Lingvisticæ Investigationes - Volume 18, Issue 1, 1994

In this study, it is shown that the "category changing" property of morphological rules of conversion is unable to account for compound words, for formal and semantic reasons. Several convergent facts demonstrate that the compounding process is syntactic in nature. Consequently, it is argued that X-bar theory must be involved in compound word formation because of the "lexical function" of the syntax. Empirical data are mainly focused on French Noms Composés à base Verbale, or NCV, as tire-bouchon (cork screw),porte-parole (spoke person), gagne-pain (job), etc., which are analyzed as base generated "quasi-VPs" embedded in a NP. Thus, the NPWP exocentric dominance instantiates a "syntactic conversion" at the D-structure level. Such a categorial hierarchy is based on the "distribution changing" property of X-bar theory rather than on the "category changing" property of structuring morphological rules. Therefore, the high productivity of NCVs in French and other Romance languages is due to their morphology, which allows SPEC\HEAD agreement and VERB RAISING movement. The licensing of exocentric X-bar structures in grammar depends on several semantic principles of lexical interpretation, which are relevant to hyperonymy, hyponymy, meronymy, etc.. Thus, the model is dispensed with a superfluous component of "peripheral" rules of compounding. In conclusion, exocentricity of syntactic structures leads the author to claim that X-bar schema is primitive in grammar and that a given phrase is not the necessary projection of its head.

The Romance languages, with the exception of French, admit two ways of expressing indefiniteness in the plural, the bare plural (- N) or the plural introduced by a special determiner (Det N). We show that Det N is selected when reference is made to particulars, as opposed to mere categorization by naming the species, the latter being indicated by means of the bare plural. We are then left with the paradox that Det N is the form normally showing up in metaphorical expressions after the copula, a typically non referential position. We argue that this peculiar behaviour is due to the fact that Det N is basically a form of quantification whereas the property of referentiality is derived and contextually bound.

The notion of support verb enables us to construct a coherent lexicon of predicative nouns. It also enables us to describe their syntax and to provide a strong basis for their semantic description. This article gives a description of the construction of the support verb faire where the predicative noun relates to sport and musical activities. We also point out the privileged semantic and syntactic relation between the support verb subject and the predicative noun. The major property is the possibility for the determiner to commute with other determiners, whether or not accompanied by a modifier. This description enables us to divide the whole of these predicative nouns into five units. Each unit is homogeneous as far as its semantics and syntax are concerned.

In this paper the author presents his lexicon-grammar of frozen adverbs in Italian, in a general way. One of the important conclusions that emerges from the author's research concerns the existence of similar syntactical structures that are classified by Maurice Gross in his lexicon-grammar of frozen adverbs in French.The two group classifications are therefore similar: the same classes and almost the same syntactical-semantical properties and results. At this point a vast comparative work is justifiable.The author therefore wants to take particular account of the contrastive analyses effectuated on two frozen adverb classes. With the outcome of the corresponding situations, the author has been able to make a definate percentage of the semantic, lexical and syntactical equivalences and differences.

The goal of this paper is to give an answer to the following questions concerning the French de-existential construction as illustrated in (1): il y a une place de libre (often described as a 'gallicisme'): i) does the existential predicate govern a small clause (SC) of the form NP-de-AP? ii) which is the assigner of a 0-role to the NP: the AP within the construction or the existential predicate? iii) what Case-assigner licenses that NP? iv) what is the precise categorial status of the item de, and its precise role in the configuration?Our answer to those questions will be that: i) the NP-de AP string in (1) is an SC constituant functionning as an internal argument of the (inaccusative) existential predicate; ii) the AP is the 0-role assigner of the embedded NP; iii) the Case-assigner of that NP is de, which appears to be a (partitiive) quantifier; therefore de, and not the matrix predicate, is the Case licenser of the NP. De is the functional head of the SC, which has to be categorized as a QP. It is shown that the Case assignated by de to the Spec of the SC is Partitive, so that the 'gallicisme' supports a unification of the status of the SC subject and of the internal argument NP in the French existential construction.Other de-AP-constructions are then analysed as sharing basic licensing properties with (1).

Cet article expose les principaux problèmes que pose la syntaxe et la sémantique de la négation explétive en français contemporain. Il fait état des analyses récentes à cet égard, particulièrement celle de Muller (1991), en faisant ressortir leurs qualités et leur insuffisances.