Discreteness and the musilanguage continuum
Presenter Name:Elizabeth, Phillips
Co-Authors:Sam Kim, Steven Brown
Vocal music is the oldest and most universal pitched music, but even it must have had an evolutionary precursor. The “musilanguage” model of music evolution (Brown, 2017) posits that music and speech both evolved from a joint precursor that was characterized by many of their shared features, including affective prosody, combinatoriality, and an imprecise levels-and-contours pitch system. From this precursor, the innovation of lexicality could have led to the evolution of speech, whereas the innovations of precise metric and tonal organization could have led to the evolution of music. However, speech and solo singing still have many degrees of overlap, such as the ordered pitch relationships of tonal languages, the melody-like prosody of recitation and oration, and the speech-like quality of chants. As such, we propose that, setting aside linguistic content, speech and music lie on a continuum ordered by pitch- and interval-class discreteness. The more precisely-organized pitches and intervals are, the more song-like an utterance will sound. To test this, we developed several computational measures of pitch- and interval-class discreteness, and applied them to 76 recordings from a cross-cultural corpus of infant- and adult-directed speech and song (Hilton et al., 2022). The results of a regression analysis across recording type indicate that pitch- and interval-class discreteness is indeed a distinguishing factor along the musilanguage continuum.