Introduction: Towards a linguistic anthropology of AI Introduction : vers une anthropologie linguistique de l'IA

Abstract

This essay introduces the themed cluster of articles, ‘Towards a linguistic anthropology of AI’. The advent of artificial intelligence (AI), especially in large language models capable of producing coherent discourse mimicking conversational interaction, is exerting unprecedented pressure on prevailing concepts of language, personhood, and the human. This provides a challenge and opportunity for reflection, particularly for linguistic anthropology, that sub-field centred on language as medium/practice of meaning-making in social life. Moving beyond AI as communicating machine to explore communicative processes that structure AI as ideological imaginary and technological infrastructure, the articles in this cluster draw on linguistic anthropology to examine different facets of AI as people take it up, such as the features that invite users to endow chatbots with personhood (Keane), attempts to use AI to unlock animal communication (Handman), and research seeking linguistic biomarkers predictive of psychopathology (Semel). To contextualize the articles, this introduction offers some basic semiotic distinctions developed within linguistic anthropology needed to parse AI, with a focus on the concept of semiotic ideology. We then situate AI within a set of ideological genealogies – from Enlightenment visions of universal languages to the representationalism of modern linguistics – before turning to themes of transduction and semiosis beyond the human.