Proceedings of the Annual Meeting of the Cognitive Science Society

This study explores the cognitive processes behind authorship perception in text-based communication, examining how humor and emotional expressions influence the ability to distinguish human- from AI-generated texts. Drawing on theories of language processing and social cognition, we investigate whether humor and emotional tone serve as cues for attributing human-like qualities. Through an experiment, 212 participants evaluated text messages varying across humor (present/absent), emotional expression (positive, negative, neutral), and authorship (human/AI). Using Likert scales and open-ended responses, participants assessed human authorship likelihood and described the cognitive strategies guiding their judgments. Findings reveal how emotional and humorous content shapes authenticity judgments, offering insights into cognitive mechanisms underlying human-AI interaction. This research bridges classical cognitive theories with future challenges, highlighting the role of emotional and pragmatic cues in evolving digital communication contexts. Implications for cognitive science and the study of language processing are discussed.