From the Linguistic Turn to the Probabilistic Turn: Ontological Conditions of Meaning, Normative Sedimentation, and the Dual Status of Artificial Intelligence

This paper develops the thesis of a probabilistic turn in contemporary philosophy and social theory. While the linguistic turn established language as the fundamental mediation of social reality and transformed our understanding of collective dynamics, knowledge, and action, recent developments in artificial intelligence reveal that it is possible to produce socially recognisable forms of meaning without lived semantics or intentional comprehension, operating exclusively on probabilistic regularities of use. This finding compels a philosophical displacement: from the question of what meaning is to the question of what conditions make something count as meaning within a social world. The article argues—on the ontological plane, with consequences for empirical research and social theory—that social reality is grounded in plausibility equilibria: inferential and institutional structures that stabilise expectations and coordinate action through dynamics which, analysed at their deep structure, reveal probabilistic patterns. However, it contends that genuine normativity—obligation, justification, responsibility—cannot be derived from such stabilisation. To account for this limit, the concept of a normative exterior is introduced, understood not as a transcendent instance but as the active residue of historical sedimentations that enter into dynamic tension with new plausibility equilibria. Artificial intelligence occupies a dual status within this framework: as an epistemological revealer of structures that have always operated in the production of meaning, and as a historically unprecedented disruptive factor capable of accelerating, challenging, hybridising, cementing, or opening new normative configurations.