Prosodic skewing of input and the initiation of cross-generational sound change

This paper addresses a proposal about how the seeds of sound change are planted during cross-generational transmission in the particular case of persevering vocalic chain shifts, that is, changes that appear to span many generations. Specifically, we explore the idea that the realization of vowels during child-directed speech may set up young learners to construct their own vowel space in slightly but consistently different ways from those of their caretakers, a process we call ‘prosodic skewing’. If this view is correct, it reveals a particular way that social and structural factors interact in sound change, where cultural norms (how caretakers talk to children) systematically bias the structural input to learners. We draw evidence from a cross-generational study of three American dialects where vocalic chain shifts are believed to be underway.