The relationship between label-based and speech-based perceptual evaluations: The case of Enshi Mandarin regional varieties

The current study explored the relationship between participants’ label-based evaluations of six regional varieties in Enshi Prefecture, China, and their speech-based evaluations of talkers from these varieties using a label ranking task and a speaker evaluation task. The results revealed that under correct identification of talker dialect, participants’ evaluations of real talkers based on speech samples were different from their evaluations of ‘imagined’ county-based dialects, suggesting that speech-based talker evaluations are not solely governed by ideological values associated with dialects. Focusing on a small, understudied community in China, this study contributes to our understanding of the local language attitudes, and language use and maintenance in Enshi Prefecture. An integrated approach is needed to build a model of talker evaluation, which must include a complex set of linguistic, social cognitive, and situational objects. The current results suggest that the object(s) that primarily drives talker evaluations is not the talker’s dialect itself.