Despite decades of research and the development of numerous psychiatric medications, widespread suffering continues among people who hear voices, particularly when voices are uncontrollable, dominating or frightening (Daalman et al., 2011). While responses to voice hearing vary cross-culturally (Luhrmann et al., 2015; National Institute for Health and Care Excellence [NICE], 2014), in the US, first-line treatment is typically medication, with the goal of suppressing or stopping voices (Sommer et al., 2012). Many voice hearers, however, find medication ineffective in reducing their distress or intolerable due to its adverse effects (Lepping et al., 2011). Although psychological interventions like cognitive behavioral therapy and mindfulness training are beginning to be offered more widely for voice hearers (Styron et al., 2017), meta-analytic studies suggest that they are only moderately effective at reducing voice-related distress (Smailes et al., 2015).
Current literature suggests that voice hearing is phenomenologically diverse (Woods et al., 2015), and numerous individual and group therapeutic interventions have been developed to address various aspects of the experience (Ruddle et al., 2011; Styron et al., 2017). The peer-support groups established by advocates of the Hearing Voices approach represent an important alternative, allowing voice hearers to engage in an in-depth exploration of their experiences while receiving support from others coping with similar challenges, toward the goal of “living with voices” in more positive ways (Dillon & Hornstein, 2013; Hornstein, 2018; Romme et al., 2009; Styron et al., 2017). Even though many members of these Hearing Voices Groups (HVGs) report profound life changes as a result of their participation, and HVGs have now spread to 30 countries on five continents (Corstens et al., 2014), these groups have been little studied (Branitsky, Longden & Corstens, in press).
Previous literature suggests numerous distinct features and benefits of attending HVGs. Being able to share openly with peers without fear of consequences enables participants to form social connections, learn new ways of understanding and coping with voices, and feel less distressed by their experiences (Beavan et al., 2017; Dos Santos & Beavan, 2015; Longden et al., 2018; Oakland & Berry, 2015; Payne et al., 2017). As a result, participants report feeling more empowered and self-confident (Longden et al., 2018; Oakland & Berry, 2015) and require fewer hospital admissions and crisis interventions (Beavan et al., 2017). Other interventions derived from the Hearing Voices approach – like Maastricht Interviewing and creating a Voice Construct – have also been associated with reduced voice-related distress and an ability to connect voices to previous life events (Schnackenberg et al., 2018).
There are, however, significant methodological challenges in studying HVGs, one reason for the limited research in this area. Individuals do not attend a group for a fixed period, so a standard baseline-treatment-outcome design cannot be used (e.g. as in evaluating a 10-session CBT group). People come to HVGs for many different reasons and with diverse needs, so simple measures of satisfaction are inadequate to assess effectiveness. Further, since HVGs were explicitly developed as an alternative to interventions focused on voice cessation, their success cannot be measured by the extent of “symptom reduction” or “absence of hallucination.” We therefore sought a new way of studying HVGs that could do justice to their complexity and distinctiveness.
Understanding precisely how HVGs work is crucial, since the current lack of knowledge about their mechanisms of action contributes to the skepticism many professionals have about these groups. This is particularly true in the US, where HVGs are spreading rapidly but often dismissed as not “evidence-based.” Accordingly, we focused on identifying the processes within HVGs that enable significant transformation, i.e. how exactly do these groups function so as to be seen as effective by so many voice hearers? Further, we asked: What specific features distinguish HVGs from other kinds of support or clinical groups?
Data collection, participant recruitment and sample characteristics
To answer these questions, we recruited a national sample of HVG members across the US, to whom we administered a detailed, mostly open-ended questionnaire inquiring into voice-hearing history, experiences in HVGs, comparison of HVGs with other interventions and life changes (if any) outside the group as a result of participation. A subset of the sample also participated in follow-up interviews to enable further insight into the nuances of their experiences. All materials and methods were designed collaboratively by a team of researchers, voice hearers and clinicians, who worked together at every step to ensure that a full range of HVG experiences would be included. Because previous research has focused largely on experiences within groups and not attended to the distinctive and highly varied nature of members’ voice-hearing experiences themselves, we deliberately included both questionnaire and interview items that would elicit information about participants’ relationships to their voices and their voice-hearing history more generally, on the view that the specific nature of these experiences would inevitably inform the way each individual sought to understand and manage what was happening to them.
Participants were recruited from across the US (where no previous studies had been done). Facilitators of all groups approved for listing on the HVN-USA website were contacted, provided with both electronic and hardcopy access to the questionnaire and asked to inform their members about the research. We emphasized that they should not pressure anyone to participate nor administer the questionnaire during regular group meetings. However, in instances where members were unsure how to access or complete the materials, the facilitator or a member of the research team provided technical assistance. In addition, to enable HVG members (or prior members) to volunteer directly on their own, we posted recruitment notices (with an online link to the questionnaire) on websites likely to reach US voice hearers (e.g. Intervoice, Mad in America, HVN-USA) as well as social media forums (e.g. Facebook groups); these notices were also circulated by peer organizations and distributed at conferences and facilitator trainings. Since groups continue to launch across the US, active data collection continued for more than a year to ensure that members of new groups could participate.
A total of 161 responses were submitted, of which 113 were included in the final analysis. Of the 48 submissions that were excluded, 24 came from respondents who had not attended a HVG, 14 provided insufficient data to analyze and 10 were found to be duplicates.
The final sample of 113 participants was highly diverse. Participants were asked to self-identify their gender, age, race and/or ethnicity and asked about their engagement with both HVGs and traditional mental health services. A total of 104 participants reported their age; the range was 19–79 (M = 42.3); demographic information for all other categories is displayed in Table 1.
Responses to the questionnaire were kept anonymous, with consent forms stored separately. On the last page, respondents were asked if they would be open to contact for a potential follow-up interview; from those who volunteered and could be reached, 15 were ultimately selected for interviews on the basis of their cultural diversity and range of voice-hearing experiences (e.g. brief, long term, traumatic, spiritual, etc.). Interviews were conducted via telephone by a research team member who is a clinical psychologist, and participants were given the option to review their questionnaire responses prior to the call. Interviews lasted 30–60 minutes, were audio-recorded (with participants’ permission) and transcribed verbatim. The interviewer employed a semi-structured style, asking all participants the same set of questions about their voice-hearing experiences and how their group functioned, but also encouraging a free-flowing interaction to allow for narrative coherence. Participants were identified in transcripts and data analyses with pseudonyms, and all procedures, materials and methods were approved by the host institution’s IRB.
Methods of analysis
Aside from the demographic and psychiatric history questions, all data from the questionnaires and interviews were qualitative. They were analyzed in a variety of ways by members of the research team, working collaboratively: (1) an individual profile was written for each participant, synthesizing their questionnaire responses; then (2) a first-level interpretation was constructed, consisting of a distilled one-to-three paragraph summary of that individual’s history of voice hearing and experience in the group, highlighting its phenomenologically distinctive features (following Fischer & Wertz, 1979). Two researchers completed these first two steps independently, then resolved any minor disagreements (mostly of emphasis) between them. Next, (3) employing a grounded theory framework, analytic memos were written to identify key themes across participants; and finally (4) these were integrated via a thematic analysis into a general model that could describe the precise mechanisms by which HVGs work. During these last two steps, three researchers independently analyzed the data and excluded from the final results the few instances of disagreement; in general, there was a remarkably high level of consensus. We chose to include both individually focused, phenomenological analyses (of each participant) and general structural analyses (of HVGs themselves) in order to identify characteristics of these groups seen as effective by a wide range of voice hearers. To triangulate methods and increase robustness of the findings, examples from individual participants were incorporated into the analytic memos, and the constant comparison method (Charmaz, 2006; Strauss & Corbin, 1997) ensured that core elements of the model applied equally across individuals.
This study did not allow for a “member check” of results since questionnaire respondents were anonymous. However, key members of the research team were themselves long-time HVG members or facilitators, which helped to ensure that findings applied across a range of voice-hearing experiences.
Our data make clear that the most important characteristic of voice hearing is its variability; thus, for any intervention to be effective, it must be responsive to a wide range of needs and experiences. Some people have heard voices all their lives; for others, the experience began recently. Some people hear their voices all the time; others, only in certain circumstances. Some can identify their voices; others cannot. Some hear one voice; others many. Some hear voices that are frightening or make threats; for others, voices are reassuring or helpful. Some people are distressed by their voices; others have voices that enrich their lives, serving as companions or guardians. And some people do not hear voices per se; they have visions, or anomalous olfactory sensations, or they sense presences, etc. Indeed, despite the nomenclature of the groups, the Hearing Voices approach is explicitly intended to include a wide variety of anomalous experiences, subsumed under the category of “hearing voices” as a shorthand for capturing perceptions that are both atypical and often stigmatized. Consequently, people with widely varying experiences come to HVGs, and it is the complex constellation of elements in these groups, and the processes by which these elements interact, that make HVGs uniquely responsive to disparate needs and effective in meeting them. Our large sample and in-depth data from each participant allowed us to explore the full range of members’ experiences and to identify their key features.
The thematic analysis revealed two general classes of elements characterizing HVGs: (a) those related to the style of interaction and (b) those related to the content of meetings. The elements concerning style of interaction include: the group’s intentionally non-judgmental attitude; its non-hierarchical role structure (facilitators do not “run things”); use of curiosity to foster a detailed articulation of experience; sense of community, belonging and connection among participants; opportunities to help others as well as to be helped; sense of safety amidst vulnerability; no behavioral targets/outcomes; shared responsibility for meetings; freedom to speak in organic, unstructured dialogue on any topic and potential for authentic relationships – those not constrained by professional roles or codes of practice – both inside the group and beyond it.
The elements concerning the content of meetings include: welcoming multiple perspectives; no pathologizing or use of formulaic language; sharing coping strategies in non-prescriptive ways; valuing expertise by experience; enabling anomalous experiences to be made meaningful according to personal frameworks; appreciating variability in both experience and in modes of explanation. All these elements work together in a unique structure that allows fundamental change to occur.
To summarize the transformational process in HVGs, we propose a three-phase model (illustrated in Figure 1 and described below). This model is not idealized; it emerged directly from thematic analyses of our data and represents key commonalities across participants. The specific content of the voice-hearing experience and the particular ways it is understood and dealt with will naturally vary across individuals, but the basic processes that allow for transformation in HVGs are generalized, and these are the focus of our model. We arrived at this model using the constant comparison method of grounded theory, combined with the features of a “general structure” in Giorgi’s phenomenological approach (Fischer & Wertz, 1979; Giorgi & Giorgi, 2003). Only those elements of HVGs identified as essential across our varied participant group are included. Three researchers each drafted initial versions of the model, and those elements for which there was complete consensus were included in the final version.
A three-phase model of transformation in hearing voices groups.
The three phases characterize the sequence of transformation in HVGs. These phases are not prescriptive, and a given individual may move back and forth among them or not go through the whole sequence. We propose the model not as a means of assessment or evaluation but as a useful framing of how HVGs operate, i.e. their mechanisms of action, which hold across groups regardless of their demographic variations (e.g. rural/urban, midwestern/northern, long-established/newer, etc.).
Phase 1: discovery
During this first phase, group members learn (often for the first time) how widespread and variable voice hearing actually is. From their initial meeting, they are exposed to people with a range of anomalous experiences that may or may not resemble theirs. They meet voice hearers who are not disabled by their voices and those who understand their experience using a range of frameworks other than illness models. They learn that voices in adulthood can sometimes be linked to trauma or to childhood sexual and emotional abuse (Read et al., 2005; Sheffield et al., 2013; Varese et al., 2012). They discover that there are many ways to respond to voices, that they can be made sense of, and it is possible for voice hearers to live lives far beyond what they, their doctors or families had thought possible.
Further, members discover a unique context, in which people respond in a curious and non-judgmental way to whatever experiences are discussed, no matter how perplexing or non-consensual. Few rules apply: no one is required to speak, the range of potential topics is unconstrained (i.e. not limited to “voices”) and members can interact naturally (i.e. with no prohibition on “crosstalk,” but rather an encouragement of dialogue as it naturally occurs). There is no assumption that members are “too fragile” to talk about challenging feelings or to participate in discussions where such experiences are shared; neither is there any pressure to reveal anything. One participant described the group as “a space of radical acceptance, non-judgment, support, and safety.” Another said the HVG offered “effortless understanding”; for another, it was “freedom, mutually created, like a sacred space.”
Consequently, members report a sense of wonder and relief that experiences long-kept secret can be discussed openly. This atmosphere allows for greater curiosity about their own and others’ experiences, less fear about acknowledging or exploring them, and a sense of belonging, connection, and hope (often for the first time and after years of feeling alone, hidden, hopeless and strange). In striking contrast to what many had repeatedly been told – that blocking out voices and ultimately having them stop ought to be their only goals – members discover a landscape of possibilities for understanding and change, even for experiences that have long puzzled or distressed them.
Phase 2: reframing
In this second phase, these discoveries about voice hearing and the functions of HVGs become the grounding for a new kind of exploration. The group’s modeling of self-awareness and self-determination encourages each member to articulate subtleties and variations in their own experience and to situate them within their specific life circumstances. This openness to multiple perspectives and ways of responding to voices, as well as the commitment to “the expertise of experience,” stimulate an iterative process of reframing, in which members evolve an understanding of the nuances of their own psychologies.
With the group’s encouragement, they may try out new ways of responding to voices or learn to negotiate with them, rather than feeling powerless to affect their intensity, tone, frequency or content. The non-hierarchical structure fosters self-efficacy and mutual responsibility (i.e. “it’s up to each of us to make the group work,” “everyone has something to contribute”). The group becomes a laboratory for exploring new and more authentic forms of relating to oneself and to other people.
As a result, the voice-hearing experience itself often undergoes change. Voices may stop or become more contextualized, or quieter, or clearer, or a person’s relationship with them may start to shift (e.g. become less oppositional or less controlling; empathy, rather than fear, may emerge more strongly). As one participant said: “I am no longer just a pawn to the whims of the voices … I am more in control. My voice matters, not just theirs.”
Phase 3: change
In this final phase, the discoveries, explorations and reframings of prior phases coalesce to enable a variety of forms of transformation, both within the group and outside it. Because multiple perspectives on voice hearing are always present in HVGs, and coping strategies are shared in non-prescriptive ways, members are free to try out a range of possibilities without fear of negative consequences. The group’s focus on the individuality of people’s life circumstances and the nature of their anomalous experiences encourages each member to develop the “expertise of lived experience” from which to imagine a wider range of possible selves. Further, by developing skills and modes of coping that fit their own particular forms of distress and available resources, members gain a greater sense of self-efficacy and become more willing to take new chances outside the group.
Because HVGs are not time-limited, people can attend however often (or seldom) they choose and for as long as they want. Knowing that they can always come back to the group – even after a long absence – enables members to deal with setbacks more effectively; since support or assistance is available if needed, they don’t have to abandon a strategy for change just because it doesn’t work right away.
Since, as indicated earlier, there is no one outcome sought by all HVG participants, there can be no single measure of their effectiveness. Nonetheless, 79% of our participants reported experiencing helpful changes in their voices, their relationships and/or how they understood their voices as a result of attending the group. These changes took a variety of forms across multiple domains: voices became less bullying, invasive or frightening; greater empathy and awareness of subtleties of psychological life resulted in increased confidence and efficacy and less shame; and deeper, more satisfying relationships took shape outside the group.
In addition, members reported many forms of behavioral change: reduced need for hospital or crisis services; more limited and targeted use of medication, or tapering off it entirely; better sleep; a return (or an ability now to pursue) important life projects like improved housing, employment, education and social relationships. For some members, being part of the worldwide Hearing Voices Movement instils a broader sense of purpose and belonging; many seek out facilitator training to start new groups and expand the approach. As one participant put it: “I didn’t have anyone to turn to when I started hearing voices … I want to be that person for other people.”
Since, as indicated earlier, a person may well move back and forth among the phases (e.g. trying out a new way of relating to voices may necessitate a reframing of the experience), the HVG can serve different functions at different times. This variability in members’ needs and levels of experience in the group are seen as strengths; people can learn from one another even if their experiences are quite different and use the examples of others to imagine lives that go beyond their past or current circumstances.
Being a member of a HVG can be very powerful, and participants struggled to find the words to describe how significant these effects can be. “I can’t overstate how important my group is to me,” said one man. “There were tears in my eyes the first time I came.” In her follow-up interview, one woman said movingly: “For years I had felt that I wasn’t really a human being … I became human in relation to the other members in the group, and then carried that into my relationships with my voices and with other people in my life.” Another participant said: “I just felt sane there in a way I don’t anywhere else … the group creates a context for people to be able to use their experiences instead of being tortured by them.” Another said: “We are not broken; we just have to figure out how to navigate the experience.”
We found group members to be intensely curious about the phenomenology of voices (their own as well as those of others) and appreciative of a context where they can try to make sense of these very complex experiences. HVGs embody an unusual combination of the philosophical and the pragmatic. In a given meeting, discussion may range from pharmacology to spirituality to trauma theory, from abstract explanations of voice origin to concrete strategies for coping with a specific situation. This broad range of topics means that members are continually exposed to ways of thinking different from their own while being under no pressure to adopt any of them. Further, since facilitators do not “run” HVGs, members have to assume shared responsibility for the smooth functioning and effectiveness of meetings.
Facilitators are seen as essential to well-functioning HVGs, with a distinctive role not found in other kinds of groups. They encourage members to engage one another with “more of a relationship connection as opposed to a therapeutic connection,” as one participant put it. Another, himself a facilitator, noted: “I’m not the interpreter of other people’s experience, the person set up as the ‘expert.’ In HVGs, people are the experts of their own experience … if someone comes in and says ‘I have schizophrenia,’ I’ll say, ‘Ok, what does that mean for you?’” Another participant highlighted the importance of allowing time in meetings to absorb what was happening: “The facilitator doesn’t have to fill silence … it’s ok that we sit in silence for a while as people process what’s been said.”
In both the questionnaires and interviews, we asked how groups dealt with conflict if it occurred; participants repeatedly emphasized their use of a collaborative approach. One gave this example: “Recently someone in the online group was screaming, and it was very disruptive. But instead of asking him to leave, people focused on what it was like for him to be having that experience, and the situation calmed down.”
Many members come to HVGs having been told their thoughts are delusional and their perceptions unreal, so being regarded as someone with “the expertise of experience” often comes as a revelation, contributing to increased confidence and self-efficacy. Because HVGs allow people to seek help as well as providing help to others – even in the same meeting – members’ sense of themselves as passive, disabled patients is constantly challenged. As one put it: “I don’t feel like a sick person there. I can get support but still be treated like I have wisdom. I feel like I have something to offer.”
Members use the group in different ways at different points in their own evolving process of understanding and change. Sometimes what is needed is to “sit in quiet understanding,” as one participant put it, with no focus on advice or support. “They want to be able to put words to it,” said this member, without necessarily being pressed to think a certain way about the meaning of an experience or how to respond to it.
This ongoing articulation of the subtleties of voice hearing can enable changed relationships both to voices and to the sense of self. At HVG meetings, one participant said: “You hear different strategies and a lot of it involves interacting with voices in one manner or another … To have allies who get in that foxhole with you and try to figure out strategies when things are going bad is just invaluable … I’ve waited all my life for something like that.” Another said the group enabled her to “feel and act more like a person and less like a problem.” Another said it gave her “more spaciousness in dealing with voices … and new avenues of engagement with the world.”
Some participants reported that their voices come to see themselves as group members in their own right. As one put it: “They feel heard. They are not as loud, aggressive or invasive. They enjoy the groups.” Another noted: “My voices used to yell at me and scream in pain; they’re much calmer now.” For these participants, the group acts directly on the voices during the meeting, rather than indirectly (i.e. by encouraging a kind of exploration that can then affect the voices outside the group context).
Our results support and expand upon the findings of previous researchers, who have highlighted connection with other voice hearers, exploring the meaning of voices and developing new ways of responding to them, as well as being able to share experiences without fear of judgment as crucial elements of HVGs (Beavan et al., 2017; Dos Santos & Beavan, 2015; Longden et al., 2018; Oakland & Berry, 2015; Payne et al., 2017). Our results also reveal additional key elements (e.g. non-hierarchical role structure; freedom to speak and interact with other members on any topic in naturally-occurring dialogue; prioritizing expertise by experience), and crucially, how the interactions among these elements make HVGs unique. By fostering a sense of curiosity about the subtleties of psychological life, the group can serve both as a safe “container” (Payne et al., 2017) and as a laboratory for exploring new ways of responding to voices and incorporating them into one’s life.
Our study was the first to examine the specific relations between the varied phenomenology of voice hearing and the structure of HVGs, to better understand how these groups manage to meet the needs of such a diverse range of individuals. The openness to multiple frameworks of interpretation seems to play an especially crucial role. As one participant noted: “The medical model takes away from social and cultural and political and economic contexts for people’s experiences. I found that those contexts are key to my understanding.” Another emphasized: “Our groups are about being able to sit with uncertainty … They are not about fixing each other.”
While HVGs have a distinctive ethos and set of values, they can nevertheless often be integrated into traditional mental health settings. Indeed, HVN-USA intentionally includes both Full and Affiliate HVGs: the latter may meet in clinical settings, have non-voice hearers as facilitators or restrict membership to individuals using a particular service (e.g. voice hearers on a certain unit), whereas the former must meet in non-clinical settings and be open to all voice hearers. Affiliate groups do, however, commit to upholding the values of the HVN charter, including a primary focus on connecting and sharing experience, rather than on treatment or psycho-education; having all facilitators, including mental health professionals, use non-medical language and be open to sharing their own experiences; and respecting the expertise of each individual to determine the meaning of their voices and how best to respond to them.
Limitations and future directions
Our sample was intentionally limited to the US since no previous research on HVGs had taken place there. However, this did mean it was geographically skewed by the current clustering of HVGs in certain regions of the country (e.g. Northeast, West Coast) and overall, by the smaller number of groups in the US as compared to countries (e.g. the UK) where the Hearing Voices approach is better established. However, there is no reason to think that our findings are geographically limited in applicability; data from studies in England, Australia, and now the US suggest that HVGs are remarkably similar wherever they are, especially if they explicitly adhere to the HVN charter.
The demographics of this sample represent another limitation, since, as noted above, 62% of participants were women and 66% identified as white. Further, like all studies reliant on self-report data, we cannot know whether our participants withheld or could not articulate important information. It took time to complete our questionnaire (especially for detailed responses to open-ended items) and even more to participate in a follow-up interview, so our sample is clearly limited to people who have sufficient experience in HVGs to want to describe it. But since the goal of our study was to identify the transformative processes in these groups, such individuals are the ones best placed to offer assistance. Therefore, although our sample does not include people who went to one HVG meeting and never returned, or those who for whatever reason cannot articulate their experience of being in the group, they would be unlikely to help us understand how HVGs work. We did have participants who reported no great change in their circumstances, yet still wanted to contribute their insights into the operation of the groups.
A core principle of the HVN charter is that people come to these groups voluntarily, so their effectiveness cannot be evaluated via a randomized controlled study in which some participants are assigned to this intervention and others are not. We took the fact that almost half our sample provided their contact details for a potential follow-up interview, and many participants submitted lengthy and detailed responses to the questionnaire, as evidence of commitment to the study and trust in its methods and procedures. Given the long and distressing history of psychiatric patients being recruited for research studies without their consent or for purposes not fully disclosed to them, we see our intentional inclusion of voice hearers on the team at every stage of the project as an important corrective and encourage other researchers to move toward this goal in their own work.
Our team is currently at work on additional reports of our findings, including detailed phenomenological analyses demonstrating the range of individual HVG experiences, as well as systematic comparisons of specific subgroups (e.g. new vs. experienced members, online vs. in-person participants, etc.). Despite the complexity and methodological challenges of studying HVGs, there are now a range of exciting potential directions for researchers to develop innovative methods of data collection and analysis to advance this important area of investigation.
The authors gratefully acknowledge the contributions of Jacqui Dillon, Lisa Forestell, and Andrea Weisman to initial phases of the project; Maria Narimanidze to the follow-up interviews; and Caroline Mazel-Carlton, Cindy Marty Hadge, and Sera Davidow to participant recruitment.
No potential conflict of interest was reported by the author(s).