Conceptualising the development of the native listener: What are infants attuning to when they become perceptually tuned to the "sound patterns" of native speech?

It is a common observation that late learners of a second language (or third, etc.) usually have a noticeable accent from their native language when speaking the later-learned language. What is less overtly obvious, and yet no less remarkable, is that adults also evidence a strong native "accent" in perceiving speech from later-learned, and completely unfamiliar, nonnative languages. At the beginning of the journey of language acquisition, however, young infants do not display a native accent in tongue or ear. It should probably not be surprising that the tuning of these natural instruments for spoken communication requires some amount of experience with an ambient language. What is perhaps surprising, however, is how early and in what form the perceptual attunement begins to emerge. In the first 6 months, infants universally discriminate whichever consonant or vowel contrasts are presented to them, regardless of whether or not they occur in the infant's to-be-native language. But by 8-10 months of age they show a decline in discrimination of many nonnative consonant contrasts; a somewhat different sort of shift appears in perception of many nonnative vowel contrasts by 6 months. The tuning of young children's perception to native speech contrasts undergoes further refinement by native language experience during the early years of childhood, including an apparent shift from focusing on language-specific low-level phonetic details, to beginning to recognize higher-level lexical phonological structure.

Three open questions, however, remain: 1) What is the nature of information that young infants extract from speech? 2) How does experience tune up recognition of language-specific consonant and vowel properties from the very young infant's universal sensitivity to simple phonetic properties? and 3) How can developmental changes in perception of nonnative speech provide insights about the infant's emerging native phonological system? This talk will examine several contemporary theoretical perspectives/assumptions about the basic nature of information that very young listeners detect in speech (innate phonological categories? linguistic-phonetic features? acoustic cues? stimulus evidence re: the distal source, i.e., articulatory gestures of the human vocal tract?). We will then discuss the implications of each viewpoint for understanding how infants develop a native phonological system from their experience listening to native speech (general or language-specialized perceptual mechanisms? indirect cognitive construction or analysis processes? direct perception of distal event properties?).