EARLY LANGUAGE ACQUISITION: CRACKING THE SPEECH CODE
Patricia K. Kuhl
Abstract
Infants learn language with remarkable speed, but how they do it remains a mystery. New data show that infants use computational strategies to detect the statistical and prosodic patterns in language input, and that this leads to the discovery of phonemes and words. Social interaction with another human being affects speech learning in a way that resembles communicative learning in songbirds. The brain's commitment to the statistical and prosodic patterns that are experienced early in life might help to explain the long-standing puzzle of why infants are better language learners than adults. Successful learning by infants, as well as constraints on that learning, are changing theories of language acquisition.
Summary
* Infants learn their native language quickly and effortlessly, and follow the same developmental path regardless of culture. However, it has proved difficult to understand how they do this, or to build computers that can reproduce this feat.
* An early and essential task for infants is to make sense of the speech that they hear. Each language uses a unique set of about 40 phonemes, and infants must learn to partition varied speech sounds into these phonemic categories. Young infants are sensitive to subtle differences between all phonetic units, whereas older children lose their sensitivity to distinctions that are not used in their native language. The loss of discrimination for foreign-language distinctions is paralleled by an increase in sensitivity to native-language phonetic units.
* There is evidence that infants analyse the statistical distributions of sounds that they hear in ambient language, and use this information to form phonemic categories. They also learn phonotactic rules language-specific rules that govern the sequences of phonemes that can be used to compose words.
* To identify word boundaries, infants can use both transitional probabilities between syllables, and prosodic cues, which relate to linguistic stress. Most languages are dominated by either trochaic words (with the stress on the first syllable) or iambic ones (with the stress on later syllables). Infants seem to use a combination of statistical and prosodic cues to segment words in speech.
* Social influences are important in speech learning. Infants learn more easily from interactions with human beings speaking another language than they do from audiovisual exposure to the same language material, and their speech is strongly influenced by the response of others around them, such as their mothers. The importance of social input in language learning has some similarities to social influences on song learning in birds.
* Language experience causes neural changes. One hypothesis, native language neural commitment (NLNC), proposes that language learning produces dedicated neural networks that code the patterns of native-language speech. As these networks develop, they make it easier for new speech elements and patterns to be learned if they are consistent with the existing patterns, but place constraints on the learning of foreign-language patterns. NLNC might explain the closing of the 'sensitive period' for language learning; once a certain amount of learning has occurred, neural commitment interferes with the learning of new languages so they cannot be learned as easily.
More on Kuhl's study of language acquisition can be found at:
<http://www.ed.gov/news/pressreleases/2001/07/07262001-kuhl.html>
The core point here (for lbo discussion) is that "the individual" _as a human being_ (i.e., as a language-using animal) does not exist but emerges from social relations.
Carrol