The development of phraseological complexity across oral and written modes in L2 French

Vandeweerd, Nathan [UCL] Alex Housen [Vrije Universiteit Brussel] Paquot, Magali [UCL]

Recent research has begun to provide evidence for the important role of phraseological complexity in raters’ perceptions of advanced L2 writing proficiency. In particular, Paquot (2018, 2019) showed that learner texts that were rated as more proficient were associated with higher phraseological sophistication, operationalized as the mean pointwise mutual information (PMI) for adjectival modifier and direct object and syntactic dependencies. Similar results have also been found for L2 French writing (Vandeweerd, Housen, & Paquot, 2021). Until now however, research in this domain has focused mostly on the written mode with little scant research into the development of phraseological complexity in the oral mode. In comparison to writing, speech allows for less online planning which may result in less complexity– the so-called Trade-off Hypothesis (Skehan, 1998). This planning time may facilitate the selection of rarer phraseological units, as seems to be the case with lexical complexity (Bulté & Housen, 2009). In addition, the functional differences between oral and written tasks may have an effect on the types of phraseological units that are present in the two modes. While cross-modal comparisons of phraseology are still rare, there is some evidence that oral registers make more use of verb-based phraseological units whereas literate registers make more use of noun-based units (Biber & Gray, 2013). The main aim of the current study is to compare the development of phraseological complexity across three tasks: written argumentative essays, oral narratives and oral interviews to determine the extent to which phraseological complexity develops differently across modes. The data for this study come from the LANGSNAP corpus (Mitchell, Tracy-Ventura, & McManus, 2017), a longitudinal corpus that traces the linguistic development of students during a 9-month stay abroad. Adjectival modifiers (adjective + noun) and direct objects (verb + direct object) are extracted using in-house R scripts (R Core Team, 2021) and PMI is calculated on the basis of the 10-billion word FRCOW16 corpus (Schäfer, 2015; Schäfer & Bildhauer, 2012). The results showed that for the learners of French in the study, written tasks were generally more phraseologically complex than spoken tasks. Learner productions also became slightly more phraseologically sophisticated (but not diverse) over time but developmental trends differed across tasks, which highlights the importance of considering task variables when measuring L2 phraseological complexity.