Many people in the Western world consider the experience of hearing voices to be abnormal and symptomatic of psychosis. Psychotic disorders such as schizophrenia are characterized by a break with reality. Symptoms include auditory hallucinations, (hearing voices that others cannot hear) and delusions (holding false beliefs, even in the face of incontrovertible evidence to the contrary). These disorders are believed to be caused by genetic and environmental risk factors and are often treated with a combination of antipsychotic medication and cognitive-behavioural therapy. Yet, in some cultures, hearing voices is seen as a gift. As a result, prejudice and stigmatization are not so prevalent and people are less afraid to share their experiences.

The sample

The sample comprised 31 women and 29 men from California, in the United States, Chennai, India and Accra, Ghana. The mean age of the first two groups was 42; the Ghanaians were significantly younger, (mean age of 32). The Ghanaians were all resident in a psychiatric hospital, whereas the other two groups combined both inpatients and outpatients.  All participants had experienced at least two symptoms of schizophrenia for at least one month but had shown signs of the disorder for at least six months. Most had been symptomatic for many years, and their conditions had caused significant impairment in their daily lives.

Although some of the participants were drug-users, the researchers were careful to ensure that all people included in the final sample had heard voices prior to drug-use.

The interviews

Interviews were mainly conducted in English if possible, otherwise, in the participants’ native language. The researchers carried out a structured interview. All participants were asked about the number, frequency, and familiarity of the voices, whether they engaged the voices in conversation and what the voices said. They were also asked about any distress caused by the voices and whether the voices were ever a positive influence in their lives. Finally, they were asked about what they thought was the cause of their experiences.

In all three cultures, participants experienced both good and bad voices, had conversations with the voices, heard whispering, heard the voice of God, and said they hated the voices and experienced them as an ‘assault’. There were also some significant cross-cultural differences.

The nature of the voices

In Chennai and Accra, the voices were much more likely to be positive, whereas in the US, no one reported only positive voices, although 50% said there was a positive dimension.

In the US, voices were described as violent, harsh, insulting, and hated and 70% said the voices told them to hurt other people or themselves. They were seen as an intrusion into the individual's private world and were deeply distressing. This was rare in the non-US samples, who found the lack of control over the voices less troubling.

In Chennai, voices were described as playful, entertaining, interesting, and enjoyable, although this was not always the case. Even when the voices were cruel and unkind, people seemed loath to describe them negatively, one woman said “it (the voice) teaches me what I don’t know’.

The familiarity of the voices

In the non-US samples, the voices were often familiar, e.g. a spouse, parent, sibling, or neighbour, or a human-like spirit known to the participant; there was a sense of connection/relationship, even if this was not always positive. In Chennai, participants described hearing relatives giving guidance, but also scolding them.

In the US, only 10% recognized the voices (e.g. victims of childhood sexual abuse who heard the voice of their abuser).

Speaking with God

In the US 25% reported hearing God speak, although 75% were religious. In Accra, 80% said they literally heard God’s voice. God was said to guide and protect them. God also told them to ignore the bad voices. One woman said his voice was quieter than the demons who also spoke to her, but she was guided by God.

Causes of the voices

The US sample was more likely to attribute the voices to a brain disease caused by genes or trauma and 85% used diagnostic labels including ‘schizophrenia’ and/or ‘schizoaffective disorder’; they talked about their voices in terms of a disrupted relationship between their thoughts and their mind, e.g. ‘I don’t think there’s anything there or anything. I think it’s just the way my mind works.’ In contrast, only 20% of the Chennai sample used these labels. In Accra some understood the voices as related to psychiatric illness, although diagnostic labels were rarely mentioned. Non-Western participants interpreted the voices as relationships as opposed to a sign of ‘a violated mind’.

The study concludes that the harsh, violent voices that are a common aspect of the Western experience of schizophrenia are not inevitable and that if people can be taught to engage with and connect with their voices, they may be able to modify the tone and content of what is heard, leading to a less frightening experience.

The study used structured interviews which meant all participants were asked the same questions and the responses could be directly compared.

Luhrmann controlled for religiosity and urban-dwelling both of which may have affected the experiences of hearing voices. San Mateo in California was said to maybe be less chaotic than Accra and Chennai, yet all three are densely populated urban areas. Furthermore, most of the participants were religious, so the more positive experiences of those in the non-US samples could not be attributed simply to increased religiosity.

As the interviews were digitally recorded, Luhrmann was able to transcribe the interviews word-for-word, increasing the credibility of the final transcripts and ensuring that all the data was available for analysis. She also used competent speakers of the native languages (e.g. Tamil and Twi) of the participants, again increasing credibility as the analysis was based on the actual meaning conveyed by the participants.

Luhrmann also had access to the charts and clinical observations for the interviewees. As she was working with local psychiatrists, she was able to employ method triangulation to cross-reference the credibility of what her interviewees were telling the researchers.

Because the research is that as the data was obtained via face-to-face interviews, the extent to which social desirability bias affects the data is unclear. It appears that the non-Western participants did not want to speak negatively about their voices, which were often identified as members of their family and community. This meant cultural norms around in-group loyalty may have stopped the participants from speaking freely about their experiences.

A further limitation is that the ethnicity of the US sample is not stated and therefore it is unclear how homogenous this group was. Given the nature of the study, it would have been interesting to know more about the cultural identities of these individuals.