This article is a quantitative study of the variable deletion of post-obstruent /l/ and /R/ in word-final obstruent-liquid clusters (OLC) in French (capable [kapab(l)], cidre [sid(ʁ)]). The analysis of over a thousand tokens extracted from a corpus of interviews gathered in Nancy and Rennes shows that the reduction of word-final OLCs is a stable sociolinguistic marker in northern, standardised metropolitan French. Patterns of stylistic and social stratification in age, gender, and social class and interaction with following phonological context are attested, but OLC reduction does not appear to be an ongoing change. It is argued that the data provide further evidence of variable morpheme-final consonant clusters as a ‘primitive’ feature of vernacular dialects.

L'article présente une recherche portant sur le fonctionnement de l'unité étrangement dans le discours, à partir des notions d'aspect et de stratégie discursive. L'approche adoptée vise à rendre compte du lien entre forme et sens par le biais de critères linguistiques et d'outils théoriques issus de la théorie de la polyphonie. Nous posons comme hypothèse que l'adverbe étrangement opère une sélection verbale sur la base de critères sémantiques aspectuels, et qui transparaît dans les propriétés distributionnelles des emplois. Ce faisant, il conditionne dans le même temps la structure polyphonique de l'énoncé dans lequel il apparaît.

This article explores perceptions of appropriateness in the French second-person pronoun system in relation to various sociopragmatic factors among native and nonnative speakers of French between the ages of 18 and 29. Participants completed an online survey in which a series of five social-interactive situations were presented. Analysis focused on the similarities and differences among native, near-native, advanced and intermediate speakers in their perceptions of formality, social distance, power/equality and appropriate tu/vous use, as well as correlations between the sociopragmatic factors and selection of tu or vous. Results indicate a high degree of variation within and across participant groups, with nonnatives tending to be more conservative (e.g., more formality, higher frequencies of vous) in their choices than their native-speaker counterparts. In concluding, the findings are discussed in relation to pedagogical implications.

Dans cet article, nous analysons l'influence de la valeur aspectuo-temporelle attribuée aux formes composées sur la fréquence d’accord du participe passé (APP). On sait que l'APP est variable dans les usages de la langue parlée, notamment lorsqu’il est employé avec l'auxiliaire avoir. Partant de l'observation de corpus de langue parlée, et d’une approche variationniste, nous évaluons la fréquence d’accord des participes féminins potentiellement audibles, en fonction de l'effet de sens qui est attribué à la séquence [avoir (pres) + PP], selon qu’il est évalué comme ‘résultant’ ou ‘événementiel’ selon la terminologie de Desclés et Guentcheva (2003). Les résultats de cette analyse laissent entrevoir une influence faible mais relativement constante de ce facteur sémantique sur l'APP, et soulèvent plus avant la question de la présence d’éléments spécifiques au médium oral comme facteurs d’influence sur le marquage morphologique du genre; en d’autres termes, de l'existence d’une grammaire spontanée de l'oral (Branca-Rosoff, 2005).

This study seeks to assess the competence in, and current perceptions of, the traditional regional ancestral variety, Picard, among school students in Lille. In the first part, dialect retention is evaluated by a series of language tests, administered in two rounds of fieldwork, the first carried out between 1995–1999 covering the whole of Lille-Métropole and the second undertaken in 2004–2005 focusing on the urban heartlands. The results show that competence in Picard, as measured by the tests, is highest among European or Metropolitan French subjects in the north-eastern part of the urban area (Roubaix-Tourcoing) despite significantly greater ethnic diversity among the school populations.

In the second part of the study, the subjects’ perceptions of a range of Picard and Regional French varieties, as represented by a series of sample recordings, covering most of the picardophone areas are evaluated against a linguistic analysis of ‘picardité’ as measured by the density of Picard features in the extracts concerned. The results show a lack of correspondence between the two evaluations. Subjects’ perceptions are further clarified by their metalinguistic comments regarding the value of Picard and its place in their linguistic repertoire. As this variety is rarely if ever used in peer-to-peer exchanges, relative competence does not correlate with indicators of socialisation that proved significant with regard to other aspects of the students’ linguistic repertoire and was also found to correlate negatively with measures of regional loyalty and cultural value.

The ‘real-time’ aspect of the study appears to point to some degree of ongoing erosion of awareness and confidence in recognition of the varieties concerned among the population investigated, despite some important cultural events which might have been expected to counter this trend.

This article investigates the recently developed adjectival properties of the French noun clé ‘key’, as attested in for instance un poste très clé ‘a really key position’ and Cette mesure est-elle vraiment clé? ‘Is this measure really key?’. The main purpose of this study is triple: it consists in analysing (i) which adjectival uses can be found in modern French, (ii) to what extent they are accepted by native speakers (from different geographical varieties) of French, and (iii) how they can be accounted for within the framework of Construction Grammar and Construction Morphology. It will be hypothesised that French clé is subject to categorial gradience as a consequence of an ongoing constructionalisation process.