ABSTRACT
The goal of this chapter is to better understand the neural processes underlying speech production by investigating native and non-native speech. The first section outlines several behavioral findings relevant to development and acquisition of native and non-native speech. The second section gives an outline of the potential ontogeny of speech production development within the framework of a neuro-computational model. The third section reviews results of brain imaging studies of native and second-language speech production in relation to predictions of the proposed model. It is concluded that investigation of native and non-native speech production reveals aspects of the underlying control processes and the task-dependent use of perceptual target reference domains. The results of brain imaging experiments support the following hypotheses: Separate neural processes are used depending on whether acquisition of a language occurs in infancy vs. later in life. Feedforward control is used for production of well-established articulately patterns, whereas feedback control is used for production of articulately patterns being learned. Speech planning control relies more on orosensory information than auditory information under task constraints in which the auditory target reference domain is ambiguous. Individuals with more well-established auditory target reference domains show greater activity in brain regions involved with forward articulatory-to-auditory mapping.
