Retour vers liste

Détail de la contribution

Auteur: Regine ECKARDT

What does, and doesn't shift - The rôle of speakers in free indirect speech

Abstract/Résumé: What does, and doesn’t shift Free indirect discourse requires the reader to interpret passages of text as thoughts or utterances of a protagonist. A semantics for free indirect discourse offers new access to a much-debated distinction in current literature, the one between asserted („at issue“) content and expressive („conventionally implied“) content. Consider the following example: (1) Tom sighed. He had lost, alas! In semantic interpretation, the second sentence gives rise to two propositions: (a) that he (= Tom) had lost, (b) that Tom regrets he had lost. These two propositions will add to the ongoing story in different ways. (a) conveys a thought of Tom. Hence, the story requires an update by (a’) ‚Tom thought that he had lost’ Possibly, the reader will infer from (a’) that Tom lost indeed, but in a literal interpretation, not more than (a’) should be predicted. In contrast, the propositions in (b) does not report a thought by Tom but an actual sensation of Tom. The overall story requires an additional update by (b’) ‚Tom regretted that (Tom had lost)’ It is not necessary, and not even appropriate to render alas! by anything like ‚Tom believed that he regretted that he had lost’. In my talk, I take a closer look at the management of world parameters in intensional semantics and free indirect discourse. I propose that the world parameter of expressives, modal particles, emotives and other speaker oriented items always refers to the world-of-utterance. Potentially, then, commentary content is modelled in terms of diagonalized Kaplan’s characters, whereas the world parameter of neutral lexical entries follows different patterns and requires a different treatment in intensional semantic interpretation. The resulting account locates propositional attitudes, utterances and commentaries in the overarching framework of context dependent meaning in the tradition of Kaplan. I argue that different notions of ‚meaning’ in Kaplan’s terms are needed in order to distinguish the different ways in which speakers assert propositional content and comment on it.