ALGORITHMIC DISCOURSE AND THE POSTHUMAN SUBJECT: A CORPUS-ASSISTED CRITICAL DISCOURSE ANALYSIS OF AI-MEDIATED LANGUAGE AND DIGITAL IDENTITY IN ONLINE SPACES

Abstract

The continued development and implementation of large language models (LLMs) and AI-mediated communication have altered language production, mediation, and consumption in digital contexts in significant ways. The current research endeavors to provide a critical assessment of the ways in which mediated language both constructs, constrains, and reproduces digital identity within online contexts, with specific regard to the ideological labor of algorithmic systems. Positioning itself within the confluence of Critical Posthumanism, Language Ideology Theory, and Critical Discourse Analysis, the study is a secondary data analysis that utilizes two publicly available social media corpora: the Pushshift Reddit Dataset and the Twitter/X Academic Research archive. Examining AI-generated and AI-mediated texts through a Corpus-Assisted Discourse Studies (CADS) approach, the study identifies systematic patterns of standard language ideology, the algorithmic marginalisation of non-standard English varieties and the rise of a (per)formative posthuman identity, as proposed in the present study. Findings add to the development of posthuman linguistics as an emerging sub-discipline of applied linguistics and to current applied sociolinguistic debates on the intersections of AI, linguistic justice and digital selfhood.