Speaker
Description
We are surrounded by voices (both natural and synthetic) of varying quality, which raises the question: does the brain process those voices differently? Building on research revealing the timing of neural correlates for voice characteristics (within 100 ms for physical traits such as gender and age, and after 350 ms for social attributes, Lavan et al., 2025), we hypothesized that perceived humanness might have its own neural signature. Forty-seven native German speakers evaluated the gender, trustworthiness, and humanness of 96 vocalic sounds. Using representational similarity analysis, we examined correlations between these behavioral ratings and neural activity from Lavan et al.’s existing EEG dataset across an 800-ms window. Humanness ratings correlated with neural activity from 78 ms post-stimulus onset, peaking at 120 ms, closely paralleling physical trait processing and contrasting with later social trait correlates. These findings demonstrate that perceived humanness is processed rapidly and automatically alongside basic physical voice characteristics.