INDIVIDUAL VARIATION IN INFANT SPEECH PROCESSING: IMPLICATIONS FOR LANGUAGE ACQUISITION THEORIES

To what extent does language acquisition recruit domain-general processing mechanisms? In this dissertation, evidence concerning this question is garnered from the study of individual differences in infant speech perception and their predictive value with respect to language development in early childhood. In the first experiment, variation in the processing of a linguistic unit at six months was found to predict vocabulary development at around 2 years of age, whereas processing of a non-unit did not. In the second experiment, one possible source for that variation in linguistic performance was assessed, namely information processing abilities. Infants were tested on the same linguistic task as in Experiment 1, and on a well-researched task that yields a measure of information processing in infancy. No covariance was found between measures gathered in the linguistic and the information processing tasks. In a third experiment, the impact of variation in the infants’ input on their speech processing was investigated. Correlations between infants’ performance in a speech sound discrimination task and acoustic characteristics of their primary caregivers’ speech were investigated. Two types of acoustic characteristics were measured; some were not relevant to the speech sound being tested, but are known to influence infants’ attention and learning (pitch and pitch modulations); others were specific to the contrast tested. Results suggested that only those characteristics relevant to the contrast being tested affected infants’ speech processing. In sum, these three experiments and extensive literature reviews suggest specific ways in which domain-general factors (such as attentional mechanisms) are involved in infants’ development of linguistic knowledge. While these factors appear to play a role in the learning of phonological units, their influence may not be evident once linguistic categories are already established.