EXPLORING METAPHOR IN HUMAN-AI DIALOGUE: A COGNITIVE LINGUISTIC APPROACH

Abstract

Metaphor forms a deep cognitive mechanism on the basis of which humans are able to conceptualize abstract things through familiar experiences, thereby influencing both reasoning and communication. Metaphors gain particular significance while interpreting human-AI dialogue since throughout the interactions with the AI, the end users often embellish it as "assistants," "partners," or even "minds," which actually refer and reflect quite deep-seated models of intelligence, agency, and sociality. However, a fundamental split arises here since most AI goes through language in a literal, mechanistic way. This study interrogates the pivotal role of metaphor in human-AI interaction through a cognitive linguistic lens. In a conceptual metaphor theory orientation, it systematically analyzes conversational data for dominant metaphorical patterns and assesses their effect on user trust, expectations, and perceptions about artificial intelligence competency. The findings, from this data analysis, testify that metaphor structures not just human understanding but through it also actively configures the relational positioning of AI. Thus, such research highlights an urgent call on systems that dynamically recognize and adapt to metaphorical language input toward conceiving more natural, ethically responsible, and cognitively aligned human-AI interactions.