Is low-arousal laughter a reliable cue for irony? Individual differences in integrating speech and laughter meanings

Irony can be cued by a variety of verbal and non-verbal markers, all sufficient, but none necessary to identify irony. The current study investigated the impact of low-arousal laughter in irony identification in interaction with verbal context. Participants (N = 119) listened to short stories including a brief context and a final target utterance which they evaluated as ironic or not. We manipulated the degree of incongruity between the context and the target utterance (No/Weak/Strong Incongruity) and the Presence/Absence of laughter before the target utterance. Overall, contextual incongruity with the target utterance was the strongest predictor of ironic interpretations regardless of laughter. Laughter had a significant effect only in the Weak Incongruity condition when considering individual variability. A cluster analysis revealed the emergence of three patterns of performance reflecting different laughter and speech semantic integrations: for some subjects, laughter cued towards ironic interpretations; for others, laughter reduced the probability of an ironic interpretation; while for a small group, laughter did not influence responses. This study demonstrates the importance of going beyond a speech-centered approach and including non-verbal behaviors when modeling language processing, as well as considering individual variability when exploring pragmatic processes. • Laughter influences irony comprehension when contextual cues are ambiguous. • We observe inter-individual variability in laughter interpretation in ambiguous contexts. • When context is not ambiguous laughter does not influence irony identification. • Need to go beyond a group level analysis when exploring pragmatic processes considering individual differences. • The integration of speech and non-speech elements is needed for language and dialogue modelling.