Listeners as innovators, speakers, and members of speech communities
Patrice Speeter Beddor
John C. Catford Collegiate Professor of Linguistics, University of Michigan
Contrasts between speech sounds are conveyed by multiple, often redundant phonetic properties, making it possible for listeners to differ systematically from each other in the weights they assign those properties when deciding what a speaker has said. In this talk, I view individual listeners' novel perceptual weights as having the potential to contribute to new norms in a speech community. I discuss how perceptual innovation is made public and consider social factors that might contribute to—or impede—such innovation.
The main empirical basis for my discussion of these issues is American English and Afrikaans listener-speakers' perception and production of coarticulatory vowel nasalization. (Coarticulatory nasalization is the usual synchronic "raw material" for the historical change from vowel+nasal sequences to contrastive vowel nasalization.) Speakers of each language differ considerably from each other in the extent of produced coarticulation; in addition, though, nasal coarticulation in Afrikaans is socio-ethnically indexed, with speakers of the less prestigious variety producing less extensive coarticulation than speakers of the more prestigious variety. Results show that listeners' perceptual weights for coarticulation are mirrored in their productions: participants who attend closely to vowel nasality in perception also produce earlier onset of coarticulatory nasalization. However, for Afrikaans, this pattern held only for female participants. These findings lead me to propose that (i) a potential source of a novel variant is innovative listeners who are also (co-)articulatory innovators but (ii) which listeners mirror perception in their productions is conditioned by the social structure of coarticulatory variation in the speech community.