Auditory and Acoustic Correlates of Vocal Stereotypes of Good Guys and Bad Guys

Twenty-seven voices of characters in Japanese animation were analyzed auditorily using a modified version of Laver’s framework for voice quality description. It was found that long-term pharyngeal states (i.e., constricted or expanded pharynx) play an important role in differentiating the voices of heroes and villains. F0 and vowel formant frequencies were measured for the same samples. In an experimental setting, subjects listened to content-masked speech excerpts of the same voices and rated their impressions of gender, age, physical and personality traits, emotional states and vocal characteristics. The correlations among the auditory and acoustic analyses and perceptual experiment results were calculated. It was found that pharyngeal constriction negatively correlated with a number of trait items in the perceptual experiment, including positive personality traits and vocal characteristics. Vowel formant frequencies and mean F0 negatively correlated with some items that were pertinent to physical strength (e.g., “big,” “strong,”).