Research Article
What Is a Service Animal? A Careful Rethinking
Margaret Price, Ph.D.
Ohio State University
Abstract: I argue that the discursive tactics used to maintain a clear boundary between “legitimate” and “illegitimate” service animals rely on a set of assumptions that perpetuate unequal relations of power, and ultimately harm others (human and nonhuman alike). In support of this argument, I outline my theory of crip spacetime, which draws upon the material feminist notion that disability is an intersectional and emergent phenomenon, becoming (rather than being) through intra-active environments. Thinking through the ontology of service animals and their human companions in terms of crip spacetime demands that we apply what Christine Kelly (2016) has called accessible care in relationships.1

Keywords: service animal; ethics of care; spacetime
      I am returning home from the 2016 National Women’s Studies Association (NWSA) conference in Montreal, Canada. I arrive at the Columbus International Airport, where my partner Johnna will pick me up in a few minutes. Ordinarily I would take a car service home from the airport, but this evening I am utterly undone, in tears, hardly able to find my way out of the building. I texted Johnna in the middle of the night, telling hir what happened, and now ze is on hir way to meet me.1 
       
      As I step through the outer doors by the baggage claim, into the smell of car exhaust and cigarette smoke, I see a sign on the wall through which I’ve just passed:


Image description: A greenish-white stenciled sign on a glass wall. It comprises three sections, arranged vertically. The first section, accompanied by a stencil of a medium-sized dog with a circle around it and a slash through the circle, reads, “No Animals except seeing eye / comfort dogs or those being transported by air. Proper restraints required.” The second section also has a graphic, this time of a handgun and a knife being crossed out by a circle and a hash mark. The text in the second section reads, “No weapons or joking with weapons like toys. Violators will be prosecuted.” The third section’s graphic is an upright stick figure in profile in a wheelchair, accompanied by text reading, “This facility is completely accessible to all persons with disabilities.”

            I offer this sign as a starting point to explore a question that’s been on my mind for many years: What is a service animal? I started asking the question because I was personally concerned about my own service dog, Ivy—which means I was also concerned about myself. But as I continued to ponder Ivy’s and my ways of moving in the world, I realized there are larger questions to consider. These include the meanings of “service” and “care”; what it means to make care into a site of labor performed by animals; what it means when we combine animals’ capacities and disabilities with our own; and what all that might teach us about the intersecting qualities of animal, human, and world. In this essay, I explore these questions through a theory I call “crip spacetime.” Described in more detail below, crip spacetime draws upon both structural and new materialism to explain the ways that disability “becomes” through space, time, objects, texts, and organisms. Particularly, crip spacetime can be discerned in the tensions, affinities, and violences between and among those—for example, as a wheelchair becomes both a gendered person and a body part (see Belser, 2016), or as the mineral lead becomes racialized (see Chen, 2012).
      My overall argument is that the discursive tactics used to maintain a boundary between “real” and “fake” service animals rely on a set of assumptions that perpetuate unequal relations of power, and ultimately harm others (human and nonhuman alike). However, this project carries risk, because any attempt to re-define “service animal” has the potential to place people and animals at risk of harm in other ways. I take seriously what it means to question this definition. Handlers’ lives, not to mention livelihoods, often depend on their service animals. The appearance of poorly behaved animals who are claimed as service animals does harm to other teams, who are already subject to constant comments, harassment, and discrimination (see Siler). Thus, the stakes are high, and I am rethinking with care. 
No Animals
      A different, but more common, starting point could be the legal discourse that surrounds and constitutes service animals. According to the Americans with Disabilities Act (ADA), a service animal is a dog or miniature horse “individually trained to do work or perform tasks for a person with a disability.”2 Service dogs and service miniature horses can enter restaurants, stores, schools, and public buildings; fly on airplanes and ride buses; live in housing where pets are not allowed; and so forth. Legal restrictions are few. If the animal “compromises legitimate safety requirements that are necessary for safe operation” (e.g. is too heavy for a small plane) he may be legally turned away.3 Also, if she isn’t housebroken or well controlled by the handler, she can legally be refused entrance to a public space such as a restaurant. U.S. law specifically forbids asking about service animals if their functions are “readily apparent.” Even if the function is not readily apparent, questions are strictly limited, and in most cases, no documentation or proof of certification is required.
      Looking outward from legal definitions, it’s well documented that animals have been providing service to humans for thousands of years.4 Rebecca J. Huss (2010) notes that “dogs were kept at healing temples in ancient Greece,” and that pictures of dogs leading blind people have been found in thirteenth-century Chinese scrolls (pp. 1166-1167). This history is repeated in many articles about service animals, sometimes with additional details about early examples of animals being specifically trained to assist disabled people (Wenthold & Savage, 2007), and gaining more detailed focus in the period following World War I, when dogs were first trained to serve as “seeing eyes” for blind veterans (Eustis, 1927; Eames & Eames, 2001). As the twentieth century progressed, the forms of care deemed “service” proliferated. Animals had been formally trained to work with deaf and Deaf people at least by the 1970s, and shortly thereafter were documented as providing service for “depression, panic disorder and post-traumatic stress disorder” (Huss, 2010, p. 1169; see also Eames & Eames, 2001). With the 1990 passage of the Americans with Disabilities Act (ADA), and its subsequent amendment in 2010, the legal definition of “service animal” in the U.S. was codified. However, as a flood of reports from scholarly, personal, and popular news sources show, the interpretation and enactment of this definition have been anything but simple.
      I mentioned that I have a service dog, Ivy. She’s about nine years old. She also weighs five pounds, cannot walk fast due to patellar luxation, and is almost completely blind. When she accompanies me to work, on errands, to doctor’s appointments, or on trips, she rides in a carrier that hangs by my right hip. Details of her breeding and puppyhood are unknown. I do know that she lived the first part of her life in traumatic circumstances (mostly neglected, sharing a house with 27 other dogs), after which she was rescued by an organization in north Georgia. I adopted her in 2010.
      The details in the previous paragraph may have already disqualified me as a legitimate handler of a legitimate service dog in some readers’ opinions. It’s often assumed that “service” by an animal requires substantial size and strength. Popular accounts about fake service dogs often seem to get in digs about the suspicious nature of small size, as in this 2009 article quoting a grocery-store worker: “Like when you see little Foo Foo in someone’s purse, you know that’s not a service animal” (Yardley, 2009). The crudeness of this attitude is not shared by most advocacy and/or training organizations, though, which are well aware that service animals come in many sizes, and that their human companions’ disabilities may not be immediately apparent. For example, Anything Pawsable, a news and information magazine for the service-animal community5, features an item titled “How Can I Tell if a Service Dog is Legitimate?” which argues that the most accurate way to identify a service animal is not through size or breed, but through behavior (n.d.). However, Anything Pawsable also includes dozens of items aimed at drawing a sharp distinction between “legitimate” and “illegitimate” service animals; in fact, the category “Fake Service Dogs” occupies its own page on the site, along with “Gear and Equipment” and “Training Tutorials.” The reasons for maintaining such a sharp distinction, according to Anything Pawsable, are many: for instance, fake service animals may hurt or frighten members of the public because of a lack of training, which “casts a shadow on the entire [service animal] community”; they may distract legitimate service animals from doing their jobs; and they may undermine the seriousness of a service animal’s purpose in the first place (“The Hidden Complications”).
      In other words, even though Anything Pawsable recognizes that service animals and their human companions are extremely diverse, it still emphasizes the importance of maintaining a boundary between the real and the fake. That same policing impulse—to draw sharp lines between “real” and “fake,” “legitimate” and “illegitimate”—is shared by many service animal breeders, trainers, and handlers.6 This process is what Ellen Samuels (2014) would call biocertification—that is, constantly looping bureaucratic processes which take as their goal “the determination of the ‘truth’ of disabled bodies” (p. 123). As Samuels emphasizes, these looping processes are never about disability in isolation, but always caught up with other often-policed categories, including race, gender, and citizenship. Correspondingly, the tactics used to maintain the boundaries around “service animal” draw from many intersectional discourses. In the next section, I explore a number of those discursive moves, which invoke topics as broad-ranging as health, safety, fitness, obedience, excellence, measurement, and dirt. These disparate topics all call forth a similar public anxiety about bodies out of place: bodies that don’t look the way they are supposed to, function according to standards of “excellence,” excrete in ways considered “dirty” rather than properly contained, and so forth. My purpose is to question the governance of animal as well as human bodies, and especially the governance of animal-human relations.
Proper Restraints Required
      I carry a piece of paper everywhere I go: a letter signed by my psychiatrist. The letter specifies the tasks Ivy performs, and that I have been diagnosed with “more than one serious mental illness listed in the Diagnostic and Statistical Manual of Mental Disorders 5.” It also lists the nature of the limitations on major life activities that my disabilities cause, and describes how Ivy’s tasks mitigate those limitations.
      I rarely mention this letter, even if Ivy and I are challenged as we enter some space. Usually, I’d prefer to take a minute to educate the questioner, both for my own and for other service-animal companions’ sakes. But if the negotiation becomes lengthy or if I’m particularly debilitated that day, I may say something like, “I can show you a letter from my doctor,” which almost always resolves the conversation in my favor. For some reason the mere invocation of the letter serves as its own evidence.
      Crip spacetime as a theory attempts to map the world in terms of the knowledges that disabled people have been communicating—to each other, and increasingly, to others—for quite some time now. It is not intended as a model, but I do offer it as a corrective to some of the concerns about the social model that have led over the past 10-15 years to crip theory, cripistemology, and critical disability studies. These concerns often point to a series of overly simple divides posited by various versions of the social model, including person/environment and impairment/disability, as well as the attempt to collapse all disabled bodies into a single (implicitly white, male, cisgender) figure. In response, crip spacetime turns its focus away from the human individual to focus on the spatial, the relational, the non-human animal or object, and the group. In crip spacetime, there is no clear distinction between “person” and “environment”; rather, disabled life (and death) become through a complex dance of space, time, objects, texts, and organisms. In this way, the theory of crip spacetime draws upon Karen Barad’s (2007) new-materialist theory of intra-activity. However, crip spacetime also draws upon the more structural-materialist point of view of writers like Nirmala Erevelles (2011, 2014), who insists that the violences of inequality—for instance, racial and global inequalities—must be not only “recognized” but must be at the center of what we understand as the becoming process (pp. 28-29). I am not calling for a both/and move, but rather, drawing upon Christine Kelly (2016), for an ongoing and hopefully useful ambivalence between these two approaches.
      Thinking through situations in terms of crip spacetime demands that we apply what Kelly (2016) has called “accessible care” in relationships. This form of care recognizes that intimate relationships are always emerging in the context of larger systems of power and violence; that we cannot choose sides among independence, dependence, and interdependence, but rather must constantly navigate the tension among these concepts; and that we must be willing to dwell with a certain amount of ambivalence. Kelly argues, “Ambivalence provides breathing room by allowing some of the seemingly irresolvable debates to simply remain irresolvable” (p. 40). Accordingly, crip spacetime refuses the desire to purify disability into a nugget of information—e.g., “My disability is this, so access (or service, or care) will look like that.” Rather, crip spacetime asks us to question notions of consistency, individuality, functionality, and coherence when they are applied as evaluative tests for who and what should be valued, and in what ways. Moreover, it demands a focus on both the particulars of intra-active situations as they unfold, as well as the systems that produce and justify unequal, violent relations based on human exceptionalism (Taylor, 2017; see also Ferguson, 2013). In the following sections, I apply crip spacetime to a specific example, my own service dog, by analyzing the discourses that seek to define “real” service animals—as well as other discourses, which resist those boundaries by seeking a more capacious way of recognizing service and care from/with animals.
      Ivy would be instantly ruled out as a potential service animal by most breeders and trainers. Trainer Kea Grace’s (2016a) article “10 Things That Make a Dog Unsuitable for Service Dog Work” names “structural imbalances” and “vision or hearing problems” as two of the dealbreakers; Ivy has both. Grace also lists a range of issues Ivy does not have, but which similarly resonate with disability and fitness7: “genetic illness,” being “overweight or obese,” and “timidity.” The emphasis on health continues in another article by the same author (Grace 2016b): not only should a service animal be housetrained, but any illness causing gastric upset should be “very, very, very, very rare.” (This sentence made me pause and wonder about illness that might cause only “very, very” or “very, very, very” rare occasions of upset.) The importance of near-flawless health and behavior is repeated over and over again, not only in Anything Pawsable, but in myriad other publications dedicated to explaining what service animals are and are not. Assistance Dogs International (ADI), for instance, puts it this way:
      Assistance dogs have to work to the highest possible standards of health, welfare and task work. The standards are a critical guide for all ADI member programs and are vital in defining what an assistance dog is. 
With the phrase “what an assistance dog is,” ADI ties excellent behavior and health to the very definition of service animal. If a dog is badly behaved or unsound, it’s not just a bad service animal—it’s not a service animal at all.
      To be clear, I’m not saying service animals should be incontinent, riddled with anxiety, or disruptive. Rather, I’m trying to unearth the assumptions surrounding service animals—assumptions about health and fitness, of genetics and breeding, and of contamination and safety (including frequent discussion of excrement). These themes reflect the arrangements of life and matter identified by Mel Chen (2012) as “animacies”: a hierarchichal ordering based upon “understandings of lifeliness, sentience, agency, ability and mobility in a richly textured world” (29). Following Chen, I’m interested in how this ordering works, and also how it is tied to other orderings having to do with consistency, productivity, and safety. If we considered the possibility of a service animal-human relationship that was more mobile, even more fragmented—less coherent—than its conventional form, what new possibilities might emerge for ways to think about disability (and human animals, and animals)?
      Thinking in terms of crip spacetime means recognizing the possibility that a five-pound, bow-legged, blind rescue dog could be fully fit to participate in a relationship of care that qualifies as “real” service. But it also means recognizing that such an investigation may do harm as well as good. The presence of service animals in public space is usually fraught, and there are good reasons to uphold an image of service animals as near-perfect organisms. Companions of service animals are constantly challenged, questioned, and discriminated against; meanwhile, the animals themselves are subjected to uninvited greetings and touching. Emily K. Michael (2016) tells of moving through the world with her service dog York, routinely deflecting people “digging for medical details or distracting York with high-pitched puppy talk.” Sometimes, such encounters turn violent, as in Stephen Kuusisto’s (2016) “Denied a Cab Ride, Grieving for Who We Are.” In this essay, Kuusisto details the “contempt and mean-spirited bullying” he experienced when trying to arrange a cab ride from Detroit to Ann Arbor. The story has many painful turns, including this one, which occurs just after Kuusisto mentions that he writes for the New York Times:
      He [the cab driver] began shouting that Donald Trump had won the presidency and “you people” (apparently meaning blind New York Times readers) “don’t matter anymore.” He was absolutely vicious and crowing about how people like me don’t matter. 
Elsewhere in the same piece, Kuusisto describes the driver as “sneering” and “abusive.” And his story is not an anomaly. Advocacy sites and message boards within the service-animal community offer thousands of other examples. Moreover, harassment may be the least of a companion’s problems: her life, not to mention her livelihood, may depend upon her service animal remaining undistracted and unthreatened. Little surprise, then, that disabled people are sometimes among the fiercest defenders of a clear line between “real” and “fake,” “fit” and “unfit” service animals.
      And yet it is also disabled companions of service animals—including Kuusisto (1998; forthcoming) and Michael (2016)—who write about this relationship with the most subtlety and the deepest attunement to “becoming” in company. Rod Michalko (1999) calls the relationship “The Two-in-One”:
      At one time, I am master; at another, [guide dog] Smokie is. Now I am handler, now he is. … This is a fluid relation that does not apply when leader and follower are understood as static and completely separable entities. Thus mastery, handling, and ownership are situated phenomena and not ontological ones. … We move together as one, touching and imagining both each other and our world. (p. 185) 
The relation Michalko describes is strikingly like the intersectional process of “becoming” discussed by Erevelles (2011, 2014). Note that Michalko describes himself and Smokie as “imagining both each other and our world.” Moreover, though it was published almost 20 years ago, Michalko’s description of himself and Smokie as bringing one another and the world into being resembles Barad’s (2007) theory of intra-activity. Pre-existing human and pre-existing animal do not come together in a pre-existing environment; rather, the relation constitutes human, animal, and the matter of the world.
      But becoming is not a neutral process (Erevelles, 2011, 2014). It unfolds through and because of unequal, often violent, histories. For animals, those histories include subjugation, torture, and neglect, as well as (in the case of dogs) a long entanglement with humans through which the two species became mutually reliant (see Taylor, 2017). For humans, those histories include colonial violence, ableism, and racism, as Erevelles explains, as well as a struggle to figure out how we understand ourselves, other animals, and objects as members of what Kelly Oliver (2016) calls “the moral community” (p. 248; see also Ferguson, 2013).
      I have been wondering “What is a service animal?” (and “What is service?” and “What is care?”) because these questions demand that I also ask what it means to be (or rather, become) a disabled human in the world. Oliver (2016) points out that attempts to police boundaries between humans and animals through criteria such as functionality or sentience lead to “the nonproductive type of line-drawing” that inevitably seems to end in declaring a hierarchy of ways to exist (p. 253). I want to take up Oliver’s suggestion that we understand people and animals through a version of feminist care that emphasizes “an ethics of proximity” (p. 242; see also Ferguson, 2004). This form of care might move beyond an individualistic model of “service”—one animal, one human—to consider more complicated forms of collective care. But it would also consider the material costs of doing so.
Comfort
      Mid-November 2016. The NWSA conference-goers, thousands of us, arrive in Montreal reeling from Donald Trump’s election just a few days before. We invoke it in our presentations, we tell stories about how we learned the news, we organize resistances, we cry. We hold onto each other longer than usual when we meet and when we say goodbye. Many of us avoid the windowless session rooms with their rows of front-facing chairs, sitting instead in small groups, talking, touching each other. We seek comfort.
      Ivy is not with me, for complicated reasons. I hesitated over the decision. I had to weigh it against the stress that the trip would have placed upon both of us—particularly since it involved crossing an international border—and I ended up deciding to leave her home. (One of the ironic things about being accompanied by an animal whose job is to ameliorate panic is that the constant questioning and periodic harassment that flare up in her presence may actually bring on panic.) Only after I arrive at NWSA, when my friends express sorrow at Ivy’s absence, do I realize that she could have comforted them as well.
      The term “comfort animal” is archaic, but still crops up with surprising regularity—for instance, stenciled on the glass wall of the Columbus, Ohio airport. It vibrates with notes of domesticity and home culture (“comfort food”); of violent transnational and sexual histories (“comfort women”); of animals themselves (“creature comforts”). If the term didn’t make me so uncomfortable, I might even claim it, since comforting is one of Ivy’s most important jobs.
      But comfort is not supposed to be a service animal’s job—or “task,” as it is defined by U.S. law. A service animal, according to C.F.R. § 36.202, must “do work or perform tasks” for a person with “a physical, sensory, psychiatric, intellectual, or other mental disability.” Comfort is, in fact, explicitly outlawed, according to Title III:
      The crime deterrent effects of an animal’s presence and the provision of emotional support, well-being, comfort, or companionship do not constitute work or tasks for the purposes of this definition. 
This emphasis on “work or tasks” is frequently used to govern the difference between “legitimate” and “illegitimate” service animals. Joan Esnayra, founder of the Psychiatric Service Dog Society, notes that ADA lawsuits may be won or lost on the word “task”: “If you say ‘comfort,’ ‘need,’ or ‘emotional support,’ you’re out the door,” she is quoted as saying in a 2009 New York Times article. “If you talk about what your animal does in terms of ‘tasks’ and ‘work,’ then you stand a chance” (Skloot, 2009).
      The distinction between service animals and emotional support animals (ESAs) is similarly defined in terms of work or tasks, and comprises a curious web of overlaps and differences. While service animals may enter more spaces, they must be dogs or miniature horses; by contrast, an ESA may be any species. The variety of species permissible as ESAs has led to a number of sensationalized news stories reporting on claims made on behalf of ferrets, iguanas, ducks, pot-bellied pigs, monkeys, cats, goats, boa constrictors, kangaroos, and parrots. Patricia Marx (2014) wrote a first-person account that describes obtaining a letter from a “therapist type” and then entering various public spaces with (one at a time) a turtle, a snake, a turkey, an alpaca, and a pig. Marx’s piece is written for comic effect, but her tone is often dismissive, if not sneering. For example, she notes that the National Service Animal Registry recorded 11,000 emotional-support animals in 2013, then asks, “What about the mental well-being of everyone else?”
      Although pieces like Marx’s are meant to be taken lightly, their appearance signals a growing sense that the proliferation of many different kinds of ESAs (including reptiles and rodents), combined with legal confusion over which kind of animal qualifies for what, may be undermining and endangering people with “legitimate” service animals (Kogan et al., 2016; Skloot, 2009; Teitell, 2013). As a result, organizations that advocate for service animals and their companions have doubled down on their exacting definitions of “task,” “work,” and “support,” in an effort to debunk widespread charges of fakery. For example, Please Don’t Pet Me, an advocacy site founded and maintained by service-animal handlers, offers a dedicated article on “Understanding the Differences Between Tasks and Work,” while Anything Pawsable offers a full article to distinguish between “tasks” and “natural behavior.”
      Attempts to define “task” often link this concept to another one: trainability. Anything Pawsable quotes trainer Susan Lilly Grace: “Any behavior offered by the dog that isn’t directly trained and linked to a cue is considered a natural behavior” (McCormack, 2015). Thus, if a disabled person begins to shake from anxiety, and her dog jumps on her lap in a comforting way through his own initiative—but not because he was trained to do so on that specific cue—that is a “natural behavior,” not a “task.” This trainer also mentions that “companionship” and “emotional support” don’t count as tasks, because they cannot “easily be verified on a specific cue.”
      That argument isn’t hard to agree with if the task in question is something that can be physically observed, such as jumping on a person’s lap. However, if we think about less easily observable behaviors, where does “task” end and “natural behavior” begin? For example, let’s return to Susan Lilly Grace’s point about the “natural” behavior of a dog trying to comfort his human companion. What if the comfort required is not periodic (and thus cue-able), but constant? I clicker-trained Ivy, and I know that in order to cue a dog, you need at least two things (besides the clicker): you need an occasion on which the dog is doing the behavior (the “task”), and you need an occasion on which the dog is not doing it. As the trainer, you must click at the very instant the behavior occurs. Your dog learns that hearing a “click” means Right! Yes, that! That thing you did at that exact moment! So if I click for Ivy just as her hindquarters touch the floor on a sit, the sound tells her Hindquarters on floor=you met the goal. According to clicker-training methods, a reward should follow the clicking, so that your animal remains invested in the training. But the click itself is not the reward; rather, it’s a signal of when the rewardable behavior occurred, and an assurance that a reward will be forthcoming at some point.8
      But at what point does providing support or comfort become a task? The desire not to engage this question leads to murky, sometimes rather tortuous, use of language. For example, the ADA National Network explains that ESAs “sometimes help with depression, anxiety, and certain phobias, but do not have special training to perform tasks that assist people with disabilities.” (As a trainer, it is tempting to imagine perverse versions of cue-able “tasks” that would transform this “help” into “service”—for example, if an animal does silly tricks on cue, thus making the depressed or anxious person laugh and feel better, is she now a service animal?) The ADA National Network document attempts to clarify the “task/support” distinction by explaining that a psychiatric service animal’s tasks might include “reminding the handler to take medicine, providing safety checks or room searches, or turning on lights for persons with Post Traumatic Stress Disorder, interrupting self-mutilation by persons with dissociative identity disorders, and keeping disoriented individuals from danger.” Some of these are easily recognizable as individual tasks that can be trained and cued; others, such as “keeping from danger” might be more difficult to pick out. As someone who is routinely disoriented with brain fog and cognitive delays, I can testify that one of the most disorienting features of disorientation is that you often don’t realize you’re disoriented, or to what extent. Ivy certainly helps keep me from danger in such situations, but not usually because I’ve given her a specific cue.
      Tasks performed by psychiatric service animals seem to present particular difficulty to those attempting to interpret the “thicket” of laws and regulations surrounding service animals (Lipka, 2011). An article aimed at helping university counselors form policy, for example, explains that a psychiatric service animal might “prevent” compulsive or destructive behavior (Kogan et al., 2016). Is “preventing” something a task, if that task must be cue-able? It’s not hard to see that one could make an argument for “preventing” in terms of observable, identifiable moments—e.g. pulling someone’s hand from their mouth if they’re biting their fingers, or preventing them from stepping into a busy street if dazed. But it also seems evident that “keeping from,” “preventing” (as well as “soothing” or “calming”) might not be easily identifiable, and cues might not be easily demonstrable. Or, to put it another way—at what point does “help[ing] with depression” become “preventing self-harm or suicide”? If the presence of an animal causes self-harm or suicide not to begin in the first place, where is the task?
      Perhaps in an effort to de-emphasize the complicated “task/natural behavior” division, many advocacy sites suggest that service animals’ realness is related to the length of the training process. For example, Please Don’t Pet Me notes that “Service dogs receive hundreds of hours of socialization, advanced obedience training and formal training to perform the jobs for which they were intended.” No matter how well-trained a pet might be, this article explains, it almost certainly has not been through those hundreds (or thousands) of hours of learning how to work. But a certain amount of slippage persists, unremarked, in that argument. For example, what if the human companion’s needs are not so complicated, and training could be briefer? Does that mean the needs themselves are less urgent, the disability itself less severe? Do all human companions need to be disabled in ways that require excellence from our animals? The argument for excellence is compelling, since it is tied to the more general public attitude toward service animals and companions (who, as noted earlier, often experience discrimination and harassment). Many publications, including Please Don’t Pet Me, remind the reader that inappropriate behavior on the part of a service animal creates a “ripple effect” that has negative consequences for all service-animal teams.
      And yet, I want to ask—are those really the terms on which we want to defend ourselves, going forward? Is unswerving excellence (except on “very, very, very, very rare” occasions (Grace, 2016b)) an acceptable price for being allowed to travel, go to school or work, receive medical care, and participate in cultural events? Is that how we want to pursue the argument about our own value as disabled people? For that matter, has it ever worked for any oppressed group to try to maintain perfection as the price of admission? It makes sense to me that one should not visit a restaurant accompanied by a dog who habitually craps on the floor.9 But is it crucial that the animal be flawless, or nearly so? There seems to be an alarmingly eugenic discourse at work, through which the hardworking, genetically optimal, perfectly behaved service animal is held up as a standard for all disabled people and their animal companions.
      The current system of defining service animals assumes that humans have clearly recognizable disabilities, which open up specific needs, which in turn can be met by the work of a service animal (its tasks). But if we think about the complex, caring relationship between service animals and their human companions in terms of crip spacetime, the bright lines between human-disability-need-animal-task dissolve. In fact, every service-animal companion I’ve consulted with in the writing of this article acknowledges a subtler, more nuanced relation between themselves and their animals than legal or even public definitions would allow. The lived experience, as my friends and acquaintances describe it, makes room for ambivalence. Insistence on clear lines between “real/fake” or “natural behavior/task” tends to arise in response to the external violence of discrimination that is the daily experience of most service animals’ companions. In other words, most people with service animals whom I know personally tend to regard the relationship as intra-active: emergent, fluid, a becoming-together. But they also recognize the violence that is part of that becoming.
No Weapons
      I stand on the sidewalk outside the Columbus airport, breathing the damp, chilly air filled with car exhaust. I am almost nonfunctional. That’s why I noticed the sign in the first place: I am moving slowly, not processing anything aural, and staring at things I wouldn’t ordinarily notice, like the lines between the squares of linoleum on the floor, or signs painted on glass walls.
      I am in this state—drifting and dull, post-panic-attack—because the person in the seat behind me on the plane was drunk and violent. He spent the flight rocking in his seat, occasionally punching my seat, and saying “Fucking cunt” about the flight attendant when she refused to serve him any more alcohol. The flight attendant departed the back of the plane for a safer area, while the man continued to swear and occasionally punch my seat. I froze, lost my breath, and didn’t move or turn around. After the plane landed, I waited until he was gone. Then I asked the flight attendant if I could file a complaint.
      I won’t detail the process of attempting to make that complaint. I talked to the gate agents, then their red-jacketed supervisor, then (fifteen minutes later) the police officer they said had to be called because the airline (Delta) couldn’t do anything. By the time the police officer arrived, the drunk man had left the gate, the flight attendant had climbed back onto the plane and flown away, and the officer laughed at me and said, “Well, what do you want me to do about it?” I looked at his badge and he said, “You can look at my badge number all you want, it won’t do you any good.” Then I decompensated and was unable to speak for a while. I remember the officer asking (not kindly), “You want me to call you an ambulance?”, to which I shook my head and walked away. I remember leaning against a wall and texting Johnna to come pick me up. I remember that I couldn’t see very well, but I found the escalators and the moving sidewalk that led me to the baggage claim. I stepped on the pad in front of the automatic doors, moved outside into the fume-filled air, turned back to look at the glass wall, and saw the sign.
      Johnna arrives with both our dogs in the car. Maybe I wouldn’t have decompensated if Ivy had been with me, or maybe I still would have. Maybe the officer would have taken me more seriously if I’d had a marked service animal by my side—or maybe less so. Maybe the presence of a small dog would have made the drunk man less violent, or maybe more so. I don’t know. I do know this: When Johnna arrives with the dogs, I am not in need of my service animal specifically. I just need anyone—preferably everyone—in my family.
      On the sign whose image I showed at the beginning of this article, discourses of violence (“No weapons or joking with weapons like objects. Violators will be prosecuted”) are tucked between the top statement (“No Animals”) and the abruptly utopian turn of the bottom statement (“This facility is completely accessible to all persons with disabilities”). Interestingly, only the middle statement—with its images of a gun and a knife, and its unsubtle reminder that even “joking” about weapons is grounds for detainment—mentions actual prosecution. Although the top statement also warns against illegal activity, it does not mention any consequences. In fact, the subject of those first two sentences is entirely absent, from a grammatical point of view. But in the statement about weapons, potential lawbreakers are named: they/we are “Violators.” Paraleptic non-mentions of race, nationality, violence, and policing appear between the lines of this sign as “shadows”—affective elements present but unnamed (Adams & Erevelles, 2015).
      Those same shadows hang over all animals in U.S. airports, including the pets and ESAs and service animals waiting with their companions in the security line; the police dogs moving quietly beside their handlers; and the thousands of human animals thronging the halls and trying to get somewhere else. Some shadows are cast from hundreds of years ago, some from the present day: dogs used as weapons and tools of surveillance; Lynndie England holding a leash while dragging a man by the neck in Abu Ghraib prison; the comparisons made between Michael Vick’s appearance and that of a pit bull (Weaver, 2015). All human-animal relations are shadowed by the cultural process Barad calls “thingification” (2003, p. 812). For example, as Sunaura Taylor (2017) documents, European and American sideshows of the 19th and 20th centuries featured acts such as Ape Girl and Jo-Jo the Dog-Faced Boy, which “played out the various colonial and scientific dramas of their time.” In these shows, Taylor recounts, “animality was front and center—with the most demeaning of animal comparisons being reserved for people of color and for intellectually disabled people” (p. 104). The medical-scientific logics used in the becoming of race and disability were also logics that transformed people into animals, and animals into objects, and objects into despised, unworthy matter. Indigenous and feminist versions of object-oriented ontology both point out and resist these violent processes, while simultaneously emphasizing the importance of recognizing agency in nonhuman actors (see Kristen Arola, 2017; Kim TallBear, 2011). If one is attuned to these histories, it is eerie to note the implications in the words of Dorothy Harrison Eustis’s Saturday Evening Post article from 1927: “The dog must have perfect obedience and yet he cannot be a machine.”
      Animals and objects cannot ethically be used as stand-ins for that which we wish to posit as less-than. Rather, we must acknowledge humans, animals, and objects as differently valuable entities (Gibson, 2006; Haraway, 1991, 2008; Kafer, 2013; Kim, 2015, 2016). In some cases, we may observe that objects have animal-like or human-like qualities; in others, that we ourselves are animal-like and object-like, without allowing that observation to signal “a necessarily degraded status that licenses violence” (Kim, 2016, p. 141. See also Belser, 2016; Taylor, 2017). My call to crip care in the service animal-human relationship is fueled by more than a general sense that it’s the right thing to do. It is also fueled by my conviction that if we do not work our way toward a more capacious, more fully relational understanding of what a service animal is (and thus, what we ourselves are), we will continue to perpetuate the violences that accompany the hierarchization of animacies (Chen, 2012). We will literally be saying that some animals are more equal than others.
      Is there room to recognize, both theoretically and materially, that perhaps we could move away from the assumptions of unswerving excellence, cleanliness, and hyper-ability that characterize most legal and institutional definitions of “service animal”? Can we take up the possibilities, and also the costs, of a less controlled proliferation? I am inspired by Emily K. Michael’s investigation into the many encounters she manages every day as she moves through the world with York. Michael points out that, although people’s endless questions can get tiresome, they also reveal “patterns of thinking about the more-than-human world.” She notes that many people comment sadly that they miss their own dogs, and rather than taking this as an opportunity to emphasize the difference between the carefully-trained York and others’ (presumably) not-as-thoroughly trained dogs, Michael takes a different turn. She asks: What if humans did have animals accompanying us more routinely into public spaces? How might our spaces change, and how might humans’ relations to one another and to animals change as well?10
      At the close of her book Beasts of Burden, Sunaura Taylor (2017) states that that for several years she had a service dog, Bailey, who then became disabled because of spinal surgery. Now, she writes, “[My husband] David and I are undoubtedly Bailey’s service humans” (p. 223). And yet, Taylor points to this relationship not as the end of Bailey’s work as her service animal, but as an extension of it:
      There is a sense of something appropriate—beautiful actually—about being a gimped-up, dependent, inefficient, incapable human supporting and being supported by my inefficient, dependent, and gimped-up dog. … Awkwardly and imperfectly, we care for each other. (p. 223) 
Taylor would be the last person to suggest that this awkward, imperfect relationship should be taken as a generic ideal. Bailey might not, of course, be the right “fit” for a different disabled person. But Taylor’s relationship with Bailey fits her, fits him, fits their world, and is a form of care that should be valued—in legal and institutional as well as personal ways. There is room for many kinds of service animals in the world. Acknowledging this means not only recognizing a different kind of care in human-animal relations, but caring about those relationships differently as well.
Margaret Price, Ph.D. teaches rhetoric, composition, and disability studies at The Ohio State University, where she also serves as the Director of the Disability Studies Program. Her book Mad at School: Rhetorics of Mental Disability and Academic Life (University of Michigan, 2011) won the Outstanding Book Award from the Conference on College Composition and Communication. Her scholarly and creative essays have appeared in Ms. magazine, Disability Studies Quarterly, the Journal of Literary and Cultural Disability Studies, Profession, Bitch magazine, Kairos, enculturation, Composition Studies, the Disability Studies Reader, and others. She is at work on a book titled Crip Spacetime.
References
Arola, K. (2016). Creative repetition, rhetorical sovereignty, and the ‘electric pow wow’. Presented at Cultural Rhetorics Conference, East Lansing, MI.
Adams, D. & Erevelles, N. (2015). Shadow play: DISCRIT, Dis/respectability and carceral logics. In A. S. Annamma, D. Connor, & B. Ferri (Eds.), DisCrit: Disability studies and critical race theory in education (pp. xx). New York: Teachers College Press.
Assistance dogs. (n.d.). Retrieved from http://www.assistancedogsinternational.org/standards/assistance-dogs
Barad, K. (2003). Posthumanist performativity: Toward an understanding of how matter comes to matter. Signs, 28(3), 801-831.
Barad, K. (2007). Meeting the universe halfway: Quantum physics and the entanglement of matter and meaning. Durham & London: Duke UP.
Belser, J.W. (2016). Vital wheels: Disability, relationality, and the queer animacy of vibrant things. Hypatia, 31(1), 5-21.
Brennan, J. & Nguyen, V. (2014). Service animals and emotional support animals. ADA National Network. Retrieved from https://adata.org/publication/service-animals-booklet
Chen, M. (2012). Animacies: Biopolitics, racial mattering, and queer affect. Durham: Duke UP.
Chen, M. (2014). Brain fog: The race for cripistemology. Journal of Literary and Cultural Disability Studies, 8(2), 171-184.
Eames, E. & Eames, T. (2001). Bridging differences within the disability community: The assistance dog movement. Disability Studies Quarterly, 21(3). Retrieved from http://dsq-sds.org/article/view/293/335 
Erevelles, N. (2011). Disability and difference in global contexts: Enabling a transformative body politic. New York: Palgrave Macmillan.
Erevelles, N. (2014). Crippin’ Jim Crow: Disability, dis-location, and the school-to-prison pipeline. In Ben-Moshe, L., C. Chapman & A.C. Carey. Disability incarcerated: Imprisonment and disability in the United States and Canada (pp. 83-99). New York: Palgrave Macmillan.
Eustis, D.H. (1927, Nov. 5). The seeing eye. The Saturday Evening Post. Retrieved from http://www.saturdayeveningpost.com/2016/04/14/history/post-perspective/post-article-launched-seeing-eye-program.html 
Ferguson, K. (2004). I ? my dog. Political Theory, 32(3), 373-395.
Ferguson, K. (2013). What was politics to the Denisovan? Political Theory, 42(3), 167-187.
Garland-Thomson, R. (2011). Misfits: A feminist materialist disability concept. Hypatia 26(3), 591-609.
Gibson, B. E. (2006). Disability, connectivity and transgressing the autonomous body. Journal of Medical Humanities, 27(3), 187-196.
Grace, K. (2016a, Jan. 22). 10 things that make a dog unsuitable for service dog work. Anything Pawsable: News & More for Service & Working Dogs. Retrieved from http://www.anythingpawsable.com/10-things-make-dog-unsuitable-service-dog-work/#.WLi48hiZMdV 
Grace, K. (2016b, Oct. 1). Things service dogs in public should and should not do. Anything Pawsable: News & More for Service & Working Dogs. Retrieved from http://www.anythingpawsable.com/things-service-dogs-public/#.WLi4iBiZMdU 
Haraway, D. (1991). Simians, cyborgs, and women: The reinvention of nature. New York: Routledge.
Haraway, D. (2008). When species meet. Minneapolis: University of Minnesota Press.
How can I tell if a service dog is legitimate?. (n.d.) Retrieved from http://www.anythingpawsable.com/service-animals/how-can-i-tell-if-a-service-dog-is-legitimate/#.WLiz_xiZMdU 
Huss, R. J. (2010). Why context matters: Defining service animals under federal law. Pepperdine Law Review, 37(4), 1163-1216.
Kafer, A. (2013). Feminist, queer, crip. Bloomington & Indianapolis: Indiana UP.
Kelly, C. (2016). Disability politics and care: The challenge of direct funding. Vancouver/Toronto: U of British Columbia P.
Kim, E. (2015). Unbecoming human: An ethics of objects. GLQ, 21(2-3), 295-320.
Kim, E. (2016). Curative violence: Rehabilitating disability, gender, and sexuality in modern Korea. Durham: Duke UP.
Kogan, L. R., Schaefer, K., Erdman, P., & Schoenfelt-Tacher, R. (2016). University counseling centers’ perceptions and experiences pertaining to emotional support animals. Journal of College Student Psychotherapy. 30(4), 268-283.
Kuusisto, S. (1998). Planet of the blind: A memoir. Crystal Lake, IL: Delta.
Kuusisto, S. (2016). Denied a cab ride, grieving for who we are….Retrieved from https://stephenkuusisto.com/2016/12/03/denied-a-cab-ride-grieving-for-who-we-are/ 
Kuusisto, S. (forthcoming). Have dog, will travel: A poet’s journey with an exceptional Labrador. New York: Simon & Schuster.
Lipka, S. (2011, Oct. 11). Federal case over banning a student’s therapy dog illustrates thicket of disability rules. Chronicle of Higher Education. Retrieved from http://www.chronicle.com/article/Federal-Case-Over-Banning-a/129349/
Marx, P. (2014, Oct. 20). Pets allowed. The New Yorker. Retrieved from http://www.newyorker.com/magazine/2014/10/20/pets-allowed
McCormack, T. (2015). Why natural behaviors aren’t trained service dog tasks. Anything Pawsable: News & More for Service & Working Dogs. Retrieved from http://www.anythingpawsable.com/service-dog-tasks-must-be-trained/#.WLi9nxiZOu4 
Michael, E. K. (2016, May). Working resonance: Concerto for guide dog, handler, and world. The Hopper. http://www.hoppermag.org/working-resonance/ 
Michalko, R. (1999). The two in one: Walking with Smokie, walking with blindness. Philadelphia: Temple UP.
Misrepresenting pets as service dogs. (n.d.). Retrieved from http://pleasedontpetme.com/pets.php 
Oliver, K. (2016). Service dogs: Between animal studies and disability studies. PhiloSOPHIA, 6(2), 241-258.
Understanding the difference between tasks and work. (n.d.) Retrieved from  http://pleasedontpetme.com/taskwork.php 
Samuels, E. (2014). Fantasies of identification: Disability, gender, race. New York: NYUP.
Siler, W. (2017, Aug. 29). Stop faking service dogs. Outside Magazine.
Skloot, R. (2008, Dec. 31). Creature comforts. New York Times Magazine.
Taylor, S. (2017). Beasts of burden: Animal and disability liberation. New York: The New Press.
TallBear, K. (2011, April 24). “Why interspecies thinking needs Indigenous standpoints.” Theorizing the Contemporary. Cultural Anthropology. Retrieved from https://culanth.org/fieldsights/260-why-interspecies-thinking-needs-indigenous-standpoints 
Teitell, B. (2013, Sept. 18). Service dogs barred, doubted, and deeply treasured. The Boston Globe. Retrieved from https://www.bostonglobe.com/lifestyle/2013/09/18/the-growing-number-dogs-assisting-people-with-invisible-conditions-causing-conflict-and-some-cases-confrontation/igPnUBYHa97K07ccBGJJVJ/story.html 
The hidden complications of fake service dogs. (n.d.) Retrieved from https://www.anythingpawsable.com/fake-service-dog-complications
Weaver, H. (2015). Pit bull promises: Inhuman intimacies and queer kinships in an animal shelter. GLQ, 21(2-3), 343-363.
Wenthold, N. & Savage, T. (2007). Ethical issues with service animals. Topics in Stroke Rehabilitation, 14(2), 68-74.
Yardley, W. (2009, Sept. 3). Oregon wants ‘dog friendly’ to be less so. New York Times, pp. xx
___

Endnotes
1. My thanks to the human companions who have helped me develop and revise this piece: Kennan Ferguson, Emily Michael, Rod Michalko, Bethany Stevens, Sunaura Taylor, Tanya Titchkosky, and Melanie Yergeau. Thanks also to Ryan Sheehan for his extensive research support, and to the reviewers and editors who have made this piece so much better. Any remaining errors of fact or judgment are mine.
2. Laws governing service animals in the U.S. include C.F.R. § 36.202; the Americans with Disabilities Act (ADA), including its 2010 amendment; the Fair Housing Amendments Act (FHAA); and the Air Carrier Access Act (ACAA). The Department of Transportation’s Guidance Concerning Service Animals in Air Transportation (2003) is also often cited. Laws outside the U.S. vary; Assistance Dogs International (ADI) and the International Association of Assistance Dog Partners provide information about international law.
3. When referring to unspecified service animals in the singular, I use the pronouns “he” and “she” by turns.
4. Unless otherwise specified, “animal” in this article refers to “nonhuman animal.”
5. Eames & Eames (2001) document the emergence of a subculture of disabled people who are partnered with service animals, particularly dogs.
6. In addition to the potential harms to animals/handlers noted earlier, the appearance of animals in public space may harm those who are debilitated by phobias, allergies, or other problems that make it difficult or impossible to be around some kinds of animals. In exploring what it might mean to crip the definition of service animal, I do not mean to argue that its potential problems are unimportant. Harm and pain are important, and should receive more attention in new-materialist theories.
7. By “fitness,” I mean a metaphorical fitting/misfitting, as defined by Rosemarie Garland-Thomson (2011): a “material arrangement” through which a body/object is more or less congruent with the surrounding, and shifting, circumstances (p. 594).
8. This description is a very brief overview of one positive-reinforcement training approach. My thanks to Sharon Wachsler and Caroline (CMoore) Moore, who introduced me to this method, helped me find books and resources on the subject, and offered feedback on my training videos.
9.  Since I’m in the habit of thinking about limit cases, even this example doesn’t seem so terribly extreme when I consider the fact that I am routinely expected to swim in public pools that babies and children have both urinated and defecated in, to touch doorknobs and elevator buttons teeming with fresh bacteria, and to shake hands with people who have recently been sneezing, coughing, wiping their faces, and picking their teeth. In general, human bacteria and environmental toxins are a greater threat to me than a dog taking a shit in my vicinity. For more on the complications of dirt, contagion, and toxicity in public space, see Mel Chen (2012, 2014) and D Adams and Nirmala Erevelles (2015).
10. As noted above, an abundance of animals in public space would not be an unqualified good. People with phobias or allergies, for example, might be harmed. My point is not that we should create a sort of hipster utopia, with dogs sitting on barstools beside their human companions (as much as that image might charm me personally). Rather, I want us to re-think the ways animals are allowed to be part of our systems of care, particularly when those systems insist on rigid definitions of disability and need.
      
      REVIEW OF DISABILITY STUDIES: AN INTERNATIONAL JOURNAL
Volume 13
 Issue 4
      
      

Page 1