Integrating critical corpus and AI literacies in applied linguistics: a mixed-methods study

This study examines the integration of corpus literacy and critical AI literacy in language education. As part of a sequential mixed-methods design, undergraduate learners from a southern European university (n = 37) participated in structured workshops. These workshops focused on the use of the corpus analysis software, Sketch Engine and the Generative AI tool, ChatGPT for vocabulary and grammar learning. Quantitative and qualitative data from pre-workshop surveys, hands-on workshop activities, and a post-workshop focus group reveal that while learners initially lacked familiarity with corpus tools, they came to recognise their value in developing corpus-mediated linguistic awareness. Conversely, ChatGPT was widely used by the learners but elicited mixed evaluations regarding reliability. The findings suggest that corpus literacy fosters critical engagement with language data, while generative AI tools present challenges related to over-reliance and a lack of truly personalised learning. Consequently, the study argues that developing corpus literacies can act as a means of developing critical AI literacies, which can enhance language learning for learners using AI. Specifically, these findings underscore the value that corpus approaches bring to language learning and the need for explicit training in AI tools used in autonomous language learning. Future research should explore scalable pedagogical models for this integration.