Glossolalia Against the Model: Artaud, AI, and the Limits of Language

Abstract

This paper examines the epistemological and ontological implications of large language models (LLMs) through the lens of Antonin Artaud's theory of expression. Drawing on Artaud's concept of glossolalia and its refusal of semantic coherence, the study critiques the smoothing, optimization, and datafication of language in AI systems. Engaging with theorists including Gilles Deleuze, Félix Guattari, Byung-Chul Han, Mark Fisher, Georges Bataille, and Shoshana Zuboff, the paper argues that LLMs instantiate a necrosyntax, a metaphysical condition wherein language circulates without living intensity. The work calls for new expressive paradigms that affirm linguistic rupture, waste, and corporeality as sites of resistance to machinic rationalization.

Keywords: Glossolalia, Antonin Artaud, Large Language Models (LLMs), AI and Expression, Machine Language, Posthuman Ontology, Algorithmic Culture, Ontology of Language