The effect of speaker-identity on dialect processing
This study investigated the combined role of video- and voice-familiarization on dialect processing. Six white actresses were audio- and video-recorded reading the same monologues and pseudo- and real-words, in two dialect guises: Southern and Standardized US English. These recordings were used in a lexical decision task (LDT), where for a given list an actress was presented as mostly Southern, mostly Standardized, or was presented evenly in both guises. Southern and non-Southern participants were assigned to one of 4 conditions, which differed in whether there was video as well as audio, and whether participants were familiarized with the actresses’ accents through monologues before LDT. The video with monologue condition most affected listener’s performance, and these effects were most robust in Southern listeners. These results suggest that listeners can use speaker identity to form expectations about a speaker’s dialect, and that some dialect processing difficulties are about dialect expectations.
