Retour vers liste

Détail de la contribution

Auteur: Yan HUANG

Unarticulated Constituents (UCs) in Neo-Gricean Pragmatics

Abstract/Résumé: In recent years, the concept of unarticulated constitutes (UCs) has generated heated debates in both the philosophy of language and linguistic semantics and pragmatics (see e.g. Recanati 2002, Stanley 2002, Marti 2006). By UCs is meant a propositional or conceptual constitute of a sentence that is not linguistically explicitly expressed in that sentence. Stock UC examples include (i) It’s raining [in where], (ii) John is ready [for what] and (iii) John went to a bookshop and [then] bought a novel. The aim of this paper is twofold. Firstly, while examples like (iii) is traditionally analysed as containing a conversational implicature (see e.g. Grice 1989, Horn 2004, Levinson 2000), examples like (i) and (ii) have never received any formal treatment in neo-Gricean pragmatics. To fill this gap, I’ll provide a neo-Gricean account of the type of examples like (i) and (ii) in terms of Horn’s (2004) R- or Levinson’s (2000) I-principle, taking into account the Gricean distinction between generalised conversational implicature (GCI) and particularised conversational implicature (PCI). Secondly, following Huang (2007), I shall defend the neo-Gricean position that the pragmatic enrichment of UCs in these examples is nothing but a conversational implicature. The reasons are threefold. In the first place, the pragmatic enrichment that is required to recover the semantic content of UCs is engendered largely by the same Gricean pragmatic mechanism that yields a conversational implicature. Secondly, currently there is no reliable test either in theoretical pragmatics or in experimental pragmatics that can be used to distinguish alleged explicature, as in Relevance theory, the pragmatically enriched said, as argued by Recanati or impliciture, as argued by Bach, from conversational implicature. Thirdly, given the metatheoretical principle known as ‘Occam’s razor’, a unified, implicature analysis is methodological preferable, because it postulates fewer theoretical categories or representational levels than its competitors.