Retour vers liste

Détail de la contribution

Auteur: Giuliana GIUSTI

Is AP a phase?

Abstract/Résumé: Assuming that CPs, vPs and DPs are phases, as extensively claimed in the literature, and given that APs can be predicates in the clause, parallel to vPs, the following two questions arise: • Are predicate APs phases, or do they need to be embedded into a vP in order for the lower clausal phase to be instantiated? • Are adnominal APs phases? And if so, is there a difference between direct and indirect modification, the latter being more similar to a predicate (Cinque 2010)? We base our argument on Hinzen’s (2012:325) threefold semantic ontology: a) objects have reference in space; b) events have reference in time, c) propositions have reference in discourse. In this framework, phases are the syntactic counterpart and the cause of the following interpretive tripartition: vPs refer to events, CPs refer to propositions, DPs refer to objects (apparent event nominals are conceptualized as individual (instances of) events. For a phase to be embedded into another phase, its left edge must have an uninterpretable feature that relates the lower phase to the upper phase (Case as uT in DP, some analogue uF in embedded CPs). APs refer to properties, they certainly do not refer to either objects or propositions. As a consequence, this framework can attribute phasal nature to APs if they are shown to refer to the state of having the denoted property. It is well known in the literature (cf Baker 2007) that contrary to verbs, adjectives across languages do not agree for the person features of their subject; in many languages they agree for other features such as number, gender or noun class (if they are predicates) or with the noun they modify (if they are adnominal). Furthermore, languages vary as to whether predicate adjectives have verbal, non-verbal or mixed inflection (Stassen 2011). In European languages adjectives do not inflect for tense and do not bear clausal negation (opposite to phrasal negation), for this reason they occur with a verbal element: The paper argues that the extended projection of AP is not a phase. As a consequence, in order for the adjectival expression to refer to a state, it needs to combine with a stative V that project a vP. This relates the properties above with other properties of adjectives: • As do not receive Case, but can only agree for Case • Direct and indirect modification can be related to the absence/presence of a null vP-predicate structure • Direct modification adjectives cannot select an argument