Project Summary/Abstract Speech is an incredibly complex cognitive function of the human brain that distinguishes us from other species. It is the major aspect of human communication that allows us to exchange ideas and convey emotions, therefore impaired speech can lead to devastating effects such as social isolation and anxiety. According to NIDCD reports, between 6 and 8 million individuals in the United States have some form of speech deficit, however our current understanding of the brain mechanisms that support speech is still limited. The specific mechanism that we plan to address in our research is auditory feedback control of speech production. Accurate and fluent production of speech critically depends on hearing oneself. Auditory feedback from hearing one’s own voice allows for real-time monitoring and adjustment of vocalization in order to produce intended speech. The aim of this project is to elucidate the sensory-motor network interactions in the human brain that enable this process. The sensory-motor interactions during speech production is usually modeled as a feed-forward and feed-back system: Vocal-motor regions in the frontal cortex sends a corollary discharge signal to the auditory cortex during vocalization to suppress responses to the predicted auditory outcome. When produced speech is different than intended speech, an error signal encoding the mismatch is fed from the auditory cortex back to the vocal-motor regions for the online correction of produced speech. Supporting these ideas, electrophysiological studies have demonstrated suppressed responses in the auditory cortex during speech production. Moreover, enhanced auditory responses were reported when auditory feedback was altered artificially. However, the spatial topography of these different responses, detailed network dynamics between the auditory and vocal-motor regions as well as how these dynamics relate to speech behavior remain unknown. In this project, we will address these gaps in our current understanding of speech production by using electrocorticography (ECoG) in neurosurgical patients, which directly measures neural activity across large- scale networks in the human brain with unprecedented spatial and temporal resolution. We will employ a delayed auditory feedback (DAF) paradigm to artificially disrupt speech fluency and apply novel multivariate autoregressive techniques to infer causal interactions within the speech network. Abnormal sensory-motor network interactions have been implicated in various disorders including stuttering, aphasia, Parkinson’s disease, autism and schizophrenia. Contributing to the mission of NIDCD, the results of this project will provide an empirical basis for understanding speech deficits and will help patients by guiding better therapeutic interventions ranging from speech therapy to assistive devices that will improve fluent speech production.