Abstract
Infants can learn and generalize phonetic categories through speech sound frequency distributions. Nevertheless, previous research with varying participant ages and testing paradigms reported incongruent findings regarding the effect of distributional learning of phonetic contrasts. The current study examines infants’ distributional learning of non-native tones using electroencephalography. 5-6-monthold Australian infants were exposed to an 8-step continuum of a Mandarin Chinese high-level vs. high-falling tonal contrast. The bimodal condition had frequency peaks near the two ends of the continuum (steps 2, 7) whereas the peak was at the midpoint of the unimodal condition (steps 4, 5). Before and after listening to their corresponding distribution, both groups were tested on the same sounds (steps 3, 6) in a passive oddball paradigm. The unimodal group (N = 8) showed strong sensitivity to the sound distinction at post- but not pre-distributional learning. The bimodal group (N = 8), no significant neural sensitivity or difference was observed in pre- or post-distributional learning. The finding that unimodal exposure enhances infant perception is novel and is explained by their acoustic sensitivity to peak location, highlighting the role of the magnitude of the acoustic distinction in the stimuli when prior training and exposure is insufficient to establish phonetic categories.
Read more at Research Direct