This study investigates variable positioning of adverbs in compound verb tenses in vernacular Quebec French using the sociolinguistic framework of Variation Theory (Weinreich et al. 1968; Labov 1969). While variable adverb placement is addressed in both the prescriptive and linguistic literature, whether their explanations for it hold in practice remains to be determined; quantitative research of this phenomenon in usage-based corpora is limited, and rare in French. The research objectives are therefore to determine the productivity of variable adverb placement in French in these verbal contexts, to uncover the linguistic and/or social factors which constrain it, and to evaluate the extent to which current treatments of this variable in the literature accurately reflect what occurs in speech. Data is thus extracted from a corpus of spontaneous discourse, is coded for several linguistic and social factors, and is quantitatively analysed using standard variationist methodology (Poplack & Tagliamonte 2001). Overall rates of variant use suggest that variable adverb placement is robust, with adverbs occurring just slightly more frequently after the past participle than between the auxiliary and the participle; placement at the beginning of the sentence is rare. The results of the distributional and multivariate analyses largely confirm the purported conditioning effects of the tested linguistic factors, suggesting that prescriptive and theoretical linguistic approaches are generally correct in their accounts of this phenomenon. However, closer investigation reveals these effects to be sensitive to the lexical identity of the adverb, namely, their particular placement preferences; once these positioning predilections are taken into consideration, the conditioning effects of the linguistic factors essentially disappear. Sociodemographic factors are also found to be mildly implicated in variable adverb placement, and these too are sensitive to the influence of the lexical identity of the adverb. Ultimately, it is argued that this variable is primarily lexically-constrained, a finding which can be only minimally and indirectly inferred from the relevant literature. Taken together, the results of this study provide new and vital insight into the mechanisms underlying variable adverb placement in French, and also highlight the importance of quantitatively investigating such variable language phenomena in corpora of vernacular speech.

This study investigates variable positioning of adverbs in compound verb tenses in vernacular Quebec French using the sociolinguistic framework of Variation Theory (Weinreich et al. 1968; Labov 1969). While variable adverb placement is addressed in both the prescriptive and linguistic literature, whether their explanations for it hold in practice remains to be determined; quantitative research of this phenomenon in usage-based corpora is limited, and rare in French. The research objectives are therefore to determine the productivity of variable adverb placement in French in these verbal contexts, to uncover the linguistic and/or social factors which constrain it, and to evaluate the extent to which current treatments of this variable in the literature accurately reflect what occurs in speech. Data is thus extracted from a corpus of spontaneous discourse, is coded for several linguistic and social factors, and is quantitatively analysed using standard variationist methodology (Poplack & Tagliamonte 2001). Overall rates of variant use suggest that variable adverb placement is robust, with adverbs occurring just slightly more frequently after the past participle than between the auxiliary and the participle; placement at the beginning of the sentence is rare. The results of the distributional and multivariate analyses largely confirm the purported conditioning effects of the tested linguistic factors, suggesting that prescriptive and theoretical linguistic approaches are generally correct in their accounts of this phenomenon. However, closer investigation reveals these effects to be sensitive to the lexical identity of the adverb, namely, their particular placement preferences; once these positioning predilections are taken into consideration, the conditioning effects of the linguistic factors essentially disappear. Sociodemographic factors are also found to be mildly implicated in variable adverb placement, and these too are sensitive to the influence of the lexical identity of the adverb. Ultimately, it is argued that this variable is primarily lexically-constrained, a finding which can be only minimally and indirectly inferred from the relevant literature. Taken together, the results of this study provide new and vital insight into the mechanisms underlying variable adverb placement in French, and also highlight the importance of quantitatively investigating such variable language phenomena in corpora of vernacular speech.