Dans les dictionnaires usuels du français le mot ‘boîte’ est défini à partir d'une valeur de base correspondant à un prototype matéeriel (ou perceptu). Les très nombreux composées (introduits de manière alternante par les prépositions A ou de) qui ne se conferment pas à ce prototype (par exemple, ‘boîte aux lettres’, ‘boîte d';essieu’, ‘boîte de nuit’) sont décrits sous les rubriques ‘fourre-tout’ de ‘spécialité’, ‘analogie’, parfois aussi ‘métaphore’. A ce mode de présentation, une approche non dictionnairique, mais de sémantique lexicale, suggère d'opposer le principe d'un stéréotype ‘intensionnel’ (ou concept) formulé en termes exclusivement fonctionnels (contenir X. et produire/fournir Y.). Mettant a priori les (quasi-) innombrables usages du mot sur un pied d'égalité, cette description permet de rendre justice à cette variété et à cette productivité grâce à une mécanique flexible d'ajustement entre l'instruction stéréotypique générique et les contraintes matérielles spécifiques.

This study provides an acoustic profile of the prosody of right-dislocations in French, using the computer hardware and software package known as CECIL. Twenty-eight right-dislocations occurring in a corpus of natural French discourse provide the data for the acoustic analysis. While right-dislocations appear to fulfil various functional roles in discourse, no correlation appears between functional type and the acoustic properties studied (pause, frequency, amplitude).

Nous étudierons l'hypothèse selon laquelle le E caduc précédé d'une seule consonne — (V) C_C — serait en train de se stabiliser dans le français non-méridional, surtout en syllabe initiale de mot, mais aussi en monosyllabe et en syllabe médiane. A l'aide de deux corpus de français parlé parisien, rassemblés en 1972–1974 et en 1989, nous montrons d'abord que cette stabilisation ne peut pas être confirmée dans la langue parlée de Parisiens cultivés. Le E caduc étudié semble être une variable sociolinguistique stable, variant avec le style et avec 1'âge des locuteurs. La deuxième partie de notre analyse fait apparaître en outre un conditionnement lexical important en syllabe initiale (0% de maintien dans semaine, 100% dans relation). Nous examinerons le rôle joué par la fréquence du mot, par la présence du préfixe re-, et par la relation articulatoire entre les consonnes qui entourent le E caduc.

Les jeux de langage, en tant que manifestations de la langue, peuvent constructivement participer à la réflexion linguistique. Ce sont non seulement des instruments de validation d'hypothèses émises sur la langue source, mais aussi des outils permettant de mettre à jour des phénomènes phonologiques non visibles en dehors des manipulations qu'ils occasionnent.

Ainsi, l'étude phonologique que nous avons menée sur le verlan nous a conduits à appuyer certaines hypothèses phonologiques sur le français, concernant en particulier la constituance syllabique, et à attirer l'attention sur d'autres aspects plus originaux.

The passé surcomposé régional is widely recognised as a feature of the régional French of the langue d'oc andfranco-provençal areas. It is also generally regarded as a highly stigmatised form in relation to the prescriptive norm. Less clear, however, is its précise function in the tense System of speakers who use it. In this paper, we investigate the temporal, aspectual and expressive properties of the régional double compound, taking into account semantic/pragmatic, sociolinguistic and syntactic factors. Our analysis suggests that in addition to the temporal and aspectual properties traditionally attributed to this form, the expressive properties of the surcomposé régional are central to any description of its function in contemporary spoken French, and fundamental to the opposition between the passé surcomposé and the passé composé.

This paper examines the position of negative adverbs such as mie, pas, point and jamais in Middle and Classical French infinitival clauses. Instead of linking the movement of the infinitival verb to the strength of functional categories such as AGRℴ, I propose to link it to a parametric change of NEGP from strong to weak. Up to Classical French, the infinitival verb can move to AGRℴ because NEGP is strong; this movement of the infinitival verb to AGRℴ allows the movement of negative adverbs, which are base-generated in VP initial position. At the beginning of Classical French, a parametric change affected the strength of NEGP, from strong to weak. As a result of this parametric change, movement of the infinitival verb to AGRℴ becomes more limited. Moreover, the difference between pas and other negative adverbs is due to a change in the nature of pas, from an adverb base-generated in VP initial position which can move, to a fixed adverb base-generated in the specifier position of NEGP.

This is a personal delineation of part of a methodology for the History of the French Language, aiming to combine the methodology of linguistics with that of history proper. Both traditional and modern methods of ‘historical linguistics’ fail to take account of a real time dimension, whereas ‘language history’ often resembles institutional, cultural and social history. We ask how we identify the ‘event’ and the ‘object’ of linguistic history, and how we distinguish variation, innovation, shift and change. We ask also what the linguist can contribute to the historian's reconstruction of the past.

The article argues that, contrary to a widespread view (e.g. Haiman, 1985; Palmer, 1984), agreement in those languages which exhibit it is not a purely redundant, semantically empty and grammatically predictable phenomenon, but performs several important functions at the level of discourse.

Taking French as the example language, I will argue (section 2.1) that agreement signals the function-argument interpretation to be assigned to pairs of expressions of various kinds; and second, that it may also code anaphorically the high-focus status of particular discourse referents (section 2.2). Section 3 compares certain written errors in agreement marking made by advanced learners of French, with certain other interpretative errors in their reading of French articles - errors based on agreement relations and leading to the mis-assignment of reference to an agreement target or personal pronoun. Finally, section 4 argues that third person personal pronouns should be treated differently from the (essentially predicative) agreement targets discussed in sections 2 and 3, claiming that they do not participate in agreement stricto sensu.

The article examines the variable distribution of word-final consonant devoicing (=WFCD) among working-class speakers in the Roubaix district, with respect to phonological conditioning and speaker characteristics. WFCD is shown to affect coronals, labials and velars in that order, and to be favoured by pre-pausal position. Among speakers over forty-five WFCD is primarily associated with female speakers, and to a lesser degree with male speakers under thirty. This sociolinguistically unusual distribution of a strongly vernacular variant may plausibly be attributed to language/dialect contact consequent on the immigration of Flemish-speaking textile workers. Such language contact would have tended to reinforce an already existing characteristic of Picard patois rather than introducing a totally new feature as the brief review of other Picard varie-ties would suggest.