NON-NATIVE SPEECH SOUND PRODUCTION CHANGES EVEN WITH PASSIVE LISTENING TRAINING

Abstract

<p>The difficulty of non-native speech sound acquisition depends on several factors in addition to the central role of the mother tongue sound system. Age of exposure, amount of exposure as well as the type of training provided have been shown to have an effect on learning outcomes. In addition, theories of speech perception suggest that perception and production may either be strongly dependent upon each other as to the degree that perception triggers motoric patterns as well, or that activation of the auditory system is enough for perception so that the production system remains inactivated. In order to study whether mere exposure to auditory stimulation would result in production changes, we selected two groups of Finnish learners in a two-day listen-and-repeat training protocol. Both groups were auditively exposed to a non-native speech sound contrast embedded in a semi-synthetically produced pseudo-word context /ty:ti/ - /tʉ:ti/. While the passive listening group merely listened to the stimulus pairs without any motoric actions, the production group actively produced the stimulus words according to the provided model. We performed acoustic analyses and extracted the values of the two lowest resonance frequencies, formant 1 (F1) and formant 2 (F2) from the productions. The results indicated no statistically significant differences between the groups, neither in the formant values nor in their standard deviations. However, as a function of training, both groups showed clear changes in the standard deviation values thus indicating changes in production performances. This suggests that both training protocols have an effect on production learning, and more importantly, that the motoric commands seem to alter on the basis of mere auditory stimulation. This further suggests that the motoric system is activated even in perceptual tasks.</p>