关键词: 语音感知, 统计学习, 感觉-运动联结, 婴儿

Abstract: Background. During the first year of life, infants’ speech perception undergoes a remarkable reorganization: their attention shifts from broadly discriminating between native and non-native contrasts to focusing on distinctions within their native language. How do infants develop this sensitivity to speech sounds? The acquisition of speech perception is theorized to rely on two complementary mechanisms: statistical learning and sensorimotor coupling. Statistical learning emphasizes infants’ sensitivity to the distributional regularities of speech input, whereas sensorimotor coupling highlights the modulatory role of articulatory motor feedback in speech processing. The present study examined the contributions of these two mechanisms during the sensitive period for speech perception in infancy and aimed to determine whether speech perception arises from a single mechanism or from the interaction of multiple mechanisms. Method. The present study examined the contributions and boundaries of statistical learning and sensorimotor coupling in 80 Mandarin-learning 8-month-old infants. An alternating/non-alternating visual fixation paradigm was employed across two experiments. In Experiment 1, we manipulated the distributional properties of the familiarization input to test the effect of statistical learning on speech perception. Infants were familiarized with either a unimodal or bimodal frequency distribution of speech tokens and subsequently tested for their ability to discriminate the Hindi dental–retroflex contrast (/ɖ/-/d̪/). In Experiment 2, we manipulated articulatory motor feedback by using a teething toy that either restricted or did not restrict tongue-tip movement, following the paradigm of Bruderer et al. (2015). No additional auditory familiarization was provided, allowing us to assess the causal role of motor feedback in speech perception. Both experiments included speech and non-speech stimuli. The speech stimuli were the Hindi dental–retroflex contrast (/ɖ/-/d̪/), while the non-speech stimuli were acoustically matched synthesized tones based on piano timbre, designed to parallel the temporal and spectral properties of the speech stimuli. Results. Across all distributional input and motor manipulation conditions, infants’ looking times decreased progressively with repeated stimulus presentations, indicating a habituation effect. (1) Experiment 1 showed that after exposure to the bimodal distribution, infants exhibited significantly longer looking times to alternating speech streams than to non-alternating speech streams, indicating successful speech discrimination. In contrast, under the unimodal distribution condition, looking times did not differ significantly between alternating and non-alternating speech streams, suggesting no effective speech discrimination. When the stimuli were non-speech sounds, neither the bimodal nor the unimodal distribution elicited discrimination behavior. (2) Experiment 2 showed that under the unrestricted tongue-tip movement condition, infants demonstrated significantly longer looking times to alternating speech streams than to non-alternating ones, indicating speech discrimination. In contrast, when tongue-tip movement was restricted, no effective speech discrimination was observed. For non-speech stimuli, neither movement condition elicited discrimination behavior. Conclusion. Statistical learning and sensorimotor coupling jointly support the development of early speech perception, revealing a multi-mechanistic architecture in infant language acquisition. Both mechanisms exhibited domain specificity for speech stimuli, indicating that speech processing gains early priority in infancy. This domain specificity may stem from infants’ selective attention to speech and the unique activation of sensorimotor couplings by speech signals. Moreover, The role of environmental input in speech acquisition appears to be multilevel: while the sensorimotor mechanism can generate initial discrimination sensitivity in the absence of specific speech experience, the statistical learning mechanism refines and stabilizes this sensitivity through continuous exposure to the ambient language environment.

Key words: speech perception, statistical learning, sensorimotor coupling, infants