This thematic issue of JFLS takes as its subject the connected linguistic phenomena of levelling and diversity in the French of the Hexagon. Diversity in language is obvious enough, is indeed part and parcel of all living languages, since social difference finds expression in linguistic difference. Levelling is, however, harder to define in a way that applies across languages and even more difficult to adapt across languages: the French term nivellement is spontaneously understood as ‘elimination of all differences’ i.e. ‘standardisation in the strong sense’, whereas the English term levelling has the sense rather of ‘standardisations partielles’ (the plural is important) and could perhaps be better translated in this way.

Using data from an obsolescent dialect situation in northern France, this paper questions the view that dedialectalization is a process of level-by-level attrition which leaves a linguistic residue in Regional French (the ‘Structured Obsolescence Hypothesis’). Comparison of dialect index scores for a number of variables reveals significant variation in rates of attrition within levels, with some phonological and morphological variants showing greater vitality than others, but no consistent relationship between levels as the model would predict. An alternative model is proposed, based on the relative learnability of different variants, and it is further argued that rejection of the Structured Obsolescence Hypothesis calls some other assumptions about Regional French into question, notably the view that it can be considered an intermediate variety between dialect and standard, and that it is necessarily ephemeral in nature.

This article is situated within the recent strand of SLA research which applies variationist sociolinguistic methods to the study of the acquisition of sociolinguistic variation by the L2 speaker. Whilst that research has tended to focus on the study of morphological and morphosyntactic variables, this article aims to investigate a number of acquisitional trends identified in previous research in relation to phonological variation, namely the variable deletion of /l/ by Irish advanced L2 speakers of French in both an instructed and study abroad environment. Based on quantitative results using GoldVarb 2001, the study further illuminates the difficulty that the acquisition of sociolinguistic variation poses to the instructed L2 speaker, who is found to make minimal use of informal sociolinguistic variants. In contrast, contact with native speakers in the native speech community is seen to allow the L2 speaker to make considerable sociolinguistic gains, not only in relation to the acquisition of the informal variant in itself, but also in relation to the underlying native speaker grammatical system as indicated by the constraint ordering at work behind use of the variable.

Notre objectif est de décrire l'évolution de la locution adverbiale quelque part de son emploi spatial standard à des emplois non spatiaux en vogue à l'heure actuelle, mais non encore analysés, comme ceux qu'exemplifient les séquences:

Moi, ça m'embête quelque part de dire à un enfant euh non tu parles pas comme ça. Quelque part, il y a eu un abandon de la structure éducative (propos d'un député UMP sur le rôle des parents et de l'école dans la prévention de la fugue et de l'école buissonnière)

Je sais que Brian n'est pas mon enfant, mais quelque part il l'est aussi (propos d'une mère dont la fille de 15 ans a eu un enfant et que la mère aide à élever, 27/01/03, émission TV)

(Quelque part, ça m'interpelle/m'attriste/me dérange/me chagrine, (relevé à l'oral)

Dans une première partie, nous analyserons le ‘sens’ spatial en mettant plus particulièrement en relief les deux propriétés qui le caractérisent, notamment le trait de ‘partition’ généralement omis dans les descriptions. Après avoir distingué les emplois non spatiaux des emplois spatiaux, nous essaierons dans notre deuxième partie:

1. de faire ressortir deux types d'emplois non spatiaux,

2. d'expliquer ces emplois, dans une perspective de grammaticalisation, à partir des deux traits sémantiques postulés pour les emplois spatiaux,

3. de décrire les effets de sens auxquels ils donnent lieu et qui expliquent en partie leur succès.

This article presents a synthesis of the sociolinguistic situation of what the authors refer to as the ‘contemporary dialects’ of French in the France of today. The introduction emphasises the methodological and conceptual problems attending any such definition and evaluation, attempting to clarify the complex situation and to identify the various kinds of ‘dialects’, ‘uses’ and ‘speakers’. We then concentrate on the regional ‘dialects’ of French in continental France, urban and rural, and summarise a series of important recent studies, concentrating on local variation. We also distinguish the sociolinguistic situations of the northern and the southern parts of the country. Even though France is known to be a highly centralised country, whose linguistic policy has been aiming at monolingualism for the past two centuries, the article offers some possibly surprising and nuanced results that show more variation than established opinion would generally admit about contemporary France.

Anglo-Norman (AN) showed a tendency to lose Old French conjugation and gender inflectional distinctions, but is thought to have largely maintained the syntax of Old French. This study considers whether in the early 14th century AN syntax continued to follow continental French (CF) by moving towards new word-order patterns, namely XSV order and subject-verb inversion after et, which were to typify Middle French. Using corpora of CF and AN historical writing, especially chronicles, it is found that AN to some extent shadowed developments found in later 13th and in 14th century CF. In both AN and CF, XSV order was widespread with time adjuncts, but avoided with place adjuncts and direct and indirect objects. This dissociation was not calqued on Old/Middle English subject-verb inversion, which showed a different dissociation, i.e. inversion of verb and nominal subjects, but not pronominal subjects; AN showed no influence of this contrast. Inversion after et was found in AN, but only with unaccusative verbs, whereas in CF by the late 13th century it was spreading to other verbs as well, having initially shown a similar limitation as in AN. It is concluded that underlying syntactic processes of change began to affect AN as well as CF, but that they were interrupted by the switch away from French in England in the later 14th century.

En dépit de leur nature et de leur fonctionnement syntaxique distincts, le pronom y en emploi spatial se trouve en concurrence, sur les plans sémantique et pragmatique, avec les deux adverbes là et là-bas. En termes d'accessibilité cognitive, il sera montré que les référents de y jouissent d'une plus grande accessibilité que ceux des deux adverbes. Les trois marqueurs ont tous des emplois anaphoriques. Si là et là-bas s'emploient également de manière déictique, tel n'est pas le cas pour y. Là-bas est le seul à pouvoir marquer la distance par rapport au locuteur. Quant à la progression thématique, les trois marqueurs peuvent exprimer le thème, mais y n'est jamais rhématique, contrairement aux deux autres. Enfin, y et là-bas sont plus aptes que là à marquer un processus spatial dynamique.

Dans cet article, nous examinons la structure en de quoi + infinitif qui peut être régie par un verbe:

(i) Il y a de quoi s'interroger!

ou ne pas l'être:

(ii) Jamais la France du très haut ne s'est aussi bien portée que sous Jospin. De quoi irriter l'ex-Premier ministre

Nous envisageons deux possibilités d'analyse, selon la valeur de de. Soit celui-ci est considéré comme un article partitif. Mais alors de quelle façon l'infinitif s'articulet-il à de quoi? Soit il constitue une préposition. On se rapproche alors d'énoncés tels que:

(iii) Il a de qui tenir

mais ici, de quoi prend une fonction inhabituelle, celle de circonstant, alors que dans un exemple tel que (iii), de qui remplit la fonction de complément essentiel indirect par rapport à l'infinitif. Dans un dernier temps, nous avons étudié les emplois en discours de cette structure, selon que la proforme quoi anaphorise ou non une structure prédicative de l'énoncé précédent et selon les infinitifs utilisés.

This article examines levelling and diversity in northern urban French pronunciation through the optic of folk (= non-linguists') perceptions of variation. These perceptions are investigated by the identification of authentic voice samples (rather than other instruments widely used in perceptual dialectology such as mental mapping): respondents from the Pays de la Loire region of north-western France heard extracts of scripted speech from Nancy and Rennes, and were asked to identify the speakers' regional background and say whether they were of urban or rural origin. The results of this test show that while some difference between the two speaker location groups was accurately perceived, the informants also formed some inaccurate judgements, partly based on social stereotypes. Overall there is some confirmation of accent levelling, and of general social psychological tendencies such as stereotyping, annexation and time-lag in perceptions of regional–social linguistic variation.

We know very little about the French vocabulary that is learned in school and this paper reports a study which measures learners' vocabulary size and progress in secondary school. The methodology for estimating vocabulary size in French is comparable with vocabulary size testing in other foreign languages, and this makes comparison with vocabulary learning in French and other languages possible. Results suggest that learners learn about 170 words per year up to GCSE and about 530 words per year in ‘A’ level study and are influenced by word frequency. On average, learners take GCSE with under 1000 words of French vocabulary and ‘A’ level with about 2000 words. These results appear modest compared with historical data and when compared with other language exams pitched at the same CEF levels as GCSE and ‘A’ level. Vocabulary size predicts ‘A’ level grade particularly impressively. There is a worrying period where progress, even of the best learners, appears to halt for several years.

This article presents a corpus analysis designed to determine the extent to which noun endings in French are reliable predictors of grammatical gender. A corpus of 9,961 nouns appearing in Le Robert Junior Illustré was analysed according to noun endings, which were operationalised as orthographic representations of rhymes, which consist of either a vowel sound (i.e., a nucleus) in the case of vocalic endings or a vowel-plus-consonant blend (i.e., a nucleus and a coda) in the case of consonantal endings. The analysis classified noun endings as reliably masculine, reliably feminine, or ambiguous, by considering as reliable predictors of grammatical gender any noun ending that predicts the gender of least 90 per cent of all nouns in the corpus with that ending. Results reveal that 81 per cent of all feminine nouns and 80 per cent of all masculine nouns in the corpus are rule governed, having endings that systematically predict their gender. These findings, at odds with traditional grammars, are discussed in terms of their pedagogical implications.

L'article propose de questionner la production discursive des normes (les discours tenus sur les façons de parler d'autrui ou de soi-même) sous l'angle de la multipolarisation des espaces de référence. Il rend ainsi dans un premier temps compte d'enquêtes épilinguistiques effectuées dans trois configurations urbaines distinctes: 1. le discours sur l'accent de Rouen, forme linguistique identitaire et en même temps stigmatisée, 2. le discours sur le cauchois d'Yvetot (le cauchois est une des variétés du normand), et 3. le discours sur les langues à Rennes (entre autres, autour de la langue française et des langues bretonnes). Dans un second temps, après avoir rappelé les propositions théoriques concernant la spatialité en sociolinguistique urbaine, il présente une modélisation inédite de l'espace urbanisé pour tenter de mieux saisir les spécificités des dynamiques discursives impliquées par la culture urbaine. Il commente, en regard avec le concept de nivellement, les conceptualisations qui sont à la base d'une telle approche, et notamment la centralité linguistique comme processus producteur de normes et son corollaire quasi dichotomique, le confinement linguistique, comme processus vecteur de contre-norme.

Using evidence from a variety of sources (dialectological and sociolinguistic studies, written and oral history and works of literature), this study seeks to describe how, in a period of rapid industrial expansion to which immigrant labour was a crucial contributing factor, large numbers of Belgian migrant workers (the majority of whom were Flemish-speaking) were assimilated into the local Romance-speaking community. In an area often characterised as diglossic (French-Picard), the influx of large numbers of Flemish speakers gave rise to a three-way language-contact situation. While charting some of the most important changes in the vernaculars of Lille, the study seeks to explain why an alloctonous group of such significant proportions living so close to their homeland apparently assimilated so readily.

Notre article porte sur le français de jeunes habitants de quartiers pluri-ethniques et multilingues de grandes villes françaises. Il présente la poursuite d'une approche exploratoire amorcée en 2005 par Jamin et Trimaille consistant à la base en un rapprochement d'études sociolinguistiques réalisées à Grenoble (Trimaille, 2003a) et en région parisienne (La Courneuve et Fontenay-sous-Bois, Jamin, 2005), mais qui s'élargira ici à Marseille (Gasquet-Cyrus, 2004). Nous proposerons que dans un contexte de relative uniformité du français urbain et de tendance convergente du ‘sud linguistique’ vers les normes plus neutres du français d'oïl (pour une variable au moins, la rétention/effacement du schwa), ce nivellement ne se fait pas seulement dans la direction du français décrit comme standard. Des formes non standard similaires sont en effet attestées dans plusieurs centres urbains géographiquement assez distants. Par un ensemble de données quantitatives et qualitatives issues des trois études mentionnées supra, nous proposons d'interpréter la présence dans les trois sites de recherche d'une de ces formes, la palatalisation/affrication des occlusives dentales et vélaires, comme l'indice d'un certain degré de ‘convergence dans la divergence’ par rapport au français standard. Nous nous intéresserons ici aux facteurs de diffusion de cette variante, et plus particulièrement aux rôles respectifs de la variation stylistique et des représentations sociolinguistiques.

Cet article discute la dichotomie cas structural vs non structural par rapport au datif français. Plusieurs faits semblent plaider en faveur du caractère non structural de ce datif, comme le marquage des datifs nominaux par une préposition (cf. cas oblique) et l'existence de datifs non argumentaux (cf. cas sémantique). Néanmoins l'examen des différents types de datif dont dispose le français montre que l'apparition de ce datif est soumise en premier lieu à des facteurs configurationnels. Dans la grande majorité des cas la présence du datif n'est en effet pas possible sans celle d'un autre argument interne dans le VP, qui a un rôle thématique inférieur à celui du datif.

This article seeks to assess in detail the extent of levelling of regional varieties in the so-called collateral-language areas (Langue d'Oïl and Franco-Provençal) of France over the course of the 20th century. By comparing the ‘classic’ accounts of regionally marked pronunciation among speakers born in the first half of the century (Martinet, 1945; Walter, 1982; Carton, Rossi, Autesserre and Léon, 1983) with studies based on informants born since 1965, I seek to characterise and remap the varieties that still show divergence from the supra-local or levelled variety, which has been referred to as Oïl French, in the ancestral Langue d'Oïl and Franco-Provençal regions and to define the areas beyond these ancestral collateral-language areas, where speakers of 40 years or under at the time of writing may be fairly characterised as conforming to the supralocal norm.