Is vocal emotional processing fine-tuned to speaker gender, identity and phonetic content? - Evidence from...

The perception of vocal emotions requires efficient and flexible processing of auditory input. A hallmark of this flexibility is perceptual adaptation, which enables listeners to tune into the current auditory context (e.g., angry voices), thus sensitizing the system to sudden changes (e.g., a shift to fear). However, how vocal emotion adaptation is modulated by either stable speaker characteristics such as gender and identity, or dynamic cues such as phonetic content is insufficiently understood. This gap was addressed in three experiments using a paradigm inducing simultaneous opposite aftereffects: during adaptation, emotions were systematically tied to a voice feature to test if adaptation aftereffects could be induced in opposite directions along a fear-anger continuum. For example, after adaptation to angry-male and fearful-female voices, male targets should be more often perceived as fearful and female ones more often as angry. Based on theoretical considerations, such simultaneous opposite aftereffects were predicted for speaker gender (Experiment1) and identity (Experiment 2) but should not appear for phonetic pseudoword content (Experiment 3). Indeed, the expected pattern was observed for speaker gender and pseudoword. Evidence for speaker identity was not fully conclusive but offered promising hints that listeners may exhibit nuanced adaptation to different speakers as well. Together, these findings show that vocal emotion adaptation is partially fine-tuned to stable speaker characteristics, whereas this does not seem to be case for dynamic phonetic content. This provides important empirical support for recent calls that a comprehensive understanding of vocal emotional processing requires a closer look at the speakers expressing them.