The presence, nature and role of formulaic sequences in English advanced learners of French : a longitudinal study

Abstract

The present study is a longitudinal investigation of the presence, nature, and role of formulaic sequences (FS) in advanced English learners of French. The learners investigated are in their second year of an undergraduate degree in French at the onset of the study, and are tested before and after a seven-month stay in France. FS are defined psycholinguistically as multiword units which present a processing advantage for a given speaker, either because they are stored whole in his/her mental lexicon (Wray 2002) or because they are highly automatised. The construct of FS is particularly relevant to investigate key linguistic issues such as the dynamism of linguistic representations, their idiosyncratic nature as well as the relationship between the lexicon and grammar. FS have been shown to be frequent in the oral productions of native speakers. They also play an important role in first language acquisition as well as in the initial stages of instructed second language (L2) acquisition. However, very little is known about their presence and role in advanced L2 learners, as most studies dealing with them have not adopted a psycholinguistic approach and have focused on L2 learners’ knowledge and use of idioms and idiomatic expressions. Conversely, this study seeks to evaluate and characterise the presence of psycholinguistically-defined FS in advanced learners as well as examine their longitudinal development in relation to the development of the learners’ fluency and lexical diversity. It seeks to determine whether FS use can be said to play a role in the development of fluency and lexical diversity and if it does, describe the underlying mechanisms that account for this role. Data from five learners performing five oral tasks (an interview, a story retell and 3 discussion tasks), repeated before and after their stay in France, was elicited and transcribed. FS were identified through the hierarchical application of a range of criteria aiming to capture the holistic nature of the sequences. The necessary criterion used for identification was fluent pronunciation of the sequence, and additional criteria were applied such as irregularity, holistic mapping of form to meaning or holistic status of the sequence in the input. Fluency was operationalised through 4 measures (phonation-time ratio, speaking rate, mean length of runs and articulation rate) and lexical diversity was measured using D. The results show that psycholinguistically-defined FS represent about 27% of the language of advanced learners of French. The typology of the identified sequences shows that they are mostly grammatically regular but that despite the advanced level of the participants, some present non-nativelike characteristics. Individual differences in the learners’ repertoires of FS as well as task effects are also found. Between time 1 and time 2, across the group of 5 subjects, there is a general and statistically significant increase in FS use, fluency and lexical diversity. Significant correlations are found between FS use, fluency and lexical diversity. The qualitative analysis suggests that FS use plays a role in increasing fluency by allowing longer speech runs, contributing to the reduction of pausing time as well as the speeding up of the articulation rate. At the internal level of processing mechanisms, the results suggest that FS play a facilitating role not only in the formulation stage of speech production but also in the conceptualisation and articulation stages. Significant correlations are also found between FS use and lexical diversity, which suggests that FS, by lightening the processing burden and freeing some attentional resources, might facilitate the acquisition of new vocabulary. The analysis of the development of the learners across all variables shows a single developmental path with similar processes of automatisation but with different rates of acquisition, as the learners vary in how efficient they are at proceduralising their language. Because of this, it is suggested that the year abroad is more likely to be beneficial for a given subject if their language has already reached a certain level of automatisation pre-time abroad.