# Published by Blackwell Publishers Ltd. 2001 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden MA 02148, USA. Journal of Sociolinguistics 5/4, 2001: 453±474 Real-time changes in the paradigm of personal pronouns in Montreal French 1 He Âle Áne Blondeau University of Ottawa, Canada Based on the results of three variationist studies on personal pronouns used in Montreal French, this article shows how real-time data can shed light on `apparent time interpretation' and increase our understanding of morphosyn- tactic changes. Longitudinal data for a 24-year period from three corpora of spoken French are used to discuss cases of variation: variation between the clitics on and nous, alternation between on and tu/vous, and variation among non-clitic plural pronouns. While the ®rst case illustrates a change occurring over a long period of time, the other variables show changes observable in the 24-year time span under study. In particular, analysis of the variation among non-clitic pronouns suggests a socio-stylistic specialization of the variants at a certain stage in the grammaticalization of compound forms. In sum, this article uses longitudinal evidence to show how the variables are involved in a process of restructuring the paradigm of the French pronouns. KEYWORDS: Pronoun, morphosyntax, French, longitudinal, linguistic change 1. INTRODUCTION Some modi®cations in language use at the morphosyntactic level have been shown by many variationist studies on the paradigm of the personal pronouns used in Montreal French (Laberge 1977; Sanko 1981; Thibault 1991a). As this article illustrates, even though the researchers interested in variation within the pronominal system of Montreal French have not studied exactly the same variables across time, there is sucient interrelation between their studies to enable us to better understand the dynamic occurring in that area of morphosyntax in a real-time perspective. The studies cited here come from the same large body of data on Montreal French collected longitudinally between 1971 and 1995. In particular, the main real-time study cited here concerns the use of the compound forms of the non-clitic pronouns by a cohort of speakers interviewed at three dierent times in their life span: in 1971, 1984 and 1995. In this paper, the methodological distinction between studies in apparent time and real time is used (Labov 1972, 1994) to take into account the temporal