The development of cross language speech perception : the influence of age, experience, and context on perceptual organization

Previous research (Werker, Gilbert, Humphrey, & Tees, 1981) in which we compared English infants, English adults, and Hindi adults on their ability to discriminate two pairs of Hindi (non-English) speech contrasts indicated that infants without prior specific language experience can discriminate speech sounds according to phonetic categories, whereas adults may lose this ability as a function of either age and/or specific language experience. The present work was designed to answer several questions that emerged from that earlier research. First, experiments focussed on delineating the time course of the "decline" in non-native speech discrimination abilities across childhood. Second, experiments examined the generality of developmental change between infancy and adulthood by looking at cross-language speech perception of a new (Thompson, an Interior Salish Native Indian Language) non-English speech contrast. Third, speech perception performance was examined in relation to cognitive and linguistic development to try to determine why the decline occurred at one rather than another point in ontogeny. Finally, an attempt was made to clarify the nature and implications of the apparent loss of non-native speech perception abilities by varying both the discrimination procedure and the perceptual set conditions used in adult testing. The results of these experiments replicate our original findings (Werker, et. al., 1981) showing that infants can discriminate the universal set of phonetic contrasts, and that there is a decline in this ability as a result of specific linguistic experience. This decline occurs within the first year of life. The data tentatively support the notion that a certain level of memory development (enabling an early form of representational ability) may be necessary before specific experience can modify initial infant abilities. In addition, the results show a difference between phonetic and phonemic (meaning based) perception in adult subjects, with the phonemic being the most robust and the most easily demonstrated. It is suggested that phonemic perception may reflect the structuring of cognitive/perceptual categories. Speculation as to the form (prototypical) and the format (initially enactive, later symbolic) of the representation of these categories is offered.Arts, Faculty ofPsychology, Department ofGraduat