Although much is known about the acoustic characteristics of children's speech, there is comparatively little known about the mechanisms available to a child for producing sounds that can be recognized by listeners as speech, and how those mechanisms may differ from those of an adult system.
The specific aims of the proposed research have been designed to address this gap in knowledge by studying the sound production properties of a child-like speech production system based on a computational modeling approach. The research will be guided by three hypotheses: 1) children develop a sound production system in which the laryngeal sound source interacts nonlinearly with the vocal tract filter to generate harmonic energy; 2) children develop the ability to produce vowels and consonants based on the acoustic sensitivity of the vocal tract; and 3) the fundamental frequency of the voice source and the formant frequencies of the vocal tract are tuned by the child speaker to maximize the generation of harmonic energy in a pattern that expresses the acoustic characteristics of the vocal tract in the speech signal. The significance of the project is that it will provide new information about the acoustic, aerodynamic, and kinematic factors that guide speech development. This could lead to new explanations of aspects of children's speech production such as the timing of vocal tract gestures and the coarticulation of speech units. The proposed research may also impact theories of speech production and speech motor control, and facilitate the next generation of speech synthesizers. In turn, the models developed could serve as an empirical framework for testing hypotheses regarding speech impairments in children.
The small size of a child's speech production mechanism generally results in 'high-pitched' speech. The characteristics of these sounds are quite different from the speech produced by adults, even though children can transmit the same clear message to a listener. The long-range goal of this research is to understand how a child develops a vocal sound production system capable of generating intelligible speech at an early age and then continually adapts to the anatomic changes due to growth.
