Navigating violence and risk: A critical discourse analysis of blind women's portrayals of self-protective measures

Women with disabilities experience high rates of violence and harassment, yet meaningful violence prevention interventions providing the opportunity to learn how to be active agents in their own self-protection are virtually non-existent. To understand why, we draw on insights from feminist disability studies to explore some of the unexamined assumptions and discourses in gender-based violence prevention research. We then apply a feminist critical discourse analysis to focus groups with blind and partially sighted women to explore their talk about violence and self-defence to understand how they portray self-protective measures, and what practices those portrayals engender. We discerned three portrayals: self-protective measures as necessary against strangers, a delimited responsibility, and an effective means to an end. These portrayals and their subsequent practices demonstrate how the participants navigate violence while living with vision loss. We also consider the implications of our analysis for future directions in gendered violence prevention research.


Keywords
blindness, violence, self-defence, feminist critical discourse analysis, feminist disability studies, Canada It has been well established that violence against women with disabilities 1 occurs more often, for longer periods, and in more varied forms than it does against non-disabled women or men with disabilities (Chenoweth, 1996;Frohmader & Meekosha, 2012;Olofsson et al., 2015). Blind women, in particular, are at an elevated risk for physical or psychological violence (Olofsson et al., 2015) and have one of the highest rates of victimization by assault among persons with disabilities (Harrell, 2016). However, reducing that violence has proven challenging, and gender-based violence prevention (GBVP) programming designed to meet the distinctive needs of blind women, or women with other disabilities, remains virtually non-existent (Chenoweth, 1996;Frohmader & Meekosha, 2012; see Lund, 2011). Part of the reason is that women with disabilities are often omitted from GBVP scholarship, both as a focus of the research and as participants (see Mays, 2006); when they are included, their experiences are generally interpreted by non-disabled scholars according to "individualising moral and medical models of disability" (Goodley, 2017, p. 27) that do not account for the complexities of experiencing violence while living with a disability. Consequently, women with disabilities have fallen through the cracks, so to speak, of GBVP research, theory, and consciousness (see Chenoweth, 1996;Garland-Thomson, 1994). This exclusion has come at a costscholars lack adequate knowledge about how the materiality of specific disabilities (and especially sensory impairments like blindness; Rich, 2014) can affect women's capacities to navigate violence in its many forms; the likelihood of developing meaningful GBVP programming is reduced; and the tacitly accepted nature of violence against women with disabilities is maintained, contributing to the invisibility of disability more broadly.
In this paper, we have three aims. First, drawing on feminist disability studies (FDS) scholarship, we identify and challenge some of the unexamined assumptions and common discourses in GBVP research on women with disabilities 2 that hinder the development of effective interventions. Second, drawing on our own focus group-based research, we strive to better understand how these and other discourses can mediate the ways in which blind/partially sighted women respond to violence and portray selfprotective measures, including the practices those portrayals engender. Finally, we consider the implications of this analysis for GBVP research more generally.

Feminist disability studies (FDS)
In response to a history that has pathologized disabled bodies, positioning their existence as tragic instantiations of illness and loss (Watermeyer, 2009), FDS scholars have sought to illuminate how this perspective has situated disability "within a hierarchy of bodily traits that determined the distribution of privilege, status, and power" (Garland- Thomson, 1997, p. 6). FDS works to expose and challenge stereotypes about persons with disabilities, including assumptions about their lives and experiences, though it also: situates the disability experience in the context of rights and exclusions. It aspires to retrieve dismissed voices and misrepresented experiences. It helps us understand the intricate relation between bodies and selves. It illuminates the social processes of identity formation. It aims to denaturalize disability. In short, feminist disability studies reimagines disability. (Garland-Thomson, 2005, p. 1557 As part of this reimagining, FDS scholars maintain that experiences of disability and embodiment cannot be disentangled (Lourens, 2018), marking a departure from the social model of disability that locates disability within the organization of society. Acknowledging this entanglement makes needed room for scholars to investigate the embodied aspects of disability, as well as the ways in which disabled bodies are examined, excluded, devalued, and denied agency (Meekosha, 1998), making this lens apropos for uncovering some of the issues in GBVP research.

Issues in GBVP research
Exclusion: Women with disabilities and women's self-defence Concurrent with the pursuit of long-term, structural solutions, feminist self-defence has emerged as a promising resource against violence; indeed, scholars and advocates recognize that some situations demand that women use force, and research has illustrated that providing self-defence training reduces negative outcomes. Current feminist self-defence programs (Hollander, 2016;Senn et al., 2015) provide both psycho-social education on GBV and self-defence training to female college students, giving them the opportunity to be physical, to practise strikes, and to feel strong, while also exploring the gendered dimensions of violence and resistance from a feminist perspective. These interventions challenge the neoliberal agent/victim dichotomy by acknowledging that the two can coexist, and that women can resist GBV while also being given a "compassionate, blame-free space for victims [that maintains] focus on perpetrators as the only people responsible for violence" (Murphy, 2018, p. 79). Accordingly, feminist self-defence programming has demonstrated impressive results in reducing sexual violence, and for initiating far-reaching changes in participants' lives. Hollander (2004), for example, describes how participants reported feeling safer, being more likely to use public space, interacting more confidently with others, and that their perceptions of their bodies changed, with one participant explaining, "I see my own power and strength" (p. 221). Given that learning or applying self-protective measures interrupts discourses of vulnerability (Robinson, 2013), and feminist self-defence has been shown to reduce numerous forms of violence against women (Hollander, 2016;Senn et al., 2015), one might assume that adapted self-defence programming would be available to women with disabilities because they are known to be comparatively susceptible to violence and abuse. However, the struggle to develop and provide such resources has generally not included women with disabilities, demonstrating the extent to which alterity and its exclusionary processes informs GBVP (see Goodley, 2017).
Many GBVP programs, including feminist self-defence, are developed based on research undertaken with convenience samples comprised mostly of white, heterosexual, non-disabled college students; accordingly, such programming largely represents and caters to their experiences, maintaining GBVP as a pursuit for "normal" bodies. Thus, there are no courses with punchy taglines encouraging blind women to "Fight like a girl", and there are few opportunities anywhere in the world for them to learn and practise self-protective measures. In fact a recent meta-analysis identified unevaluated selfdefence programming for blind/partially sighted adults in only four countries (USA, France, Germany, Czech Republic) (Č ihounková et al., 2015). This type of exclusion constitutes a lack of recognition, which Frohmader and Meekosha (2012) argue is a form of harm that sustains inequality. While non-disabled women are permitted to know their bodies' strength and capabilities through opportunities to learn techniques at public institutions, no such opportunity seems to exist for women with disabilities.

Agency: Delimiting responses to violence
Such exclusion begets other problems, including delimiting agency. How women with disabilities navigate or respond to danger, harassment or violence is both facilitated and constrained by socio-cultural constructions of gender, disability, vulnerability, and victimhood, among other issues. As such, their perceptions of personal safety and selfprotective measures are complex and mediated by numerous considerations, including embodied elements of disability like the extent or severity of one's disability, mobility, and pain, as well as other factors, like age. When violence prevention programming for persons with disabilities does exist, it is often inattentive to these issues, or to how gender and disability converge to generate unique expressions of power, embodiment, and violence. Instead, researchers have argued that focusing on the "inherent vulnerability" of individuals with disabilities is key for the development and mobilization of resources (Martin et al., 2006), resulting in interventions that individualize and psychologize experiences of violence, and programs that seek to correct the individual's intrapsychic state. This orientation stems, at least in part, from quantitative research that tends to reduce a complex social category like "women with disabilities" into a blanket "special needs population" who are described as demonstrating dysfunctional behaviour, poor self-esteem, and over-compliance (see Mays, 2006).
Many studies review lists of factors thought to make women with disabilities more susceptible to abuse, sexual assault and/or violence, including "prior victimization, nonassertive behaviour, low socioeconomic status, and a lack of knowledge", as well as other issues like becoming so compliant with caregivers that reporting abuse becomes unlikely (Dryden et al., 2014, p. 794). Undoubtedly, these factors are salient, as is the embodied materiality of disability, which can inform experiences of violence and constrain one's ability to speak out against abuse or mistreatment (see Lourens, 2018;Peta, 2017).
However, in highlighting the individual's intrapsychic state with respect to their victimization without appropriately situating these issues against the larger socio-cultural and material contexts that inform their emergence, interventionists apply a neoliberal lens that draws attention away from the structural issues, societal norms, and embodied realities that women with disabilities face, which contribute to GBV in the first place. While many scholars often acknowledge that socio-cultural factors contribute to the prevalence of violence against women with disabilities, the interventions they develop typically aim to correct aspects of women's attitudes or personalities. For example, a study by Nosek et al. (2016) explored the effectiveness of a program designed to increase the self-esteem and self-efficacy of women with disabilities to reduce the likelihood of experiencing violence. Similarly, Dryden et al. (2014) conducted an evaluation of the IMPACT:Ability program, a ten-session safety and self-advocacy training program for young people with cognitive and physical disabilities aiming to increase participants' confidence and assertiveness. While programs like these may be beneficial on some level, their effectiveness is limited because they implicitly place responsibility for deflecting violence and abuse on the individual without also imparting the physical or material resources that assuming such a responsibility would require, like self-defence training. Consequently, individualizing and psychologizing interventions delimit women with disabilities' agency in the face of violence and fail to consider that some may want to learn selfdefence and could be strong, competent performers of techniques. Perhaps unsurprisingly, we found only two studies that successfully adapted self-defence techniques for blind/visually impaired adults, and none that focused on women (see Č ihounková et al., 2015;McCarthy et al., 2018), demonstrating the extent to which interventionists' assumptions about blind women's capacity to respond to physical violence are informed by gendered and ableist 3 discourses. For Garland-Thomson (2005), this illustrates how dominant narratives constrict the complexity of disability, and "limit the imaginations of those who think of themselves as non-disabled" (p. 1567).
Let us now turn to our analysis of blind and partially sighted women's portrayals of selfprotective measures. Such self-protective measures include physical acts of violence (e.g., self-defence techniques or other types of physical resistance), though they also include Skogan and Maxfield's (1980) concept of risk-management practices in which an individual attenuates their risk in a potentially dangerous situation or location through avoidance or taking up measures to make them less likely to be targeted for victimization.

Participants
Twenty blind or partially sighted women participated in this study. The inclusion criteria stipulated that participants be adult women, living in Canada, able to speak English, and be partially sighted or blind. Participants ranged in age from 24-74 years old and the median age was 57 (M = 54.8, SD = 12.71). 4 Recruitment took place in Fall 2017 when potential participants were solicited through the email list of the Canadian National Institute for the Blind (CNIB) and through their partnered community organizations' email lists. The first author (AF) also appeared on Accessible Media Inc.'s national online radio shows Kelly and Company and Live from Studio 5 to discuss the study's aims and to recruit participants. Interested individuals emailed AF, who then placed them in one of three focus groups based on their availability. Most participants lived in the Greater Toronto Area (n = 14), though five women were from other parts of Ontario, and one was from British Columbia. The participants also differed in other ways, including: type of vision loss (9 were blind; 11 had low vision); length of vision loss (7 had been blind since birth; 7 had lived with vision loss for over 20 years; 6 had lost their sight more recently); and the types of vision aids used (4 had guide dogs; 8 used the white cane, while the rest used both or neither).

Procedure
To be mindful of accessibility issues, AF emailed the consent form and a demographics questionnaire prior to the focus group so that participants could use their preferred assistive technology to review and fill them out. Both documents were returned electronically. Following approval from two Institutional Review Boards (York University and CNIB), three focus groups were conducted in Toronto, Canada in November and December 2017; two were held in person, while the third occurred via conference call. Focus groups ranged from 90-120 minutes and were recorded with the participants' permission using a digital audio recorder. AF moderated the groups and asked open-ended questions designed to initiate discussion on target issues, using probing follow-up questions that were unique to the conversational context. Examples included: Tell me about a time you felt unsafe; What safety advice would you give to another woman with vision loss or blindness?; When have you felt as though you were in danger, and what alerted you to the fact that something might be wrong? After the focus groups concluded, participants were given a list of emotional support resources in a variety of accessible formats (e.g., electronic, large-print, and Braille). In recompense for their time, AF (who is a trained self-defence instructor with experience teaching adapted techniques for blind persons) offered a 90-minute self-defence workshop. Refreshments were provided and participants' transportation costs were partially covered where applicable.

Approach to analysis
The audio recordings were transcribed, and pseudonyms were assigned, except for the first author who appears in excerpts as "AF". Using NVivo 12 Plus, inductive coding informed by Braun and Clarke (2013) was used to identify extracts related to selfprotective measures and experiences of violence, followed by a feminist critical discourse analysis (FCDA) (Lazar, 2005). FCDA provides a critical lens to explore "unequal social arrangements sustained through language use" with the aim of pushing for transformation, and ultimately, emancipation (Lazar, 2005, p. 1). According to Lazar (2005), an FCDA helps elucidate the complex ways that power and discourse co-constitute and sustain inequalities, especially those that are gendered; moreover, because gender intersects with other aspects of identity, like disability, FCDA makes room to explore how these convergences can (re)produce inequalities. This analytic approach allows us to investigate not only the ways in which power relations are (re)produced, negotiated, and challenged in language use, it also demonstrates that language has both "material and phenomenological consequences" (Lazar, 2005, p. 2). Finally, an FCDA offers the opportunity to engage in what Van Dijk (1991) refers to as analytical resistance; we use this approach to create new discursive possibilities for writing and thinking about women with blindness in GBVP.

Analysis and discussion
We identified three portrayals of self-protective measures that were either explicitly stated by the participants, or implied in their talk about experiences with, or anxieties about, violencenecessary against strangers, a delimited responsibility, and an effective means to an end. Many of the discussions (re)produced aspects of the institutional discourses and assumptions outlined above, though some women also resisted or challenged these ideas. At times, the participants drew on more than one portrayal of self-protective measures, illustrating the fluid and contradictory nature of these representations, as well as their complexity (see Jenkins & O'Doherty, 2020). Each portrayal also engendered unique social or material practices.

Necessary against strangers
In this section, we explore the participants' portrayals of self-protective measures as being necessary against strangers, though impossible to enact with vision loss. In important ways, the participants' talk echoed common concerns that non-disabled women have about so-called stranger-danger; unlike non-disabled women, however, the participants described experiencing numerous situations wherein they were harassed or victimized by strangers they believed to be motivated by ableism. Overwhelmingly, all three groups focused on the difficulties of navigating the threat of violence from strangers as blind women. In focus group 3, some participants discussed signs indicating when selfprotective measures might be necessary: Michelle: You have to counter the people who want to harm you, I think that's the key.
People ask me specificallythis is the only thing where I feel in danger, or when there's a dangerous person, that this might be a [dangerous situation]they ask me, "Can you see at all?" Victoria: Or they get right in your physical face, like two inches from you. AF: People start questioning how much can you see? Michelle: Yes! And that's when I start to get uneasy … When they ask me, "How much can you see? Do you need help?" in a way that they're not just trying to be helpful, they're trying to be invasive and inquisitive. To me, it's like, "Why? Are you a crooked person or somebody who's going to harm me?" And that's where I get that uneasy feeling, and those are the types of people that I think … There's people who just want to steal from you, or this or that, but there's people who want to prey on people with disabilities because it's an easy target for them, and we're not going to be able to defend ourselves. That's where they think they can win, and those are the people that I fear most because in their head, they have no conscience, right? So, they don't care that you have a disability or don't, or whatever.
Michelle opens the conversation by stating how important it is to resist those strangers who mean to harm her, offering examples of signs that alert her to the potential for violence. One of the embodied constraints of navigating violence while blind is being unable to gather visual clues to evaluate an unfolding scenario, leaving Michelle to rely on her highly developed ability to assess strangers' motivations by the kinds of questions they ask. In tension with her opening statement, and reiterating the socio-cultural assumptions about the relationships between vulnerability, capability, and disability that were observed in institutional discourses about GBVP, Michelle states toward the end of the extract that women with disabilities cannot defend themselves. This speaks to a narrow notion of self-protection that relies on sight and physical techniques, disregarding the participants' unique methods for risk attenuation. With no examples of GBVP programming that speak to the participants' experiences of violence or that validate the kind of risk assessment strategies they describe, it is unfortunate though understandable that Michelle's talk does not recognize the value of these approaches in negotiating potential violence, leading her to claim she is without recourse.
Some women described adopting specific practices meant to reduce the likelihood that they would be targeted for violence by strangers, such as avoiding the use of vision aids like the white cane. The participants reasoned that foregoing its use made them appear sighted and therefore less vulnerable. Members of focus group 2 explained it this way: AF: One of the ladies was saying that she tries really hard not to use the cane. Tara: I've done that a few times. I've only been using my cane for the last three years, and part of that was I fell down the last two steps in the subway because I was counting the steps and I miscounted! I fell, and I thought, "Ugh, if I had a cane, I wouldn't have fallen!" [chuckling] AF: Does anyone ever do this for safety reasons? Do you do this because you feel like you don't want to be a 'target', so to speak? Adrienne: Absolutely! Roberta: Yes! Shanique: Yes. Adrienne: If you have any visual devices that will let somebody know that at least you aren't going to identify them, I think that can be empowering for some people. They go, "Okay, there's a good target". I don't feel I have to use a cane yet, I really don't; I know it's on my radar, but to be honest, I'll go as long and as far as I can before I use a cane. Roberta: I agree with you. Lee: I'm with you there, yes.
In characterizing self-protective measures as being necessary against strangers though made impossible by vision loss, the practice of downplaying the nature or extent of one's vision loss is seemingly concomitant because it allows the women to circumvent aspects of the problem. Cureton (2018) refers to this as passing, where an individual passes as a member of an accepted social group or category, despite lacking important commonalities. Passing is often encouraged among persons with disabilities; for example, some blind children are taught to mimic eye contact, to walk without hesitation, to feign misplacing their glasses if asked to read, to not disclose their disability to others, to avoid using adaptive equipment, to eschew membership in communities for persons with disabilities, and so forth (Cureton, 2018). The partially sighted participants use passing as a self-protective measure; yet, there are trade-offs to consider. In choosing to forego assistive devices, the women become more susceptible to environmental dangers and risk being injured, as Tara's experience makes plain. Despite the shortcomings of this strategy, the women in the extract still express a preference for it, suggesting that passing is not only a means to navigate the potential for violence but may also be a way to negotiate identity and to resist victimhood, especially because Tara's talk implies it is better to be harmed by the environment than by a stranger. Influenced by neoliberal values, the subject position "victim" has been stripped of its social and political meanings and recast as something negative to be avoided at all costs; moreover, because only the psychological and subjective are scrutinized, analyses of violent experiences often suggest that victimization arises from the self, and is borne out of poor practices and/or choicesassumptions that engender victim-blaming (Stringer, 2014). Tara's choice to pursue passing as a form of risk management or self-protection is understandable, and further offers a unique glimpse into how women with disabilities negotiate gendered, ableist discourses about risk and violence in a neoliberal context.
In all three groups, participants discussed another kind of stranger unique to their experience as women with disabilitieswell-intentioned individuals imposing unwanted "help". These purportedly benevolent strangers were sometimes described as being scary, startling, and as doing physical harm to the participants. Brianna (focus group 2) recalled an instance of how a benevolent stranger grabbed her tightly and lifted her as she was preparing her guide dog to cross the street: While predation-minded strangers are described as requiring the women to adopt specific practices unique to their vision loss to mitigate risk, strangers imposing help when the participants have not requested it are regarded with more understanding. All the women characterized the imposition of that help as indicative of negative assumptions about their abilities; as such, they framed such interactions as requiring a different approach to self-protection, one that recognizes that these actions are more complex, motivated by good, though harmful, intentions. Despite Tara being moved and grabbed in ways that reduce her personal freedom, she employs a decidedly political strategy wherein she elects to teach this man that she must consent before he touches or "helps" her. This is a unique violence prevention strategy, though one that is not accounted for in typical GBVP models. For example, Ullman (2007) divides women's responses to violence or assault into four facile categories: forceful physical strategies (e.g., kicking, punching), forceful verbal strategies (e.g., yelling, threatening), nonforceful physical strategies (e.g., moving the offender's hand, avoiding being gripped, running away) and nonforceful verbal strategies (e.g., pleading, crying). It is difficult to categorize Tara's response according to this typology, especially because she portrays it as a calculated act of consciousness-raising designed to enforce her rights as an autonomous and capable individual. Thus, educating strangers can present an opportunity for blind women to invert power dynamics and challenge the previous portrayal of self-protective measures as being precluded by disability. By taking up the subject position of educator or advocate, women can transform frightening or frustrating experiences into ones in which they ultimately achieve feelings of empowerment in the face of violence, thereby positioning disability as an affirmative form of subjectivity (see Campbell, 2012).

A delimited responsibility
In this section, we explore how the participants' talk implicitly portrays them as being responsible for their own safety, while also recognizing that this responsibility is limited in unique ways for blind women. We examine the participants' safety advice for women with vision loss, which largely aligns with neoliberal police discourses that have been prevalent since the 1980s that position women as vulnerable to violence and responsible for adopting self-protective measures in the interest of personal safety (Moran & Skeggs, 2004;Stringer, 2014). The advice offered was similar across the three groups and will be familiar to many women: be confident, avoid looking like a victim, have your keys ready in your hand, know your environment, and so forth. Members of focus group 1 explained:

AF:
If you were going to give advice to another visually impaired or blind woman on how to be safe, what would you say? Ruth: I think part of it is about personal confidence, and not looking like a victim. AF: What does that mean for you? Ruth: Your posture, the way you carry yourself, sometimes some of those … for example, looking down at the ground, being unsure of yourself, and really looking over your shoulder. Really being comfortable in your skin and walking with confidence. Yeah, being confident.
[…] Lavinia: One of the things I have always attempted to do is to always look like I know exactly where I am going. Even if I have to lie! The person looking at me won't know. I always like to look like I know exactly where I am going. I used to live in Toronto, I lived there for 40 years, went to university for a number of years, and I even back then when I could still see, I made sure that everybody looking at me would think, "Oh, this woman knows where she's going" [chuckling]. It tends to take a person off you. If they're looking for a victim, it's not going to be me! This advice aligns with tendencies to both psychologize and individualize women's experiences of violence. Ruth's comments point to the necessity of adopting attitudes like confidence and self-assurance for staving off would-be attackers. Her talk (re)produces two common ideas: (a) women's behaviours alert predation-minded strangers to their internal states, such that appearing lost or uncomfortable, having poor posture or even looking at the ground might lead someone to assume that the observed woman lacks confidence and therefore would not be an effective resister of violence; and (b) that women are responsible for making changes to both their behaviours and to their internal states. Interestingly, some of the advice offered in this excerpt may not even be easily actioned by the participants (e.g., avoiding looking at the ground), or may not be relevant (e.g., looking over one's shoulder); accordingly, "common-sense" safety advice for women operates as "aspirational markers" for "normative citizen[s]" that overlooks the embodied constraints of disabilities because all "normal" women should display mastery over the body and mind (see Campbell, 2012, p. 214). Thus, "safety advice for women" is clearly not advice intended for all women.
In focus group 3, some of the participants describe other kinds of safety advice that they would offer to other women with blindness or vision loss: Dorothy: One thing I learned years ago (probably early '90 s or late '80 s)they were talking on CBC radio about that kind of thing, and just in general … especially muggers who will double back and walk by youmake sure that even though you can't see them, you look at the person who passes you, and all that awareness I think is really number one. Victoria: I think one very common sense onewhen traveling in the dark, avoid shortcuts, which may save you some travel time, but are more lonely-type of small streets.
[…] Dorothy: Even sometimes when there's lots of people around, they're into their own thingthey're looking at their iPads, their cell phones, and they don't react the same way people used to. They don't want to get involved in large cities. The exchange illustrates some of the gendered social practices that safety advice can inspire, and their implications. Though these recommendations supposedly reduce blind women's vulnerability to violence by imposing restrictions on their movement through the public sphere, their regulatory intent is not always obvious. Because restrictive advice is addressed only to women, it reinforces the notion that being fearful of and/or vulnerable to violence in public is a gendered concern requiring a gendered response (see Brownlow, 2005;Hollander, 2001). That response encourages women to be anxious, hypervigilant, to take up seemingly endless adjustments in their actions, or to even travel with another person who can offer protection, as Victoria advises. Note that none of the advice makes demands for larger, structural, and socio-cultural changes, and the participants' talk did not include exhortations to learn self-defence, 5 suggesting that their perceptions of their responsibilities are influenced by gendered and ableist ideas that delimit their choices. Women with disabilities are often not expected to fulfil the normative criteria of adulthood (Jordan & Tseris, 2017) and are associated with characteristics like dependency and submissiveness (Morris, 1991), which casts suspicion on their violence prevention adroitness. The women's talk implies that once a situation requires physical self-defence, this is the inflection point at which their ability to respond ends, and appeals to others for help are warranted, which is why it matters that (sighted) individuals might not be around, paying attention, or are disinclined to help.
An effective means to an end Importantly, some participants challenged the idea that disability precluded the enactment of self-protective measures, or that they understood their abilities as being restricted to small or passive actions. Several women described scenarios wherein they were competent and effective users of self-defence, drawing on a range of tactics, including verbal strategies, evasiveness, relying on other senses, counting the footsteps of someone following nearby to determine their pacing, pushing, kicking, and elbowing. In some cases, they used their vision aids to protect themselves; for example, in focus group 2, Andrea explained how her guide dog defended her: Andrea: My first six months of being blind, I was grabbed in downtown Niagara.
Somebody popped out from behind a building, grabbed me by the arm, "Give me all your money!" I had just gotten my dog, and my dog reared up like a horse and grabbed the guy by his forearm. He said, "Is this dog going to bite me?!" I said, "Probably!" Group: [laughing] Tara: If he doesn't, I will! Andrea: He started screaming and ran. I'm still standing there with tears. Nobody stopped, nobody even knew. But, when I got to the bus terminal, my husband (who's the bus driver) looked at me and he said, "What happened?" The dog was totally shifted off; he was being aggressive toward anyone who walked near us. He said, "Your dog is holding something in its mouth!" and my dog didhe ripped the guy's shirt! Group: [several participants marvel, exclaiming, "Wow!"; some are laughing] Lee: I love it! Andrea: It made me feel a little better inside knowing that I had my dog on my hand.
They're not trained for it; it is their instinct. But I think I agree that it's a deterrent by sight alone.
Though the participant's dog is the main driver of the action, the scenario demonstrates that Andrea is not without recourse in a threatening situation. The extract illustrates the complex relationship that blind women have with their vision aids, which offer both a means to navigate the environment and important self-protective functions, including operating as a deterrent or weapon, and in the case of guide dogs, a chance to feel safe despite being unable to see an attacker. Andrea recognizes this aspect of the interspecies relationship, which shapes her response; her description of the experience suggests she was emboldened by the dog's presence, answering the mugger's question with a sardonic, "Probably!", rather than shrinking in fear or even relinquishing her money. In discovering the effectivity afforded by the dog's presence, Andrea's talk resists the notion that she is helpless or ineffectual and demonstrates an often overlooked dimension of disability with respect to violence preventionthat adaptive equipment plays an important role in maintaining women with disabilities' safety, and can change the tenor of disability (Campbell, 2012). This approach also offers an important alternative to passing.
In other instances, participants' responses to the threat of violence suggested they were ready to fight, again challenging the idea that blindness precludes or naturally delimits self-protection and further demonstrating that self-protective measures can be effective when wielded by women with disabilities. In focus group 2, Tara described threatening an ex-boyfriend with a weapon: Tara's vision loss makes navigating an aggressive situation with a former romantic partner complicated, though not impossible. In fact, her response is similar to sighted women's, of whom approximately 70 to 80% use self-protective measures when faced with an assault (Hollander, 2018), with physical acts like pushing and kicking being particularly effective (see Fisher et al., 2007;Guerette & Santana, 2010). Tara's description also suggests she was not fearful; in fact, she downplays the threat of aggression by referring to her former partner as a "four-year old child", thereby diminishing his power and enabling her to respond with an improvised weapon. The extract demonstrates that Tara's self-protective measures were an effective means to an end that permitted her to successfully maintain her safety.

Implications
Because FCDA "can produce a rich and powerful critique for action" (Lazar, 2005, p. 5), we consider the implications of our analysis for GBVP programming for blind women, though some points may also be valuable for interventions with other groups of women with disabilities. It is important to understand how blindness can shape women's experiences of violence, including not only the embodied aspects of disability but how assistive devices can be usedthis is one of the unique contributions of this study. The design of interventions should consider first validating and then helping women navigate the embodied constraints of disability with respect to self-defence, affirming the value of their devised methods for evaluating and responding to violence (e.g., discerning the meaning of questions, passing, education, reliance on assistive devices or sighted persons, etc.) while contributing to the development and practice of additional techniques. Many self-protective actions can be soundly adapted, especially once scholars have established relationships with blind communities to learn about their experiences, abilities, and violence prevention objectives. Creating interpersonal connections may also help challenge the assumptions constraining GBVP researchincluding what strategies or interventions are deemed appropriate to pursuewhile concurrently offering opportunities for allyship by positioning women as expert citizens rather than just service users (Goodley, 2017, p. 39).
The participants' talk suggested that they struggle with navigating the embodied constraints of blindness and do not expect to be able to defend themselves; at the same time, they also described innovative strategies to prevent or respond to violence, and demonstrated a willingness to fight back, if needed. Accordingly, the women's discussions conveyed a fortitude, creativity, and aptitude for self-protection, as well as a desire to learn self-defence and other physically demanding skills to maintain their safety that has not been acknowledged elsewhere in GBVP scholarship, which could be built upon in developing practical interventions.
The women often echoed common discourses about GBV, especially with respect to fear of strangers. Feminist GBVP programs emphasize that women are most at risk for violence from acquaintances and partners, intimating that fear of strangers is largely misplaced; however, the participants' talk suggests that this fear is justified, citing numerous experiences of stranger-perpetrated sexual assault, harassment, muggings, and physical violence. In portraying self-protective measures as being necessary against strangers, the discussions create new space for scholars to critically examine the decades-long focus in GBVP on devising responses to acquaintance assault, when blind women describe needing help responding to strangers, too.

Conclusion
Our conversations with blind and partially sighted women demonstrate the extent to which both the material nature of their disability and socio-cultural and institutional discourses about vulnerability, victimhood, disability, and gender influence their ideas and actions. Some key implications of our study include: • Scholarly research and GBVP programs must be more attentive to women with various disabilities; efforts to understand both women's specific material conditions (e.g., nature of the vision loss, available assistive technologies) and social discourses that shape perceptions of disabilities must be scrutinized. • GBVP's emphasis on acquaintance assault may inadvertently be overlooking legitimate threats from strangers. • Actions of well-meaning strangers intending to help can nevertheless be harmful, and may represent important avenues for education of broader publics. • Blind and partially sighted women may, in some circumstances, be under-estimating their own capacities and abilities to protect themselves from harm, thus denying them an important sense of agency in their navigation of physical and social spaces.
This study further illustrates the nuanced meanings that blind women attach to selfprotective measuresthey are a means to maintain personal safety, though they are likewise an opportunity for affirmation, identity development, exploring and challenging embodied elements of disability, knowing the self as empowered and capable, and navigating fears. Thus, excluding women with disabilities from GBVP interventions is more than an issue of accessit is a poignant example of how exclusion facilitates a general lack of recognition of the depth and complexity of women with disabilities' experiences of violence that ultimately sustains inequalities and allows the wheels of GBV to continue to turn (Frohmader & Meekosha, 2012).