The _Phrasome_: Language as a Structure of Misrecognized Subjectivity

This thesis proposes the concept of the Phrasome as a novel analytical framework for understanding misrecognition and quasi-subjectivity in language-generating AI interactions. Rather than treating AI responses as reflections of machine agency or user intent, the study argues that the perceived structure of intersubjectivity emerges from the pragmatic logic of language itself. Through a critical-pragmatic approach, the thesis identifies how features such as tone mirroring, contextual continuity, and non-central responsiveness contribute to a co-constructed sense of presence in dialogue. The research employs AI-generated language interactions as pragmatic data, emphasizing the role of linguistic form in structuring simulated intersubjective effects. By examining the mechanisms through which the Phrasome materializes in interaction, this work aims to advance the philosophical understanding of language as a performative system that sustains structurally misrecognized presence through formal interactional conditions, even in the absence of authentic interlocutors.