International Journal of Arts and Humanities

Beyond Machine Legibility: Artificial Intelligence, Ontological Relativity, and the Epistemic Politics of Indigenous Language Preservation

Abstract

The swift expansion of Artificial Intelligence (AI) is reshaping how knowledge is produced, classified, preserved, and accessed, generating new possibilities for documenting and revitalising endangered Indigenous languages while simultaneously raising urgent questions of epistemic justice, cultural representation, and knowledge ownership. This article argues that the crisis of Indigenous language preservation in the age of AI is not primarily technological but epistemic and ontological: it concerns who defines what a language means, what knowledge counts, how that knowledge is classified, and whose worldview becomes computationally legible. Language is treated here as an epistemic and ontological medium through which communities articulate histories, cosmologies, moral concepts, ecological knowledge, and culturally specific understandings of reality. When Indigenous languages are digitised through AI systems designed within dominant Western linguistic and technological frameworks, preservation efforts may paradoxically reproduce epistemic marginalisation by privileging external categories of classification and intelligibility. Employing critical analysis as its methodology, the article interrogates the epistemic assumptions, power relations, and technological structures underlying AI-mediated language preservation. Its theoretical framework draws on Digital Humanities and ontological relativity, the latter illuminating how distinct linguistic systems embody and organise different conceptions of reality, being, and human–world relations. The article contends that digitisation is never epistemically neutral: the categories, metadata, algorithms, and computational architectures used to represent languages shape what aspects of Indigenous knowledge become visible, intelligible, or excluded in digital environments. It therefore advances a shift from technological preservation toward ontologically sensitive digital preservation. The article concludes that the future of Indigenous language preservation should not merely render Indigenous languages legible to machines; rather, digital technologies must become capable of accommodating the diverse ontological worlds these languages encode. AI can contribute meaningfully to epistemic justice only when preservation moves beyond linguistic form to safeguard the knowledge-worlds, meanings, and ontological possibilities that languages sustain.

Keywords: Artificial Intelligence; Indigenous Knowledge Systems; Epistemic Justice; Indigenous Languages; Language Preservation; Digital Humanities; Ontological Relativity; Digital Epistemology; Indigenous Data Sovereignty; Linguistic Diversity; Algorithmic Representation; Cultural Knowledge.