Vocal activity in verbal hallucinations.

Vocal and forearm EMGs were recorded from 19 hallucinating and 22 nonhallucinating psychiatric inpatients. A microphone placed close to the lips was used to detect subvocal speech. The subjects' EMG and subvocal responses to pre-recorded statements were also assessed. Subjects who experienced hallucinations during assessment were asked to estimate the frequency of their hallucinations and rate phenomenological of their most recent hallucination. Results essentially replicated studies that found increased vocal potentials in hallucinators, but also showed that these increased potentials were nonsignificant when nonvocal measures are included in the statistical analysis. Subvocal speech and coincident increases in vocal EMG with reports of hallucinations, and with reports of pre-recorded statements, were not found. A significant negative correlation was found between the mean vocal potential of hallucinators and the perceived location of their most recent hallucination. The pre-recorded statements were generally perceived to be louder, clearer and more outside the head than the most recent hallucination.