Voice Onset Time Imitation in Teens Versus Adults

We compare teens' and adults' imitation of sentences with shortened and lengthened voice onset time (VOT), in order to test whether purported age-based advantages in phonetic acquisition may be due to differences in imitative ability.

Teens (M _age = 13, n = 39) and adults (n = 31) completed an explicit imitation and discrimination task on pairs of sentences characterized by canonical and manipulated (shortened or lengthened) VOT. We assessed extent of imitation using two acoustic metrics (ΔVOT and Proximity), accuracy on the discrimination task, and correlations between imitation and perception.

Both age groups modified VOT when imitating stimuli with both lengthened and shortened VOT. Adults, however, showed significantly more lengthening than teens (i.e., higher ΔVOT), as well as VOT values that were slightly but significantly closer to the target stimulus values (i.e., lower proximity). Both age groups showed above-chance discrimination accuracy, and a significant relationship between individual perception and production performance was found for lengthened-VOT sentences.

We found no evidence that teens have a greater imitative ability than adults; in fact, adults showed significantly more imitation based on both acoustic metrics. Both age groups showed robust imitation of VOT manipulations in both directions, in contrast to previous work showing lack of imitation for shortened VOT. Extent of imitation was predicted by individual perceptual performance, but only to a limited degree, underscoring the importance of other factors in explaining individual variation in imitative ability.