Irony vs. Cover-up? Facial Dynamics of Positive Utterances in Negative Contexts in Japanese Speakers

Abstract The present study examined how Japanese speakers express verbal irony through prosody and dynamic facial behavior. Native Japanese speakers performed short-scripted dialogues that ended with the same positive utterance with positive (sincere) or negative intentional contexts. The speaker’s intent in the negative context was further classified, based on post-recording interviews, as irony or a socially smoothing “cover-up” stance. We extracted prosodic features (utterance duration, fundamental frequency, and intensity) and facial features (frame-by-frame facial action units and gaze direction). Prosodic measures primarily distinguished sincere productions from negative-context stances,with sincere productions showing higher pitch and greater intensity; irony productions were longer than their corresponding sincere productions. Facial measures revealed clearer differences among the production groups. Irony was associated with sustained brow-related activity and enhanced mouth-opening activity, together with an early downward-right gaze tendency. Sincere utterances showed stronger smile-related activity and a tendency toward upward-right gaze, whereas cover-up had an intermediate and heterogeneous profile. These findings provide a time-resolved account of multimodal stance marking in Japanese and suggest that positive utterances in negative situations may reflect not only ironic intent but also socially smoothing speaker intentions.