Categorizing Emotion in Spoken Language: An Analysis of Semantic and Prosodic Contributions to Emotional Communication

The current study aimed to replicate and expand upon research conducted by Bagley, Abramowitz and Kosson (2009) to examine categorization of emotional sentences among non-psychopathic individuals. 36 monolingual English-speaking undergraduate participants categorized spoken English sentences (produced with neutral prosody but containing semantic cues to emotion) and French sentences (produced with appropriate prosody but with no semantic cues to emotion) into one of five emotion categories: happiness, sadness, anger, fear, or neutral. By isolating the semantic and prosodic information available to listeners, we determined that categorization accuracy was higher among sentences expressing anger in the prosodic condition. Accuracy was higher among sentences expressing all other emotions in the semantic condition. Overall, the lowest categorization accuracy was found for sentences expressing fear in the prosodic condition. Across all emotion categories and both presentation conditions, reaction time was longest for sentences expressing fear in the prosodic condition. Although all participants in the current study had normative scores on the Self-Report Psychopathy Scale, those with relatively high scores displayed lower categorization accuracy for semantic sentences expressing happiness, anger and fear than lower-scoring participants. An extension of the current study comparing this normative sample to a group of individuals with psychopathy will need to account for possible implications of subclinical psychopathic characteristics on vocal affect categorization accuracy.