During a study of inclusive education in middle school for students considered to have significant disabilities, I asked Jacob, an eighth grader who types to communicate, if there was anything he would like to tell his teachers. He typed, "TELL THEM I AM SMART." In another interview he reiterated this statement saying the worst part of middle school was "WHEN THE TEACHERS DONT THINK I'M SMART." It is clear from these comments that Jacob clearly wanted me to convey his desire to be seen as competent to others. He desperately wanted to change his teachers' perceptions and present a different construction of himself. One of the primary goals of my research agenda is to bring to the forefront the perspectives of persons too often marginalized and silenced within their schools and communities. I aim to "give voice" to their joys, their struggles, and their stories as a way to provide an alternative to the dominant discourse of disability and competence.

Burton Blatt (1981) wrote, "A person is defined by the stories he tells about himself as well as by the stories that are told about him." Too often, stories are told about people with disabilities and the emphasis is on the voice of the professional, not the voice of the person (Biklen, 1988; Biklen & Duchan, 1994; Blatt, 1981; Bogdan & Taylor, 1976, 1994; Kliewer & Biklen, 1996). Qualitative research that directly involves individuals considered to have significant disabilities as participants is extremely limited. Most interview studies focus on the perspectives of caregivers or teachers instead (Carlsson, Paterson, Scott-Findlay, Ehnfors & Ehrenberg, 2007; Lloyd, Gatherer & Kalsey, 2006). There is a gap in the research for the voices of individuals who experience disability (Jones, 2007). One of the goals of disability studies scholarship is to create opportunities for traditionally marginalized perspectives to be heard. This begs the question, "How do we as researchers in disability studies work to create those spaces?" In this paper, I describe the importance of "giving voice" or "facilitating voice" within critical, qualitative research, while posing challenges to a narrow interpretation or conceptualization of voice. I illustrate the importance of conducting research that aims to give voice, while also problematizing this concept for individuals who do not use speech as their primary mode of representation. In other words, how do we ensure a space within a critical, qualitative framework for the inclusion of "voices" that do not speak?

Bogdan and Biklen (1998) describe giving voice as "empowering people to be heard who might otherwise remain silent" (p. 204) or who have been silenced by others. Britzman (1989) writes, "A commitment to voice attests to the right of speaking and being represented." In other words, voice is the right and the ability to make oneself heard and to have ones experiences and perspectives available to others; to participate in the construction of the self and to decide how to represent that self to others. Or as Mazzei and Jackson (2009) state in the introduction to their critique of voice in qualitative methods, "Qualitative researchers have been trained to privilege this voice, to 'free' the authentic voice from whatever retrains it from coming into being, from relating the truth about the self " (p. 1). As I consider the meaning and implications of a commitment to voice, I relate this discussion to research with individuals who challenge normative constructions of typical speech and voice.

Research Methods

This article was born out of a critical qualitative study of the experience of middle school for students considered to have significant disabilities, specifically adolescents with labels of autism and intellectual disability. In this study, I attempted to make visible the taken for granted assumptions regarding disability and competence in the context of middle school, through interpretivist qualitative research (Collins, 2003). Critical qualitative research is based on several assumptions: (a) that all meaning is created through discourse, (b) that all discourse and thought is "mediated by power relations, which are socially and historically situated" (Kincheloe & McLaren, 2000, p. 291), and (c) that certain groups are inherently privileged over others. In other words, I conducted this study with recognition that we live in an inherently ableist society. One goal was to give voice to the experiences and perspectives of individuals considered to have significant disabilities in an attempt to make visible and, hopefully, to disrupt those systems of power and privilege that serve to marginalize. My consideration of this topic emerged as I reviewed the data and my own comments on the research process through a methodological lens and confronted the problem inherent in an unproblematized construction of voice in qualitative methods.

Data sources

While initially there were five student participants, I focus this article on one student. Jacob is a young man that loves hiking, swimming, watching movies, and spending time with his sister. He receives special education services in his school under the educational label of autism and he uses supported typing or facilitated communication1 as his primary means of expression. I utilized the data collection techniques of participant observation in school and community settings and in-depth interviewing of the students, their parents, and school personnel. I observed Jacob 18 times, in over 10 different class settings. In addition, I interviewed Jacob formally in his home three times. Informal conversations and interviews were included in the participant observation field notes. During interviews with Jacob conducted at his home, his mother Hope acted as his communication facilitator, providing physical and emotional support. During informal school conversations, his one to one teaching assistant provided communication support. For the purpose of this article, I focus primarily on interview data as is it was the participant voice I aimed to bring forth. While the focus of this analysis is my experience with Jacob, at times I will refer to interview data from other qualitative studies involving research participants who type to communicate.

Data analysis

Keeping with the nature of qualitative research, data analysis proceeded inductively. Rather than beginning with any preconceived notions and then attempting to prove or disprove my theories, the analysis has developed from the ground up. I conducted ongoing coding of the data to look for emerging theories and developing themes during the data collection process. However, upon completion of data collection, I coded all of the data again, developing both primary and secondary codes. This form of data analysis is generally referred to as developing grounded theory (Strauss & Corbin, 1998). My interest in issues around voice and methodology emerged early in the data collection process and I wrote frequent memos about the methodological implications of this work for individuals who do not use speech.

At the completion of data collection I returned to the raw data and considered it through a more methodological lens with a specific interest in issues of voice and representation. I used the following research questions as guides. What does it mean to "give voice" within the context of critical qualitative research in the feminist tradition? How are notions of "giving voice" complicated when the research participant does not use verbal speech? Elsewhere, I have reported on the findings related to inclusive education in middle school (Ashby, 2010). For the purposes of this paper, I address only issues of voice and qualitative research.

Problematizing Voice and Method

One of the earliest and most fundamental goals of the feminist movement was to "find voices" for women (DeVault, 1999). Early feminists argued that the perspectives of women were not being considered or represented in research and began work to change that. This often took the form of scholars who were women interviewing other women, in an attempt to bring to light feminist concerns within a larger political context.

The equation of feminist commitments with qualitative work — interviewing in particular — is not especially recent. The notion that women interviewing women, aligned with a distinctive sensibility for women's lives and experience, confers an especial affinity between feminism and qualitative research is a relatively long-standing one. (Atkinson, Coffey & Delamont, 2003, p. 80)

Despite the continued focus on the centrality of women, feminist principles of voice have been applied to critical research with other marginalized populations and their interactions with oppressive systems. Whether those systems are oppressive for their inherent sexism, classism, heterosexism, and racism or, in the case of my research, ableism, feminist concerns with power and voice are clearly relevant and useful. How are particular voices silenced? Whose perspectives are buried (Carlsson et al, 2007; Charlton, 2000; Jones, 2007). Individuals with disabilities have been historically marginalized in society, their differences marked as deviant in a hierarchy that clearly privileges an "able-bodied" existence (Wendell, 1996; Thomson, 1996). Just as feminist researchers explore issues through a lens of gender inequality, those same principles can inform how we think about research with individuals who have been oppressed due to disability in an ableist world (Ashby, 2010; Hehir, 2002, 2005). But unlike feminist methods, it is often nondisabled researchers interviewing individuals considered to have disabilities.

While my goal as a researcher has always been to give voice to the perspective of individuals, particularly students, considered to have significant disabilities, the further along I got into the research process, the more uncomfortable I became with an unproblematized use of the term giving voice. Was I really giving voice? Was it mine to give? Whose voice is it really? Who benefits from the telling? Is spoken voice preferable? I began to question the implications of giving voice and what that means methodologically, particularly with regard to research participants often presumed to have cognitive disabilities.2

One critique of giving voice is that the practice can actually reinforce the very systems of oppression that it seeks to redress. Regardless of the intentions of the researcher, hierarchies of power and privilege are re-inscribed when the researcher presumes to give voice to someone else. A desire to "give voice" can assume several troubling "truths." One, it assumes that the person or group being researched has no voice and therefore, needs someone else to bring their experiences to light. It denies that these individuals have their own voice and can (and do) choose to exercise it, although admittedly students with disabilities are often denied the opportunity to do so. In order for students with disabilities to be heard and listened to, they become the objects of research, where their voices are mediated and interpreted. I cannot assume that by attempting to give voice, I am free from the problematic nature of interpretation. While it is their experiences I am aiming to bring forth, I do so recognizing that like all qualitative researchers I bring my own perspectives to bear on the data. It is not just the students' experiences that are being brought to light; it is my interpretation of that experience (Kincheloe & McLaren, 2000). It is also important to remember that while the aim of giving voice is emancipatory, the researcher often benefits more from the telling than the researched.

Another assumption is that only some people have to "give voice." Orner (1992) poses the question, "Why must the oppressed speak?" Am I in such a privileged, dominant position as a white, academic, nondisabled scholar that my position and my experience go unquestioned? Am I benefiting from the unmarked status that comes with being part of a dominant social group? As a researcher attempting to give voice, "We ask for revelations from others, but we reveal little or nothing of ourselves; we make others vulnerable, but we ourselves remain invulnerable" (Behar, 1993, p. 273). Again, those revelations and vulnerabilities should lead to positive changes through dissemination of knowledge, but how soon and to whose benefit. Also, should students with disabilities be responsible for explaining to others how to "deal with difference" or should teachers and school personnel have to do that important work themselves?

In addition, the very act of constructing a research project aimed at investigating the experiences of individuals considered to have significant disabilities implies that those experiences and perspectives are inherently distinct from those of their nondisabled adolescent peers. There is an "othering" at work that runs the risk of essentializing the experience of disability as something fundamentally different from other ways of being in the world. By developing a study that specifically focuses on the experiences of middle school for students with significant disability labels I am treating those students as a separate group from the non-disabled population. That segregation defies the inclusive nature of education I would hope to see and could perpetuate distinctions between students with and without disabilities. Do attempts to give voice to marginalized and oppressed groups continue to set them apart in the research process? But how can that be avoided if we are to ensure that these voices get heard?

The challenges inherent to an uncritical construction of giving voice are further exacerbated when the research participants are individuals who do not use verbal speech as their primary means of expression. What is the meaning of giving voice when that voice does not present itself as spoken speech? In the following section I explore those questions and challenges in greater detail, focusing on the intersection between critical qualitative research, disability studies and individuals who do not speak. The following key themes have emerged from this analysis: a) the question of competence for individuals who do not use speech, b) hearing silence, c) agency and voice, and d) broadening the conceptualization of "voice" beyond speech.

Including Voices that Do Not Speak in Research: The Question of Competence

"TELL THEM I AM SMART." This plea was a frequent refrain in my conversations with Jacob. He desperately wanted his teachers and peers to see him as a competent, capable young man. Unfortunately, he faced frequent challenges to that competence. For example in a conversation with Carla Baldwin, Jacob's eighth grade math teacher she provided the following assessment of his challenges. "What is really wrong with Jacob is that Mom is setting him up [for failure] at home. She seems to think he is a genius and she keeps telling him that. No wonder he is frustrated." She described her frustration with Jacob's participation and their difficulties getting him to complete the work and be part of the class. She told me how they have seen no independent work from Jacob; he produced nothing that was not totally prompted by adults.

Jacob described the difficulties of his inclusion in middle school centering on teachers failing to recognize his competence. He typed often of wanting to be heard and seen as "smart." He even asked me to tell his teachers he wanted harder homework. Carla, on the other hand, located the difficulties Jacob was experiencing within unrealistically high expectations, perpetuated by his mother. The juxtaposition of these two statements demonstrates the importance of a consideration of voice in research. Whose voice is privileged? Whose authority counts? Typically in schools, it is the voice of the professional that merits greater saliency and consideration. As a critical qualitative researcher, I chose to highlight the quiet voice, the one with less power. This disjuncture also highlights the complicated nature of voice and competence, which can be seen in the controversy over supported typing or facilitated communication.

In facilitated communication or supported typing, a facilitator provides physical and emotional support to a person for whom speech is not reliable and who cannot point independently (Institute on Communication and Inclusion, 2010, 2000). Typically, the facilitator provides physical support in the form of backward resistance to the typing hand or arm to help slow or stabilize the movement. This person may also provide sensory or physical support to the body to help with regulation and motor control. Other types of support are provided as well, including prompting the communicator to keep going or providing verbal feedback about the message. All of this is intended to help the individual move through the issues of motor planning, sensory regulation challenges and difficulties with initiation experienced by many individuals with autism. The ultimate goal of this communication support is that the person will learn to type with increasingly less physical touch and/or learn to speak aloud what is typed.

This form of mediated communication has been controversial. Critics of the practice allege that the words typed are not coming from the communication aid user him or herself, but instead from the person providing the physical support. Inherent in this critique is the assumption that the communication aid user is not competent enough to represent her or his ideas and therefore those ideas must be coming from the facilitator. After several research studies demonstrated that facilitator influence of the message was possible or failed to confirm authorship of typed messages (Bebko, Perry, and Bryson, 1996; Bomba, O'Donnell, Markowitz, & Holmes, 1996; Cabay, 1994; Eberlin, McCannachi, Ibel, & Vlope, 193, Klewe, 1993; Shane & Kearns, 1994; Wheeler, Jacaobson, Palieri & Schwartz, 1993), many organizations denounced the practice publicly, making it difficult to ensure access to this communication method. While other studies using a variety of methodologies have successfully demonstrated authorship of the typed communication (Cardinal, Hanson & Wakeman, 1996; Emerson, Grayson & Griffiths, 2001; Janzen-Wilde, Duhcan & Higginbotham, 1995; Niemi & Karna-Lin. 2002; Sheehan & Matuozzi, 1996; Tuzzi 2009; Tuzzi, Cemin & Castagna. 2004; Weiss, Wagner & Bauman, 1996; Zanobini & Scopese 2001), or demonstrated that individuals can learn to type with minimal or no physical support or to read aloud what they have typed (Broderick & Kasa-Hendrickson, 2001; Kasa-Hendrickson, Broderick & Hanson, 2009; Rubin, Biklen, Kasa-Hendrickson, Kluth, Cardinal & Broderick, 2001) the practice remains controversial.

At the heart of the critique around supported typing is the question of whether or not individuals, who do not speak and are presumed by many to be incompetent, could possibly produce such coherent and sophisticated typed communication. "Are people who have been identified as cognitively disabled competent (or incompetent) to represent themselves? Is it possible that these people can have observable physiological, cognitive, or behavioral disabilities, but also exhibit behavior and thinking that could be termed "normal" — that is rational?" (Erevelles, 2005, p. 4). The critics of the method seem to challenge, "Whose voice is it?" What is the burden of proof of authorship for individuals who require support to communicate (Ashby & Causton-Theoharis, 2009)? Other individuals with autism who type to communicate have articulated challenges to their competence based on challenges to their methods of communication (Blackman, 1999; Mukhopadhyay, 2000; Sellin, 1997).

Jacob's typed messages and assignments produced at home were often met with skepticism and doubt. He had far more difficulty typing open-ended text with his in-school facilitator, who had substantially less training and experience, and was rarely supported to share in class. For students like Jacob who type to communicate, not only are the content of their messages scrutinized, but often the validity of the words. When validity and authorship are challenged, clearly voice is challenged as well. Take, for example, the typed words of Josh, a young man who was a participant in a study conducted with individuals who type to communicate in higher education settings (Ashby & Causton-Theoharis, 2011).

This forced me to consider the credibility of voice and whose voices are acknowledged and recognized as legitimate. Jacob and Josh have often been silenced, both by lack of access to the support they need to communicate effectively, but also by the persistent challenges to their competence.

What happens when voices are silenced not from without, but from within? As qualitative researchers, how do we make sense of times when our participants do not engage in a line of questioning that seems emancipatory to us? In the next section I explore the question of reading silence from a methodological perspective.

Hearing silence

Another important question concerns how to "read" silence or those moments when a research participant does not engage in a line of questioning that seems vital. If one goal of research is to help the individual give voice to his experience and provide an opportunity to talk back to the technologies of power that oppress, how can one make sense of silence? How should I interpret that seeming refusal to engage? Is it the result of internalized oppression and more evidence of the rigid power structures at play? In the case of this study, perhaps Jacob chooses not to "speak" out of an awareness of not being safe. Ellsworth (1989) describes how some situations are discursively dangerous. Sometimes is it not safe to speak.

I posed the following question to Jacob during one of our interviews, "What is your favorite class at Carter Middle School this year?" He responded that it was Health and provided the following follow-up information.

However, when I changed direction and asked him about school experiences that were not so positive, it was much more difficult to solicit explanatory information.

It took several minutes before I got any response to the question at all. When he did answer the question, the response was much more halting.

Jacob produced an extended string of sounds, with lots of consonants.

I was not able to get Jacob to return to that line of questioning and tell me what the teachers were yelling at him for. I made it clear to all participants that I would not pursue a line of questioning or a topic of conversation that made them uncomfortable and that they could always change the subject. However, in this case I so desperately wanted to hear what Jacob had to say. His perception of his teachers' treatment of him in class seemed so vital to understanding what school is like for Jacob. But, he would go no further. He would not take the next step of elucidating the situation any more for me. On the contrary, he took an almost apologetic stance and asked me to tell his teachers that he is trying.

How can I make sense of his silence on a topic that seems so visceral and so central? Did he not feel safe to speak? Was it because he felt scared to say anything negative about anyone at the school because he needs them? Was it because he felt he did not have the power or the privilege of dissent? Was he worried about getting someone else in trouble? Despite the many dehumanizing ways I observed him being treated in school throughout this initial study — talked about as if he was not there, forced to sit through classes with no way to respond, spoken to in an infantilizing manner, surrounded by adults at all times, ushered in and out of classrooms seemingly at the whim of the adults — was it possible that he did not feel he had the option of complaining? As a person who requires physical support to communicate and needs adaptations to class work and activities if he is to participate, he is in the precarious position of needing the minimal support he is provided at school, even if he knows it is not enough. But as I tried to make sense of these silences, I realized that I was once again imposing my own interpretations on the data.

Or perhaps, Jacob chooses not to "speak" at times because his voice has been so discounted in the past (DeVault, 1999; Lather, 2009). "[S]ilencing, the more active form of the term, is taken to refer not just to quieting, but also to censorship, suppression, marginalization, trivialization, exclusion, ghettoization, and other forms of discounting" (Devault, 1999, 177). If you do not believe you will be heard, why bother? It is also necessary to consider what will happen with the words once they are spoken or shared. This involves both what the researcher does with the words and what the eventual reader does with them once they are published. How are those words received? What use gets made of people's voices once they are shared? (Orner, 1992) Even the most conscientious researcher cannot guarantee that the published words will not be interpreted or used in a manner antithetical to the original intention of the work.

However, there is another alternative version of silence that is more empowering. Perhaps not speaking, or not typing, is also a way to subvert systems of power that limit and marginalize. "Voice clearly happens in spoken utterances 'voiced' by our participants, but does it not also happen when they/we fail to audibly voice an opinion with words and instead happen voice displeasure, discomfort, or disagreement with silence?" (Lather, 2009, 45). Should researchers be attending to what is unspoken as well as what is voiced or typed? Perhaps silence can be speech or at minimum a part of speech worth attending to closely (MacLure, Holmes, Jones, & MacRae, 2010). There may be agency in choosing not to speak that is more resistance than oppression (DeVault, 1999).

Agency and voice

In an earlier study on developing greater physical independence for individuals who type to communicate, Jacob had indicated a goal of learning to type with less physical support and perhaps speak aloud what he had typed. But about halfway through the strategy sessions designed to work toward this goal he typed, "independence to me is only pleasing myself and not others." The speech language therapist working with him on this project responded, "Are you trying to tell me that when you are here it is too hard to say no to anything I suggest?" to which he replied, "Yes. Independence isn't doing your ideas." He went on to explain in future sessions that there are times when independence seems very important and times when he prefers more support if it enables him to communicate more easily with his peers. This study (Rossetti, Ashby, Arndt, Kasahara & Chadwick, 2008) troubled notions of independence in communication and called for an expression of agency that recognizes the performative, contextually bound and dynamic nature of voice and choice.

Giving voice seems to imply that there is an authentic voice for the researcher to uncover and bring forth. If so, the job of the critical qualitative researcher is to "do the work of excavation" (DeVault, 1999). When I envision excavation, I picture someone with a pickaxe digging a hole, working to uncover a hidden, but fully formed, treasure. If the researcher just keeps picking away, eventually they will discover the "truth" of the person. There are two elemental problems with the assumption of an authentic voice. One, it denies the fact that all data is subject to interpretation.

While qualitative research provides readers with access to the world of people they would not otherwise know and to some extent allows these people's stories to be told, the subject never actually tells his own story (Bogdan & Biklen, 1998, p. 204).

The researcher brings her own interpretation to the data, situated within her own social location, which may or may not accurately reflect the participant's intended meaning. Also, the data presented are still dependent on the questions asked or the settings observed. The researcher may never have queried the topics the participant most wanted to share. While that challenge always exists with qualitative interviews, this is exacerbated with the participant does not spontaneously share verbal information. If individuals with disabilities struggle with initiation, expecting them to interrupt a line of questioning and interject a new idea is perhaps unreasonable. This places an even greater burden on the interviewer to consider alternative lines of questioning and to read subtle body cues that you may need to probe in a different direction. The presence of a facilitator within the interviewer/interviewee dynamic is yet another element to consider. Even if the facilitator is providing no physical support, she or he often repeats phrases or parts of questions posed by the researcher as a prompt to initiate a response. The words stressed, the tone of voice and the pauses all change the context of the conversation and perhaps the response.

The problematic nature of interpretation and voice is exacerbated when the research participant has limited expressive communication. In a typical interview with one of the non-disabled parents or teachers in this study of middle school inclusion, a one-hour interview would yield 20 to 25 pages of single spaced typed transcript. An interview with one of the other student participants might yield only a few pages and an interview with Jacob might only produce a few lines of typed text due to the time it takes to produce a response and the particular sensory challenges that need to be supported. There is just not as much data to work with; less data means a greater possibility of misinterpretation. With fewer supporting details there are many more blanks left to be filled in. There is a temptation as a qualitative researcher to want a "clean story," to find what Ellsworth (1997) refers to as a "beautiful fit" or the interpretation that helps the voice make sense with the rest of the narrative. We have to fight against the urge to seek consistency and cohesion in the data. Lather (2009) argues that over-empathizing and making assumptions based on our own sense of mutuality can actually serve to erase meaningful understandings of difference. Rather than stressing the ways in which we are "really all the same," perhaps we need to more adequately attend to the disjunctures and diversities of experience.

In relation to research methodology, is an attempt to access the "authentic voice" of the individual even possible or do assumptions about voice deny the performative nature of identity? Our knowledge of the world, and our existence in it, is always fluid and changeable depending on context and situation. Our identities, and therefore our voices, are always discursively constructed (Alcoff, 1995; Denzin, 1995). Similarly, the search for a unified, authentic voice denies the subject discursive agency (Alcoff, 1995). We craft an identity for ourselves, both consciously and unconsciously, that we present to the world. This fluidity and permeability of identity makes it impossible to imagine that one could uncover an authentic voice, which lies outside of discourse. Rather than discovering the "truth" about a person or uncovering an authentic voice, critical qualitative researchers should aim at working with the person to facilitate their construction and presentation of self.

How might we think about facilitating agency in the research process rather than giving voice, recognizing the performative, often interdependent, and contextually bound nature of voice? "Expressions of agency require contexts in which communication partners believe that an individual has something to say, provide communication supports, and possess the listening skills to 'hear' them" (Rossetti, Ashby, Arndt, Kasahara, & Chadwick, 2008). How do researchers support the agency of people with disabilities in the research process? How do we create space and opportunity for individuals with disabilities to express their own ideas?

Broadening the conceptualization of voice beyond verbal speech

Does "voice" assume speech? Orner (1992) states that the "discourse on voice assumes a fully conscious, fully speaking, 'unique, fixed and coherent self'" (p. 79). Atkinson et al. (2003) adds

The celebration and representation of voices, rather, implies endowing speaking subjects with a special significance. Giving voice to subjects includes the representation of individuals and groups who have been muted and marginalized. It implies the expression of their unique experience, usually through the reproduction of personal testimony and narrative.

The authors describe celebration of voice as "endowing speaking subjects with a special significance." While the message is clearly about empowerment and representation of marginalized perspectives, this passage also highlights a key challenge. Representation centers on speech. Spoken voice is clearly privileged in American culture. Where, then, does that leave individuals who do not speak or for whom speech is not always reliable? Consider the following example.

During an observation in Jacob's eighth grade English class, Jacob was scheduled to present a poster and report on a famous nineteenth invention with his partner Diana. When it was their turn to present to the class, Jacob stood up and walked to the front with Diana. She handed Jacob the paper and he held it out in front of himself. Diana read the paper and turned back and forth between the class and Jacob who was standing to her right. Diana talked about the invention of the bicycle in 1890 and stated that it was important because they did not have bicycles before that. Jacob laughed and clucked quietly. As the presentation continued, there was muttering and snickering from students in the class. Diana read another page about the invention of the washing machine and then they were done. One of the male students in the class sarcastically shouted out, "Good job Jacob." Several other students cheered as Diana and Jacob took their seats.

Jacob never spoke or participated in the presentation in any way other than holding the paper on which the material was written. I am sure he participated in the development of the project and the writing of the text, but that was not evident during the oral presentation. With no means of complex expression available to him in class, he was effectively silenced, rendered voiceless and unable to present himself as a thinking, capable member of the class. In an attempt to have Jacob present "without support" he was unable to present at all. His ideas and perspectives were never accessible to his peers. Normative conceptions of performance, participation and independence clearly impact the opportunities provided to individuals with disabilities. Some people with disabilities will always need the support of another person to communicate and make themselves heard. Traditional liberal theory leaves individuals labeled with cognitive disabilities outside the ranks of "citizen" (Erevelles, 2002). This often results from an emphasis on independence and utilitarian principles of liberal theory. In keeping with the idea of a self-sufficient, rational being as the ideal of American individualism, disability studies scholars have recognized that "the autonomous individual is imagined as having inviolate boundaries that enable unfettered self-determination, creating a myth of wholeness" (Thomson, 1996, p. 32). In western culture, independence is prized over interdependence and social good comes from "one's individual utility, intrinsic ability and personal performance with society being enriched when individuals attain their personal ends" (Kliewer, 1998, p. 3).

People considered to have cognitive disabilities, some of whom may never be able to function in ways that are considered independent according to Western traditions, are often seen or constructed as less then fully human. I would argue, however, that none of us operates truly independently and that the idea of inviolate boundaries is a myth for everyone. Unfortunately, if you require support from others to dress, or move or communicate those dependencies become justification for exclusion from the ranks of the American ideal and perhaps from the ranks of those who have a "voice" that merits attention. If we continue to conceptualize voice as speech, certain individuals with disabilities will always be constructed as being without one. If we only listen to a normative voice (Mazzei, 2009), one that looks and sounds familiar, we will recreate that which we already "know." However, if we think of voice more broadly as the ability to express oneself and be heard by others, that can encompass typed text, non-verbal communication, gestural communication or silence, which leaves the door open for a more expansive conceptualization of participation and engagement. Mazzei (2009) argues for an orientation of "listening in the cracks," hearing voice in all aspects of interaction, from the words spoken — or typed — to the gestures and silences. Listening for the unvoiced and the differently voiced can be messy, uncomfortable work, but it creates the opportunity for a fuller, richer understanding.

Facilitating Agency in Research Methods

The issues raised in this article have implications and import both to educators and researchers. Broadening the conceptualization of voice in the classroom is emancipatory for people who do not speak. The emphasis on verbal speech leaves many non-speaking people with disabilities further marginalized in their school and classroom communities. When speech is considered the only, or at least, preferred, way to express one's wants and ideas, that can limit the access and support provided for those individuals who communicate in nontraditional ways. Instead, teachers need to recognize all forms of communication, including assistive technology, sign language, and supported typing, and provide access to these systems and supports.

However, I want to focus on the methodological implications of a call to reconceptualize voice in disability studies research. Perhaps instead of attempting to "give voice" we need to "facilitate voice and agency," further leveling the researcher/researched hierarchy. This notion of facilitating agency recognizes the interdependent and dynamic nature of voice and critical qualitative research. Voice is not mine to give or uncover. I am not discovering a hidden reality and bringing it to light. Rather, voice is discursively created through the process of research. And if voice is discursively created, we have to acknowledge our own internalized discourses of normalcy and subjugation. "I cannot unproblematically bring subjugated knowledges to light when I am not free from my own learned racism, fat oppression, classism, ableism or sexism. No teacher [or researcher] is free of these learned and internalized oppressions (Ellsworth, 1989, p. 99).

If the goal is to facilitate voice in all of its forms, and to support individual agency, we must make every effort to include individuals with significant disabilities in the research process, not just as objects of research, but also as researchers. We need to engage people in all phases of the research process from project development to data collection to dissemination, such that we are not continuing to tell stories about people with disabilities, but facilitating the process by which individuals with disabilities can tell their own stories. We should aim to conduct research with people rather than on them. We need to thoughtfully consider how best to support individuals with disabilities to represent themselves in the research process, rather than always being the objects of someone else's representation.

When conducting research involving individuals with expressive communication differences, it is important to incorporate multiple sources of data. For example, pairing qualitative interviewing with participant observation allows for richer access to the experiences and perspective of the individual with a disability. Videotaping interviews can allow easier access to the body language, facial expressions and eye gaze of the participant, all of which can shed additional light on those "voices in the cracks" (Mazzei, 2009). Even if the person's verbal expression is limited such that asking them to comment at length on your data would be difficult, reviewing the data and then asking for a yes/ no response that you have "accurately" reflected their perspective is an important step in the data collection process. Too often, claims of accuracy imply a positivist orientation that assume there is one real truth of events and denies the contextual and performative nature of voice. The idea here is not to ensure the "truth" of what transpired, but rather to, as accurately as possible, reflect the participants intended meaning. It can also be useful to bring together the multiple data sources and consider them together in an interviewing session. For example, after conducting an interview in Jacob's class I would often meet with him at his house and talk about some of what I observed. Then he could comment on his interpretation of the events or provide his reaction to the situation.

During the interview sessions, it is also important to consider the sensory and movement needs of the participants. Conduct the interview in a location where the person feels comfortable and take frequent breaks to allow for movement and release. Responding through typing can be physically demanding and time consuming. Therefore, attention to fatigue levels and adequate wait time is essential. It also may be necessary to conduct the interview over several sessions. As individuals who type to communicate typically provide shorter responses than those provided verbally, it is helpful to return to the same topic a few times to develop a more in-depth response. It can be helpful to provide prompts such as "You told me last time that you do better in classes where there is more small group work. Can you give me an example of what that looks like?" Or, "You described being 'cast into a cold pool.' I am not sure what you mean by that. Can you explain it another way?" In my experience, participants who type have the greatest difficulty communicating emotionally charged information. Just as speaking individuals may find it hard to get the words out when they are upset, many individuals who type experience greater motor planning difficulties when they are anxious or stressed. It may be necessary to provide more breaks or support during those times.

I often struggled with knowing how far to push into uncomfortable territory. I often sensed that Jacob, and the other students in the study, had so much more they wanted to share but did not, or could not. Perhaps I needed to be more open about those observations, rather than focusing only on the data. If I am calling for a broader conceptualization of voice, perhaps that entails attending to those awkward moments more closely and theorizing together about their meaning.

However, it is also important to be cautious with typed communication and not give those typed words undue weight. The spoken word is transient. Someone can make a comment, laugh and then easily add, "Only kidding" or "What I really meant to say is…" It is too easy to attribute gravity to text when it is in black and white, especially if it fits with a theme we are seeing emerge in the data. As qualitative researchers committed to making space for marginalized voices, we need to also provide space for dissent and contradiction. It is important to frequently stop and say, "What I hear you saying is… Am I on the right track?" We have to remember how often these voices that do not speak have been overlooked, dismissed or even discounted as invalid. This increases our responsibility to proceed with care.

Finally, we have to carefully consider what is done with the research and who benefits from the telling. In the case of my project with Jacob, many of the students and their parents consented to be included in the project because they were displeased with the educational experiences they were afforded. They wanted a way to speak back to the system. I hope my work will bring about change in the lives of future students considered to have significant disabilities. However, I fear that some of the participants desired a more immediate return on their investment, a timely improvement on the circumstances of inclusion. Critical qualitative researchers need to be very clear about the goals of the project and the reasonable potential outcomes. Rather than assuming that my research project can itself be the force that brings about change, I need to see my role as standing with my participants and making space for their voices, whether spoken or typed, to be heard. Critical disability studies scholars have a responsibility to make central the voices of individuals too often relegated to the margins. We need to actively participate in the de-centering of professional or medical knowledge in favor of listening deeply to the voices and the silences of individuals considered to be disabled. But we have to proceed cautiously and humbly, always remembering to ask, "Whose voice is it anyway?"