Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.philol.msu.ru/~otipl/new/mscl/2008/abstracts/ramaglia.pdf Дата изменения: Wed Mar 19 02:19:58 2008 Дата индексирования: Thu Apr 8 09:36:02 2010 Кодировка:

Nominal modification at the Syntax-Semantics Interface: comparing attributive adjectives in Italian and English
In his seminal work, Bolinger (1967) discusses an asymmetry between two functions that adjectives can perform, namely attributive and predicative modification; furthermore, he points out that the syntactic asymmetry between the pre- and the post-nominal position of a number of adjectives in English is related to a semantic difference between reference- and referentmodification, respectively. In later works, it has often been argued that a similar interpretive asymmetry can be found in many typologically unrelated languages and that it can be connected with different morpho-syntactic properties of the adjectives themselves. As far as Italian (as well as other Romance languages) is concerned, it has been observed that the position of an adjective w.r.t. the head-noun usually induces a specific interpretation. Consider, for example, Bernstein's (1993) account of the asymmetry between pre-nominal (non-restrictive) and post-nominal (restrictive) modification, and the large number of semantic properties discussed in Cinque (2007), distinguishing "fuctional" (i.e., non-predicative) and "lexical" (i.e., predicative) adjectives (cf. (1) and (2), respectively): (1) Un vecchio amico a old friend `An old (= long-standing) friend' (2) Un amico vecchio a friend old `An old (= aged) friend' (cf. L'amico Х vecchio `The friend is old (= aged)') This work aims at providing a comparative analysis of adjectival modification in Italian (as a representative of Romance) and English within the Minimalist framework. Assuming the abovementioned dichotomy between functional and lexical adjectives, I propose that their semantic asymmetry is correlated to a specific difference in their syntactic structure. In particular, I argue that the Merge position of a modifier is determined by the semantic relation it establishes with the head-noun. In the spirit of Bouchard (2002), I take the semantics of a common noun to be a network of interacting features, and I propose that each of them is encoded in a dedicated functional head within the nominal structure. With this in mind, I claim that functional adjectives are merged as Specifiers within the extended projection of the NP (cf. Cinque 1994): in particular, every adjective of this kind establishes a local (i.e., Spec-Head) configuration with the functional head encoding the specific semantic feature related to the adjectival class to which it belongs; this kind of structure is interpreted as the intersection between the denotation of the adjective and the single semantic property encoded in its corresponding functional head, thus deriving its intensional semantics. For instance, assuming that X° in (3) encodes the specification of the time interval in which the function `president' holds, my proposal is that the AP futuro `future' (merged in Spec,XP) only modifies the X° head (not its complement NP): as a consequence, the relevant DP is interpreted as denoting an individual characterized as a president at an interval of time in the future. (3) Il futuro presidente [DP il [XP [AP futuro] X° [NP presidente]]] `The future president' On the other hand, as far as lexical adjectives are concerned, I propose that they are merged as DP-internal predicates: following Rebuschi (2005), I assume the existence of a predicative ConjP projection encoding an intersective relation between the denotation of the adjective and that of its subject (corresponding to the NP in (4)); this structure therefore derives the extensional semantics of this kind of adjectives: (4) Il libro rosso [DP il [ConjP [NP libro] Conj° [AP rosso]]] the book red `The red book'


This analysis allows for a compositional approach to the semantics of adjectival modification (cf. Bouchard 2002), including the case of the so-called "non-intersective" adjectives (cf. (3)). Indeed, the interpretive difference between a lexical and a functional adjective does not depend on the fact that the former, but not the latter, is interpreted intersectively. Rather, in both cases the denotation of the adjective intersects with the set determined by the modified item. This corresponds to the whole network of semantic features constituting the noun in the case of lexical (i.e., extensional) adjectives (cf. (4)), and to a limited part of its semantic properties in the case of functional (i.e., intensional) adjectives (cf. (3)). On the basis of this proposal concerning the Merge structure of the two kinds of adjectives, I will then concentrate on the different position in which they are realized in Italian and English. While in Italian it nearly always coincides with their Merge position (functional APs appearing prenominally ­ with the exception of relational and thematic adjectives, for which a specific derivation will be suggested ­ and lexical APs being post-nominal), English requires most of its lexical adjectives to be preposed to a pre-nominal position: (5) An old friend (ambiguous: `long-standing'/`aged') (cf. (1-2)) (6) The red book (cf. (4)) Following Bouchard's (2002) suggestion, I will show that the surface difference in the realization of lexical adjectives in these two languages can be ascribed to an asymmetry in the encoding of Number as an interpretable feature. In particular, I propose that in English the realization of Number on the nominal head (e.g., dog vs. dogs, man vs. men) triggers the movement of a lexical adjective to the Spec of an Agreement Phrase within the extended projection of the NP, where its agreement in Number with the head-noun can be checked: (7) The red book [DP the [
AgrP [AP

red]
Spec-Head

Agr° [ConjP [NP book] Conj° tAP]]]

As far as post-nominal adjectives in English are concerned, I will follow Bouchard (2002) in assuming that their position is due to the fact that they do not play any role in the determination of the extensity of the nominal expression in which they occur: as a consequence, they do not fall under the scope of Number, which allows them not to have to agree in Number with the headnoun. As for the derivation of lexical adjectives in Italian, I will propose that the realization of Number on the determiner in this language (e.g., il `the-SING' vs. i `the-PLUR') allows them to check the relevant feature in situ (i.e., in their post-nominal Compl,ConjP position) under an Agree relation with D°: (8) Il libro rosso [DP il [ConjP [NP libro] Conj° [AP rosso]]]
Agree

This proposal has far-reaching consequences for the analysis, as it can also derive a number of additional properties of attributive adjectives (including their ordering restrictions) and the scope relation between different kinds of modifiers.

References Bernstein, J. (1993), Topics in the Syntax of Nominal Structure across Romance, PhD. dissertation, CUNY. Bolinger, D. (1967), "Adjectives in English: Attribution and Predication", in Lingua 18, 1-34. Bouchard, D. (2002), Adjectives, Number and Interfaces, Oxford, Elsevier. Cinque, G. (1994), "On the Evidence for Partial N-Movement in the Romance DP", in G. Cinque et al., Paths Towards Universal Grammar. Studies in Honor of Richard S. Kayne, Washington, DC, Georgetown University Press, 85-110. Cinque, G. (2007), The Syntax of Adjectives. A comparative Study, ms., University of Venice. Rebuschi, G. (2005), "Generalizing the antisymmetric analysis of coordination to nominal modification", in Lingua 115, 445-459.