A Corpus-Based Multimodal Analysis of Written Assignments in the Target Language Use Domain: Implications for L2 Multimodal Writing Task Design and Assessment

This study presents a corpus-based multimodal analysis of successful L2 postgraduate written assignments, an under-researched yet developing area of multimodal communication, along with a thematic analysis of associated rubrics and feedback at UK universities. Despite the increasing multimodal nature of writing in authentic settings and higher education, most L2 writing tasks in language proficiency tests or EAP assessments involve no multimodal composing. To understand the authenticity of L2 writing tasks in English, it is crucial to analyze multimodal writing assignments that students are typically asked to do at the postgraduate level. Drawing on target language use domain analysis, theoretical frameworks of multimodality and corpus linguistics, this paper responded to the need for research in this area by analyzing 1993 multimodal semiotic resources in a corpus of 150 successful multimodal postgraduate written assignments. The present study identified frequent multimodal meaning-making patterns in terms of types, meanings and intersemiotic relations. The empirical findings and a theory-driven genre-based model developed in this study are of significance to test developers, EAP instructors and language teachers because they provide a foundation to inform the construct of L2 digital multimodal writing and to develop rubrics in both low and high-stakes assessment contexts.