That trauma can play a significant role in the onset and maintenance of voice-hearing is one of the most striking and important developments in the recent study of psychosis. Yet the finding that trauma increases the risk for hallucination and for psychosis is quite different from the claim that trauma is necessary for either to occur. Trauma is often but not always associated with voice-hearing in populations with psychosis; voice-hearing is sometimes associated with willful training and cultivation in nonclinical populations. This article uses ethnographic data among other data to explore the possibility of multiple pathways to voice-hearing for clinical and nonclinical individuals whose voices are not due to known etiological factors such as drugs, sensory deprivation, epilepsy, and so forth. We suggest that trauma sometimes plays a major role in hallucinations, sometimes a minor role, and sometimes no role at all. Our work also finds seemingly distinct phenomenological patterns for voice-hearing, which may reflect the different salience of trauma for those who hear voices.