Auditory hallucinations:  Psychotic symptom or dissociative experience?






Andrew Moskowitz, Ph.D.

Dirk Corstens, M.D.



  The Journal of Psychological Trauma, 2007








While auditory hallucinations are considered a core psychotic symptom, central to the diagnosis of schizophrenia, it has long been recognized that persons who are not psychotic may also hear voices.  There is an entrenched clinical belief that distinctions can be made between these groups, typically on the basis of the perceived location or the ‘third-person’ perspective of the voices.  While it is generally believed that such characteristics of voices have significant clinical implications, and are important in the differential diagnosis between dissociative and psychotic disorders, there is no research evidence in support of this.  Voices heard by persons diagnosed schizophrenic appear to be indistinguishable, on the basis of their experienced characteristics, from voices heard by persons with dissociative disorders or with no mental disorder at all.  On this and other bases outlined below, we argue that hearing voices should be considered a dissociative experience, which under some conditions may have pathological consequences.  In other words, we believe that, while voices may occur in the context of a psychotic disorder, they should not be considered a psychotic symptom.



“Once you hear the voices, you realize they’ve always been there. 

It’s just a matter of being tuned to them.” 


(Mark Vonnegut, The Eden Express, 1975) 



Mark Vonnegut’s simple insight, expressed in his compelling autobiographical account of becoming psychotic in America in the 1960s, supports the position put forth in this paper, namely, that hearing voices should be considered a dissociative experience and not a psychotic symptom.  This position is opposed to that of most clinicians and researchers, who believe that auditory hallucinations[1] (AH) are best considered a psychotic symptom, or that some AH should be considered psychotic and others dissociative.  Nonetheless, our position is supported by both clinical and research evidence, including the evidence that many persons who hear voices do not show any (other) symptom of a mental disorder, and that no experienced characteristic of voices adequately distinguishes between those persons and others who are designated mentally ill.  Accordingly, we propose that AH should, under no circumstances, be considered a psychotic symptom, despite the fact that they sometimes occur in the context of a psychotic disorder.


Throughout recorded history, AH have been understood in a number of ways.  At various times, and under various circumstances, voices have been viewed as religious or spiritual phenomena (voices of gods, demons, angels), supernatural or psychic experiences (indicative of ghosts or telepathy), psychological experiences (post-traumatic, dissociative, psychotic), or entirely normal (voice of conscience, one’s own thoughts, creative inspiration, grief experiences, hypnopompic and hypnagogic experiences).  Over the past 150 years, considerable effort has been expended in attempting to distinguish between these allegedly different forms of hallucinations, particularly along the lines of determining which voices merit the attention of physicians or mental health professionals and which do not. 


Typically, these attempts have been mounted around certain perceived characteristics of auditory hallucinations, most commonly whether the voices appear to be emanating from outside the voice hearer or from inside their head (or some other part of their body).  Classically, the latter, particularly when combined with ‘intact’ insight, have been referred to as ‘pseudo-hallucinations’, while only the former, i.e., ‘external’ hallucinations, have been accorded the label ‘proper hallucinations’ or ‘true hallucinations’ and considered a psychotic symptom (Jaspers, 1963/1913).  However, insight is technically not an experienced aspect of the hallucination but a subsequent evaluation by the voice hearer (essentially equivalent to a secondary delusion).  While it appears likely that there are some – as yet unidentified – personality or psychological factors which determine whether voices are heard internally or externally (and help to explain why one person can hear both internal and external voices and why some voices’ perceived source of location may change), there is no evidence that perceived location (or any other experienced characteristic of AH) map onto diagnostic categories or relevant clinical variables (such as treatment response).  Indeed, a recent study pointedly titled ‘On the non-significance of internal versus external auditory hallucinations’ examined this issue in a large cohort and concluded, “(T)he clinical relevance of location is not confirmed, and the conceptual clarity and clinical utility of the pseudohallucination is undermined” (Copolov, Trauer & Mackinnon, 2004, p. 5).  Berrios and Dening (1996), in an exhaustive historical review, likened the concept of ‘pseudohallucinations’ to a “‘joker’ in the diagnostic game” whose fluidity allowed clinicians to “call into question the genuineness of some true hallucinatory experiences that do not fit into a pre-conceived psychiatric diagnosis” (p. 761).


If this is so – if there is no reliable way to distinguish hallucinations experienced in persons diagnosed with schizophrenia from those with other disorders or even from the general non-psychiatric population – what then do we make of auditory hallucinations[2]?  While AH have long been considered a core symptom of psychosis, if their clinical utility remains limited, should that assumption be re-evaluated?  Should the concept of auditory hallucinations be decoupled from the concept of psychosis?  And if that occurs, what becomes of the concept of psychosis?


Historical overview

Why some people do and others do not hear voices, and how to understand those voices, has been the subject of debate throughout the centuries.  The two questions are closely related, as assumptions about voices have usually (but not always) been made by persons who do not hear voices, and their judgments have often been strongly colored by class, gender, and race beliefs.  But was there a time when hearing voices was considered entirely normal?  While the research data reviewed below suggests, currently, as many as 8-10% of the general population hear voices at some point in their life, one theorist believes that, millennia ago, all people heard voices.  Julian Jaynes (1976), in a popular and controversial book, has argued that, until a few thousand years ago, humans did not possess self-reflective consciousness, and heard voices they attributed to gods, which guided their decision-making in small and large matters.  He called this ancient mental structure the ‘bicameral’ mind, and argued that people who hear voices in modern times experience a relapse to this form of mind, primarily under the influence of extreme stress.  Jaynes also predicted that hearing voices would be associated with activity in the right temporal lobe, analogous to the left hemisphere speech recognition/understanding center known as Wernicke’s area.  While Jayne’s highly speculative theory has received little attention in psychiatric circles, there is now some evidence, reviewed below, in support of his contentions.


For much of recorded human history, hearing voices was associated with divine inspiration, or Satanic possession; as such, those who decided the source of others’ voices (often on nebulous criteria) were typically associated with the Christian Church or other religions (Sarbin & Juhasz, 1967).  Then, in the mid 16th century, Teresa of Avila, fearful that her nuns would be persecuted by the Inquisition for the voices they heard, argued that some voices should not be seen as divine or devilish, but as the result of more mundane physical problems.  She wrote that some voices were due to melancholy, a ‘weak’ imagination, or drowsiness, sleep or sleep-like states (Sarbin & Juhasz, 1967).  Importantly, Teresa wrote that such individuals should be treated ‘as if’ they were sick, but Sarbin and Juhasz note that the ‘as if’ was quickly dropped as the phenomenon of hearing voices became medicalized.


The medical perspective became the dominant mode of understanding hallucinations (in all sensory modalities) in the 19th century, a term given its modern definition by Esquirol in 1832.  Esquirol distinguished hallucinations from illusions on the basis that the latter involved genuine sensory perceptions (plus distortions); he stated that a person hallucinating “ascribes a body and an actuality to images that the memory recalls without intervention of the senses” (Esquirol, 1832, cited in Bentall, 1990, p. 82).  While Esquirol believed that hallucinations were invariably pathological, a dispute developed in subsequent decades as to whether this was so.


This question culminated in a series of debates at the Société Médico-Psychologique in Paris, foreshadowing many of the debates still held today.  As summarized by Berrios & Dening (1996), these discussions considered “whether all hallucinations were abnormal; and whether location (internal versus external) and insight (present versus absent) were relevant factors to the definition of pathological hallucinations” (p. 756).  Berrios & Dening (1996) note that these important debates were overlooked by Karl Jaspers, in his highly influential General Psychopathology (1963/1913), who attributed the distinctions to an important figure from the 1880s, Victor Kandinsky.


Insert Table 1 about here


Kandinsky was important not only because, unlike the French theorists, he himself experienced hallucinations, but also because he claimed to experience both ‘true’ (TH) and ‘pseudo’ (PH) hallucinations, and endeavored to distinguish between the two[3].  Kandinsky argued that only TH involved the activity of a subcortical perception center (just like genuine perceptions), while PH involved only the activity of a center for apperception (a term rarely used now, which means the state of being conscious of perceiving) along with a ‘center for abstract, unconscious images’ (Berrios & Dening, 1996).  Thus, Kandinsky decided that PH were as sensorially vivid as TH, but lacked the ‘external objectivity’ of TH because in only the latter was the subcortical perception center activitated.  He also believed that PH were internally localized.   


Building on Kandinsky’s distinctions, Jaspers (1963/1913) developed a concept of PH which was dominant throughout the 20th century. Jaspers claimed that PH were similar to TH in that they both had ‘full, fresh sensory elements’ and could not be voluntarily controlled, but differed from the latter in that they were experienced in ‘inner subjective space’ and did not appear concretely real.  Admittedly, Jaspers concept of ‘reality’ here is vague and has led to differing interpretations.         


Jaspers’ view of PH held sway for half a century, until a series of papers out of Britain began to reconsider the concept.  In three papers all published in 1966, Sedman (1966a, 1966b, 1966c), on the basis of phenomenological research, rejected the equation of PH with ‘inner’ voices, arguing that location was irrelevant, but that PH were ‘ego-syntonic’ and lacked ‘publicness’ (i.e., the person did not expect the voice to be heard by others), while TH were ‘ego-dystonic’ and perceived as public phenomena.  However, echoing Jaspers, he claimed that TH were exclusively related to schizophrenia or schizoaffective disorder.  Hare (1973) went further, arguing, somewhat tautologically, that any voices heard by non-psychiatric patients should be considered non-pathological, and that the term PH should be reserved for voices heard by psychiatric patients but interpreted non-delusionally (i.e., a voice of conscience, hearing one’s thoughts). TH were all voices experienced by psychiatric patients and interpreted delusionally.  Finally, Taylor (1981) attempted to bring some clarity to the discussion by suggesting that PH had been used in two different ways over the past 100 years, one, which he called ‘imaged’ PH, linked more to German authors and the other, which he called ‘perceptual’ PH, linked to British authors.  The key difference, Taylor (1981) suggested, was that imaged PH, as propounded by Kandinsky and Jaspers, involved imagery, not perceptions, described as being less vivid and ‘real’ than TH, and internally localized, while perceptual PH, as propounded by Sedman and Hare, were not phenomenologically different from TH, but were accompanied by intact insight, in contrast to the latter.  Taylor did not endorse one approach over the other, but insisted that authors should be clear as to which definition they were using. 


This entire history has been thoroughly reviewed by Berrios & Dening (1996), who as noted above, strongly questioned the validity of the concept of pseudohallucinations.  However, in a parallel paper, Dening & Berrios (1996) argued that many clinicians continue to believe that not all hallucinations have the same clinical import.  This desire to ‘separate out’ some forms of AH from others can be clearly seen in the various 20th century attempts to ‘parse out’ hallucinations and relate them to different conditions and diagnoses.


Auditory hallucinations and differential diagnosis

Auditory hallucinations were not central features of either Kraepelin’s Dementia Praecox or Bleuler’s Schizophrenia.  For Bleuler, AH, while occurring frequently in schizophrenia, were derivative of the central disturbance of loosening of associations, and were also common in other disorders.  A much closer link between AH and schizophrenia, however, was forged by Kurt Schneider (1959), whose position underpins the diagnosis of schizophrenia in the DSM-IV (American Psychiatric Association, 1994) and ICD-10 (WHO, 1993). Schneider felt that certain AH, particularly those commenting on an individual’s thoughts or behaviors or two or more voices conversing with each other, were pathognomic (only one symptom required for a diagnosis) for schizophrenia; he also thought that hearing one’s thoughts aloud was characteristic of the disorder.  The first two symptoms were adopted by the American Psychiatric Association for the diagnosis of schizophrenia, and now form two of the three pathognomic symptoms (the third, also problematic, is ‘bizarre’ delusions[4]). Interestingly, while the APA had specifically linked external AH to schizophrenia in early DSMs, this link was eliminated (presumably for lack of empirical evidence) in the DSM-IV (APA, 1994). 


There is, however, considerable evidence that AH, particularly those considered pathognomic for schizophrenia, are not only not unique to that disorder but occur more frequently in dissociative identity disorder (Honig et al, 1998; Kluft, 1987; Ross et al, 1989; Ross et al, 1990).  In light of this observation, several authors have argued for new dissociative diagnostic categories (‘Dissociative hallucinosis,’ Nurcombe et al, 1996; ‘Dissociative subtype of schizophrenia’, Ross, 2004) in which AH are featured prominently.


Other authors have taken a different tack, addressing AH directly by proposing new classificatory systems.  Thus, Van der Zwaard & Polak (2001), after reviewing the concept of pseudohallucinations, argue that PH should be broken up into several categories, namely: a) nonpsychotic hallucinations (e.g., ‘isolated’ nonpsychotic hallucinations, such as occur in grief reactions, AH arising from sensory deprivation, and ‘vivid internal imagery’, typical in dissociative disorders), b) partial hallucinations, such as ‘fading hallucinations with increasing insight’, and c) transient hallucinations (such as ‘short lapses’ in reality testing in persons diagnosed with borderline personality disorder).  This classification appears to offer little improvement over the simply PH/TH dichotomy; it seems difficult to justify calling AH in dissociative disorders ‘imagery’ and ‘partial hallucinations’ appears to be a nonsensical concept.  Further, there is evidence that AH in borderline personality disorder are not transient, as has been generally believed, but in fact are ‘ongoing and pervasive’ (Yee et al, 2005).  In light of this, Yee et al (2005) has offered an alternative classification, namely that AH be split up into: a) normative, b) traumatic-intrusive, c) psychotic, and d) organic hallucinations.


But is there any evidence that AH can be successfully carved up in this (or any other) manner? 


Research evidence


A number of relevant research studies have now been conducted, examining the prevalence of hallucinations in the general population, contrasting voices heard by patients with non-patients or, within clinical populations, assessing whether certain characteristics of voices predict clinical or outcome variables.  Some studies have also demonstrated clear links between AH and dissociative experiences.


AH in non-psychiatric patients


The studies assessing the prevalence and nature of auditory hallucinations in non-psychiatric populations are illustrated in Table 2.


Insert Table 2 about here


Six large-scale studies, based primarily in the UK, the Netherlands, and the U.S., have found rates of between 0.6% to 8.2% annual or lifetime prevalence of hallucinations in the general population (excluding, to a greater or lesser degree, identified psychiatric patients)[5].  Three studies of selected non-psychiatric samples, primarily university students or medical patients, have reported slightly higher rates, ranging from 2%-13%.


The first large-scale study to assess hallucinations in the general population was carried out primarily in England between 1889 and 1892, and involved interviews with 17,000 adults (Sidgewick, Johnson, Myers, et al., 1894).  This study, known as the Sidgewick study after the first author, was conducted by the Society for Psychical Research (SPR), an organization dedicated to researching psychic phenomena, such as telepathy.  As such, a number of the interviewers may have been biased toward discovering hallucinations in non-psychiatric patients.  However, the study has been described as well designed, and incorporated a number of relevant exclusion criteria (Tien, 1991).  Sidgewick et al found approximately 10% of their participants to have experienced visual, auditory, or tactile hallucinations over their lifetime, just under 7% when corrected for sleep-related experiences.  Most of those hallucinations were visual. Those experiencing auditory hallucinations decreased from 2% of those in their 20s, to just under 1% for those over 30.  In addition, a gender bias was recorded, with women experiencing more hallucinations than men.


The results of the Sidgewick study were broadly similar to those found in a large-scale US study 90 years later.  The NIMH Epidemiological Catchment Area (ECA) project took place in the early 1980s, and involved structured interviews with over 18,000 adults in five major metropolitan areas (Tien, 1991).  While the lifetime prevalence for any hallucination was slightly higher in the ECA study – 13% – this figure, which also included olfactory and gustatory hallucinations, did not significantly differ from the Sidgewick result. Across the age groups, auditory hallucinations averaged about 2% lifetime prevalence, with the exception of 18-19 year olds, for whom more than 3% endorsed such experiences.  Importantly, while the ECA study was designed to assess the prevalence of mental disorders in the community, the majority of those endorsing auditory hallucinations did not appear to be mentally ill, as most claimed that these experiences were not a cause of distress or led to any impairment in their functioning. 


Interest in this area has accelerated over the past few years, with four groups of researchers publishing population-based study results since 2000.  Jim van Os and his colleagues in the Netherlands (Van Os et al., 2000) assessed a random sample of over 7000 adults, finding a lifetime prevalence of 8.2% for hallucinations (primarily auditory); as in the ECA study, the vast majority of those who responded positively reported that they were not bothered by their hallucinations and had not sought treatment for them. 


In a UK study, over 8000 individuals were screened for mental health problems as part of the Fourth National Survey of Ethnic Minorities (Johns et al., 2002).  One question on the screening was about hearing voices saying “quite a few words or sentences when there was no one around that might account for it”.  While just over 1% of the ‘white’ sample responded affirmatively to this question, almost 3% of the ‘Caribbean’ sample did likewise (but only 0.6% of the Asian group).  The figures represent annual prevalence rates. Again, only 25% of this group met criteria for a psychotic disorder when further assessed (by telephone), and there was no relation between a reported history of diagnosis or treatment for psychosis and hearing voices.  This is particularly notable as hearing, “quite a few words or sentences” reflects a fairly high level of AH (Posey & Losch, 1983-1984). 


Similar results were found in a Dutch study of 914 adolescents (Dhossche et al., 2002), who were followed up eight years later (86% of initial cohort successfully contacted).  Five percent of the initial cohort, and two percent of those followed-up, endorsed AH on a self-report questionnaire.  While AH were associated with depressive and substance abuse disorders (determined by a structured interview), there was no statistical association with any psychotic disorder.


Finally, one study did find a link between AH and psychotic diagnoses in the community (Ohayon, 2000).  This study involved telephone interviewing of over 13,000 persons aged 15 or older in the UK, Germany, and Italy, using an instrument designed to assess sleep-related disorders and experiences.  Non-sleep-related AH, found in 0.6% of the participants, were considerably less common than in the prior studies.  Participants who heard voices at least once a week had significantly higher levels of psychotic and depressive disorders, but the strongest associations were with anxiety and bipolar disorders.  However, this was not true for persons hearing daytime voices on a less than once a month basis.  The lower levels of population-based AH in this study may have been due to methodological issues, including the use of telephone interviews, an instrument not primarily designed to assess psychiatric disorders or symptoms, and an inflexible interview schedule.    


Three studies, also displayed in Table 2, have found auditory hallucinations to be commonly experienced by university students.  In Posey & Losch’s (1983-1984) study, close to 10% of 375 students reported hearing a “comforting or advising” voice and 5% reported conducting conversations with the voice(s).  They concluded that “clinical finding of hallucinations of a mild sort… should not be taken as suggestive of psychopathology” (p. 111). 


Barrett & Etheridge (1992) replicated Posey & Losch’s (1983-1984) study, with a larger sample (at the same university).  Between 6.1% and 12.6% of the sample endorsed hearing voices that were not (apparently) their thoughts, their name being called, or sleep-related. Barrett & Etheridge did not report on percentages of participants conversing with their voices.  Importantly, they also found that psychopathology, as rated by either the MMPI (Hathaway and McKinley, 1967) or the SCL-90-R (Derogatis, 1994) was unrelated to AH.  Echoing Posey & Losch, they concluded, “(R)eports of verbal hallucinations in the general population cannot reasonably be explained as the result of psychopathology” (Barrett & Etheridge, 1992, p. 385).         


Further evidence of the frequency of hallucinations in student populations was provided by a research study of over 600, primarily Israeli, university students, conducted by Barrett and Joseph Glicksohn (Glicksohn & Barrett, 2003).  Using a modification of the Barrett & Etheridge (1992) measure, they found between 2 and 9% of their sample endorsing AH consistent with those endorsed by students in previous studies.  But are these AH similar to those experienced by psychiatric patients, or can they be characterized as ‘pseudo-hallucinations’ and distinguished from the hallucinations of psychiatric patients?


Phenomenology and clinical utility

As was noted in the historical overview of this topic, the feature of AH most often put forth as central to distinguishing ‘true’ from ‘pseudo’ hallucinations is the perceived location of the voice.  While recent reviews have criticized this position, many in the psychiatric field continue to believe that external voices are characteristic of psychotic AH, and internal voices indicative of dissociative or ‘normal’ voices.  There have now been a number of studies which have addressed this issue, the results of which can be seen in Table 3.  While most of the studies include only psychiatric patients, two directly compare the AH of psychiatric patients with those of non-patients. 


Insert Table 3 about here



A cursory glance at the table offers little support for the clinical adage described above.  While between 31% and 61% of persons diagnosed schizophrenic hear exclusively external voices, 56% of non-schizophrenic psychiatric patients, 27% of those with dissociative disorders and 40% of non-patients do as well (the medical patients in Mott, Small & Anderson’s (1965) study are not considered here for reasons described in the table).  Exclusively internal voices are reported by 22-43% of those diagnosed schizophrenic, 31% of those with psychotic diagnoses other than schizophrenia, 33% of those with dissociative disorders and 47% of non-patients.  The most significant variation occurs in the percentage of psychiatric patients reported to hear both internal and external voices, which varies between 6% and 42%; however, it appears likely that at least some of this variation can be attributed to methodological issues.  


Thus, while the AH of psychiatric patients and non-patients do not appear to be different with regard to perceived location, there are differences in the relationship and reaction to the voices.  Honig et al (1998) reported that non-patients heard fewer

 ‘negative’ voices than both sets of patients (53% compared to 93% and 100%), but all groups heard high levels of positive voices.  The most significant differences related to responses to the voices – more than 3/4 of the patients were afraid of the voices, but none of the non-patients was.  Likewise, all of the patients, but only 20% of the non-patients (NP), reported that the voices disturbed their daily life, and all of the NP but only 7-12% of the patients felt that they could control the voices.   


Several studies have attempted to assess the experienced or rated characteristics of voices in psychiatric patients, including perceived location, and relate them to clinical variables.  Aggernaes (1972) found that AH were rated as highly similar to ‘real’ perceptions in every way except for ‘publicness’; that is, only a third of his primarily chronic psychiatric patients felt that their voices could be heard by others.  Oulis et al (1995) carefully examined internal AH and found, in contrast to Jaspers (1963/1913), that they were experienced as ‘real’ as external AH[6].  Yee et al (2005) came to similar conclusions.  However, in Oulis et al’s (1995) study, internal AH were rated as less hostile and were accompanied by significantly more insight than external AH. 


A rich phenomenological study of AH in 100 psychiatric patients was conducted by Nayani & David (1996).  As Nayani and David found more of the ‘newer’ psychiatric patients to experience external AH than the more ‘experienced’ patients, they inferred that AH might move from external to internal over time, with increasing voice engagement by the patient (consistent with Havens, 1962 and Romme et al, 1992).  This view was supported by their finding that internal AH were more grammatically and syntactically complex than external AH.  Perceived location was unrelated to diagnosis, but internal AH were associated with increased insight. 


In a recent cross-sectional study with a large sample (n=197), Copolov et al (2004) found few relationships between internal AH and other characteristics of the voices, other than increased insight.  Diagnosis, positive or negative content, and level of medication usage were all unrelated to perceived AH location.  Copolov et al’s conclusion was that the perceived location of AH lacked clinical utility.


However, as none of the above studies screened for dissociative disorders, it may be premature to conclude that perceived location of voice(s) is ineffective in predicting diagnosis, only at differentiating between the various psychotic diagnoses included.  Even so, Honig et al’s (1998) study, which did carefully compare persons diagnosed schizophrenic with those with dissociative disorders, found no significant differences in this regard.


A number of authors have commented on voice hearers’ relationships with their voices.  Nayani and David (1996) found that multiple voice hearers had distinct relationships with their voices, and appeared to know intimate details of their life.  Some consider such ‘omniscience’ to be an almost universal characteristic of voices (Chadwick & Birchwood, 1994).  Nayani and David (1996) also found that the voices frequently spoke with different accents from those of the patients, but used ones which were always personally and/or culturally relevant to them.  Others have noted strong linguistic and affective similarities to relationships found between persons in the ‘real world’ (Hoffman et al, 1994; Miller, O'Connor, & Di Pasquale, 1993); the extended duration of AH episodes – 30 minutes or more (Aggernaes, 1972; Oulis et al, 1995) – is also suggestive of this.  Indeed, Benjamin (1989) reported that all patients in his study had “meaningful, integrated and interpersonally coherent relationships with their voices” (p. 309).  Perhaps of most importance, Nayani and David (1996) found that patients who were unable (or unwilling?) to converse with their voices were considerably more distressed than patients who could dialogue with them.


Much of the above, including the findings of the perceived ‘omniscience’ of voices (Chadwick & Birchwood, 1994), are consistent with Jaynes’ thesis that hearing voices is a normal part of human evolutionary history, and that contemporary AH derive directly from voices of ‘gods’ heard millennia ago.  Jaynes believed that voices developed around the need to make novel decisions when human societies become more diversified and complex.  His linking of AH with decision-making is echoed by Nayani and David (1996), who conclude that AH “bear a strong resemblance to patterns of thought that are part of the normal experience of making decisions” (p. 184).  Such a view is also shared by Marius Romme and his colleagues in Maastricht, who attempt to determine early on, when working with a person who hears voices, what ‘problem’ the AH represents.  Even Jaynes’ speculation about the role of the right hemisphere in AH has received some support.  Anthony David, after reviewing recent neuroimaging work, concluded that AH may involve “right auditory regions which are implicated in the processing of the prosodic aspects of speech” (David, 1999, p. 101).  Clearly, Jaynes’ theory, while not without its flaws, goes a long way toward explaining why AH remain so common today, and appear similar whether heard by psychiatric patients, grieving partners, traumatized children or tribal elders.


AH and dissociative experiences

While AH are clearly not unique to persons diagnosed with dissociative disorders, is it possible that dissociation is the mechanism underlying all forms of AH?  It has long been acknowledged that dissociative symptoms are strongly related to traumatic experiences. Dissociative symptoms make up a core part of the diagnosis of Post-traumatic Stress Disorder (PTSD) and particularly Acute Stress Disorder (APA, 1994).  Traumatic or highly stressful experiences precede the development of AH in the vast majority of voice hearers (Heins, Gray, & Tennant, 1990; Romme & Escher, 1989) and childhood trauma appears more strongly related to AH than to other ‘psychotic’ symptoms (Ensink, 1992; Read & Argyle, 1999; Ross, Anderson, & Clark, 1994).  While these studies link trauma with dissociation or with AH, direct evidence relating psychotic symptoms in general, and AH in particular, to dissociative experiences has now been found.


Strong links between psychotic symptoms, including AH, and dissociative experiences in a wide range of clinical and non-patient populations have been demonstrated in a number of studies (reviewed in Moskowitz, Barker-Collo and Ellson, 2004) and psychotic symptoms, particularly those considered pathognomic for DSM-IV schizophrenia, are commonly found in persons with DID (Kluft, 1987; Ross et al, 1990).  Four studies have demonstrated robust links between AH and dissociative experiences.  In a population of non-psychotic adolescents, primarily from a residential treatment center, Altman, Collins & Mundy (1997) found that dissociative experiences, as measured by the Dissociative Experiences Scale (DES; Carlson & Putnam, 1986), strongly predicted AH, even after controlling for mood disturbance and schizotypal thought processes.  Similarly, Kilcommons & Morrison (2005) found scores on the DES (particularly the depersonalization subscale) to significantly predict AH in psychotic patients (even after controlling for severity of trauma), and Morrison & Peterson (in press) found DES scores to correlate .71 with predisposition to AH in adult non-patients.  Finally, Glicksohn & Barrett (2003), also looking at adult non-patients, found 25% shared variance between DES scores and measures of both AH and predisposition to AH.  All of the above authors concluded that dissociation, in one form or another, was a ‘predisposing factor’ for AH.


Predictors of psychotic symptoms and mental health system contact

If AH are indeed linked to dissociative experiences, as the evidence reported in the prior section indicates, what determines why some people, but not others, develop delusional explanations for these experiences and enter the mental health system?  A longitudinal study in the Netherlands, following children and adolescents positive for hearing voices over a three-year period, has gone some way toward addressing this question.  Sandra Escher, along with Marius Romme and their colleagues at the University of Maastricht, located 80 children and adolescents who acknowledged hearing voices when first contacted via media channels.  Most of the participants no longer heard any voices by the end of the three-year period (Escher et al, 2002a).  Significant variables which distinguished children who continued to hear voices from those who did not, along with factors determining who developed delusions and received mental health system contact, were sought.  Escher and colleagues found, consistent with Morrison’s (1998) model of voice persistence, that clinician-rated anxiety and depression were the strongest predictors of voice continuation; they were also the best predictor of who developed delusions and received professional mental health care  (Escher et al, 2002a, 2002b, 2004). Other factors predicting voice persistence included clinician-rated severity and frequency of AH at baseline, dissociation scores (children who continued to hear voices reported more dissociative experiences than children whose voices stopped), and lack of clear spatial and temporal triggers (Escher et al, 2002a). Significantly, neither mental health contact nor the development of delusions was related to voice continuation.


With regard to mental health system contact, Escher and colleagues found the voices of children who received care to be rated as more severe, and having more impact on their emotions and actions, than children who did not enter the mental health system.  Not surprisingly, this group also reported more ‘problem behavior’ than children who did not receive care (Escher et al, 2004).  In addition, delusions were associated more strongly with external than internal voices in the older children (i.e., 13 years or above) but not in the younger ones (Escher et al, 2002b).  Dissociation scores were unrelated to the development of delusions in this population (Escher et al, 2002b).  As noted above, they were also unrelated to mental health system contact, but did predict voice persistence (Escher et al, 2002a, 2004).


Finally, if AH appearing in persons designated psychotic do not, in fact, appear different from AH appearing in other persons, can they be treated similarly?


Clinical approaches to AH

It is becoming increasingly acknowledged that AH do not typically respond directly to psychiatric medications; if they are effective, it is argued that this occurs through decreasing the distress experienced by the voice hearer (Kapur, 2003).  This has also become the approach advocated for psychotherapeutic methods of dealing with AH (Chadwick, Birchwood & Trower, 1996; Kingdon & Turkington, 2005).  Such approaches do not attempt to address AH directly, but rather to deal with the responses to and assumptions about AH. 


However, some clinicians, such as Colin Ross (2004), are proposing that AH be dealt with in a manner similar to that long considered standard in the field of dissociative disorders.  As typical treatment of DID often involves direct therapeutic engagement with many if not all of the personality parts or alters (Putnam, 1989), might it not be possible for a therapist to directly engage with the AH heard by the voice hearer?  This proposal is not as radical as it might sound, as having the person directly communicate with his or her voices in the presence of the therapist is now being advocated in cases of PTSD in which AH are experienced (Brewin, 2003).


The second author of this article, working closely with Marius Romme and Sandra Escher, has extensive experience in engaging with voices in the manner described above.  Over the past 10 years, several dozen voice hearers in the vicinity of Maastricht in the Netherlands have allowed their voices to be directly worked with, in a variant of the Voice Dialogue approach (Stone and Winkelman, 1989).  In this approach, the therapist asks permission from the voice hearer to directly speak with his or her voices, and then proceeds to – quite respectfully – ‘interview’ the voice as one might a new acquaintance.  The timing and reasons for the voice coming into being, its relation to the person, and what it ‘wants’ are all explored.  This is repeated for all voices (that allow contact), with the person usually taking a different chair in the room for each voice.  The Maastricht approach also includes other psychosocial interventions.


Typical results can include a decrease in the perceived ‘destructiveness’ of the voices, including a transformation from ‘negative’ to ‘positive’ voices with increased understanding from the person, along with an increased capacity (and willingness) to dialogue with the voices.  No person has so far become more psychotic, and quality of life is sometimes substantially improved.  A key variable in the effectiveness of this treatment appears to be whether or not the person can (or believes they can) dialogue with their voices.  If they cannot, then the therapist also cannot work with the voices.  The following clinical vignette illustrates this approach to working with AH.


Case illustration

‘Karen’ presented for treatment shortly after leaving a psychiatric hospital where she had been living for the prior four years, struggling with hostile and command AH and suicidal ideation.  During her contact with the mental health system, she had been diagnosed (at various times) as suffering from schizophrenia or borderline personality disorder, and received antipsychotic and other medications which helped reduce her feelings of anxiety and fear but did not directly impact on the voices.  Karen had heard voices (from outside her head) since she was a child, but functioned well – working as a secretary – until around 20 years old when she joined a religious sect. When the sect members were informed about Karen’s voices, she was told that they were ‘instruments of the devil’ and instructed to get rid of them.  From that point on, the four male voices became more negative and disturbing, commenting on Karen’s behavior and her thinking and telling her to kill herself[7].  The only ways she had of coping with them were to ignore them, go to sleep, isolate herself or argue with them when she could no longer bear it.  None of these strategies worked particularly well. 


At the first session, the therapist asked Karen for permission to talk to the voices.  She agreed and each voice was interviewed in turn. They all told the same story.  They reported that they came into Karen’s life when she was about four, when she was feeling quite lonely (and around the time, it later emerged, that she had been sexually abused), and served as her companions.  Their job was to help Karen feel less lonely and overcome difficult moments.  Before she joined the sect, Karen accepted the voices and they felt acknowledged. However, afterwards everything changed.  As Karen’s attitude to the voices changed, as she began to reject them, so they too rejected her – becoming very negative and telling her to kill herself.  Her initial admission to the hospital was precipitated by her finally giving in to the voices and attempting to kill herself.


In speaking with the therapist, the voices requested that Karen again accept them, as she had in the beginning.  Karen, who was able to ‘overhear’ the therapist talking with the voices, agreed and began setting aside time in the evening to engage with the voices.  By the next session, two of the voices had disappeared, and those that remained were easy to ignore.  The voices became more positive, no longer criticizing Karen or exhorting her to kill herself.  After a few more, largely supportive, sessions the therapy was completed.  When she contacted the therapist four years later, she reported that she was happy (despite being divorced) and was living with her two young children in a new city.  She had not been psychiatrically hospitalized since the treatment, and remained on a very low dose of antipsychotic medication.  Karen now heard only one voice, but it was one she liked to talk to it because it was positive and supportive.  


Discussion:  AH as non-pathological dissociative experiences

On balance, the above historical, clinical, and research evidence does not suggest that AH appearing in psychotic disorders can, on the basis of their experienced characteristics, be distinguished from AH reported by persons with no mental health system contact.  Instead, a diagnosis of schizophrenia or another psychotic disorder and psychiatric treatment in persons with AH appears to be most strongly related to the level of depression and anxiety those voices foster (Close & Garety, 1998; Escher et al, 2002a, 2002b, 2004; Morrison, 1998).  Persons who are able to cope well with the voices, particularly those that are able to engage the voices and reach “some sort of peaceful accommodation and acceptance of the voice as ‘part of me’” (Romme & Escher, 1989, p. 44), either do not enter the mental health system at all or exit the system successfully (Nayani & David, 1996).  In this way, AH can be conceptualized as behaving similarly to a trauma (an argument that has previously been made, e.g., Frame & Morrison, 2001; Meyer et al, 1999) in that it is the person’s response to the trauma or AH, and in particular the extent to which they can incorporate it into their existing schemas (or successfully modify them) that determines whether they will draw the attention of a mental health professional[8].


While others have previously argued that AH should be considered essentially ‘normal’ phenomena (Morrison, 1998; Romme & Escher, 1989), no one, as far as we are aware, has taken this position to its logical conclusion:  that AH should no longer be considered psychotic symptoms, but simply one of a number of experiences (including various traumas) that can, under certain circumstances, induce delusions and lead to entry into the mental health system.  Further, we believe that the evidence is strong enough to argue that, not only are voices non-pathological phenomena, but that they are dissociative in nature.  As such, our position goes further than that of authors who have argued that dissociative experiences provide a predisposition or diathesis for AH (Allen & Coyne, 1995; Glicksohn, 2004); it is closer to the positions of Merckelbach et al (2000) and Watson (2001), who question whether the concepts of dissociation and psychosis are, in fact, distinguishable.


It is almost axiomatic at this point to assert that AH result from the ‘misattribution’ of internal experiences to external sources (Bentall, 1990; Morrison, 1998).  Various cognitive mechanisms have been posited to explain such mistakes (drive to maintain self-esteem (Bentall, 1990); cognitive dissonance (Morrison, 1998)), but such explanations seem unsatisfactory.  Perhaps they can explain why some thoughts are experienced as ‘intrusive’ or ‘inserted’ (i.e., not belonging to me), but they are less successful in explaining how such extruded thoughts become ‘heard’.  Further, we would argue that dissociative mechanisms are needed to explain the central experience of AH – that of an ‘other’ relating to one’s ‘self’.  It is hard to see how one can have a relationship with an extruded or projected thought; engaging in a relationship is consistently how the experience of AH is described.


However, conceiving of AH as dissociative phenomena does not solve the enigma of voices.  Several questions about diagnostic and clinical utility remain, calling for future research: 

1)    Is perceived location completely irrelevant?  Despite all of the research reported in this paper, more thorough investigation needs to be done before this question can be put to rest.  Do AH really move from external to internal with successful coping or treatment, as several have suggested (Havens, 1962; Nayani & David, 1996; Romme et al, 1992)?  What about voice hearers who have both internal and external voices?  Are the external voices more ego-dystonic than the internal ones?

2)    Is the public/private dimension more important than the internal/external dimension?  While these dimensions often overlap, they are not the same.  A confusion (or diffusion) between public and private worlds seems to bespeak a high level of pathology.  Most people who hear voices, internal or external, do not believe they can be heard by others (Aggernaes, 1972; indeed, for many years, the standard clinical inquiry about AH was, “Do you hear voices other people can’t hear?”), and some people believe that their thoughts are projected and can be picked up by others.  Perhaps significantly, this symptom of ‘thought projection’ or ‘thought broadcasting’ (a delusion by definition) is very rare in dissociative disorders and borderline personality disorder, in contrast to other Schneiderian symptoms such as thought insertion and thought withdrawal (which can simply be accurate descriptions of unusual internal experiences; Kluft, 1987; Ross et al, 1990; Yee et al, 2005), but is quite common in persons diagnosed with schizophrenia (Peralta & Cuesta, 1999; Tandon & Greden, 1987).  This suggests that the public/private dimension might have diagnostic implications, and could be a more fruitful area to explore than the internal/external dimension.

3)    What determines why some people can engage with their voices and others cannot – a factor seemingly so important to their recovery?  Is it that persons cannot engage or simply are afraid to?  Typically, benevolent seeming voices are engaged with and ‘malevolent’ voices resisted (Chadwick & Birchwood, 1994; Close & Garety, 1998), but this doesn’t seem to be the whole picture.  As seen in the clinical vignette, voices that are initially benevolent can turn malevolent when they are ignored.  Further, it may be that the ability to engage with voices is a skill, related perhaps to dissociative capacities.  Recall that Escher et al (2002a) found that dissociation correlated with voice persistence, but was unrelated to psychotic diagnosis or mental health treatment.  Is it necessary to ‘cultivate’ the voice, to nurture it in a sense, in order to engage with it?  Here is what Mark Vonnegut (probably highly dissociative as he experienced amnesic episodes, uncommon in schizophrenia; Steinberg, 1995) had to say about his initial AH experiences: 

At first I’d had to strain to hear or understand them.  They were soft and working with some pretty tricky codes.  Snap-crackle-pops, the sound of the wind with blinking lights and horns for punctuation.  I broke the code and somehow was able to internalize it to the point where it was just like hearing words (Vonnegut, 1975, p. 106). 

Over time, his AH experiences was transformed.

The voices weren’t much fun in the beginning.  Part of it was simply my being uncomfortable about hearing voices no matter what they had to say, but the early voices were mostly bearers of bad news.  Besides, they didn’t seem to like me much and there was no way I could talk back to them… But later the voices could be very pleasant.  They’d often be the voice of someone I loved, and even if they weren’t, I could talk too, asking questions about this or that and getting reasonable answers.  They were very important messages that had to get through somehow.  More orthodox channels like phone and mail had broken down. (Vonnegut, 1975, p. 106). 


Finally, where does de-coupling the concept of AH from psychosis leave the latter concept?  We would suggest – in tatters.  The DSM-IV does not have an adequate diagnosis of psychosis; all it does is refer back to symptoms considered psychotic.  Indeed, it offers three different definitions, progressively broader: (1) delusions and prominent hallucinations experienced without insight, (2) delusions and prominent hallucinations (regardless of insight), and (3) the four Criterion A positive symptoms of schizophrenia (i.e., delusions, hallucinations, disorganized speech, and grossly disorganized behavior or catatonic behavior). 


The arguments for considering disorganized speech and disorganized/catatonic behavior are weak and have been discussed elsewhere (Moskowitz et al, submitted).  Briefly, either symptom may be due to speech or behavior that is relevant to a context of which the listener/observer is unaware (and as such, is not disorganized); one only has to consider the speech and behavior of a person experiencing a trauma flashback, and imagine how their behavior would be interpreted if the observer was unaware of their history, to see this.


Thus, we are left with only delusions as the paradigmatic psychotic symptom.  As such, it is difficult to see what the concept of psychosis adds to the concept of delusions.  Perhaps this concept should simply be dispensed with, and persons experiencing delusions should be referred to simply as – delusional.  The questionable validity of the concept of psychosis has obvious implications for the validity of the current concept of schizophrenia, as the two concepts are closely wedded in the DSM-IV.  That, however, is beyond the purview of this article.


In conclusion, the evidence supports our view that AH: a) are best conceptualized as dissociative experiences which appear in individuals predisposed, for reasons not yet clear, to hear voices when under stress, b) require ‘cultivation’ or ‘nurturing’ to make their meaning clear, c) resolve when appropriately engaged with by the individual, possibly moving from externally- to internally-perceived in the process, and d) are consistent with Jaynes’ evolutionary theory of the bicameral mind.  Accordingly, we call on our mental health colleagues to no longer designate AH as a psychotic symptom, to conduct research on the remaining areas of confusion in AH (such as determining those factors influencing engagement with voices), and to explore alternative ways of working with the voices heard by many of us. 



Aggernaes, A. (1972). The experienced reality of hallucinations and other psychological phenomena. Acta Psychiatrica Scandinavica, 45, 220-238.

Allen, J. G., & Coyne, L. (1995). Dissociation and vulnerability to psychotic experience: The Dissociative Experiences Scale and the MMPI-2. Journal of Nervous & Mental Disease, 183(10), 615-622.

Altman, H., Collins, M., & Mundy, P. (1997). Subclinical hallucinations and delusions in nonpsychotic adolescents. Journal of Child Psychology & Psychiatry & Allied Disciplines, 38(4), 413-420.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association.

Barrett, T. R., & Etheridge, J. B. (1992). Verbal Hallucinations in Normals .1. People Who Hear Voices. Applied Cognitive Psychology, 6(5), 379-387.

Benjamin, L. (1989). Is chronicity a function of the relationship between the person and the auditory hallucination? Schizophrenia Bulletin, 15, 291-310.

Bentall, R. P. (1990). The illusion of reality: a review and integration of psychological research on hallucinations. Psychological Bulletin, 107(1), 82-95.

Berrios, G. E., & Dening, T. R. (1996). Pseudohallucinations: A conceptual history. Psychological Medicine, 26(4), 753-763.

Brewin, C. R. (2003). Posttraumatic stress disorder:  Malady or myth. New Haven, Connecticut: Yale University Press.

Carlson, E. B., & Putnam, F. W. (1986). Development, reliability, and validity of a dissociation scale. Journal of Nervous & Mental Disease, 174, 727-733.

Chadwick, P., & Birchwood, M. (1994). The omnipotence of voices: A cognitive approach to auditory hallucinations. British Journal of Psychiatry, 164, 190-201.

Chadwick, P., Birchwood, M., & Trower, P. (1996). Cognitive therapy for delusions, voices, and paranoia. Chichester, U.K.: John Wiley and Sons.

Close, H., & Garety, P. (1998). Cognitive assessment of voices:  Further developments in understanding the emotional impact of voices.British Journal of Clinical Psychology, 37, 173-188.

Copolov, D. L., Trauer, T., & Mackinnon, A. (2004). On the non-significance of internal versus external auditory hallucinations.Schizophrenia Research, 69, 1-6.

David, A. S. (1999). Auditory hallucinations: phenomenology, neuropsychology and neuroimaging update. Acta Psychiatrica Scandinavica, 99, 95-104.

Dening, T. R., & Berrios, G. E. (1996). The enigma of pseudohallucinations: Current meanings and usage. Psychopathology, 29, 27-34.

Derogatis, L. R. (1994). Symptom Checklist-90-R (SCL-90-R) administration, scoring, and procedures manual (3rd ed.). Minneapolis: National Computer Systems.

Dhossche, D., Ferdinand, R., van der Ende, J., Hofstra, M. B., & Verhulst, F. (2002). Diagnostic outcome of self-reported hallucinations in a community sample of adolescents. Psychological Medicine, 32(4), 619-627.

Ensink, B. J. (1992). Confusing realities: A study on child sexual abuse and psychiatric symptoms. Amsterdam: VU Press.

Escher, S., Romme, M., Buiks, A., Delespaul, P., & van Os, J. (2002a). Independent course of childhood auditory hallucinations: A sequential 3-year follow up study. British Journal of Psychiatry, 181 (suppl. 43), s10-s18.

Escher, S., Romme, M., Buiks, A., Delespaul, P., & van Os, J. (2002b). Formation of delusional ideation in adolescents hearing voices: A prospective study. American Journal of Medical Genetics, 114, 913-920.

Escher, A., Morris, M., Buiks, A., Delespaul, P., van Os, J., & Romme, M. (2004). Determinants of outcome in the pathways through care for children hearing voices. International Journal of Social Welfare, 13, 208-222.

Esquirol, J. E. D. (1832). Sur les illusions des sens chez alienés [On the sensory illusions of the insane]. Archives Generales de Medicine, 2, 5-23.

Frame, L., & Morrison, A. P. (2001). Causes of posttraumatic stress disorder in psychotic patients. Archives of General Psychiatry, 58, 305-306.

Glicksohn, J. (2004). Absorption, hallucinations, and the continuum hypothesis. Behavioral & Brain Sciences, 27(6), 793-794.

Glicksohn, J., & Barrett, T. R. (2003). Absorption and hallucinatory experience. Applied Cognitive Psychology, 17(7), 833-849.

Hare, E. H. (1973). A short note on pseudo-hallucinations. British Journal of Psychiatry, 122, 469-476.

Hathaway, S., & McKinley, J. (1967). Minnesota Multiphasic Personality Inventory manual revised. New York: The Psychological Corporation.

Havens, L. (1962). The placement and movement of hallucinations in space: Phenomenology and theory. International Journal of Psychoanalysis, 43(426-435).

Heins, T., Gray, A., & Tennant, M. (1990). Persisting hallucinations following childhood sexual abuse. Australian & New Zealand Journal of Psychiatry, 24, 561-565.

Hoffman, R. E., Oates, E., Hafner, J., Hustig, H. H., & McGlashan, T. H. (1994). Semantic organization of hallucinated 'voices' in schizophrenia. American Journal of Psychiatry, 151, 1229-1230.

Honig, A., Romme, M. A. J., Ensink, B. J., Escher, S. D. M. A. C., Pennings, M. H. A., & deVries, M. W. (1998). Auditory hallucinations: A comparison between patients and nonpatients. Journal of Nervous & Mental Disease, 186(10), 646-651.

Jaspers, K. (1963). General Psychopathology (J. Hoenig & M. W. Hamilton, Trans.). Manchester, UK: Manchester University Press. (Original publication date 1913).

Jaynes, J. (1976). The Origin of Consciousness in the Breakdown of the Bicameral Mind. Boston: Houghton Mifflin.

Johns, L. C., Nazroo, J. Y., Bebbington, P., & Kuipers, E. (2002). Occurrence of hallucinatory experiences in a community sample and ethnic variations. British Journal of Psychiatry, 180, 174-178.

Kapur, S. (2003). Psychosis as a state of aberrant salience: A framework linking biology, phenomenology, and pharmacology in schizophrenia. Am J Psychiatry, 160(1), 13-23.

Kilcommons, A. M., & Morrison, A. P. (2005). Relationships between trauma and psychosis: an exploration of cognitive and dissociative factors. Acta Psychiatrica Scandinavica, 112, 351-359.

Kingdon, D. G., & Turkington, D. (2005). Cognitive therapy of schizophrenia. New York: Guilford Press.

Kluft, R. P. (1987). First-rank symptoms as a diagnostic clue to multiple personality disorder. American Journal of Psychiatry, 144, 293-298.

Merckelbach, H., Rassin, E., & Muris, P. (2000). Dissociation, schizotypy, and fantasy proneness in undergraduate students. Journal of Nervous & Mental Disease, 188(7), 428-431.

Meyer, H., Taiminen, T., Vuori, T., Aijala, A., & Helenius, H. (1999). Posttraumatic stress disorder symptoms related to psychosis and acute involuntary hospitalization in schizophrenic and delusional patients. Journal of Nervous & Mental Disease, 187(6), 343-352.

Miller, L. J., O'Connor, E., & Di Pasquale, T. (1993). Patients' attititudes toward hallucinations. American Journal of Psychiatry, 150, 584-588.

Morrison, A. P. (1998). A cognitive analysis of the maintenance of auditory hallucinations: Are voices to schizophrenia what bodily sensations are to panic? Behavioural and Cognitive Psychotherapy, 26, 289-302.

Morrison, A. P., & Peterson, T. (in press). Trauma, metacognition and predisposition to hallucinations in non-patients. Behavioural and Cognitive Psychotherapy.

Moskowitz, A., Barker-Collo, S., & Ellson, L. (2004). Replication of dissociation-psychoticism link in New Zealand students and inmates. Journal of Nervous & Mental Disease, 193, 722-727.

Moskowitz, A. K., Read, J., Farrelly, S., Rudegeair, T., & Williams, O. (submitted for publication). Are psychotic symptoms traumatic in origin and dissociative in kind? In P. F. Dell & J. O'Neil (Eds.), Dissociation and the Dissociative Disorders:  DSM-V and Beyond.

Mott, R. H., Small, I. F., & Anderson, J. M. (1965). A comparative study of hallucinations. Archives of General Psychiatry, 12(6), 595-601.

Nayani, T. H., & David, A. S. (1996). The auditory hallucination: a phenomenological survey. Psychological Medicine, 26(1), 177-189.

Nurcombe, B., Mitchell, W., Begtrup, R., Tramontana, M., LaBarbera, J., & Pruitt, J. (1996). Dissociative hallucinosis and allied conditions. In F. R. Volkmar (Ed.), Psychoses and pervasive developmental disorders in childhood and adolescence. Washington, DC: American Psychiatric Press.

Ohayon, M. M. (2000). Prevalence of hallucinations and their pathological associations in the general population. Psychiatry Research, 97, 153-164.

Oulis, P. G., Mavreas, V. G., Mamounas, J. M., & Stefanis, C. N. (1995). Clinical characteristics of auditory hallucinations. Acta Psychiatrica Scandinavica, 92(2), 97-102.

Peralta, V., & Cuesta, M. J. (1999). Diagnostic significance of Schneider's first-rank symptoms in schizophrenia. British Journal of Psychiatry, 174, 243-248.

Posey, T. B., & Losch, M. E. (1983-1984). Auditory hallucinations of hearing voices in 375 normal subjects. Imagination, Cognition and Personality, 3(2), 99-113.

Putnam, F. W. (1989). Diagnosis and treatment of Multiple Personality Disorder. New York: Guilford Pres.

Read, J., & Argyle, N. (1999). Hallucinations, delusions, and thought disorder among adult psychiatric inpatients with a history of child abuse. Psychiatric Services, 50(11), 1467-1472.

Romme, M. A. J., & Escher, D. M. A. C. (1989). Hearing voices. Schizophrenia Bulletin, 15, 40-47.

Romme, M. A. J., Honig, A., Noorthoorn, E. O., & Escher, A. D. M. A. C. (1992). Coping with hearing voices: An emancipatory approach. British Journal of Psychiatry, 161, 99-103.

Ross, C. A. (2004). Schizophrenia:  Innovations in diagnosis and treatment. New York: Haworth Press.

Ross, C. A., Anderson, G., & Clark, P. (1994). Childhood abuse and the positive symptoms of schizophrenia. Hospital & Community Psychiatry., 45(5), 489-491.

Ross, C. A., Heber, S., Norton, G. R., & Anderson, G. (1989). Differences between multiple personality disorder and other diagnostic groups on structured interview. Journal of Nervous and Mental Disease, 177(8), 487-491.

Ross, C. A., Miller, S. D., Reagor, P., Bjornson, L., Fraser, G. A., & Anderson, G. (1990). Schneiderian symptoms in multiple personality disorder and schizophrenia. Comprehensive Psychiatry, 31(2), 111-118.

Sarbin, T. R., & Juhasz, J. B. (1967). The historical background of the concept of hallucination. Journal of the History of the Behavioral Sciences, 5, 339-358.

Sedman, G. (1966a). A comparative study of pseudohallucinations, imagery and true hallucinations. British Journal of Psychiatry, 112(482), 9-17.

Sedman, G. (1966b). "Inner voices."  Phenomenological and clinical aspects. British Journal of Psychiatry, 112, 485-490.

Sedman, G. (1966c). A phenomenological study of pseudohallucinations and related experiences. Acta Psychiatrica Scandinavica, 42(1), 35-70.

Schneider, K. (1959). Clinical Psychopathology (5th ed.). New York: Grune and Stratton.

Sidgewick, H., Johnson, A., Myers, F. W. H., et al. (1894). Report on the census of hallucinations. Proceedings of the Society for Psychical Research, 26, 259-394.

Spitzer, M. (1990). On defining delusions. Comprehensive Psychiatry, 31(5), 377-397.

Steinberg, M. (1995). Handbook for the assessment of dissociation: A clinical guide. Washington, DC: American Psychiatric Press.

Stone, H., & Winkelman, S. (1989). Embracing our selves: The Voice Dialogue manual. Novato, California: Nataraj Publishing.

Tandon, R., & Greden, J. F. (1987). Schneiderian first rank symptoms: Reconfirmation of high specificity for schizophrenia. Acta Psychiatrica Scandinavica, 75, 392-396.

Taylor, F. K. (1981). On pseudo-hallucinations. Psychological Medicine, 11, 265-271.

Tien, A. Y. (1991). Distributions of hallucinations in the population. Social Psychiatry and Psychiatric Epidemiology, 26, 287-292.

Van der Zwaard, R., & Polak, M. A. (2001). Pseudohallucinations: a pseudoconcept? A review of the validity of the concept, related to associate symptomatology. Comprehensive Psychiatry., 42(1), 42-50.

Van Os, J., Hanssen, M., Bijl, R. V., & Ravelli, A. (2000). Strauss (1969) revisited:  A psychosis continuum in the general population?Schizophrenia Research, 45, 11-20.

Vonnegut, M. (1975). The Eden Express. New York: Praeger.

Yee, L., Korner, A. K., McSwiggan, S., Meares, R. A., & Stevenson, J. (2005). Persistant hallucinosis in borderline personality disorder.Comprehensive Psychiatry, 46, 147-154.

Watson, D. (2001). Dissociations of the night: Individual differences in sleep-related experiences and their relation to dissociation and schizotypy. Journal of Abnormal Psychology, 110(4).

Wiles, N., Sammit, S., Bebbington, P., Singleton, N., Meltzer, H., & Lewis, G. (2006). Self-reported psychotic symptoms in the general population. British Journal of Psychiatry, 188, 519-526.

World Health Organization. (1993). The ICD-10 Classification of Mental and Behavioural Disorders.  Diagnostic Criteria for Research. Geneva: WHO.


[1] Technically, the term auditory hallucinations refers to any sound heard in the absence of appropriate external stimuli, not only voices.  For example, during post-traumatic flashbacks, a wide range of sounds related to a trauma, such as police sirens, gunshots, etc., may be heard.  However, for the purposes of this article, auditory hallucinations will be used exclusively to refer to the hearing of voices.  While voices are sometimes referred to as ‘verbal’ hallucinations, the term auditory hallucinations is more familiar and as such is preferred here.

[2] We do not take a position in this paper as to whether AH may, under some circumstances, represent a ‘genuine’ spiritual experience.  Even if that is allowed, however, adequate means to distinguish such experiences from AH better explained by psychological (dissociative) mechanisms remains to be established.  In other words, at least some AH interpreted by the voice hearer as spiritual may actually derive from a disowned aspect of themselves.  At this point, we cannot adequately distinguish between the two on the basis of reported experience.

[3] The term ‘pseudo-hallucination’ was coined in 1868 by Hagen to refer to hallucinations with less sensory ‘fullness’ than perceptions (or ‘true’ hallucinations), which derived from an ‘organ of apperception’, and which were due to the spontaneous activity of memory (Berrios & Denning, 1996).  Almost all theorists have considered ‘pseudo-hallucinations’ to be essentially normal phenomena.

[4] Examples given of bizarre delusions are beliefs of loss of control of mind or body, and include delusions of control, and delusions of thought insertion and thought withdrawal (APA, 1994).  Both of these sets of delusions, along with voices commenting and conversing, also delimit three of the four pathognomic symptoms of ICD-10 Schizophrenia (WHO, 1993).  Some question whether these should be called delusions at all, as opposed to adequate descriptions of unusual/anomalous experiences (Spitzer, 1990).  All are common in DID, where they are typically understood as deriving from the influence of one part of the personality on another (Kluft, 1987; Ross et al., 1989; Ross et al., 1990). 

[5] A seventh, Wiles et al (2006), is not reported here because incidence and not prevalence data were used.  Of interest, however, is that just over 5% of almost 2000 persons, without a psychiatric history and not meeting criteria for a psychotic disorder, reported at least one of four narrowly defined psychotic symptoms (thought insertion, paranoia, ‘strange experiences’ and hallucinations; Wiles et al, 2006). 

[6] Oulis et al (1995) reported that patients had little difficulty in identifying the location from which their voice(s) appeared to emanate, a finding confirmed by most other researchers.

[7] As Karen also heard two or more voices conversing with each other, her voices fulfilled both of the pathognomic AH criteria for DSM-IV schizophrenia (i.e., voices commenting and voices conversing).

[8] Indeed, Morrison’s (1998) model of how ‘normal’ AH lead to psychopathology, in which he argues that AH may be ‘misinterpreted’ as “threatening the physical or psychological integrity of the individual,” leading to decreased mood, hyperarousal and hypervigilance, along with avoidance behaviors (p. 296), almost exactly mirrors current conceptions of PTSD.