Séminaire de Recherche: Giuliana Giusti (U. de Venise)

On Concord and Projection - Mar 11 Mai - 12h15 - L208

This talk starts with a presentation of previous work of mine (cf a.o. 2008, 2009) according to which feature sharing should be analyzed as the result of at least three different processes: Agreement, Concord and Projection. I will briefly argue that:

• Agreement is the transfer of the Person features of the subject in the clause and in the NE. In the latter case the subject is what is generally call “possessor”. This is triggered by an EPP feature of specific which is merged in the left periphery of the intermediate layer with the effect that a theta-related constituent merged in the lexical layer is made visible to the uppermost layer of the constituent in which the interpretation of the constituent is computed at LF.

• Quite differently, Concord is the transfer of a bundle of features, (e.g. Number, Word Class, and Case specifications present in the functional projections of a NE) from a functional head onto its specifier (e.g. an adjective phrase with its functional structure). This is triggered by a non overt subject position inside the modifier.

• Finally Projection is what constitutes the spine of an extended projection in the sense of Grimshaw 1991, it remerges the bundle of functional features associated with the lexical head to all intermediate projections of the constituent and is at the base of feature sharing in Concord. The core of the talk focalizes on the interaction of Concord (the relation between the AP in SpecFP and F°) and Projection (the remerge of Niφ ) in (1). AP stands for a larger constituent with its functional structure

I will claim that in Romance languages Concord on As is overt due to their inflectional morphology which is constituted by a bundle of u-features (number and gender) to be deleted by Concord against the values of parallel i-features of the head remerged in the nominal projection, while in Germanic languages the adjectives cannot delete the u-features with overt inflection and the Concord can either be non-overt (as in English), or it can cause overt realization of part of the remerged bundle of nominal features, which then appears encliticized on the adjective (as is the case of Dutch, German and Scandinavia languages). I will show that this is not a parameter in the current sense but a property of each lexical item to be merged in Spec, in fact even in Italian we observe one adjective (pronominal be-l) and two determiners (the distal demonstrative que-l and the partitive determiner de-l) which are formed by enclicization of part of the functional bundle in F° onto its specifier, according to joint work with Anna Cardinaletti.

References Cardinaletti, A. & G. Giusti 2008-9. Optional Concord in nominal expressions. Talks delivered CIDSM, Pescara July 4-6, 2008; at IGG, Siena February 26-28, 2009; Symposium ACLC research group 'the DP/NP in Germanic and Romance languages' May 19, 2009; Convegno SLI Sept. 24-26, 2009. Giusti, G. 2008. Agreement and Concord in Nominal Expressions, In The Bantu-Romance Connection. C. De Cat & K. Demuth (eds.), 201-237. Amsterdam: Benjamins. Giusti, G. 2009. On feature sharing and feature transfer. Univeristy of Venice Working Papers in Linguistics. http://lear.unive.it/handle/10278/1376

Actualités (archive)