The search for predictors of individual differences in VOT imitation
We investigate the effects of individual differences measures on individuals’ imitation behaviors. 43 participants were recorded reading aloud 84 voiceless stop-initial words twice, once with written text and a second time after auditory words from two talkers with artificially lengthened VOT. Participants then completed surveys measuring Autism Quotient, Big Five, working memory and social network size. Participant VOTs were extracted and normalized by subtracting the model fits accounting for word, word frequency, trial number, phoneme, rest-of-word duration and participant speech rate. The difference between shadowed and baseline normalized values is Imitation. An LMER model was conducted to predict Imitation with talker, by-talker word-specific VOT, and the scores and subscales of the aforementioned individual differences measures. We found significant effects of talker and word-specific VOT, but no effects of individual differences measurements. We discuss the replicability of predictors of individual linguistic behaviors and their consequences for theories of language change.
