Search

A given predicate is defined by a set of properties which combine and which automatically generate all the sentences it allows. Among them, we note the number and the semantic class of the arguments which characterize it, the adjectival and adverbial modifiers which can be added tothe scheme of arguments as well as all the transformations which affect each of these units. The speaker is responsible for attributing to sentences the set of all the forms that language allows him to generate. What has just been said can be considered as a definition of syntax.But this situation is far from exhausting the description of a language. J. Dubois and especially Maurice Gross have devoted large-scale work to fixed expressions, that is to say, to the restrictions relating to the combinatorics usually observed around a given predicate. These studies have focused on the limitations of grammar rules as they are generally described. These two authors have drawn up lists of tens of thousands of “fixed” verbs and have highlighted the limits of this fixing. However, they made an observation without highlighting the causes of the fixing, which is a much more complex linguistic fact than this work suggests. The purpose of this article is twofold. On the one hand, it emphasizes what can be called discursive equivalences: in a given situation, the same idea can be translated by expressions which have no obvious link between them, as in: con comme la lune, con comme un balai, con comme une baleine, con comme une bite, con comme une valise. Another example: voici belle lurette, voici longtemps, voici un temps fou, voici une paille, voici une paye. It goes without saying that the speaker is not master of these expressions, because they are written in the language. This article shows that these equivalences are very numerous. On the other hand, I. Mel’čuk initiated important work on pragmatemes. Again the “regular” syntax is defective. All these cases are in fact examples of pre-constructed sequences, of which this article attempts to make a first classification. These sequences are explained by specific communication conditions as seen with these examples:a) Doubt or reluctance in the face of information that one can hardly believe:à d’autres !, à d’autres mais pas à moi !, à d’autres mais pas à nous ! b) Criticism of a work that is considered null and uninteresting:c’est de la bouillie pour les chats, c’est de la bricole, c’est de la briquette, c’est de la couille,c’est de la merde, c’est de la piquette, c’est du flan, c’est du pipeau, c’est du vent.This is long-term work, which allows us to renew certain theoretical perspectives.

This study aims at accounting for one of the properties of circumstantial complements which has been unjustly neglected in linguistic studies, i.e. their capacity to be thematized as subjects of a main verb. This reconfi guration mobilizes a certain number of appropriate verbs which cannot be described without reference to this phenomenon. The article then focuses on complements of place, time, cause, goal, condition, concession as well as on certain qualitative constructions.

Dictionaries of synonyms used in word processors can be shown to be inadequate. To remedy this, the analysis of synonymy should be based on the concept of usage and to this end a description grid is proposed; this grid is made up of a dozen parameters, the main ones being those where the argument domains are defined (syntactic features and object classes); in cases of polysemy, each usage is associated to the synonyms which are peculiar to it. This description does not simply make it possible to improve the synonym function of word processors, it also opens the way for automatic processing of synonymy., Gross Gaston. Pour une véritable fonction « synonymie » dans un traitement de texte. In: Langages, 32ᵉ année, n°131, 1998. Les classes d'objets, sous la direction de Denis Le Pesant et Michel Mathieu-Colas. pp. 103-114.

G. Gross : Classes d'objet et description des verbes This article aims at clarifying the concept of « object classes » in natural languages and its usefulness in describing verbal predicates. Syntactic features such as human, concrete, abstract are not accurate enough to account for the polysemy of the operators. Object classes can be put to other uses, some of which are outlined here., Gross Gaston. Classes d'objets et description des verbes. In: Langages, 28ᵉ année, n°115, 1994. Sélection et sémantique. Classes d'objets, compléments appropriés, compléments analysables, sous la direction de Jacqueline Giry-Schneider. pp. 15-30.

It has been an empirical fact that almost all the words are polysemous. A standard dictionary such as the Petit Robert lists 60,000 entries which correspond to 300,000 meanings. Thus, in this particular dictionary one word is paired with five different senses on average. Moreover, what is being dealt with here is no more than a general reference work, designed for a daily use. It contains only most frequent and general items, but disregards all other available meanings. In what follows, contextual properties will be demonstrated to be able to set apart each instance of polysemy, thus offering an effectual tool likely to do away with ambiguities.

The goal of this article is to propose a model of semantic description of connectives (here of causal connectives). For a given connective, we proceed first by giving factorization of the morphological forms, thus avoiding redundancy in the description and putting forward equivalencies. A connective is considered to be a predicate, defined by the semantic nature of its arguments which are represented by the predicates of the main and subordinate clauses. We are thus in a position to enumerate the connectives that have the same arguments from a semantic point of view, Gross Gaston. Sémantique lexicale et connecteurs. In: Langages, 33ᵉ année, n°136, 1999. Sémantique lexicale et grammaticale, sous la direction de Yvette Yannick Mathieu. pp. 76-84.