INTEGRATING DATA-DRIVEN LEARNING AND CORPUS LINGUISTICS TO ENHANCE ACADEMIC VOCABULARY ACQUISITION

Abstract

Abstract: Academic vocabulary the relatively stable set of words that recur across disciplinary written genres regardless of subject matter poses a persistent challenge for learners preparing to study in a second language. This article examines how Data-Driven Learning (DDL), an approach in which learners consult corpus data, typically in the form of concordance lines, to inductively discover lexical and grammatical patterns, can be integrated with corpus-derived academic word lists to enhance the acquisition of academic vocabulary. Drawing on a review of corpus-linguistic and DDL research, the article traces the development of academic word lists, from the Academic Word List to the more recent Academic Vocabulary List, and synthesizes empirical evidence on the effectiveness of DDL for vocabulary learning, including a large-scale meta-analysis reporting large effect sizes for corpus-based instruction. The findings indicate that concordance-based, inductive engagement with authentic academic corpora supports deeper lexical processing and more transferable word knowledge than single-example or definition-based instruction alone, although gains vary with proficiency level, task design and the amount of guidance provided. The article proposes a staged pedagogical model for integrating word-list-informed corpus selection with DDL tasks in academic vocabulary instruction.