Language for General Purposes and Data-Driven Learning

Data-Driven Learning (DDL) is a pedagogical approach to foreign or second language learning and teaching which employs authentic language data from corpora to engage learners in guided discovery tasks, fostering deeper understanding of linguistic patterns and usage. Historically focused on error-correction and short-term gains, recent advances expand DDL’s application to discourse, pragmatics, and rhetorical functions, enhancing its effectiveness in language for general purposes (LGP) contexts. Integrated with LGP, DDL draws on key theoretical frameworks like constructivism and sociocultural theory, supporting learner autonomy and critical thinking. Methodologically, DDL employs tools such as concordance lines, frequency analysis, and collocations to analyse authentic language data. Challenges include data complexity and teacher preparation, necessitating pedagogic transfer for educational adaptation. Indirect and direct DDL approaches offer distinct benefits. Indirect DDL simplifies learning through corpus-informed materials without direct corpus visibility, maintaining a traditional teaching approach. Direct DDL involves learners directly with corpus extracts or activities like concordance use, promoting autonomy. Empirical evidence supports DDL’s efficacy, illustrated by recent case studies. Future trends in DDL for LGP include integrating large language models (LLMs) and mobile apps for enhanced accessibility and real-time linguistic analysis. While LLMs represent a paradigm shift, traditional DDL methods remain valuable, particularly in indirect approaches where linguistic analysis informs learning materials. As these technologies evolve, they will complement DDL, enabling learners to explore and understand target languages more effectively. Future research should apply DDL in diverse educational settings, develop accessible tools, and provide educator training for optimized use across learner profiles and contexts.