Intact Representation of Vocal Smile in Autism: A reverse correlation approach

Atypical emotional prosody production and perception have been reported in autism. However, it is unclear whether these particularities are associated with unusual mental representations of vocal emotions. The objective of the current study was to explore the mental representation of vocal smile in autistic adults. Twenty-nine autistic (ASD) and 29 neurotypical (NT) adults performed an auditory reverse correlation task, that affords the opportunity to extract acoustic features of mental representation and their variability. Most ASD participants (17) based their representation of vocal smile on similar acoustic features as NT participants and no difference in the level of internal noise was observed. However, comparisons between groups revealed a more typical representation in NT than in ASD. Subsequent cluster analysis revealed that the difference of typicality was explained by a small subset of ASD participants displaying different representations. A correlation analysis also revealed that the typicality was positively correlated with the empathetic level within both groups. While most autistic adults have a preserved mental representation of vocal smiles, a subset shows less robust and typical representations, which is linked to lower levels of empathy. This study highlights that the perception of vocal smiles in autism is more nuanced than previously reported, with empathy playing a substantial role in shaping these mental representations.