Audio-visual integration of emotion expression.

Regardless of the fact that emotions are usually recognized by combining facial and vocal expressions, the multisensory nature of affect perception has scarcely been investigated. In the present study, we show results of three experiments on multisensory perception of emotions using newly validated sets of dynamic visual and non-linguistic vocal clips of affect expressions. In Experiment 1, participants were required to categorize fear and disgust expressions displayed auditorily, visually, or using congruent or incongruent audio-visual stimuli. Results showed faster and more accurate categorisation in the bimodal congruent situation than in the unimodal conditions. In the incongruent situation, participant preferentially categorized the affective expression based on the visual modality, demonstrating a visual dominance in emotional processing. However, when the reliability of the visual stimuli was diminished, participants categorized incongruent bimodal stimuli preferentially via the auditory modality. These results demonstrate that visual dominance in affect perception does not occur in a rigid manner, but follows flexible situation-dependent rules. In Experiment 2, we requested the participants to pay attention to only one sensory modality at a time in order to test the putative mandatory nature of multisensory affective interactions. We observed that even if they were asked to ignore concurrent sensory information, the irrelevant information significantly affected the processing of the target. This observation was especially true when the target modality was less reliable. Altogether, these findings indicate that the perception of emotion expressions is a robust multisensory situation which follows rules that have been previously observed in other perceptual domains.