Summary Auditory verbal hallucinations (AVH) are a pervasive debilitating symptom of psychosis. Understanding the pathology and pathophysiology underlying AVH would provide scientific knowledge about brain abnormalities in psychosis, and reveal a target neurobiological system for novel interventions such as brain stimulation. Yet, it is not known where in the brain disorders of structure or function lead to AVH, what role gray matter (local information processing) and white matter (information flow between distributed gray matter areas) abnormalities play, whether distinct patterns of pathophysiology underlie different types of AVH, when neurobiological abnormalities emerge in the prepsychotic prodromal phase, and how pathology and pathophysiology progress in the early disease course after psychosis and AVH emerge. This project aims to understand the neurobiology of AVH by examining structure and function in a left-hemisphere dominant language circuit thought to be involved in the genesis of AVH cross-diagnostically during initial stages of psychosis comprising the clinical high risk stage, the first psychotic episode stage, and longitudinally into the early psychosis stage. Individuals will express a range of severity of auditory perceptual aberrations, beginning with no symptoms, progressing through attenuated psychotic auditory misperceptions and then AVH of increasing intensity and duration, culminating in severe constant AVH. Our overarching hypothesis is that AVH are caused by a final common pathway of overactivity in left Wernicke's area that leads to overstimulation of upstream auditory sensory cortices and downstream semantic language stores. We will examine neurophysiology (simultaneous EEG and MEG) and hemodynamic measures of brain activity (MRI-based arterial spin labeling) to assess resting level activation and network functional connectivity in a distributed frontal (Broca's area), temporal (Wernicke's area and primary auditory cortex), and limbic system (putamen) language-related circuit, gray matter volumes in these areas (structural MRI) and the integrity of the white matter tracts connecting them (MRI-based Diffusion Spectrum Imaging), and EEG & MEG during a verbal fluency task. Our pilot data indicate that worse AVH are associated with increased verbal fluency at all disease stages. Further AVH of voices commenting appear to arise due to hypo-connected Broca's and Wernicke's areas, whereas AVH of other (typically negative) voices are associated with hyper-connectivity with the limbic putamen. We hypothesize that abnormalities in this language system will be detected in the prodromal and at risk stage that will correlate with auditory perceptual aberrations, in the first episode stage that correlate with AVH intensity, and that the pathology and pathophysiology will worsen during the first year post-psychotic break, particularly in the white matter integrity and the network functional connectivity dynamics. Regarding therapeutic...