-centric. This prediction is not quite borne out: proper names can also be narrator-centric (Maier'12), and some pronominal (gender) features are protagonist-centric (Schlenker'04). Finally, language shift examples are problematic for this double context account in that they force us to introduce "the language of utterance" as a shiftable indexical parameter. In light of the language shifting in particular, I propose that FID is really a form of quotation. I apply the semantics of mixed quotation, whereby '"X"' means 'what x refers to as X'. This way, language and indexical shifting is entirely to be expected. The fact that pronouns and tenses are usually not shifted to the protagonist is captured by unquoting them: She was angry. "Oh, how they misunderestimated [her]" Eckardt 2012. Particles as speaker indexicals in free indirect discourse. Maier 2012. Quotation and Unquotation in Free Indirect Discourse. Maier 2013. Mixed Quotation: The grammar of apparently transparent opacity. Shan 2011. The character of quotation. Schlenker 2004. Context of thought and context of utterance. Sharvit 2008. The puzzle of free indirect discourse.