PART I
The experience of illness
Chapter One
‘One gives bad compliments about me, and the other one is telling me to do things’ – (Im)politeness and power in reported interactions between voice-hearers and their voices 1
Zsófia Demjén, Agnes Marszalek, Elena Semino and Filippo Varese
1 Introduction
Hearing voices, also known as auditory verbal hallucinations (AVHs), involves the perception of verbal content in the absence of an appropriate external stimulus. While AVHs can occur for people without a mental health condition, they are seen as a characteristic symptom of schizophrenia-spectrum disorders, with approximately 70 per cent of individuals with such diagnoses reporting hearing voices (McCarthy-Jones, 2012). AVHs also cut across diagnostic groups, as voice-hearing can occur in bipolar disorder, depression, post-traumatic stress disorder, dissociative disorders and obsessive compulsive disorders, among others.
While for some voice-hearers these experiences are a source of extreme distress and impairment, leading to a need for care, the actual phenomenology of voice-hearing is extremely heterogeneous, in terms of both form (sensory and conversational qualities, such as loudness, pitch) and content, with variations in terms of their affective impact. As a result, a sizable minority of voice-hearers cope well with their voices (Jenner et al., 2008). This is because distress is generally not caused by the mere presence of voices, but depends on a number of factors, including what the voices say, and how; the relationship that voice-hearers establish with their voices; how they make sense of their voices; their perceived control over the voices; and their ability to live the life they want to live, particularly in relation to control over their goals in life (e.g. Mawson et al., 2010; Varese et al., 2017). In this chapter, we present insights from a pilot study exploring what a particular linguistic approach, namely the study of (im)politeness in interaction, can contribute to understanding the relationships between voice-hearers and their voices, and how these might relate to issues of power, control and therefore distress.
Factors determining the degree of distress that voices cause are, of course, attended to in clinical psychological research and practice. Interventions often aim to bring about change in power and control dynamics, for example, by increasing the hearers’ coping and perceived control, changing the relationship with voices, challenging beliefs about the power of voices, etc. In fact, assessments of voice-related distress and the determinants of distress are often generally language-based (e.g. clinical interviews such as the auditory hallucinations subscale of the Psychotic Symptom Rating Scales, or PSYRATS-AH, Haddock et al., 1999). However, linguistic analysis is not typically used to explore individuals’ descriptions of their experiences, and most current approaches to understanding and treating distressing voices are limited in a number of ways. First, they involve methods that capture explicit processes, that is, questionnaires, interviews and psychometric assessments (e.g. PSYRATS) that rely both on voice-hearers’ conscious awareness of the nature of their relationships with voices and their willingness to disclose them. Second, current approaches often lack a consideration of the phenomenology or ‘lived experience’ of voice-hearing (Thomas et al., 2014; Woods et al., 2014). Finally, the ways in which the relationship between voice and hearer is established and maintained (including power dynamics related to control) remains poorly understood.
We begin to address these shortcomings here by exploring how ten voice-hearers with schizophrenia-spectrum diagnoses use language to describe their interactions with their voices, in the course of interviews with one of the authors (FV) (see Varese et al., 2016; Varese et al., 2017; Demjén et al., 2019). This means that we take descriptions of the lived experience of interacting with voices as our starting point. Linguistic analysis generally, as the chapters in this volume demonstrate, relies on systematic and theoretically based ways of investigating people’s linguistic choices when they describe their experiences. In very general terms, this involves looking at what linguistic choices are made (consciously or not) in contrast with other choices that could have been made, how such choices pattern systematically and what the implications of these choices might be. This makes it possible to develop an approach to the phenomenology of voice-hearing that also includes implicit processes, that is, processes that are less amenable to conscious monitoring and manipulation. Evidence from other areas of the psychological sciences (e.g. experimental psychopathology) suggests that such implicit processes can be influential determinants of mental health and well-being (Franck, de Raedt and de Houwer, 2007).
In linguistics, relationships between individuals, including power dynamics, are known to be reflected, negotiated, maintained and challenged in interaction (see also Brookes; Chimbwete-Phiri and Schnurr; Kinloch and Jaworska; Stommel and Lamerichs; Thurnherr et al.; Zayts and Lazzaro-Salazar, all in this volume). We assume that this, to an extent, also applies to interactions between a person who hears voices and the voices that they hear, as in the following extract from our data:
Excerpt 1
it could be one minute past midnight, a brand new day and all of a sudden the voices will say “you worthless bastard”. I am used to it now and they’ll say “you worthless bastard and it’s another day and we are gonna kick the fucking shit out of your head” and then they start on you 2 then. They call you all sorts of things “useless, pathetic, hopeless, fat, evil, nasty bastard” and all that lot and they swear at you and they say “where’s your fucking nurses now to support you and your mental health team they don’t give a fuck about you” they say “why don’t you teach the bastards a lesson and kill yourself blow the flat up, set yourself on fire, go out there later on today and walk under a bus” (Participant 6, our emphase s)
Interactions between voices and voice-hearers are clearly not the prototypical kinds of interactions that linguists normally analyse. Analysts have no direct access to these interactions except as reported by the voice-hearer. Moreover, information about voices’ identity, status or context is mediated through the hearer, or not available at all. As with reports of interactions more generally, it is highly likely that these accounts are not complete and precise records of all that the voices (or hearers) say, but simply a recollection of a particular selection deem...