The embodied simulation approach predicts that restricting facial movements disrupts emotion recognition. Such effects have been reported for facial and whole-body emotional expressions, but findings remain inconsistent, and it is unclear whether they generalize to emotional sounds. Previous work has also shown that restricting mimicry induces a positivity bias in valence ratings, consistent with the facial feedback hypothesis. Here, we tested whether restricting facial mimicry impairs recognition of emotional sounds across four emotion categories in a forced-choice task and whether it affects valence and arousal ratings. Drawing on previous findings indicating that vocal emotional expressions elicit facial mimicry, whereas instrumental emotional sounds do not, we expected to find an effect only for the former. Instrumental sounds were recognized less accurately and more slowly than vocalizations, particularly for negative and neutral expressions. Crucially, both frequentist and Bayesian analyses provided no evidence that the pen-in-mouth manipulation impaired emotion recognition or influenced valence or arousal ratings. Taken together, these findings underscore the need for further studies to systematically determine the conditions under which restricting emotional mimicry influences emotion processing, as well as the nature of such effects.
