Learning Speech

Abstract

This culminating chapter is divided into five parts. Following an initial background are sections related to time spans of phonemes, syllables, words, and phrases, respectively. A historical background, lodged in both linguistics and psychology, presents domain-specific explanations of perceptual learning of speech versus music. In contrast, this chapter continues a domain-general argument by assuming that learning of speech (as well as music) depends on grasping abstract time relationships (attractors) at various time scales in speech as well as in music. This portrait brings together many familiar dynamic attending constructs from earlier chapters. It outlines three stages of phoneme learning in infants. With age, learning develops to include learning of attractor profiles in syllables and words. Ultimately, word learning is shown to rely on both traditional and transient mode-locking.