Psychological Medicine, 1996, 26, 177-189. Copyright © 1996 Cambridge University Press The auditory hallucination: a phenomenological survey TONY H. NAYANI AND ANTHONY S. DAVID 1 From the Department of Psychological Medicine and Neuropsychiatry, King's College Hospital and the Institute of Psychiatry, London SYNOPSIS A comprehensive semi-structured questionnaire was administered to 100 psychotic patients who had experienced auditory hallucinations. The aim was to extend the phenomenology of the hallucination into areas of both form and content and also to guide future theoretical development. All subjects heard 'voices' talking to or about them. The location of the voice, its characteristics and the nature of address were described. Precipitants and alleviating factors plus the effect of the hallucinations on the sufferer were identified. Other hallucinatory experiences, thought insertion and insight were examined for their inter-relationships. A pattern emerged of increasing complexity of the auditory-verbal hallucination over time by a process of accretion, with the addition of more voices and extended dialogues, and more intimacy between subject and voice. Such evolution seemed to relate to the lessening of distress and improved coping. These findings should inform both neurological and cognitive accounts of the pathogenesis of auditory hallucinations in psychotic disorders. INTRODUCTION The auditory hallucination is a common feature of psychotic illness and has been recorded in 74% of schizophrenics in the International Pilot Study of Schizophrenia (Wing et al. 1974). Slade & Bentall (1988), who pooled data from 16 reports, yielding a total of 2924 cases of schizophrenia, determined a prevalence of 60-2 % for this symptom. The central importance of the auditory-verbal hallucination as a dis- criminating diagnostic entity for schizophrenia is framed in the DSM-III (American Psychiatric Association, 1980). Other schemes such as the Present State Examination (PSE, Wing et al. 1974) continue to emphasize the Schneiderian concept of the first-rank symptom (Schneider, 1959), which distinguishes certain types of hallucination as having special significance for the diagnosis of schizophrenia (David & Appleby, 1992). Auditory hallucinations also occur in a variety of psychiatric and neuro- psychiatric disorders in addition to schizo- phrenia (Zigler & Philips, 1961; Lowe, 1973) and, indeed, data from the Epidemiologic 1 Address for correspondence: Dr A. S. David, Department of Psychological Medicine, Institute of Psychiatry, London SE5 8AF. Catchment Area (ECA) study (Tien, 1991) show that hallucinations of all modalities occur in the general population at an annual incidence of 4-5%. Classical teaching proposes a discrimination of form from content in the explication of phenomenology. Though this distinction may be useful for epidemiological and diagnostic purposes, the implications for phenomenology are problematical. Regarding form (Larkin, 1979), features of the hallucination itself may be enumerated: for instance, duration and whether it is internal or external; while content may encompass whether the substance of the hallu- cinations is menacing, comforting or banal. But when features such as the grammatical con- struction are considered, for example second versus third person origin, or whether the 'voices' are running commentaries or single expletives, or, indeed, whether the voice is angry or robotic, it is not clear whether such attributes belong to form or content categories, or rather somewhere in between. As well as providing clues to the mechanisms and pathogenesis of hallucinations, a detailed account of their phenomenology may have practical implications for patients. Persistent 177