Retour vers liste

Détail de la contribution

Auteur: Raffaella ZANUTTINI

Co-Auteur(s): Judy B. BERNSTEIN, William Paterson University, United States of America

Referential Dependencies across Discontinuous Subjects

Abstract/Résumé: The existence of movement operations was originally postulated to account for syntactic dependencies such as the one created by wh-movement or Quantifier Raising. More recently, some referential dependencies that used to be analyzed as the result of binding relations have been reanalyzed as the result of movement, e.g. the relation between a pronoun and its antecedent (as in "John thinks he’s smart", cf. Kayne 2002), cases of obligatory control (Hornstein 1999), and the relation between a noun phrase and a clitic that doubles it (Uriagereka 1995). In this paper, we discuss a novel kind of referential dependency, attested in Appalachian English, a variety spoken in the southeastern US. We argue that it involves movement, though of a different kind from the types just mentioned. The referential dependency under discussion involves a plural pronoun and a QP that are not adjacent and yet are interpreted as the subject of a clause: (1) We don’t any of us need anything. (AppE; M&H: p. 413) (2) They didn’ nobody live up there. (our fieldwork) (3) You'll everyone have a big time. (Tiffany W, p.c.) In these examples (ungrammatical in standard English), the pronoun in initial position and the QP are interpreted as if they were unit. These subjects differ from Floating Quantifiers, since here the set of quantifiers is different, the QP contains a nominal element, and only a pronoun (no lexical DP) can co-occur in pre-verbal position. They also differ from cases of clitic doubling, which typically involve a pronoun co-referential with a definite noun phrase (e.g. Spanish "Lo vi a Juan"). Finally, they differ from cases involving Italian "ne" or French "en", ‘of them’, which do not leave a nominal element behind and do not express person distinctions. We propose that these cases involve movement of a subset of the features associated with the QP: those that express whether the domain of quantification includes the speaker, the addressee or neither, which are realized as the pronouns "we", "you" and "they". We argue for a movement relation in part because these split subjects are dependent on the presence of modals or negated auxiliaries, elements which (in English) occur in the IP layer, higher than lexical verbs, and can raise to the CP layer. Given this dependence, we suggest that it is the raising of the modal or negated auxiliary to a position higher than the canonical subject position that makes possible the splitting off of certain features from within the QP. This work offers a movement analysis for a kind of discontinuous subject that has not yet been described. If successful, it contributes to a body of work that explores the role of movement in establishing referential dependencies.