The role of metalinguistic awareness and of L2 proficiency in positive lexical transfer from English (L2) to German (L3) by French-speaking Quebeckers

Authors: Woll, Nina
Advisor: Hummel, Kirsten M.De Koninck, Zita
Abstract: It has been posited that experienced language learners develop higher levels of metalinguistic awareness (MLA), which, in turn, appears to facilitate the acquisition of subsequent languages (e.g., Singleton & Aronin, 2007). Moreover, researchers in the field of Third Language Acquisition (TLA) emphasize the positive influences of previously acquired languages in foreign language classrooms (e.g., Cenoz & Gorter, 2015), in contrast to the traditional focus on interference-based learner errors in the study of language transfer. Typological similarity and source language proficiency are known to influence transfer processes of any kind (e.g., Ringbom, 1987, 2007). However, it remains a methodological challenge both to identify the felicitous use of a target language feature as an effect of cross-linguistic influence (e.g., Falk & Bardel, 2010) and to clearly determine the role of MLA for the conscious activation of related words or constructions across languages. The present study aimed at meeting this double challenge by using think-aloud protocols (TAPs) to investigate positive lexical transfer from English (L2) to German (L3) by French-speaking Quebeckers (N = 66) after five weeks of formal L3 instruction. Participants completed a translation task consisting of 42 items selected on the basis of similarity and imageability ratings as well as word frequencies (Friel & Kennison, 2001). Participants were asked to think aloud while attempting to translate these largely unknown words from German (L3) into French (L1). Positive transfer was operationalized by correct translations that were related to an English cognate. Levels of MLA were measured by means of the THAM (Test d'habiletés métalinguistiques) (Pinto & El Euch, 2015) and complemented by the analysis of the TAPs. Proficiency levels in English were established by means of the Michigan Test (Corrigan et al., 1979), whereas exposure to the different languages as well as interest in the German language and culture were addressed in a background questionnaire. A fine-grained analysis of the TAPs revealed inter- and intra-individual variability in the conscious activation of related L2 vocabulary while allowing for an identification of distinct levels of awareness. Two independent models of logistic regressions revealed that both dimensions of MLA significantly predicted positive transfer. The first model, which featured the THAM as the exclusive measure of MLA, identified this reflexive dimension as the strongest predictor, followed by English language proficiency, whereas none of the other independent variables predicted positive transfer from English. In the second model, which included both the THAM and the TAPs as complementary measures of MLA, the applied dimension of MLA, as measured by the TAPs, was by far the strongest predictor, followed by the reflexive dimension, measured by the THAM, while English proficiency was no longer a significant predictor. Despite the fact that verbalization may have influenced overall performance to a certain extent, our observations point to the invaluable contribution of introspective data to complement findings based on language-inherent characteristics of crosslinguistic influence. Our analyses highlight the complexity of metalinguistic processes and individual strategies related to learner and context variables, in line with a dynamic view of multilingualism (e.g., Jessner, 2008a).
Document Type: Thèse de doctorat
Issue Date: 2016
Open Access Date: 24 April 2018
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
Description SizeFormat 
32604.pdfTexte26.01 MBAdobe PDFThumbnail
All documents in CorpusUL are protected by Copyright Act of Canada.