Making Space for Failure in Geographic Research

The idea that field research is an inherently “messy” process has become widely accepted by geographers in recent years. There has thus far been little acknowledgment, however, of the role that failure plays in doing human geography. In this article we push back against this, arguing that failure should be recognized as a central component of what it means to do qualitative geographical field research. This article seeks to use failure proactively and provocatively as a powerful resource to improve research practice and outcomes, reconsidering and giving voice to it as everyday, productive, and necessary to our continual development as researchers and academics. This article argues that there is much value to be found in failure if it is critically examined and shared, and—crucially—if there is a supportive space in which to exchange our experiences of failing in the field.

I n recent years, geographers have widely accepted the notion that qualitative research, and especially ethnography, is an inherently "messy" process (Agar 1986;Nilan 2002;Crang and Cook 2007;Horton 2008;Jemielniak and Kostera 2010;Jones and Evans 2011). Feminist geographers in particular have problematized the masculinist underpinnings of the discipline that deny the fundamental messiness of the field (see Billo and Hiemstra 2013;Coddington 2015). The field can be a volatile and unstable place to conduct qualitative research (Greenhouse, Mertz, and Warren 2002), which can push the researcher and the researched into problematic and potentially dangerous experiences. Yet this embracing of messiness has not been accompanied by a widespread acceptance that failure is an integral part of what we do. Although we as geographers often inhabit an untidy and sometimes chaotic research process, this is often unacknowledged when we write and speak about our research. This article pushes against this, arguing that failure should be recognized as a central component of what it means to do geographical research. We seek to use failure proactively as a resource to improve research practice and outcomes, reconsidering and giving voice to it as everyday, emotional, and necessary to our development as researchers and academics.
Despite the attention paid to the messy nature of research, in reality this is often tidied up after the fact, and is rarely reflected in research outputs for fear of looking wasteful or even "foolish" (Jones and Evans 2011, 586; see also Shore 2010). This is all the more true when it comes to acknowledgments of failure in the research process (Gill 2009;Jemielniak and Kostera 2010). The temptation to sanitize the realities of field work into persuasive chunks is an ever present and arguably necessary process, adding one more filter between what happened "on the ground" and what finds its way onto the page (Katz 1994). By "smoothing the ragged edges of the research process" (Kay andOldfield 2011, 1276), geographers can give the illusion of a linear clarity to the often frustrating fog of the research process, hiding the confusion, self-doubt, and many mistakes that are made along the way (Punch 2012). In this article we seek to pull back from this temptation, arguing that camouflaging failure is unhelpful, particularly when we as researchers strive to meet calls for greater reflexivity and honesty in research (Rose 1997;Burawoy 2013). Recent invitations for honesty in academia have stretched beyond the confines of research reflexivity to incorporate our wider role within neoliberal education systems (Mountz et al. 2015), where "failure" appears to be increasingly unacceptable. We suggest that there is much value to be found in failure if it is critically examined and shared, and-crucially-if there is a supportive space in which to exchange our experiences of failing in the field.
This article begins with a consideration of why failure is often absent from our discussions of field work and sanitized academic outputs. It then moves to contextualize failure in the contemporary neoliberal university, where academics are unable to fail and yet do so regularly. The article notes that failure is more than a banal, everyday experience, but also has the capacity to be a powerful and political tool (Halberstam 2011). The article then moves to consider three instances of our own failures in the field and what this means for researcher practice and development, before concluding with a defense of failure as everyday, emotional, and, ultimately, necessary.

Why Don't We Talk about Failure?
Although failure can undoubtedly occur at nearly any point in the research process, this article focuses specifically on failure during field research. Any account of field work requires a clear notion of where or what the field is. Here we adopt the perspective, following Kobayashi (1994) and Katz (1994), that the field is not only the location where research takes place; the field has an ability to inscribe itself on the researcher, often with implications for both researcher and researched (see Hyndman 2001;Dewsbury and Naylor 2002). In this expanded geographic sense, the field should be understood as unbounded-something that stays with researchers long after they have left the physical location of research. As a consequence, failure in the field is not something that can be easily left behind, but something that researchers carry with them as part of their experience of the field. Although the field has been critically examined as a problematic site of uneven power relations, the academic labor of field work remains a central component of how geographic knowledge is produced (DeLyser and Starrs 2001). This creates practical and ethical challenges to researchers (Nilan 2002), and exposes geographers to physical and emotional vulnerability (Caretta and Jokinen 2016).
Failure itself is a broad term, which takes in a gamut of experiences including denied or rescinded access to field sites (Moran, personal communication, 2017), uncomfortable or abandoned interviews (Nairn, Munro, and Smith 2005), failure to build rapport with participants (Roulston 2014), significant divergence from or abandonment of planned research activities (Jones and Evans 2011), and embarrassing or distressing experiences of field work (Jemielniak and Kostera 2010;Woon 2013). We use the term field work failure for an event or experience that affects the research process in a manner perceived as negative by the researcher, and diverts it away from the intended or expected path he or she had envisaged.
There are a number of explanations for the tendency of geographers to erase instances of failure from their research narratives. First, quite simply, we do not want to talk about what went wrong, for to do so is emotionally troubling and professionally embarrassing (DeLuca and Maddox 2015). We all have a cache of horror stories carefully crafted to be just embarrassing enough to share with our peers and increase our feelings of authenticity without being actively compromising. They exist somewhere along a "cringe spectrum" (Scott et al. 2012, 721) of shared research experiences. It is far rarer that we openly share the stories that leave us feeling genuinely inadequate, unprofessional, or out of our depth. Talking about these stories is uncomfortable; it leaves the researcher feeling exposed or vulnerable (Jemielniak and Kostera 2010;Stefan 2010). This is especially acute given the widespread prevalence of "impostordom" (Scott et al. 2012, 715) among early career academics, especially, and in higher education more generally (Knights and Clarke 2014;Parkman 2016). It is safer to focus on what went right during the research process, glossing over problems in favor of positive or significant results in research outputs.
Indeed the peer review process itself can be seen as a significant barrier to speaking openly about failure. This self-censorship was identified by Olson et al. (2002) as a leading cause of publication bias (meaning studies with positive results are more likely to be published than those with negative or statistically negligible results) within medical research. It is likely that a similar process occurs within geography and the wider social sciences; "after all," as Nairn, Munro, and Smith (2005) reminded us, "the academic arena is a competitive one where 'success,' rather than 'failure,' is rewarded" (222). Admitting failure in this context is perceived as a career risk by many researchers, in particular those starting out in their careers (Peters and Turner 2014). This is all the more acutely felt given the increasing pressures placed on academics within the neoliberal university, a topic that is discussed in greater detail later.

Neoliberal Failure
Critiques of the neoliberal university commonly identify a number of key characteristics that, when taken together, have had the effect of drastically increasing demands on academics' time. These include a growing audit culture (Crang and Cook 2007;Gill 2009), a decline in state funding and concomitant rise in the importance of private funding (Dowling 2008), greater individualism, competition between individuals for the holy grail of research funding (McDowell 2004), and the increasing proportion of the academic workforce made up by part-time and temporary workers (Dowling 2008;Pusey and Sealey-Huggins 2013). Many researchers have responded to these demands by "making more time" for their work (Meyerhoff, Johnson, and Braun 2011), stretching their working days to accommodate new tasks and responsibilities even as they face the increasingly stressful realization that the working day (and weekends) can only be stretched so far (Crang and Cook 2007;Mountz et al. 2015). In these conditions, admitting failures becomes nearly unthinkable, especially because under the individualism of the neoliberal university, failure (as well as success) has become uniquely personalized. The failure of a research project or grant application becomes synonymous with the failure of the academic person (B. Davies and Bansel 2005). No surprise then, that Gill (2009, 2) referred to the "toxic shame" of failure-and the fear of even talking about failure-that permeates the academic environment. To admit failure, even to colleagues or peers, is to jeopardize your position in an ever more precarious working environment. This has a profoundly isolating effect in discouraging researchers from sharing their failures for fear of being outed as "functioning at suboptimal levels" (Shore 2010, 24).
Against this backdrop it is perhaps unsurprising that geographers are unwilling to actively draw attention to the ways they fail. Ethnographers have noted that "a fear of ever making mistakes" can limit the work of researchers (Hammersely and Atkinson 2007, 92). By not interrogating instances of failure, however, an important opportunity for critical reflection and learning is missed (Jemielniak and Kostera 2010). This article situates itself in a developing body of work that acknowledges failure as being at the heart of the research process (and the academic experience more widely) and seeks to use it proactively as a resource to improve research practice. This includes well-publicized initiatives by several researchers to write and share a CV of failures (Stefan 2010;Haushofer 2016), with the aim of disrupting the (largely inaccurate) narrative of seamless progress and success suggested in the public profiles of many academics. It also includes numerous researchers (Nairn, Munro, and Smith 2005;Chatterton, Hodkinson, and Pickerill 2010;Jones and Evans 2011;Roulston 2014) who have attempted to foreground instances of perceived failure, messiness, and lucky breaks that occur in any research project, to meet the challenge of critical self-reflection, and to fulfill an ethical responsibility to share their experiences with other researchers.
The importance of the researcher's own emotions during qualitative research is well recognized (McCann and Pearlman 1990;Kleinman 1993;Young and Lee 1996;Scott et al. 2012;Calgaro 2015). The emotional labor that goes into and is produced by geographic field work leaves an affective residue that stays with researchers long after they leave the field (Drozdzewski 2015; Drozdzewski and Dominey-Howes 2015). Punch (2012) articulated the "hidden struggles" of field research and offered field diaries as a proactive method of unsmoothing the mess of research. Similarly, Hubbard, Backett-Milburn, and Kemmer (2001) not only discussed the centrality of emotion and researcher vulnerability within field work, but also stated how making such registers invisible acts to impoverish the findings. None of these researchers, however, spoke directly about moments when they fail in research. At best, failure becomes hidden in the lexicon we use to camouflage our mistakes, where messiness becomes a euphemism for failure.
This article goes further, arguing that failure is more than simply research plans going awry, but it is also inherently political. In The Queer Art of Failure, Halberstam (2011) theorized failure as: A way of refusing to acquiesce to dominant logics of power and discipline and as a form of critique. As a practice, failure recognizes that alternatives are embedded already in the dominant and that power is never total or consistent; indeed failure can exploit the unpredictability of ideology and its indeterminate qualities. (88) Halberstam drew on a number of different artefacts of popular culture, analyzing them through Marxist, queer, and radical feminist thought to interrupt the normative narrative of "success" that pervades neoliberal North American ideology (Carr 2012), also arguably present in the contemporary neoliberal higher education landscape. For Halberstam (2011), failure is therefore necessarily subversive, potentially powerful and productive, noting that "under certain circumstances, failing, losing, forgetting, unmaking, undoing, unbecoming, not knowing may in fact offer more creative, more cooperative, more surprising ways of being in the world" (2). We take up Halberstam's conceptualization of failure as a political antinarrative and potentially constructive force in academia, and in the following section of this article we reflect on three specific instances of failure that demonstrate our interconnected experiences of failure in the field. Although these instances initially present failure as a negative event or experience that affected the research process and diverted it away from its intended or expected path, they are also understood as productive experiences in light of Halberstam's notion of failure.

Landscapes of Failure
In what follows, we present three vignettes of failure drawn from our time spent in the field. We all undertook long-term ethnographic field work in challenging field sites: the recovering urban landscape of postconflict Osh in Kyrgyzstan, the nuclear geographies of the Chernobyl Exclusion Zone in Ukraine, and disabled children's orphanages in Russia. We employed a combination of ethnographic methods to record and understand our field sites. Perhaps most important, we have grappled with multiple forms of failure during our field research, and have found much value in critically discussing them on our return from the field. All three "territories of failure" (Halberstam 2011, 25) discussed here draw on field notes or diaries to (re)capture the confessional intimacy of experiencing failure in the field (Harvey 2011). Following Trigger, Forsey, and Meurk (2012, 517) and DeLuca and Maddox (2015), we use firstperson narratives to better express our individual negotiations with failure. By revealing these personal stories, we present "a vivid portrait, in miniature" (DeLyser and Starrs 2001, 7) of our field work challenges, showing how geographic research and failure can productively coexist.

Elly Harrowell's Story: Language in Osh
This vignette draws on two periods of field work totaling six months in the Kyrgyzstani city of Osh between 2013 and 2014. During this time, I carried out semistructured interviews, as well as recording ethnographic observations in a field diary (the quotes in this section are all drawn from this diary). Osh is a trilingual city in which most citizens understand all three main languages (Kyrgyz, Uzbek, and Russian), but in which language has been inextricably caught up in the identity politics that have overshadowed life in the city since the violent riots of June 2010. Although I had learned one of Osh's languages-Russian-to an intermediate level before beginning my work there, I was concerned that my lack of language skills would mean I lacked a fundamental area of expertise necessary to carry out good research. I worried that I would be unable to build a rapport with interviewees, that I would miss out on the fine detail of interactions, and that my independence and spontaneity as a researcher would be limited by the necessity of working with an interpreter.
To compensate for my lack of fluency, I decided to work with two local research assistants (RAs), one fluent in Kyrgyz and one fluent in Uzbek (both spoke fluent Russian). My field notes show the frustration and feelings of failure that my linguistic problems caused. In Week 2 I was concerned that I was missing out on "the sense of solidarity/being an insider" that comes from sharing a language with those around you; in Week 3 I was worried that the presence of an interpreter in an interview "creates a confusing power dynamic in interviews-who is leading the research? Who should the participant reply to?" By the fourth week I was concerned that I was missing out on the possibilities afforded by spontaneous encounters because "I always have to plan everything in advance with my RAs-it's really hard to just start chatting with someone in the market." By the end of my field research, my Russian language skills had improved to the point where I felt able to do some interviews on my own, with my field notes revealing that it "feels great to be able to speak and connect directly to people," and revelling in the flexibility and freedom this gave my research schedule. This feeling was short-livedreviewing the transcripts of these interviews (which had been professionally translated to ensure I did not misconstrue the information) revealed the wealth of mistakes, misunderstandings, and missed opportunities throughout these encounters. The shame of not mastering the languages of my field site followed me out of the field and into other academic encounters. Attending conferences to present my research, I felt compelled to include snippets of local languages in my presentations, microperformances of linguistic competency aimed at convincing my peers I belonged among them. I glossed quickly over questions on language and translation in my presentations for fear that I would be exposed as an imposter, a part-time ethnographer who had failed to earn her stripes by mastering the intricacies of southern Kyrgyzstani syntax.
Taking the time to reflect on these experiences with supportive colleagues led me to reassess this perceived failure, however. Looking again at the research process helped me to recognize the benefits I had gleaned from working with my RAs-the nuances they brought to my data, the insider information they provided through their explanations of phrases or customs, and the recommendations they made of places I might find interesting to visit. Had I spoken perfect Kyrgyz and Uzbek, I would not have needed to work with my RAs and might never have accessed these valuable insights. This chimes with Edwards's (1998) assertion, since taken up by a number of scholars, that we research with interpreters and not through them (see also Littig and Pochhacker 2014;Caretta 2015). Such a view acknowledges the valuable role interpreters play as coproducers of knowledge whose insights bring critical perspectives on the process of situated knowledge production and encourage the foregrounding of reflexive research practice (Edwards 2013;Caretta 2015), rather than framing them as a "necessary evil" (Edwards 1998, 199). Being "lessthan-fluent," to borrow Tremlett's (2009, 65) phrasing, also encouraged me to think carefully about language, to ask for explanations and additional information to ensure I had fully understood the point my interlocutor was making. This also had positive implications for the depth and quality of the data I gathered, even though at the time of collection I was prone to dismissing the additional questions I posed as a nuisance or failure on my part.
More generally, being honest about my level of linguistic ability contributes to addressing the pervasive silence, or mystique, surrounding issues of language and fluency in ethnographic research as decried by authors such as Borchgrevink (2003) and Gibb and Danero Iglesias (2016). Although admitting to being a nonfluent ethnographer seemed daunting, even embarrassing at first, sharing experiences with other researchers and reflecting on how my lack of fluency affected my research process has proved to be an enriching experience, reflecting Halberstam's (2011) notion of alternative ways of "being in the world" (2). Overcoming my sense of shame at this failure has enabled me to reflect on and improve my practice as a researcher, as well as sharing strategies and experiences with colleagues and peers.

Thom Davies's Story: Failing in the Chernobyl Exclusion Zone
This vignette draws on a period of ethnographic field work in the nuclear landscape of Chernobyl, in northcentral Ukraine. Between 2010 and 2014 during my field work I conducted interviews with inhabitants of the Chernobyl region as well as participatory visual methods with local farmers, border guards, evacuees, returnees, and people who were conscripted to decontaminate the Zone after the 1986 nuclear accident, known as "liquidators" (T. Davies 2013). My participants lived on the edge of the 30 km 2 nuclear Exclusion Zone, an exceptional and securitized toxic landscape into which entry is officially forbidden. Over the course of the field work it became increasingly clear that many people regularly illegally crossed into the Zone as part of their complex survival strategies, and this became a point of research focus through participant observation (see T. Davies and Polese 2015).
Most of my field work took place on the Zone's bucolic outskirts where you are officially allowed to live but are cut off from the hidden nuclear geographies across the barbed wire fence to the north. Two Ukrainian friends, who accompanied me while they worked on a photography project about Chernobyl, were with me when we decided to enter the Zone. The following quotation is a recorded "note to self": I just got arrested in the Exclusion Zone. [I'm] Just outside of the barrier place now, pretty scary actually. It was pretty scary at the time, well actually now I am still shaking if I am honest. We . . . walked into the Zone into the forest and then we saw some border guards. We tried to hide, in quite a pathetic way-obviously it was broad daylight so we were seen. Then we waited in the police car in the Zone on the track for about, well it felt like two hours. Two other police vehicles arrived; there were about twenty police there altogether. By this time it was dark and I was fucking cold. Um, some people sort of checked me out-my passport-for quite some time. Eventually we were put in a meat wagon and driven to where I am now. (Dictaphone recording) I find this note to myself particularly embarrassing and awkward. My voice is clearly nervous and uneasy-I'm stressed, "shaking," and my tone is very serious and almost quivering. I had failed. I continued talking into my voice recorder, as if it was a phone: "The long, short, and tall of it is that I have got to go to court on the twenty-second in Ivankiv." This went far beyond "the embarrassingly messy lived experience of managing fieldwork" (Scott et al. 2012, 718)-I had failed at being a professional geographer. Access had been denied and I had broken the law. The police interview I was subjected to, conducted by a member of the Ukrainian SBU (KGB), was a radical role reversal between researcher and researched-I was sitting on the wrong side of the voice recorder.
In the days before the court appearance I grew increasingly nervous. I could see everything in front of me rapidly falling apart: I would be deported from Ukraine, I feared, or worse-sent to prison. It was, after all illegal to enter the Zone without formal authorization. A small part of me imagined the humiliation of being visited in prison by my supervisor-an expert on the geographies of incarceration-only to become data in one of her publications (see Moran 2013). She would be furious, I thought, wrongly. Humbled and embarrassed, I left the field with my tail between my legs and a greater appreciation of the role the state plays in life at the margins. I had reluctantly "met the organs"-a common experience for researchers in postcommunist spaces (Gentile 2013, 426), where the state creates key methodological challenges during field research (Koch 2013). At court the judge gave my friends and me a fine, which I gladly paid, attempting to adhere to a "do no harm" ethical approach (Ellis 2007).
Failing in this way was extremely uncomfortable and unnerving, and brought with it a sense of shame. Woon (2013) discussed the importance of emotion in field work for "bequeathing the researcher with fully embodied experiences of the 'real' situation on the ground" (31), and I think the overwhelming sense of failure, and the deep embarrassment that went with it, helped me understand-albeit in a small and situated way-a little more about the lived experience of Chernobyl.
Reflecting on this field work experience with academic friends and colleagues, as well as research collaborators, allowed me to recognize the "definitive advantages of failure" (Halberstam 2011, 4). In later research visits to Ukraine I would draw on this experience to gain rapport with participants. Other people around Chernobyl had been through the same experience of being arrested in the Exclusion Zone: the temporary incarceration, the document checks, and the court appearance. Some had managed to bribe their way out of situations, others spoke of being treated unfairly by the border guardsinformation I might not have gained had I not failed months earlier. I could share a commonality with my research participants in disliking the authority of the Zone, in a country where state authorities are widely viewed as predatory and corrupt. It also helped me think more reflexively about power relations in this highly regulated toxic landscape and my unspoken privilege as a researcher. In this sense failure became a resource that helped strengthen my future research (see T. Davies 2015; Alexis-Martin and Davies forthcoming), even if it was an uncomfortable experience at the time.

Tom Disney's Story: Failing to Change the Russian Orphanage
The geographic field work drawn on in this section took place in Russia between 2012 and 2014, examining spaces of care for orphaned children. Although I visited a number of environments for children without parental care, my ethnography attempted to make sense of the operation of one institution where I worked as a volunteer-an orphanage for children with severe intellectual disabilities.
Fear of failure is a classic research problem, although arguably this fear manifests itself in different forms. Although I experienced fears about my competencies as a researcher, the greatest fear I had in relation to my research was that I would not be able to effect meaningful change. Numerous qualitative researchers have argued passionately about the political imperative to use research to combat social injustice (see Kobayashi 1994;Kende 2016); and, although I, too, pursue such aims, during my research in the orphanage such a task weighed heavily on me. The environment was a difficult one to work in, and alongside the other volunteers operating in the institution I witnessed practices that deeply troubled me. These included microscale practices of physical, medical, and emotional abuse, but also wider macroscale, systemic issues, such as the processes of institutionalization over which the children had no control, destined for a future of being involuntarily moved through various closed institutional environments (Human Rights Watch 2014; Disney 2015). As time within the field progressed, I gradually began to read these institutional processes as a form of incarceration, and became increasingly troubled by the system. My awareness of the ways in which the embedded nongovernmental organization (NGO) within the orphanage was preventing some of these processes did little to alleviate my disquiet about the system, as most of these institutions lacked such a setup and thus I understood that they were occurring unimpeded elsewhere.
I had entered the field as a researcher sharing those values of the academic community, with a desire to enact social change and improve the lives of my participants. Yet once inside the field I found myself totally powerless to shape it and increasingly I experienced the field heavily affecting my own emotional state, revealing the field as a space that "writes back with its empirical agency and its embodied effect" (Dewsbury and Naylor 2002, 256). My awareness of my inability to challenge the injustices and daily abuses of the system, as I saw them, became a recurrent theme in my field diary, as illustrated in this excerpt: Although I have some decent observational data at this point, I am feeling frustrated with the research process, I don't feel particularly like I am getting anywhere at the moment. This is also part of a wide range of emotions I have about this research at the moment; there is a general concern I have about the value of the research. What value does my PhD have to my participants, lying in beds and sitting on the floor in the orphanage every day? Even, by some massive stroke of luck, the work is published and highly cited, it will most likely have no effect on my participants.
Although I felt I was able to enact minor resistances while in the field, such as reporting instances of abuse to the NGO, who in turn could make official complaints, leaving the field constituted a major "failure" in my mind, as the system remained intact, operating much the same after my return to the United Kingdom. It was only after returning from the field that I was able to gain a sense of perspective about the scope of my research and its ability to effect change; clearly systemic change to those institutions was not within my power, even though I had felt a heavy responsibility to this while in the field. Social change is a powerful imperative for us as geographers, but in my field it became an emotional burden, and my inability to achieve it felt like a serious failure. As other geographers have found, not being able to make a difference was distressing (Klocker 2015). It took some time after returning from the field for me to accept that I was not responsible for, or capable of, changing the institutional landscape and preventing the microscale abuses and wider practices of the orphanage system. Discussions with colleagues revealed similar experiences of seeking change where this was unfeasible, and this helped to alleviate a sense of guilt at leaving behind those who I felt were effectively incarcerated in the orphanage where I had been working. It also suggests, however, that there is a wider sense of failure to effect change to our field sites, even when this is beyond our control.
The calls to effect social change leave little space for failure, and this has implications for our sense of success in the field. Massey and Barreras (2013) introduced the term "impact validity" to consider "the extent to which research has the potential to play an effective role in some form of social and political change" (616, italics in original). This approach might be repurposed to offer a sense of perspective to the limits of what is achievable in the field, that it is not always possible or feasible for a researcher to effect major change to the field. Furthermore, although I failed to drastically change the orphanage system, this sense of failure drove me forward in my research process, motivated by a need to address this; my discomfort surrounding the operation and power of the orphanage system made conversations with NGO actors in this sphere more fluid, as I was able to speak more knowledgeably about the subject and communicate my frustration surrounding my "failure" to those who shared these feelings. Often general discussions about the operation of these systems led to access to other closed field sites, or particularly candid discussions, generating considerable data and understanding to form the focus of much of my later work. A sense of failure became a realization of the need for perspective in terms of social change, and a means of facilitating research access and processes, thus mirroring Halberstam's (2011) notion that failure has productive agency.

Discussion and Conclusion
As Caretta and Jokinen (2016) noted in this journal, there is a need for more "open communication among geographers" (8) and we advance this call through our open discussion of being "fallible field-workers, negotiating challenging circumstances, not always with equal success and grace" (DeLyser and Starrs 2001, 6). Although there has been widespread acceptance of the "messiness" of research in geography (Crang and Cook 2007;Horton 2008;Jones and Evans 2011), this has not been met with an equal acknowledgment of the ubiquity and necessity of failure. Arguably the masculinist origins of geography as an academic discipline play a significant role in the erasure of failure, and although feminist geographers have critiqued these origins (Billo and Hiemstra 2013), there remains a need to acknowledge openly that failure is in fact an everyday, and indeed powerfully productive element of geographic field work. In this article we have addressed this notable absence, drawing on the work of Halberstam (2011) to illustrate the ways in which failure has political meaning, has power, and is constructive. Although we "failed" at times in our respective projects, these perceived failings facilitated our research practice and identities as academics, reflecting Halberstam's (2011) notion that failing "may in fact offer more creative, more cooperative, more surprising ways of being in the world" (2).
Reading across the three snapshots of perceived failure, a number of shared features are apparent. First, the pervasive emotion to emerge from these accounts is that of shame, akin to the toxic shame of failure identified by Gill (2009). This shame extended beyond our time in the field, as the field (and our failure there) followed us home (Drozdzewski 2015). Second, in all three cases the answer to this shame was found when we talked to each other openly and realized that we all had stories of failure to tell. Acknowledging the ubiquity of "failing well, failing often, and learning" (Halberstam 2011, 24) turned our failures into powerful and productive lessons.
There remain serious obstacles to overcome if we are to move beyond ad hoc conversations with friendly peers to a more systematic acknowledgment of the value of failure, and the way in which this can be unlocked through honest discussion. The logic of intense competition and individualism ingrained in the contemporary neoliberal university strongly discourages this kind of candor (McDowell 2004;Archer 2008). Even in writing this article we were stalked with worry about the impact of writing about failure on our own careers. Colleagues warned us to think carefully, to make sure we chose "just bad enough" stories of failure that would not damage our reputations. With academic mistakes becoming increasingly personalized, and freedom to fail becoming even rarer, embracing failure becomes a risky endeavor for geographers attempting an increasingly precarious career in academia. We suggest institutions-and funders-make space for failure, to recognize its potential value rather than treating it solely as a damning indictment of the individual researcher. Open, informal discussion sessions held before geographers enter the field where field workers of various years of experience can relay their practices, missteps, and failures might be one means of initiating the task of alleviating the shame of failure. The practice of developing and sharing CVs of failure could become more widespread, particularly by established academics, reducing the stigma when things do not go to plan. Finally, we encourage geographers to "write vulnerably" (Behar 2014, 16) in their reflexive academic work, to normalize the productive place of failure within our neoliberal institutions. It remains true that "fieldwork is not innate but learned" (DeLyser and Starrs 2001, 6), and the open discussion of mistakes, failures, and setbacks should be part of our collective learning process. As geographers, we need to not only acknowledge the place of failure within our discipline, but also find ways of "failing better" (Halberstam 2011, 24). &

Thom Davies
http://orcid.org/0000-0001-8392-8151 Tom Disney http://orcid.org/0000-0002-6080-7709 ELLY HARROWELL is a Research Associate in the Centre for Trust, Peace and Social Relations at Coventry University IV5, Innovation Village, Cheetah Road, Coventry CV1 2TL, UK. E-mail: Elly.Harrowell@coventry.ac.uk. Her research focuses on the challenge of rebuilding following serious conflict and natural hazard-related disasters, and the relationship between urban space and peace.
THOM DAVIES is a Research Fellow in the Department of Sociology at the University of Warwick, Coventry CV4 7AL, UK. E-mail: t.davies.5@warwick.ac.uk. When not failing as a human geographer, his research interests include geographies of exclusion, forced migration, nuclear geographies, and the lived experience of toxic landscapes.
TOM DISNEY is a Research Fellow in the School of Sociology and Social Policy at the University of Nottingham, Nottingham NG7 2RD, UK. E-mail: thomas.disney@nottingham.ac.uk. His research interests focus on institutional geographies and how institutions of various kinds intervene and intersect in the lives of young people and families.