Acoustic Patterns and Speech Acquisition

By far the greater part of the work at present in progress in the field of speech acquisition depends on two related descriptive tools. The first comes directly from classical phonetics and makes use of place, manner and voice descriptors and a tradition al transcription. These investigations attempt to define the sound contrasts of speech qualitatively, both in production and perception, in what are primarily productive, articulatory, terms. The second method of description uses a particular set of distinctive features (Chomsky and Halle, 1968) which are based on sub sets of these phonetic, articulatory, dimensions. These distinctive features are intended to facilitate the definition of phonological contrasts. This contribution is concerned with a complementary description of some of the aspects of speech acquisition in strictly quantifiable acoustic terms. The acoustic form of speech can be given a direct auditory as well as an articulatory interpretation and this makes it possible to arrive at a realistic appreciation of what elements in a speech sound sequence are likely to be dominant in sensory terms and how these elements must be processed in normalization for example when listening to a small as opposed to a large vocal tract so that physically different acoustic stimuli can have a common phonetic identity.