A Long History: From Universal Language to Artificial Intelligence

Our access to language is governed by various technical infrastructures, artifacts, and media. The rise of AI language models has intensified questions surrounding such technical mediation-often framed as unprecedented disruptions. While historians and STS scholars have long emphasized the sociotechnical contingencies of informatic systems, scholarship has paid less attention to how contemporary language technologies intervene in longstanding debates about language, information, and the ordering of knowledge and social life. Beginning in seventeenth-century Europe, through projects for universal languages, and extending to the mid-twentieth-century emergence of information theory, and integration of computational methods into linguistics, this article shows that the formalization, coding, and rendering computable of language were co-constituted with particular epistemic assumptions, technical circumstances, and practices of knowledge production. By situating contemporary AI systems within this longer genealogy, the article reframes language technologies as historically contingent interventions rather than radical ruptures.