Transforming Language Learning and Communication Across Multilingual and Digital Contexts

Part I – AI, Mediation and Evaluation in Multilingual Digital Communication

The first part examines the growing role of AI in shaping language use, evaluation, and communicative authority across multilingual contexts. The three chapters included here explore both conceptual and practical dimensions of AI-mediated communication, from tone and pragmatics to truth bias and automated L2 writing correction. Together, these chapters establish technological mediation as the foundational site of transformation in multilingual digital contexts.

Rosa Giménez-Moreno’s first chapter, ‘The meaning of tone variation in AI apps: Implications for ELF learners’, addresses a conceptual ambiguity that has grown more pressing with the proliferation of AI-powered writing and communication tools: the meaning of ‘tone’. Traditionally associated with phonology, tone has acquired a far broader semantic range in the context of generative AI and computer-mediated communication, including style, register, attitude, and emotional stance. Through a review of the literature and an analysis of leading AI applications such as ChatGPT and Copilot, the chapter maps the theoretical and practical dimensions of this conceptual expansion. A questionnaire administered to advanced English learners further reveals that this vagueness generates genuine confusion among language users, with tangible consequences for their development of interpersonal variation in English. The chapter makes a compelling case for greater terminological precision in AI-mediated language education.

Chapter 2 by Ángela Almela and Francisco Javier Ferre Fuertes, ‘Exploring truth bias in generative AI for deceit detection: Evidence from English and Spanish’, shifts the focus to the question of veracity assessment. Building on Truth-Default Theory and an expanding body of research on human deception detection, the chapter reports an experiment in which two GPT-4-based chatbots (ChatGPT-4o and Microsoft Copilot) and human judges assessed the truthfulness of short first-person statements in English and Spanish. The findings replicate and extend prior evidence that AI systems exhibit a robust truth bias relative to human judges, a tendency that persists across different prompt conditions. The chapter raises important theoretical and applied questions about the reliability of large language models as tools for linguistic deception detection across languages.

Chapter 3, by Sofía Moratinos-Johnston, ‘Self-efficacy in EFL writing: The use of generative AI writing tools and their perceived support’, completes this opening part by examining students’ perceptions of their own writing abilities and challenges across key dimensions of L2 writing, as well as their evaluations of a generative AI-supported instructional intervention. Through a pre-experimental pre-test–post-test design, the study highlights the varied difficulties learners face in grammar, vocabulary, and text organisation, while revealing that their perceptions of ChatGPT’s usefulness reflect a general sense of support rather than skill-specific benefits. The findings emphasise students’ positive attitudes toward GenAI for linguistic and stylistic enhancement, alongside the need for targeted lexical development and informed pedagogical guidance to foster strategic, needs-based integration of AI tools in L2 writing contexts. Closing the section, the study functions as a bridge between conceptual debates about AI-mediated meaning and concrete applications in language education.

Part II – Multimodal Input and Audiovisual Learning in Second Language (L2) Development

The second part focuses on the role of multimodal and audiovisual input in second language development, spanning first exposure, incidental vocabulary acquisition, and structured pedagogical interventions. The four chapters here share an interest in understanding how learners engage with screen-based and multimodal resources – from streaming content and captioned media to video game localisation – and what this engagement contributes to the processing and acquisition of vocabulary, pragmatic awareness, and broader communicative competence. Individual differences and metacognitive awareness emerge across the chapters as key factors mediating the relationship between input and learning outcomes. Taken together, these contributions illustrate how multimodal environments transform not only access to input but also the cognitive and affective dimensions of language development.