État présent: The Syntax of Oral French

The Syntax of Oral French

1. Introduction

Until the 1970s and 1980s, oral French was unquestionably perceived as the 'poor relation' in linguistic terms, in the sense that it is frequently unfavourably compared to its written counterpart by grammarians and linguists alike.¹ From eighteenth-century commentators through until relatively recently, it is not uncommon to find oral French equated with low-register, regional, or grammatically incorrect French, with terms such as 'français fautif', 'français populaire', 'français familier' applied directly or indirectly. Moreover, there is considerable oversimplification, to which we shall return, around the oral/written opposition, that is, a tendency to create a sharp divide which empirical evidence does not necessarily support. As Blanche-Benveniste and Jeanjean comment, 'c'est un domaine où foisonnent les malentendus, les préjugés et les mythes'.² A number of factors have helped influence the growth of research on oral French. First, the emergence of sociolinguistics in the 1970s brought a preoccupation with obtaining access to the vernacular and collecting high-quality data, and as a result, a number of spoken corpora have been established internationally, although many more exist for English than for French. Subsequent research developments, notably in discourse analysis, variationist approaches and corpus linguistics, have all advanced dramatically both the quantity and quality of research being done on oral French. This état présent focuses on syntax; lexis, morphology, phonetics and phonology would merit separate exercises.

2. Theoretical approaches

Research on oral French has drawn from a wide variety of theoretical perspectives, of which some of the most influential include variationist approaches, the approche pronominale, conversational analysis, and different schools of discourse analysis/pragmatics, including speech act theory, information processing, politeness theory and énonciation.

Variationist approaches, where data is taken from corpora which are usually stratified according to pre-determined sociolinguistic parameters such as age, gender, socio-economic status, region, and so on, have brought two major advances. First, the methodologies employed in data collection and fieldwork techniques have been refined such that authentic oral data is now much more readily accessible. The second main advance concerns insight into the relationship between speaker or situational variables and linguistic variables. So, to take one example, studies of the non-use of negative ne suggest that age is an important factor, with broad patterns from research by Ashby suggesting that younger speakers have higher non-retention rates.³ This in turn has sparked a debate on whether such patterns indicate a linguistic change in progress, that is, a change in 'apparent time', with some scholars cautioning that non-use of ne might not in fact be increasing, but rather that it could possibly constitute a stable variable, younger speakers demonstrating higher rates than older ones. Ayres-Bennett, for instance, cites seventeenth-century examples from Héroard's journal of the Dauphin's speech, as well as exploring the other syntactic, phonetic and semantic factors which have a bearing on non-use of ne.⁴ Ashby's most recent research attempts to test the same sample of speakers at a real time interval (1976–95), and does indeed appear to show a more rapid increase in rates of non-use, with the age differences less marked, perhaps because the older speakers are now also conforming to the overall tendency towards non-use.⁵ Both speaker gender and register are also shown in variationist studies to be highly relevant: according to Ashby's data, men have 'caught up' with women in rates of ne deletion, where women had been ahead in demonstrating a change in progress (this correlates with variationist findings on other phenomena), while in broadcast journalism data, corpora also suggest increased deletion, albeit at a different pace.⁶ Meanwhile, variationist statistical evidence from Canada indicates that non-use is virtually categorical in Canadian French.⁷ Such variationist studies must take into account linguistic factors such as the syntactic, lexical or phonological context...