Formulaic vs. Non-Formulaic Constructions in Second Language Writing: A MixedMethods Learner Corpus Cum Experimental Research

Written corrective feedback (WCF) has received considerable attention in secondlanguage (L2) writing research. The conducive role of WCF in developing L2 writing and second language acquisition has been corroborated by a number of theoretical frameworks, and the findings of empirical studies, meta-analyses, and research syntheses. WCF research has predominantly addressed its effectiveness in improving learners’ syntactic, lexical, and orthographic knowledge. This dissertation project extends the scope of this line of research to formulaic aspects of language and investigates the relative effectiveness of WCF targeting formulaic vs. non-formulaic constructions in L2 writing. The text-analytic descriptive aspect of this research design aimed at investigating the extent of L2 learners’ non-target-like use of formulaic vs. non-formulaic forms in L2 writing and writing teachers’ WCF treatment of non-target (non)formulaic language use. A total of 480 first drafts of essays written by 33 advanced adult English-as-a-foreign language (EFL) learners during one semester and 480 drafts of essays corrected through WCF by three EFL teachers constituted the corpus in this study. Advancing the field of learner corpus research, the findings demonstrated that whereas learners’ non-target formulaic forms outnumbered that of non-formulaic ones in their writing assignments, all three teachers provided WCF more often for erroneous use of non-formulaic forms. The quasi-experimental aspect of the research design attempts to add new empirical evidence on the L2 learning potential of accessing and processing WCF provided for formulaic vs. non-formulaic constructions in L2 writing. To this end, a total of 66 EFL learners in a Test of English as a Foreign Language preparation course participated in a pretest-posttest design, with 5 experimental groups (those who were provided with direct, indirect, direct plus metalinguistic, and indirect plus metalinguistic WCF) and a control group (those who were not provided with WCF). Maintaining a division between formulaic vs. non-formulaic forms, the findings provide empirical evidence on the interactions between types of WCF, types of linguistic targets, and the effectiveness of WCF in terms of enhancing L2 learners’ accuracy and acquisition in their revised writing and new writings in the short and long term.