Symposia: Syntax


Le document suivant n\’a pas traslation français. Ainsi, seulement la version anglaise est offerte.
Pour plus de détails, contactez s\’il vous plaît:

The Internal Syntax of the Noun Phrase

As a working hipothesis we assume the existence in a given language of a constituent named noun phrase (NP), expressed by zero, one, or more elements. In this call for papers we will leave aside relativization (already dealt with in the Amazonicas II conference), nominal morphology, and (almost) all those properties of the NP that depend on its relationship with other constituents within its domain of occurrence (position, case — except the genitive —, coordination, adposition, and the way it is accessed by the information structure of the clause). The formation of an NP responds to two motivations: characterizing a type of entity (comprehension) and delimiting a class of referents (extension). The formal aparatus that takes care of those functions may specialize in one of them or cover both. If we discuss the NP from the formal point of view, we distinguish the head from its dependents. The nature of the head conditions the internal syntax of the NP in two ways. First, in theoretical terms, the extent to which the notion of functional head (determiner) suggests the existence of a structural parallel between the NP and the clause (X-bar hypothesis complemented by the DP hypothesis). Second, in terms of the noun subclass that heads the NP. This call is centered on the more traditional notion of lexical head. Two types of lexical heads are referentially univocal and basically void of comprehension, so that they tend to form single element NPs: pronouns and names. Deverbal nouns allow for dependents, but are restricted in the way they combine with them (for instance, the recovering of arguments is prone to preempt the expression of the genitive). The absence of a lexical head may take place in two ways. The head is elided but the dependents remain (la Ø de enfrente). This interpretation involves some problems (lo mejor is not lo Ø mejor). Another situation in which the head can be thought as absent is so called “headless relatives”, a common label though somewhat improper for nominalizations in many languages. Dependents divide into arguments and modifiers. Lexical arguments are true NPs ([the king of England]‘s will), which take or not a dependency mark. The latter may occur on the head (la mesa su-pata in some varieties of Amazonian Spanish). When the pronominal and the lexical expressions of the argument co-occur (last example), the question of their mutual syntactic link arises (lexical argument and inflectional agreement versus pronominal argument and coreferential adjunct — a clue for non-configurationality). The presence of arguments pressuposes an argument structure of the head, hence valency. The structure of a genitive construction will vary depending on the valency of the head, for instance a monadic head (“alienable”) may generate a more complex configuration than a diadic one (“inalienable”). The argument may also have the shape of a finite complement clause (la sospecha de [que [nunca volvería]]). Modifiers are either lexical or gramatical. The latter, determiners, are basically articles and demonstratives. Keeping in mind that: 1) many languages lack articles, whereas others have only one (definite or indefinite) and some have two or more (in case of number and/or gender syncretism); 2) articles and demonstratives do not always form a substitution class; 3) the definitions of ‘article’ are heterogeneous: (in)definite marker versus morpheme required for a noun to instantiate an NP; 4) there is a clear diachronic relationship between demonstratives and articles. Adjectives (where they exist), adverbs (el paisano aquí), and non argumental NPs (tea for two) are the common lexical modifiers. Noun adjuncts do not always come with adpositions (Dixieland rock) and the genitive mark is used both to express arguments (interiorval2 do corpoargmt) and adjuncts (estradaval1 de terraadjnt). An adjective may project a phrase and even license case (indiferente ao sentimento, livre de obstáculos). Several categories are recruited for expressing the tight relationship between the components of the NP: agreement between head and modifiers (e.g. gender, number and case), word order restrictions (often greater than in the clause), compatibility restrictions and syncretisms (e.g. in the article), and marks of strict adjacency (linkers). In the languages we studdy, the order between head and modifier requires special attention when the latter is a noun. The occurrence of AB and BA orders does not necessarily involve indiference to word order even when one seems to describe an entity and the other a property (male chimpanzee / chimpanzee male). Allegedly, a clear lack of cohesion inside the NP is another characteristic of non-configurational languages, although one cannot discard the possibility that discontinuous NPs are mere co-referential chains in which modifiers become autonomous NPs through some sort of nominalization mechanism (common in Australia). (Something different is extraposition of relatives as in certain European languages.) The three facets of the NP as looked upon from a functional perspective are: semantic (characterizing the entity), extensional (identifying the referent) and, although less obvious, pragmatic (speaker´s strategies). The expression of the first facet is in charge of adjectives (again: where they exist) (discours frivole) and nouns (papier toilette), with a clear tendency of sequences of nouns towards lexicalization. Two aspects of possession are important: the highly diversified meaning of the genitive, and the semantic correlate of the valency of the head: alienable / inalienable possession with its boundaries varying from one language to the other. Quantification is a complex matter since it comprises, together with number proper (singular/dual/trial/paucal/ plural/total) and the distributive, the subclassification of nouns according to the quantificational properties of the denoted entity (individual, generic, collective, mass, count), as well as the gramatical tools required to switch from one subclass to the other (e.g. the singulative). Cardinal and ordinal numerals raise the question of their status inside the NP, whether modifiers or heads, specially when co-occurring with classifiers. To this topic also belong intensifiers (augmentative, diminutive) and the quantification of modifiers (la más profunda emoción). The categories of gender and class, rather similar in meaning, deserve different denominations (gender/noun clases/noun classifiers) according to their coverage of the lexicon and form of expression. Their mutual (in)compatibility in a given language has to be taken into account, as well as the basically parasitic nature of classification: speakers primarily classify for other grammatical goals, such as reinforcing the internal cohesion of the NP (agreement), tracking reference (ídem), increasing the valency of monadic nouns (“genitive classifiers”) and quantifying (“numeral” classifiers in singularization/discretization function). Extension or referentiality makes use of lexical modifiers (some adjectives, location adverbs, definite genitives — silueta de mujer / silueta de Juana) and, mainly, determiners. Languages differ with respect to the domain of application of the definite article: whether it is exclusively discourse related (foricity) or situational as well. Demonstrative inventories vary according to the deictic dimensions of person and space (sometimes time). The internal structure of the NP is not subject to much pragmatic pressure (what explains its more rigid order), without being absolutely closed to it. Besides the notion of (in)definite (speaker’s computing on hearer´s state of mind), there is room for expressing perspective (Anne arrived at 3 a.m. / Your daughter arrived at 3 a.m.), evidentiality and the modality that the former often conveys (se invitó al dizque gran profesor), the discourse hierarchization of two (or even three) third persons (proximal/obviative), and, naturally, the notional set formed by honorifics and hypocoristics. Based on the proposed topics, the participants can choose one interesting issue about the syntax of the noun phrase in the language(s) under study in order to prepare their presentations.


Francesc Queixalós

Keynote speakers:

Masayoshi Shibatani (Rice University)
Lisa Matthewson (University of British Columbia)
Henry Davis (University of British Columbia)