Top-down vocal feedback control in active hearing.

Abstract

Speaking is a sensory-motor process that involves constant self-monitoring to ensure accurate vocal production. This monitoring of vocal feedback allows a speaker to quickly adjust speech production to correct perceived errors between intended and actually produced vocal sounds. The self-monitoring in speaking is crucial for learning to speak native as well as foreign languages. An important behavior in vocal feedback control for both human speech and animal vocalizations is the compensatory change in speech or vocalizations (e.g., pitch, frequency, and intensity) when there is a mismatch between intended and perceived vocal sounds. Such a behavior requires mechanisms for continuously monitoring auditory feedback during vocal production. We have found that disruption of auditory feedback during vocalization alters coding properties of auditory cortex neurons in marmosets (a highly vocal primate species). Furthermore, when marmosets compensate for changes in vocal feedback, there are corresponding changes in their cortical neural activity. These findings suggest that the neural network underlying self-monitoring of vocal production likely consists of both sensory processing and top-down modulations via higher-cortical areas involving planning and memory.