Retour vers liste

Détail de la contribution

Auteur: Joanna BLOCHOWIAK

Towards a unified approach to logical and non-logical connectives – the example of ‘and’ and ‘because’

Abstract/Résumé: This paper defends the position according to which it is possible to adopt a unified approach to connectives based on the thesis saying that semantic meaning is minimal and non-ambiguous, and that meaning variations in use are the result of pragmatic processes (Horn 1985, Moeschler&Reboul 1994). Crucially, the hypothesis of logical part of meaning as the core semantics has been adopted for logical connectives (i.e. those which have their counterpart in logic as e.g. ‘and’) and it will be extended here for the analysis of non-logical connectives, in particular ‘because’. It has been convincingly argued that the semantics of ‘and’ is constituted by its logical meaning (Grice 1989). This meaning is minimal in the sense that the only requirement concerns the truth-value of the linked propositions while all the relations ‘and’ can express are some sort of implicature since they can be felicitously cancelled (generalized or scalar implicature in neo-Gricean approaches (Levinson 2000, Horn 1972) or an explicature in post-Gricean frameworks (Sperber&Wilson 1995,Carston 2002). This analysis explains why ‘and’ is compatible with uses conveying relations (mostly in natural language) but also with uses involving no relations whatsoever (in logic). I will claim that this kind of approach has also the power to explain main differences between ‘and’ and ‘because’ as well as to provide the complete analysis of all uses of ‘because’. Concerning the differences I will argue that in the case of ‘because’ the conveyed relations have a stronger status than implicature, as they cannot be cancelled. It means that the existence of a relation between what is denoted by the propositions linked by ‘because’ is an obligatory requirement. This phenomenon, in turn, is the consequence of the hypothesis that ‘because’ is always uttered with respect to a law or rule reflecting speaker’s world knowledge about relational regularities. Such regularities express relations that are inherited by because-sentences. I will use this hypothesis to propose the definition of basic semantics of ‘because’ which in addition to the truth-table identical with ‘and’ includes the ‘in virtue of L’ clause (in the sense specified in Greenberg 2007), where L is a non-accidental generalisation (lawlike-statement (Hempel 1948, Bromberger 1992)) that associates (e.g. causally) the types of states of affaires described by P and Q of the ‘P because Q’ sentence. Therefore, contrary to ‘and’, the fact that ‘because’ conveys relations is built in its semantics and thus these relations cannot be cancelled.