Item

Abstract

We contrast the traditional view that vowel phonemes are neutral, abstractbuilding blocks with an understanding that they convey activities of embodiedemotional facial musculature. In theory, the same facial musculature associated withvisually recognizable emotional expressions also favors the production of audito-rily recognizable sounds. We found a relational commonality in the distribution ofphonemes across perceptual, acoustic, and biomechanical metric spaces that mapswell onto emotional dimensions of valence and arousal. Specifically, /i:/-sounds (like“Gleam”) are associated with more positive emotional valence than /V/-sounds (like“Glum”), and /æ/-sounds (like “Wham”) are associated with more arousing emotionsthan /u:/-sounds (like “Womb). These trends generalize to other languages, acrossspecies, occur with both words and pseudo-words, and can be enhanced or inhibitedbymanipulatingfacialmusculature.Thesetypesofacousticassociationscomplementand integrate with other forms of embodied sound symbolism such as onomatopoeicfindings related to bouba-kiki phenomena, but the phoneme-emotion relationshipis more robust and provides a stronger functional basis for understanding languagedevelopment and evolution. The phoneme-emotion relationship provides a potentialexplanation for why humans originally evolved the ability to so finely discriminatethe acoustic phonemic characteristics upon which language is scaffolded. In short,phonemes may initially have effectively served as general acoustic emotion-detectionfeatures