A Corpus Analysis of English-French Neighbour Cognates in L1 French EFL Learner Writing

Leveraging cognates has been considered an effective way to teach vocabulary in two languages that are historically and linguistically related, as the similarity in form can help learners with vocabulary retention. However, neighbour cognates, denoting cognates with identical meanings but subtle differences in spelling, may cause negative transfer, as the deceptive similarities in spelling might, at times, impact the accuracy of students’ written output. This study aims to shed light on patterns of incorrect spelling forms so as to help learners avoid negative transfer caused by neighbour cognates. Based on a relatively large amount of data from the Longman Learners Corpus (LLC), this study examined and identified patterns of negative transfer resulting from neighbour cognates across all proficiency levels. Following a two-step screening process of 50 randomly chosen neighbour cognates, 26 were included for the final analysis. Four types of negative transfer were identified, including substitution, blends, deviation, and mixed, among which, Type 1 and Type 2 errors have been documented in Odlin (1989) and Ringbom (1987). However, deviation and mixed errors seem not to be well noticed. Besides, Type I error, which involves straightforward substitutions, was found to be the most common in learners’ written production, accounting for 77 per cent of the total errors. Therefore, intentional teaching of neighbour cognates, and designing teaching and learning materials as well as vocabulary tests should be encouraged so as to help students notice and avoid these types of errors. Keywords: English-French cognates, lexical transfer, learners’ written errors, learner corpus DOI : 10.7176/JEP/15-8-01 Publication date : July 30th 2024