A Long History: From Universal Language to Artificial Intelligence

Area of Science:

• History of Science and Technology (STS)
• Linguistics
• Artificial Intelligence

Background:

• Contemporary AI language models raise questions about technical mediation of language.
• Historical scholarship has focused on sociotechnical aspects of informatic systems.
• Less attention has been paid to how language technologies influence debates on knowledge and social order.

Purpose of the Study:

• To situate contemporary AI language systems within a longer historical genealogy.
• To reframe language technologies as historically contingent interventions, not radical ruptures.
• To analyze the co-constitution of language formalization, coding, and computability with epistemic assumptions and knowledge production practices.

Main Methods:

• Historical analysis of language formalization and computation from the 17th century to the mid-20th century.
• Examination of projects for universal languages and the emergence of information theory.
• Integration of computational methods into linguistics and the development of AI language models.

Main Results:

• The formalization, coding, and computability of language have been historically co-constituted with specific epistemic assumptions and technical practices.
• Longstanding debates about language, information, and knowledge ordering are intervened upon by contemporary language technologies.
• AI language models are reframed as historically contingent interventions rather than unprecedented disruptions.

Conclusions:

• Contemporary AI language technologies are not entirely novel but are part of a long historical trajectory.
• Understanding the historical context is crucial for analyzing the impact of AI on language, knowledge, and society.
• Language technologies are shaped by and, in turn, shape societal and epistemic frameworks.