Neural bases of speech production

Authored by: Jason W. Bohland , Jason A. Tourville , Frank H. Guenther

The Routledge Handbook of Phonetics

Print publication date:  April  2019
Online publication date:  March  2019

Print ISBN: 9781138648333
eBook ISBN: 9780429056253
Adobe ISBN:


 Download Chapter



Speech production requires the integration of diverse information sources in order to generate the intricate pattern of muscle activations that result in fluent speech. Accordingly, a large portion of the cerebral cortex is involved in even the simplest speech tasks, such as reading a single word or a meaningless syllable. Broadly speaking, there are three main types of information involved in speech motor control: Auditory, somatosensory, and motor, represented in the temporal, parietal, and frontal lobes of the cerebral cortex, respectively. The interaction of sensory and motor information across these regions underlies the learning and maintenance of motor programs that enable rapid, accurate articulation of commonly produced speech sounds. Additional areas in the medial and lateral prefrontal cortices contribute to the fluent selection and initiation of these speech motor programs. Activity in these cortical regions is modulated by two re-entrant loops through subcortical brain regions during speech production: A basal ganglia loop involved in action selection and initiation, and a cerebellar loop involved in generating the finely tuned motor commands that underlie fluent speech. These brain regions and their interconnections, along with additional subcortical nuclei in the thalamus, midbrain, and brainstem constitute the neural control system responsible for speech production. This chapter addresses research that informs our understanding of the speech production neural control system with particular emphasis on developments fostered by neuroimaging technologies such as functional magnetic resonance imaging (fMRI), positron emission tomography (PET), and magnetoencephalography (MEG). This research is discussed in reference to the DIVA (Directions Into Velocities of Articulators) model and its extension the GODIVA (Gradient Order DIVA) model. These computational neural models provide a theoretical framework for understanding and integrating past research and for developing new hypotheses for guiding future experiments.

Search for more...
Back to top

Use of cookies on this website

We are using cookies to provide statistics that help us give you the best experience of our site. You can find out more in our Privacy Policy. By continuing to use the site you are agreeing to our use of cookies.