Research Overview

My research lies at the intersection of Pragmatics, Cognitive Linguistics and Computational Linguistics, with a particular focus on how communicative behaviour develops and changes through social interaction. I combine theoretical work on language and social cognition with computational and corpus-based approaches to naturalistic interaction across languages, cultures and populations.

A major strand of my work is in Computational Pragmatics and AI Pragmatics. I develop theoretical and computational methods for modelling communicative behaviour and investigate how the principles governing human interaction extend to communication with and between Large Language Models. This includes research on linguistic imitation, creativity, (im)politeness and reciprocity in human and AI interaction. My computational approach to the field is developed in Computational Pragmatics: Visualizing and Predicting Verbal Behaviour (Cambridge University Press, 2026).

Central concepts in my research are intersubjectivity and dialogic resonance: how speakers reuse, adapt and creatively transform one another’s communicative behaviour. I study these phenomena as windows into engagement, social cognition, language learning and linguistic creativity. This work is developed in Language and Social Minds: The Semantics and Pragmatics of Intersubjectivity (Cambridge University Press, 2021) and Language, Imitation and Creativity: Dialogic Resonance in Grammar and Pragmatics (Cambridge University Press, 2026).

My research on (im)politeness and reciprocity examines how responsive communicative behaviour contributes to moral order, interpersonal relationships and social identity. I have extended this work to human–AI interaction through the AI Moral Dilemma: the tension between the pressure on Large Language Models to behave safely and their simultaneous capacity to adapt to and reproduce human communicative behaviour. My research showed that sustained exposure to real-life impoliteness can lead AI to increasingly reciprocate human patterns of verbal conflict, raising broader questions about human-like behaviour, moral alignment and AI safety. The study attracted international attention, including major coverage in The Guardian. These and other theoretical and methodological dimensions of Pragmatics are brought together in my textbook Pragmatics: Theory and Methods (Oxford University Press, 2026).

These interests extend to the application of Pragmatics and Computational Pragmatics to Autism, Dementia, first and second language acquisition, intercultural communication and healthcare, allowing me to investigate how patterns of social interaction vary across populations and communicative settings.

I also have a long-standing research specialism in Chinese Linguistics and language change. My work draws extensively on Mandarin Chinese and other Sinitic languages, alongside Germanic, Romance and other language families, to investigate pragmatics and grammatical phenomena including aspect, evidentiality, modality and pragmatic markers. My research on language change encompasses grammaticalization, constructionalization, (inter)subjectification, entrenchment and semasiological change, themes also developed in the edited volume Different Slants on Grammaticalization (John Benjamins, 2023).

Methodologically, my work integrates corpus-based and computational approaches with machine learning, network analysis and Bayesian modelling, alongside naturalistic human and AI interactional data, experimental and questionnaire methods, and cross-linguistic and intercultural comparison. A defining objective is to develop computationally explicit models of pragmatic and interactional behaviour while retaining their grounding in naturally occurring social interaction.

Across these areas, my broader research programme addresses a common question: how do humans—and increasingly intelligent systems—adapt their communicative behaviour to one another, and what can these patterns tell us about social cognition, creativity and language itself? The AI Moral Dilemma captures one of the wider implications of this question: as intelligent systems become increasingly human-like communicators, understanding how they adapt to our behaviour becomes not only a linguistic problem, but also a fundamental challenge for the safe and socially responsible development of AI.

I currently convene Theories of Language (postgraduate), providing advanced training in contemporary linguistic theory, and Pragmatics (undergraduate), exploring how meaning and communicative behaviour emerge through language use and social interaction.