Voice

A person's voice is often seen as a window into their mind, and human beings have been seen as definitively 'homo loquens' (Fry 1977). Thus, being able to talk is seen as part of being human, and this ability is often privileged over other aspects of humanness. The concept of 'voice' is being used increasingly in considering issues around rights, autonomy and inclusion in both the disability arena and in relation to children and young people (Davis and Watson 2002, James 2007). Of course it can be used in both literal and metaphorical senses. Both are important when considering the participants in this study who may be perceived as not having 'a voice' either in the conventional physical sense or metaphorically. Because of their physical impairments they have unusual and alternative voices. Conventional meanings of 'voice,' which assume naturally produced speech as its means of expression, have to be re-examined in this context. Clearly AAC users can and do have 'voices,' although these are rather different from those of natural speakers. The question then is whether these alternative voices are recognized and heard in the same ways as others. Or alternatively, whether having a different voice is treated by society as implying other kinds of 'difference' and as having implications in relation to the person's identity and the quality or value of their life (Mackenzie, Leach and Scully 2007). The issue of whether or how the 'real' voices of young people who talk in unusual ways can be represented accurately and sensitively will be addressed later in this paper. As Paterson and Hughes (1999) suggest, we live in a verbal world where those who can talk, the 'vocal bodies', are often exclusionary in the way that they structure society and 'norms of communication and norms of intercorporeal interaction reflect the carnal needs of non-disabled actors' (Paterson & Hughes 1999:604).

Augmentative and alternative communication (AAC)

Usually people who are unable to use natural speech well or at all spontaneously develop informal alternative ways to express themselves (e.g. facial expression, eye and body pointing, gestures), and these are often supplemented by more formal (and taught) low-tech modes, such as signing, picture or symbol books and boards. Many people also use high-tech modes in the form of computer-aided devices, often called 'voice output communication aids' (VOCAs). The latter produce an electronically generated voice and can be of various designs and types, matched to the physical and cognitive skills and preferences of the user.

Some children who have physical impairments develop enough speech (although perhaps very unclear or only single words), to use with family and close friends. The introduction of AAC does not prevent their use of these words, and indeed may enhance it. They gradually develop their own way of 'mode-switching' between their own speech and their aided communication systems, as needs dictate and they choose. Similarly, those with no natural speech learn to mix modes, usually using both low-tech and high-tech methods. All the participants in the study described here use more than one method of communication, and usually individuals favour some over others for pragmatic reasons.

Very often AAC users report that their low-tech systems are much quicker and easier to use than the high-tech ones, so they opt for the former when talking to people who know them well enough to understand the more elliptical and individual style of these messages. These often demand a high level of skills and knowledge of the person from the conversational partner. In contrast, high-tech VOCAs are slower and more physically effortful, but more comprehensible to strangers, and are thus essential when there is no familiar person around to mediate for the user.

Apart from natural gesture and facial expression, the most commonly used low-tech method is the 'communication book'. These are individually designed in a photo album format with up to several hundred pictures or symbols organized thematically (e.g. pages for my family, my classmates, foods, places, sports, emergencies etc). In a conversation, the AAC user indicates by eye or hand pointing, to the conversation partner, which page to turn to, and then on the requisite theme page, to the target word. Some users and their carers/assistants learn to use this system extremely rapidly, using a series of colour and position codes as shortcuts. The disadvantage is that it takes a long time to construct a sentence with grammar. Often, the AAC user depends on their conversational partner to co-construct their intended meaning from a small number of key words, e.g.

Another commonly used low-tech method is the Etran frame. The frame is a large perspex board with letters, numbers and punctuation arranged on a grid around the periphery. There is a rectangular window in the middle. The conversational partner holds the frame up so that the AAC user can 'eyepoint' at letters through the frame. This obviously requires the user to have good literacy skills, as well as the ability to concentrate well and eyepoint accurately. Once both user and conversational partner are skilled with the system, a number of shortcuts often develop, e.g. the listener may guess halfway through a word or phrase.

This takes trust and good judgment in the partner, about when to guess and when to wait. Once people know each other well, this can be a very speedy process, and again, is much quicker and less effortful than the high-tech systems. However, it is still very tiring and requires intense concentration from both parties.

There is an increasing range of high-tech communication aids (VOCAs) of different designs, and the decision about which is suitable for a particular person is complex and multifactorial. There has been a rapid improvement in the technology in the last 10 years, so that many devices are now computer-based and have huge memory capacity, so that the user can, if they have the cognitive skills to do so, access many thousands of words and phrases, as well as linking to the internet, using e-mail and a mobile phone through the device.

The quality of the electronically generated voices has improved greatly, and most models allow the user to choose from a number of voice options (e.g. age and gender appropriate), and some also allow a choice of American or British accents and some other aspects of 'style'. However it remains that, because the voices are electronic, they do lack 'naturalness' and do not have the subtle individual variations in voice quality, tone, volume and regional accent that physiological voices have. Thus, with the present level of technology they still have a 'robotic' quality, which attracts unwanted attention and is not unique to the person in the way that a natural voice is.

The way in which the user 'accesses' the letters, words or phrases on the VOCA depends on each individual's physical skills. 'Direct access' refers to systems where the person chooses buttons or squares on a screen directly with their finger (or a headpointer or foot). The majority of the participants in this study used this method. The size and bulkiness of the VOCAs varies. Smaller ones can be moved around and used on a table, on the person's lap, or even hanging around their neck or held while walking. Most, however, are bigger and are usually mounted on a tray or bar attached to the person's wheelchair. This obviously restricts the range of contexts in which it can be used. For example, if the user is sitting in another chair, lying in bed or on the floor, or of course doing sport or another activity out of their wheelchair such as swimming, their VOCA, and therefore their 'voice,' will be unavailable to them.

For most, the high-tech system is very slow compared to natural speech or most of the low-tech modes. Many AAC users, of necessity, use rather short utterances. Some go to great lengths to make these grammatically accurate, whereas others develop their own forms of ellipsis, using idiosyncratic shorthands, or sometimes just key words, relying on a mediator who knows them well to co-construct the full sentence for them. They may use pre-programmed phrases, which roughly suit the purpose, but may be rather inexact. The result is that although English is being spoken, it is almost a different dialect, or perhaps a 'creole' with its own very particular rules, logic and idioms, generated in a particular context. It can therefore be difficult for a newcomer to understand this. These shortcuts, which can be highly colloquial and individual, become part of the person's 'voice,' identity and characteristic of them as a person.

Identity

Links between thought and language have long been debated and are complex (Geertz 1993, Csordas 1994, Goffman 1959). However there is often confusion between language (as the internal process) and speech (the external physical expression), such that someone who does not speak may be assumed not to have anything to say, and thus not to be thinking either. Thus, using AAC has important implications on the identity of the person. The theoretical approach to identity drawn on here is that of Jenkins (2004), which suggests that there are broadly two complementary aspects of identity: selfhood (how you see yourself) and personhood (how others see you). These are constantly evolving, influencing each other and intertwining across contexts and during an individual's lifespan.

For people using natural physiological speech, talking is an important way of negotiating and establishing identity, as what we say is assumed to represent what we think. This has been recognized by authors who link the role of narrative to identity formation (Connelly and Clandinin 1990, Cohen 1994, Thomas 1999). Through conversation and narrating events in our lives we express who we think we are and how we would like others to see us (Ochs and Capps 2001). Thus, for AAC users whose communication systems are slow, this opportunity may be denied, unless their conversational partners adapt very specifically and expertly to the type of communication being used.

Identity is usually expressed not only through the unique content of individuals' talk, but also through aspects of style as expressed by voice quality, intonation, choice of vocabulary and grammar etc and this is context dependent (Collins and Markova 1999). Different people might be very characteristically verbose, taciturn, grammatical or slangy, but will also vary their style according to the social situations. Verbal language is also accompanied by each person's unique nonverbal communication. Potentially then, an AAC user may be denied these means of expressing their individuality, especially if as is usual, they have severe physical impairments which restrict their movements or when they are under extreme time pressure.

However, it is often overlooked that despite not having the individuality of a physiological voice, AAC users can and do express their identities through their unique way of using their AAC modes. Clearly then, AAC users do choose to adopt a range of different styles of talking to suit varying contexts, and such choices can be said to reflect aspects of identity. Familiar conversational partners, for example, are expected to be very skilled at recognizing shifts of mode (e.g. use of eye gaze or a sign in the middle of a VOCA utterance), and at interpreting even the less clear messages through processes of using prior knowledge, contextually appropriate guessing, and mediation. AAC users' ability to adapt their style and content may be limited for purely practical reasons, as an artifact of both their impairments and the technical capacities of their communication systems, but they are actually very skilled at finding verbal and nonverbal ways of expressing their individuality. Manufacturers are beginning to take notice of AAC users' desire for variation in the voice qualities on VOCAs and develop the technology to achieve this.

This overview has emphasised the central role normally attributed to language as expressed through voice, as a reflection of thought and of the person, and the way that in telling stories about their lives people bring their self-identities into being. Underlying this is the notion that agency is usually demonstrated through talk, so differences in ways of talking may bring about fundamental differences in others' perceptions about the person. As Katie (adult research advisor and AAC user) put it: 'It is tempting to think that less talk means fewer ideas!'

Both the relationship between the teenage AAC users and their VOCAs, and how they conceptualise these, were core aspects of the present study. Several participants made it quite clear that they do not want their 'machine' to be seen as more important than the person using it: "When people say 'oh what a fantastic machine', I am blunt in my reply 'the machine is boring, I am fantastic'. I do not care what people think of me for making that remark. The identity behind the machine is what is important" (Allan 2006:11).

This echoes much of the previous literature about disability and identity in emphasising the social-relational aspects of the person, rather than foregrounding their impairment or disabled status (Shakespeare 1996, Watson 2002, Thomas 2004).

Allan (2006), quoted above, who was an adult advisor on this project describes the way in which her electronic voice, produced by her VOCA is 'her voice'. Others may view this electronic voice as odd and artificial, but for her it is part of her. Here then is a potential site for interesting issues in selfhood and personhood, when people have 'non-natural' bits of their bodies. With the development of new technologies the boundaries between nature and culture are no longer secure, and as Sharp suggests: 'Constructions of the human body and of human nature itself emerge as deeply troubled' (2000:295).

The ways in which people who use AAC manage their own identity formation, and the ways in which others see their different bodies is thus a matter of interest and conjecture in fast changing times of technological advancement (Cromby and Standen 1999).

Used philosophically or anthropologically, the 'voice' is the expression of both internal and external dialogues, a representation of the person. Thus, the ways in which a person talks and moves (i.e. the latter as part of nonverbal communication) are like a 'fingerprint' of selfhood, simultaneously unconscious and conscious. This is usefully reflected in Habermas' (1984) recognition of the role of communication in intersubjectivity and Bourdieu's (1977) inclusion of it as a part of habitus.

Social communication, and particularly talk which is declarative rather than imperative, is also fundamental to the formation of social relationships (Nelson 2000, Connelly & Clandinin 1990), and especially in adolescence (James 1986, Bohanek et al 2008). Although at first glance physical production of a voice may not seem significant in relation to personhood, in a vocal culture it is important to be seen as someone who can produce words somehow, and therefore meanings. Voice, as a vehicle for language, can both generate and reflect power relations (Ng & Bradoc 1993), and thus, someone without a natural voice or someone possessing a different kind of voice may easily be disadvantaged (Forster and Caryer 1997). The way that such voices are perceived and represented by others is also of great importance. This is what this study set out to explore.

Research aims and questions

The study aimed to explore identity and the lifeworlds of teenagers who use AAC. The research questions were:

1. How do young people with severe physical and communication impairments who use AAC see themselves (selfhood)?
2. How are young people who use AAC seen by others (personhood)?
3. What kinds of social relationships do young people who use AAC have?
4. What role does the body play in the development of selfhood, social relationships and personhood for young people who use AAC?
5. What kinds of methodologies work best when doing research with young people with severe communication impairments?

Participants

Participants were 9 key young people aged 10-16 years (6 girls, 3 boys). Additional data was obtained from a further 15 teenagers aged 14-19, and from 3 adult research advisors, all with similar impairments and all AAC users. Parents, school and extra-curricular club staff also contributed. The AAC users had severe physical impairments caused by cerebral palsy. Broadly, they had typically developing cognitive skills and little or no natural speech. All used a power wheelchair and a high-tech voice output communication aid (VOCA), as well as a range of low-tech communication methods. Notably, all either currently or previously used 'communication books', either as their preferred method or when their VOCA was unavailable.

All needed physical help with daily activities such as dressing, toileting, and feeding. The teenagers attended a variety of special or mainstream schools, and had the support of many professional 'helpers' such as classroom and personal assistants and therapists. They lived with their biological families, although 4 attended residential schools.

University ethical clearance for the study was obtained. Parents gave initial written consent, however verbal consent from the teenagers themselves was also given prominence, and was ongoing, being reviewed and renewed verbally at key points during the study.

Methods

The main research method was participant observation, which took place over an eighteen-month period in the teenagers' homes, schools and extra-curricular clubs. Additionally, a series of 'extended narrative conversations' on focused topics were held with the 9 key participants. Topics included: my important people, self-description, things I love and hate, dreams for the future, friends, people who help me, my body, my communication. These conversations were supported by the researcher drawing mind maps using written words and pictures, so that the content could be reviewed, subsequently edited later in the process, and, finally, kept by the participants.

Interviews with parents and focus group discussions with school staff were also held. These contextualized the young people's data, but the aim was always to keep the young participants' perspectives and voices at the fore, as these are so often overlooked and not valued. The data then comprised ethnographic observations, field notes and transcriptions of digitally recorded conversations. All data was entered into NVivo (V7) (QSR 2006) and coded. Themes emerged iteratively and were revisited and reviewed in a constant comparison process, which started during the latter stages of the fieldwork. The initial thematic analysis of the data was presented back to the teenagers and some parents towards the end of the study. It was also discussed in depth with the 3 adult AAC users in order to get ongoing participant validation of my analysis and interpretations.

Overview of the main findings

The 5 main themes that emerged were given user-friendly names when they were presented to the teenagers for comments, and so these names have been used subsequently:

• Whose voice is that?
  (communication issues as they relate to identity)
• Being a family person and being a teenager
  (selfhood 1)
• Me myself I
  (selfhood 2, autonomous and disabled selves)
• Bodies and kit: how I see my body and my equipment.
  (the body, physical aspects of identity and the role of the kit)
• My family and others: how other people see me
  (personhood)

In summary, the results show that teenage AAC users are much more interested in the ways that they are the same as their peers, than in the ways they are different. They describe themselves variously, for instance: 'fun, clever, a laugh, beautiful, caring, helpful, thoughtful, sporty, strong, and a fighter', as well as 'singleminded, messy, stubborn, late, scruffy, and cheeky' (grouped data). They are interested in typical 'teenage' topics such as IT, music and media, fashion, sport, food, holidays and trips, going out, sexuality, group gossip and increasingly in doing things without adults.

The participants view themselves principally as 'normal teenagers,' whose families and friends are very important and influential in their lives. They and their parents used the term 'normal' quite insistently and with no hint that this in itself might be tyrannical. The normality they all highlight is social-relational, as they paint positive pictures of themselves as sociable and competent without highlighting their impairments, although they acknowledge these as part of their disabled identity. Some see their physical and communication impairments as ways in which they are 'interestingly different.' Their aspirations for life are like those of their non-disabled peers and siblings, and these are their role models. They are pragmatic about the effects of their impairments, and have expectations that these are surmountable and should not stop them from doing 'normal' things. Like previous research about disabled people and identity, they thus emphasize their social-relational selves (Thomas 2004). They want to be more socially included outside their families and to be treated age appropriately. Their main concerns were to have more friends, appropriate and reliable technology, and friendly, respectful personal assistance that would enable them to have as much autonomy as possible. They emphasise that everyone should be treated as a person who has something interesting to say.

Conclusions of the main study

The conclusions of the study were that despite the teenagers' positive and normative views of themselves, they were seen very differently by others, especially those outside their families. There was an ontological dissonance between their selfhood and personhood, which explains their frequent feeling that they were being misunderstood and that their disabled status was given too much attention. They were keen for their complete selves as socially competent and family oriented, but outward looking teenagers to be recognized.

Processes of co-construction — negotiation and collaboration

As the fieldwork progressed and I become increasingly familiar with each individual's complex and unique communication style, I reflected on the role of conversational partners as they interact with AAC users. As an anthropologist, I saw my gradual familiarization with the language and culture of AAC as similar to the classic ethnographic experience of exploring a foreign land. I moved from being a complete outsider who often needed an interpreter to help me, to an insider position, where I sometimes found myself translating what the teenagers were saying for other relative newcomers. I saw that each individual had a spectrum of conversation partners who varied in their level of skill at facilitating conversations with them. By seeing the young people in a wide range of contexts, I saw how variably successful their interactions with others could be.

With family, close friends and those working regularly with them, they were able to have fast, subtle, multimodal conversations. This was possible because these familiar people were expert mediators, who were used to interpreting the meanings of facial expressions, eye gaze, and body movements in concert with output from low- and high-tech aids. Nearly every 'utterance' needs verbalising out loud by the conversational partner to check that it has been correctly understood, and although this sounds cumbersome, it is achieved with speed and fluidity that does not disrupt the flow of conversation. In contrast, conversations with strangers or the uninitiated are much slower, more jerky and full of misunderstandings, repetitions and often some frustration on both sides.

My own skills in becoming a more skilled conversation partner were revealed in the data over time. Upon reviewing audio-recordings of earlier conversations, I saw where I had misunderstood rather telegraphic messages, and had failed to understand individuals' shortcuts or nonverbal messages. This led me to consider the skills and tasks required by both partners when one person is using AAC. These are summarized in Table 1. Some similar observations have been made by other researchers looking at the process of AAC interaction, but using different (mainly psychological or linguistic) theoretical frameworks and methodologies (Pennington and McConachie 1999, Clarke et al 2002, Clarke and Kirton 2003, Light and Binger 2003)

Some participants seemed resigned to being misconstrued, and there were various examples during the fieldwork when, in retrospect, I discovered something which revealed that I had misunderstood a previous conversation, although the young person had kept quiet about this. While some clinical researchers have described this kind of behaviour as 'passive' (Pennington & McConnachie1999), it can be reconceptualised as agentic and pragmatic. The teenagers learn to make judgments about what is worth the effort and really important to say and when to let a misunderstanding go.

Of course, all conversation and interaction is reciprocal and has an element of interpretation and clarification between the partners involved. However, the possibilities for misunderstanding and for misconstruction of the message are amplified greatly when one person uses AAC. We can see that in co-constructed conversations there is risk that the AAC message may be constructed by the conversational partner in a way that is perhaps close enough to the intended meaning for the AAC user to decide not to correct it. However, the cumulative effect of these frequent 'miss hits' might be that a quite inaccurate set of narratives about the person starts to be perpetuated. Thus, the AAC user's identity may be more actively externally constructed than is usual. Natural speakers can easily correct and retell stories in their own way, so that they have more control and authorship of their own identity construction. Some fundamental aspects of the AAC user's identity may be misconstrued because of the challenges of communication.