Retour vers liste

Détail de la contribution

Auteur: Giuliana GIUSTI

Genitive Case and the Nominal Phase

Abstract/Résumé: This paper extends Pesetsky and Torrego’s (2001, 2007) analysis of structural Case to genitive in nominal expressions. Basic questions: What is the uF parallel of the clausal Ts or To, which is present in the nominal projection and targets the possessor DP? What is the uF realized as Genitive parallel to uT of nominative and accusative case? Assuming with Rizzi (1997) that the clausal edge can split in Force and Fin, the latter being involved in Nominative Case assignment (through feature inheritance, Richards 2007), is there a parallel split in the nominal edge, and what is the feature parallel to Fin? We follow Hinzen’s (2012:325) threefold semantic ontology: Objects have reference in space. Events have reference in time. Propositions have reference in discourse. Accordingly, vPs refer to events, CPs refer to propositions, nominal expressions refer to individuals. If the phase is embedded, the left edge is interpreted in the low phase as carrying the referential value and in the higher phase as part of the higher selector. We will show that this proposal derives the necessity of subjects in propositions, and the optionality of possessors in nominals. Our research is based on the syntax of possessors in languages with and without articles, in particular we highlight the following properties: Genitive is assigned by a D-element, either as inherited by the immediately lower non-phase projection (Italian: la loro bella casa, Romanian: casa lor/Mariei fromoasă “their/Maria’s house”) or in the Left Edge (John’s home) The targeted feature of the possessor is Person, which is present in pronominal possessive adjectives (la nostra responsabilità “our responsibility”) but is not in relational adjectives (la responsabilità italiana “the Italian responsibility) in Italian. Cross-linguistic variation is found as regards the pied-piping property of D. In Italian only pronouns are pied-piped (la loro casa (“their house”) vs. la casa dei vicini (“the neighbours’ house”). In English only a subclass of lexical Ns are pied-piped (yesterday’s paper vs. the legs of the table). In Hebrew construct state and PPs are quite freely available. Genitive extraction varies cross-linguistically, but when possible, crucially respects a thematic hierarchy (cf the seminal work by Cinque 1980 and many following him) We will observe that pace Boskovic (2005 -2012), the different patterns of Genitive assignment and extraction do not distinguish DP/NP-languages but languages that assign genitive through Agree only, and languages where genitive assignment requires Concord (Spec head configuration, Giusti 2008), or even fusion (in the sense of Nunes 2004) with the selecting head, as apparent in English [DP[XPwho][D’[D‘s][NPbook]]] --> [[DPwho‘s][NP book]]].