Early neural face perception is not modulated by emotional prosody prime congruency

Faces and voices are some of the most salient sources of information for identifying others' emotions. However, little research has focused on how emotional prosody modulates the earliest stages of face perception at the neural level. This study investigated priming effects of emotional prosody on the time course of subsequently presented emotional faces using event-related potentials (ERPs). Auditory prosodic primes (spoken sentences filtered to remove semantic information) expressing happy, neutral, or angry emotion were presented and immediately followed by a face expressing a congruent or incongruent emotion. After each trial, participants (n = 58) rated the face on valence and arousal. ERPs time-locked to face onset were analyzed at all time points and electrodes using data-driven mass univariate statistics. A significant main effect of face emotion was found spanning the N170, P2, and early EPN components. A significant main effect of voice emotion was found from approximately 50-100 msec and 150-400 msec, interpreted as increased attention in V1 and other non-face visual areas due to arousing vocal stimuli, especially angry ones. Critically, these factors did not interact, suggesting that face-voice congruency did not affect early perceptual face processing in this context. Behavioural results, however, showed that voice emotion influenced ratings of face valence and arousal, suggesting that voice prime and face information are integrated only at later cognitive but not early perceptual stages. Thus, prosody primes influenced visual processing before and after facial expressions were decoded, but face-voice congruency did not modulate early face emotion processing.