When language speaks without a body: posthuman enunciation in the age of synthetic reality

Abstract

This article examines language as a posthuman practice and explores how algorithms, robots and avatars operate as agents of enunciation in the age of artificial intelligence. Drawing on Latour (2020), Barad (2003, 2007), Pennycook (2018), Morin (2025), and Lemes (2025), and based on an exploratory and qualitative methodology, the discussion advances the idea that language is not confined to human activity but is distributed across hybrid networks composed of humans and non-humans. Such a perspective challenges traditional notions of authorship, voice, and subjectivity by situating meaning-making within material, technological, and affective assemblages. Two discursive materialities are analyzed: the corporate avatar Lu do Magalu and the Morgan Freeman deepfake. Both exemplify how new communicative ecologies reconfigure trust, affect, and the perception of reality through synthetic and mediated technologies. The first materiality demonstrates how virtual agents can perform emotional proximity and stabilize digital forms of social interaction. The second exposes the fragility of perception in synthetic reality, where truth and simulation become indistinguishable, revealing the ethical and political dimensions of posthuman enunciation. The study argues that Applied Linguistics must adopt an indisciplinary and critical posture, capable of understanding language as a material-discursive event that transcends human boundaries. This approach calls for reconsideration of what it means to speak, listen, and interpret in an era shaped by algorithmic mediation. Ultimately, posthuman language practices invite an ethical and epistemological renewal, urging scholars to confront the complex entanglement between humans, technologies, and the realities they co-construct.