All-encompassing ethnographies: Strategies for feminist and equity-oriented institutional research

This article addresses the challenges inherent to conducting ethnography in all-encompassing institutions and presents strategies for equity-oriented research in such restrictive settings. All-encompassing institutions are organizations that have no separation between home, work, and play, with forms of surveillance, power imbalances, and control that create logistical and ethical challenges for ethnographers. Building on feminist orientations to qualitative inquiry, the author shares vignettes from her research in a military institution to introduce four strategies for ethnography in these contexts. The strategies are: (a) adopting a participant-centered approach; (b) attending to power dynamics; (c) negotiating consent and confidentiality; and (d) engaging participants in research presentation. These strategies are applicable to ethnographers across contexts looking to meaningfully engage interlocutors in research generation and presentation.


Introduction
When the bus pulled in behind the United States Military Academy Preparatory School 1 in West Point, we were greeted by a swarm of cadets yelling at us to move with a purpose, hurry up, and cup our hands as we were ushered into the building.It was Reception Day: the day new cadet candidates 2 reported for duty to the preparatory school and enlisted into the army, and the West Point cadets running basic training with the oversight of the tactical team were eager to shock our systems.
We moved back and forth through white linoleum hallways, in and out of rooms, up and down stairs, with no knowledge of where we were going.Our lack of knowledge about space went hand in hand with our ignorance about time; we were not told what was coming next or when it would end.The message was clear: you are not in control.We received uniforms, our height and weight were measured, and tattoos were checked for gang signs as we were processed into the institution and "trimm[ed]" to fit smoothly into "the administrative machinery of the establishment" (Goffman, 1961:16).The only pause from the chaos was the blood collection station.One candidate, Athena, told me later that she was disturbed when they took her blood without telling her what they were going to do with it: "They took my blood, they took my DNA, and now they were like, 'you're pretty much our property.'" Reception day was the beginning of my three-year 3 ethnography on military (re) socialization.I entered the preparatory school to examine how authorities worked to shape the candidates into soldiers and future officers, and how the candidates, whose civilian identities were constrained by a regimented institutional context, re-fashioned personal self-definitions and embodied resistance. 4I aimed to understand how candidates' experiences varied along gender, racial, ethnic, class, sexual orientation, and religious lines.I took an intersectional approach to see how these identity characteristics became salient, intersected, and interacted upon entrance to the institution and shaped outcomes (Crenshaw, 2017).The preparatory school was a fitting site with its explicit mission to enhance diversity at West Point, with approximately 20% of candidates identifying as women and over 50% identifying as Black, Hispanic, Asian, Indigenous or mixed race.
Upon gaining access to the preparatory school through negotiations with institutional gatekeepers and coordination between the Princeton and United States Military Academy (USMA) Institutional Review Boards, I joined the candidates in all aspects of preparatory school life to be alongside them as they navigated institutional pressures.I used autoethnography to track my own embodied changes from behaving like a soldier and power dynamics within the organization and research. 5I was with the candidates in everything from English class, rucking up the hills with 35 pounds on our backs, responding to casualties in battle drills, to lying exhausted on the barracks floor after particularly difficult training sessions.I listened to their daily stories and used my own observations and embodied experiences to ask questions. 6 Although I entered the preparatory school with a "feminist curiosity," with a commitment to reflexivity and attention to power imbalances, I was immediately struck by the unique ethical and logistical challenges that emerged as an ethnographer in the setting (Enloe, 2015:8).The preparatory school was an all-encompassing institution: the candidates lived, studied, trained, and socialized all within the long, gray building and were under near-constant surveillance.They followed a strict schedule and authorities worked to control their behavior and appearance.I grappled with how the research could become a vehicle to enhance their sense of agency in their lives, rather than exacerbate feelings of a loss of control over self-expressions.
I wrestled with how to address heightened power differentials between the candidates, authorities, and me, and the ways in which this imbalance manifested during research generation, execution, and presentation.How could I invite individuals to partake in the study without them feeling like they were compelled to do so either from peers or authorities due to formal and informal surveillance?How could I maintain anonymity and confidentiality when authorities and candidates could easily see who I spoke and spent time with on a daily basis?As I wrote up the research, I also wondered how I could promote participants' feelings of ownership over their stories in the text and facilitate their ability to speak about and back to (mis)representations. Despite my comfort improvising and adapting my methodology ad hoc in the past as an ethnographer, I soon realized that research within the preparatory school would require me to adopt specific strategies if I wanted to fulfill a feminist, equity-oriented research agenda.
In this article, I address the research challenges inherent to ethnography in allencompassing institutions and present strategies for feminist and equity-oriented research in restrictive contexts.I begin by conceptualizing all-encompassing institutions and detailing the difficulties that emerge when studying those living within them.I situate my approach within feminist and other forms of critical and postmodern inquiry that have largely shaped the methodologies and epistemologies of those working with populations in vulnerable circumstances.From there, I introduce four strategies for researching and writing about individuals in institutional settings, which are: (a) adopting a participant-centered approach; (b) attending to power dynamics; (c) negotiating consent and confidentiality; and (d) engaging participants in research presentation.These strategies support ethnographers in navigating the increased surveillance and power differentials in these contexts, as well as in creating a collaborative and reflexive research and writing plan.Although the strategies are designed for ethnographers working with people in all-encompassing institutions, they are applicable to qualitative investigators across settings looking to meaningfully engage interlocutors in research generation and presentation.

All-Encompassing institutions and research challenges
A definition of all-encompassing institutions is rooted in Erving Goffman's (1961) concept of total institutions.Goffman used the term 'total institutions' to describe organizations that break down the spatial barriers between work, home, and play (1961: xiii).In these settings, residents conduct all spheres of life within institutional boundaries under constant surveillance.They are cut off from wider society and forced to submit to institutional authorities (Cunha, 2014;Foucault, 2012;Garland, 2001;Goffman, 1961).Davies (1989) complicated Goffman's concept to distinguish between types of total institutions.She posited that total institutions differ in their level of openness and closedness, as measured by: (a) the voluntary or involuntary entrance to and exit from the institution, and (b) the internal fluidity of the institutional hierarchy (85).Ellis (2021a) further showed how many total institutions are actually far more porous than traditionally understood: their gates open daily to those working inside the space, their protocols are influenced by broader political and economic trends, and they are infused with logics from outside institutions, such as religion.Building on Davies and Ellis, I use the term 'allencompassing institutions' to describe more open total institutions like the contemporary United States military to account for the stark differences between them and more closed institutions, such as concentration camps.
All-encompassing institutions maintain many of the same characteristics of total institutions-such as an overlap between the spheres of life (work, sleep, socialize), separation between insiders and outsiders, and authoritarian control over appearance and bodily movement-but residents may leave the institution periodically and may gain privileges over time.For example, although the military contains many of the defining features of traditional total institutions, particularly in training environments and on deployment, contemporary service members are recruited volunteers that may exit after they have served their term, and over time they move up the hierarchy of the military (within their respective enlisted or commissioned tracks) (Castro, 1990;Centeno and Enriquez, 2016;Heinecken, 2015;Janowitz, 1960;Leirner, 2008). 7Soldiers can leave post with permission, wear civilian clothing when they are off duty, and maintain relationships with those outside of the armed forces.Even at the United States Military Academy Preparatory School, candidates voluntarily entered the institution and received periodic off-post privileges.Those without prior military contracts could choose to outprocess from the institution at any time.For these reasons, the military and the preparatory school in particular are more accurately characterized as all-encompassing than total institutions.Other all-encompassing institutions include settings such as boarding schools, nursing homes, monasteries, and ashrams.
Despite the comparative 'openness' of all-encompassing institutions, research on those living within them still contains specific logistical and ethical challenges.First, residents of all-encompassing institutions are hard-to-reach populations (Liamputtong, 2007: 48).Gaining access to field sites can be difficult in all social research (Feldman et al., 2003), but it is especially tricky in all-encompassing institutions because they by definition have strong boundaries between insiders and outsiders and can have "literal gatekeepers" (Ellis, 2021a: 2).Second, residents and authorities typically know details of one another's lives and observe each other across all spheres, which presents difficulties for anonymity and confidentiality both during the research and in the written work.These difficulties are exacerbated when there is a small number of residents or a few individuals with specific identity characteristics, for example, Black women (Dickson-Swift et al., 2008).Although there is debate on the merits of anonymity and confidentiality and whether it actually reduces participants' autonomy in the research process (Svalastog and Eriksson, 2010), the identification of participants to each other and authorities can lead to covert and overt consequences within institutions, including interpersonal conflicts and obstacles to career progression (Bosk, 2008).As such, I present anonymity and confidentiality as reasonable goals for institutional ethnographers with the acknowledgment that these decisions are dependent on context and are typically best made through conversations with research participants. 8 Third, the level of control that institutional authorities have over residents' lives, bodies, and discourse means that power dynamics are particularly acute in these settings.Power dynamics are embedded in all qualitative inquiry (Belur, 2014;Hamilton, 2020;Huckaby, 2011;Muhammad et al., 2015;Vähäsantanen and Saarinen, 2013); however, ethnographers have to be particularly attentive to power throughout the research process in all-encompassing institutions.For instance, residents may not feel that they can refuse participation if they believe authorities support the project or otherwise feel disempowered (Fisk and Wigley, 2000;Liamputtong, 2007).These challenges are compounded because residents are often experiencing multiple vulnerabilities.
Although scholars disagree on an exact definition of vulnerability, I align with those who define vulnerability as diminished autonomy, lack of power, limited agency or capacity to function due to physiological, psychological, spiritual, and structural factors (Liamputtong, 2007;Moore and Miller, 1999;Silva, 1995;Winfield, 2021). 9From this perspective, those within in all-encompassing institutions often experience heightened or cumulative vulnerabilities across spheres.Take for instance soldiers: they are physically at-risk due to the requirements of military training in which injury is common, and especially vulnerable to serious bodily harm during combat (U.S.Department of Defense, 2021).Psychologically, they are exposed to intentionally-high levels of stress during training to inoculate them to chaos on the battlefield and to make them docile to superiors' commands.Post-traumatic stress disorder is prevalent among those returning from war, leading to higher rates of death from suicide among service members and veterans than in conflict zones (U.S.Department of Defense, 2019).Spiritually, soldiers are isolated from their home spiritual and religious communities for long periods of time, and may have to reconcile their spiritual beliefs with the potential obligation to take a life in combat (Hansen, 2012).Structurally, military regulations control their appearances and authorities set specific guidelines for behavior and time-use (Department of the Army, 2017).Women, religious minorities, queer, and non-White service members may face further structural constraints or marginalization living within an environment in which their predecessors were excluded and authorities remain mainly White, 10 Protestant, heterosexual, upper-middle class men (Dansby et al., 2001;Hansen, 2012;Nacoste, 2001).Working with research participants who are experiencing heightened and cumulative vulnerabilities across spheres, such as those within all-encompassing institutions, requires specific methodological and epistemological orientations that are geared toward mitigating harm and increasing sense of agency for those who are facing constraints (Dickson-Swift et al., 2008;Jacobson, 2005;Liamputtong, 2007;O'Connor and O'Neill, 2004;Pacheco-Vega and Parizeau, 2018;Shaw et al., 2020).
In order to support ethnographers who may face (or anticipate facing) these challenges, I present four strategies for research within institutionalized settings.I focus specifically on addressing difficulties related to engaging participants during data collection and presentation. 11The strategies are grounded in feminist and other forms of critical and postmodern research that have largely shaped the methodologies and epistemologies of qualitative researchers working with populations in vulnerable circumstances (Fontana, 2001;Liamputtong, 2007).Postmodern scholars reject the positivist epistemology that investigators can capture an accurate representation of the 'Other' or reality because investigators' subjectivities impact data generation, analysis, and presentation.Rather, they embrace reflexivity throughout the research process in order to examine how they are "part and parcel of the setting, context, and culture they are trying to understand and represent" (Altheide and Johnson, 1994: 486;May and Pattillo-McCoy, 2000: 84).Critical scholars extend this epistemology to ethics to illuminate how power dynamics are engrained in the researcher/'subject' relationship and the ways in which 'subjects' are (mis)represented in published work (Cannella & Lincoln, 2017).They invite investigators to practice reflexivity in order to examine their relationships to structures and dynamics of injustice, both within the ethnographic process and their larger social positioning (Charmaz et al., 2017;Kincheloe et al., 2017).A critical orientation further entails "radical ethics," ethics that are actively negotiated through reflexivity and collaboration during the entire research endeavor (Cannella and Lincoln, 2017: 172-173;Denzin and Giardina, 2016;Fluehr-Lobban, 2015;Lincoln and Guba, 1989).
The strategies in this article link the existing recommendations put forth by qualitative researchers in these traditions-including attending to power differentials in the research relationship, (re)negotiating ethics throughout the project, and using methodologies and representations that allow participants to "speak back" to (mis)representations-with the unique challenges inherent in research within all-encompassing institutions (Hölscher and Bozalek, 2012;Liamputtong, 2007;Mitchell et al., 2018: 4;Shaw et al., 2020;Swartz, 2011).They are oriented toward not only minimizing harm to participants who are already in vulnerable circumstances, but also avoiding practices that may perpetuate the stigmatization of marginalized groups in the future (Flaskerud and Winslow, 1998;Mitchell et al., 2018;Paradis, 2000;Pyett, 2001).They are explicitly feminist in their focus on "embodied lived experiences" and in their aim to disrupt "systemic and structural power relations" (Hampton, 2017: 85 cited in Klostermann, 2020: 46), which helps to create conditions for equity-oriented research practices.As I will elaborate in the sections to follow, the four strategies consist of: (a) adopting a participant-centered approach; (b) attending to power dynamics; (c) negotiating consent and confidentiality; and (d) engaging participants in research presentation.
Strategy one: Adopting a participant-centered approach "Taylor!" Athena, a tall Black candidate, shouted to catch my attention.She was wearing the black physical fitness uniform with ARMY written across her chest and her long braids were pulled back into the regulation bun.
"Taylor, you need to write this down.Take out your notebook.Cadet Candidate Athena is having a mental breakdown.I've got to get off this post.I can't keep doing the same thing every day.I am going out of my mind….I think I am changing.I am changing.My body is changing but my mind is not, or my mind is changing or my body is not.Some part is stuck.I am becoming someone new, this place is changing me."I nodded along as she spoke, and we sat down together so she could further detail what was on her mind.
Athena was one of many candidates that invested herself in the research, helping me to better understand her story and the stories of others.I found that when candidates learned about my study and that I was particularly interested in the experiences of historically marginalized groups in the military, they were eager to participate in the research and share with me their experiences.They would come up to me in the barracks hallways to give me updates on their lives and invite me to shadow them for a day.Sometimes they would introduce me to other candidates whose stories "I had to hear."Candidates would ask for follow-up interviews so they could brief me on new events or experiences, and others who had not had an initial interview asked when they were going to get their chance to speak with me one-on-one.One motivation behind their desire to participate in and collaborate on the research became clear to me in an interview with Kelly, a Black candidate from a small midwestern town.She explained to me, "[w]hen you told me what you were researching, I'm like, okay so, one day someone who's in my shoes will read her book and will be like, 'this girl did it so I can do it.She's literally me.We're in the same situation.'"Participating in the study was her way to speak to and mentor future generations of Black women soldiers and create a greater sense of belonging for them in the armed forces.
Over the course of the research, I came to see my role as a midwife of the candidates' stories, helping to birth the stories into the world.In this way, my first strategy was to take on a participant-centered approach (Winfield, 2021): I trusted that participants had the skills and ability to tell their own stories and offered them unconditional positive regard as we spoke about their struggles and triumphs.I explicitly told interlocutors that I was there to help share their experiences, particularly those who had been historically excluded from the military environment in terms of participation and dominant representation.Still I was aware that, like any researcher (or midwife!), my body language (nodding, smiles, laughter), follow-up questions, expressions of solidarity, and other responses likely shaped the narratives they told me (Klostermann, 2020;Poletti, 2011).
A participant-centered approach did not mean that I was passive in the conversations or entered into encounters without a research question.Rather, it was a general orientation of openness and value toward the candidates as full people who own and are experts of their stories (Aldridge, 2014;Corbin and Morse, 2003;Rice et al., 2019;Strier, 2007;Swartz, 2011).For instance, I began interviews by asking interlocutors to speak about whatever was on their minds and allowed them to direct the pace and flow of the conversation.When there was a lull in the conversation, I would ask the next question in my interview guide, which led to interviews that would last between one and eight hours (split over several sessions during the year).
Just as the candidates' shared with me their thoughts and ideas about the research, I shared with them my reflections, theories, preliminary findings, and ultimately worked with them on the research presentation (see strategy four).My intention was to orient to participants as partners in co-creating knowledge based on their lives.Although a participant-centered research strategy is applicable to studies across settings, in allencompassing institutions or other contexts where individuals' options for personal expression are constrained, it can be a particularly powerful tool to help partakers gain a sense of control and autonomy in their situations.Similar to other social-justice oriented methodologies for working with populations in vulnerable circumstances, the strategy aims to use research as an opportunity for interlocutors to reflect on their own understandings of their contexts and explore their own agency (Mitchell et al., 2018).As such, a participant-centered approach requires a sensitivity to power dynamics both within the institution and the ethnographer-participant relationship.

Strategy two: Attention to power dynamics
During our first week of basic training, I accidentally looked around when we were walking in the barracks and made eye contact with an officer.In that moment, I had violated one of our primary rules for facial control: I had not kept my head and eyes straight forward.I quickly averted my eyes, praying that the officer did not notice or he had more pressing things on his mind.
A few hours later, when my squad was making its way into the dining facilities, the same officer got up from his table and stopped me.
"Why were you looking at me this morning, Cadet Candidate?"I immediately jumped into the position of attention and shot off one of the permitted four responses, "No excuses, sir." My squad was still moving forward so I started to inch in their direction.
"Are you going to walk while an officer is talking to you?"I halted, trembling.He moved closer, his eyes now drilling into mine.To me, he looked furious and disgusted that a measly candidate would have the audacity to look at him in the hallways.
"No, sir.No excuses, sir."When he finally let me go, I was blinking back tears.My squadmates gave me furtive looks trying to see if I was okay.When we were able to go back into our rooms in the barracks, my suitemates gathered around and asked me, "Do you know him?What happened?"Saul, one of the men in my squad, told me later that this was the moment he knew I was truly in it with them.I may have been a civilian adult woman, but I was a candidate in his eyes.
Power and control over bodily movements and discourse are part and parcel of everyday life in many all-encompassing institutions.Power and control are what afford authorities their positions, and the lack thereof is what often defines the lives of the residents.When Ed and other candidates saw that I was also subject to the discipline of authorities and with my squad when we faced "corrective action" (push-ups or other physical exercises) for misbehavior, they began to see me as their ally and in some cases, as one candidate told me, "one of us."Yet, as much as the authorities tried to treat me like a regular candidate in public and I was subject to similar (re)socialization pressures, my place within the institutional power schema was far higher.I was not an enlisted soldier and as such the military authorities had no legal control over my actions.I had full access to my company officer and sergeant to discuss my project and logistics-a project which they knew had the support of their superiors-and if I wanted to, I could just walk out the building at any time and leave.
Beyond the power associated with the researcher rather than candidate identity, I was a White, cisgender, heterosexual civilian graduate student coming from an Ivy League university.My nicknames throughout the years such as, "Princeton," "CC [cadet candidate] PhD," and "mom," off-the-cuff comments like "must be nice that you can leave whenever you want" (even though I rarely left), and the occasional "no offense, Taylor" after Black candidates spoke negatively about "White people," demonstrated the extent to which these identity characteristics were relevant to how others saw me within the setting.I was not an official military or academic authority, but my position as a researcher and several high status identity characteristics afforded me significant situational power.I may have been doing my best to adopt equity-oriented practices during the research, but there were enormous power differentials between us across several axes.As such, a close attention to how power functioned in the overall institution and within my research relationships was my second ethnographic strategy.
During informal conversations and interviews, I asked interlocutors about how they understood my role within the institution.None of my participants said they thought of me as an institutional authority-I was "too chill" or "one of us."However, they were well aware that I had access to the authorities and sometimes semi-jokingly asked me to use it for our advantage-"can you ask [sergeant or officer] if we can…," but I always refused.Some participants said I was just like any other candidate because I was them every day in the institution, whereas others emphasized I was different but it was okay because we were "both there for our education" (them for a college degree and me for my doctorate).
Often candidates shared that I was a source of comfort for them throughout the yearlike a mother, big sister or mentor.They appreciated having an "open minded" civilian who understood what they were going through with whom they could share their thoughts.Athena once told me, "I would not have been able to finish this year without you."Although her comment made me feel good as a human being, it once again underscored my situational power: I was not a formal authority, but I was an informal authority to the extent that several candidates decided that I was a good person to talk to when they wanted to share their thoughts or get life and school advice.I realized I had to be extremely careful about how to navigate an emerging therapeutic relationship, in which the need for support was exacerbated because candidates were experiencing' vulnerabilities across spheres. 12 Reflexive autoethnography further shaped my understanding of power within the institution and research (Mitchell et al., 2018;Mitchell et al., 2013).I noticed how I began to physically and discursively embody submission to authorities, and the ways in which I felt disempowered when I no longer had full autonomy over my time, physicality, or voice.I used my embodied experiences to ask candidates questions about how they felt in their own bodies and minds and gain more insight into their perceived changes.Additionally, I aimed to use my own position as a woman and religious minority in the institution to better understand the dynamics between men and women and other dyads of the powerful and the powerless that are often seen as normal in the military setting (Enloe, 2015;Hunt and Rygiel, 2006); however, I was aware from an intersectional perspective that I was still in a highly privileged position within the organization and the larger United States.
An explicit research strategy of attention to power dynamics sensitized many participants and me to the ways power worked within the institution, our research relationship, and greater national context.It went hand in hand with a participant-centered strategy that aims to elevate and affirm interlocutors through recognizing and minimizing exploitation.However, it is not possible to fully eradicate power imbalances or pressures from even ostensibly "ethically good" and collaborative studies (Durham Community Research Team, 2011;Mitchell et al., 2018: 14-15).There were likely candidates who decided to take part in the study because they 'liked me' and thought I was nice person or a good squadmate.Others may have worried that their friends might judge them if they admitted they were not interested in a project in which those friends were so invested.They may have also felt some underlying coercion from the fact that the authorities had let me into the institution.For this reason, I regularly (re)negotiated consent and confidentiality with my participants throughout the research and writing process.
Strategy three: Negotiating consent and confidentiality "You have to hear Pete's story.He's experienced the worst of it here," Athena said to me one afternoon as I sat down with her at lunch.She was seated with a few Black football players from another company whom I had not met yet.When she spotted me entering into the dining facilities and waved me over, she had explained to them who I was and about my research project.
"Yeah, write this in your notes…" Pete said nodding along.He began to share with the table stories about racial conflict with authorities during basic training and the start of the academic year.Others jumped in at certain points to share their own reflections or add details to parts of the story.At the end of lunch, when others had left the table and we were alone, I asked Pete if he was interested in officially taking part in the study and would like to do an interview together.Participation could also entail me shadowing him for a day to better understand his day-to-day life at the preparatory school, if he was interested.He nodded enthusiastically and said he thought it was important for folks to understand his experience.We arranged a time to go over the written consent form together and commence his participation in the study.
Given the heightened power dynamics in the all-encompassing institutional setting, frequent and private conversations with participants about ongoing consent and (the limits of) confidentiality were my third strategy.When inviting a candidate to take part in the study and arranging a time to review the consent document, I spoke with them one-on-one in order to minimize peer or authority pressure to participate. 13Within the allencompassing institution, person and video surveillance were nearly omnipresent so I aimed to hold these conversations in places and times where surveillance was minimal.If a candidate expressed interest in the study, we arranged a time to meet in a designated office in the academic hallway (which had far less surveillance) to discuss consent in detail. 14 The consent document, as prepared with the Princeton and USMA Institutional Review Boards (IRBs), outlined the purpose of the project, study procedures, potential risks and discomforts, and confidentiality.I emphasized that I was not able to guarantee complete confidentiality-because they were active duty soldiers, records from the study might be subject to search by army agencies.I would also be required to break confidentiality if they said or did something that suggested intent to harm self or others.I explained that I would share general findings from the study with institutional authorities with the intention to improve the preparatory school for current and future candidates, but that relevant individual disclosures would remain anonymous.Finally, the consent document reiterated that participation was voluntary and that the decision not to participate would not affect their relationship with investigators, the preparatory school or USMA, and that they were free to withdraw from the study at any time.I clarified that I was neither funded nor hired by the Department of Defense or any military agency to conduct this research, and that their participation (or non-participation) would have no bearing on their military career.
Before the candidates signed the consent document, I reviewed the material with them verbally and asked if they had any questions.For those who skimmed the document or said the common refrain, "I don't need to read it.I trust you-we did basic training together," I reminded them that it was important to read and go over together because it discussed their rights and risks as research participants and that though they may trust me as a human, they should not automatically trust me as a researcher.Although the participants only signed the consent document once, we had regular conversations about their decision to continue to participate in the study, their ability to withdraw at any time, and that they would have the opportunity to review the final written materials based on their accounts (see strategy four).
Discussions about consent brought to the surface questions about internal confidentiality. 15Although pseudonyms and changing participant details facilitate some degree of anonymity and "plausible deniability" (Reyes, 2018: 212), researchers have recognized the specific challenge of "internal confidentiality" wherein those tied to the research setting are more easily able to recognize participants and places (Bosk, 2008;Nespor, 2000;Tolich, 2004: 101). 16Internal confidentiality is far more difficult within an all-encompassing institution, in which there is enhanced surveillance and intimate knowledge of residents' lives.For example, the candidates lived, studied, trained, and relaxed together and knew the details about one another's backgrounds, experiences, and struggles.Despite my efforts to conduct interviews in private settings and associate with candidates regardless of whether they were officially part of the study (as I was part of a squad, company, and team and not all members were research participants), it still was possible for authorities and other candidates to glean who might be more active in the research.Internal confidentiality is increasingly tenuous when studying the experiences of those who are limited in number in an organization (Dickson-Swift et al., 2008).The preparatory school may have been among the most diverse officer training programsover 50% of the candidates identified as Black, Hispanic, Asian, Indigenous, or mixed race-but there were still small populations of women, openly LGBTQ+ candidates, and religious minorities.
In order to facilitate increased internal confidentiality, I conducted fieldwork at the institution over three years.This tripled the number of possible participants and added far more ambiguity to who a story of a particular person or 'character' may represent. 17For example, if there were 50 women at the preparatory school each year, my pool widened to 150 possible women candidates.This strategy is an approach to longitudinal ethnography that is specifically geared toward mitigating the risk of participant identification; though it still includes the methodological benefits of seeing how the field changes over time and facilitating the punctuated revisit in which the ethnographer can leave the institution for a period of time to regain some outside perspective and enter again with fresh eyes (Burawoy, 2003).As I will explore below, adopting a creative nonfiction form further gives researchers license to meld together the stories from multiple years of field work to increase ambiguity about who specific characters represent.In the next strategy, I show how considerations in the form and process of writing integrate the previous three strategies into research presentation.

Strategy four: Engaging participants in research presentation
One Sunday afternoon in the spring, Lisbeth and I went on a walk around post to get some fresh air.After trudging up a rather steep hill, we decided to rest for a bit on a bench.We began to talk about my research, and I mentioned that I was starting to write.
"What would you like your character's name to be?"I asked her.She knew I was writing the book as creative nonfiction and that her story would become a 'character,' as it could not fully represent her complexity as a person.
"Ooo.Hm…" She thought for a few seconds."How about Lisbeth?" "Like from the Girl with the Dragon Tattoo?" "Yeah.I like her a lot.I feel like I have a lot in common with her.Let's do Lisbeth.""Great.And for a last name?""Hm…well, we need something Asian but ambiguous.How about Sokolov?" "Sure!It's up to you."She nodded and we continued to chat about her 'character,' what details we would change in the book, and which events were important for the reader to know to understand her experiences at the preparatory school.
This encounter with Lisbeth was part of my fourth strategy for ethnography in allencompassing institutions: incorporating a participant-centered approach, attention to power dynamics, and ongoing consent into research presentation.Critical researchers have highlighted that the representation of 'subjects' in published work often maintains power in traditional locations and risks perpetuating stigmatization that can contribute to making research populations more at-risk in the future (Cannella and Lincoln, 2017;Liamputtong, 2007;O'Neil, 1996;Paradis, 2000;Pyett, 2001).As such, I aimed to carry collaboration into the writing process.Building off models of participatory art-based research presentation, I chose to present the research findings in creative nonfiction form in order to make it more accessible for participants than traditional academic works (Barone, 2008;Caulley, 2008;Mitchell et al., 2018).My goal was to take steps toward the democratization of knowledge, as well as maximize participant engagement by writing in a style that was easier to understand and contribute to for those outside of the ivory tower.
In co-creating the stories for the book, I invited my participants to choose their own character names, as names are deeply intertwined with one's identity and history (Svalastog and Eriksson, 2010).Although my interlocutors and I decided together that it was safer for their future careers in the military to use pseudonyms, choosing their own character names offered them agency in the storytelling process and ensured that the name they were called still resonated with their sense-of-self. 18When naming themselves for the book, candidates often choose the names of role models, book characters or ones that they felt gave off the correct gender and ethnic cues.
After discussions with the candidates, I also decided not to create composite characters because this blurred together distinct individuals who had their own life stories, histories, and experiences and was contrary to an intersectional approach-combining people based on one or two identity characteristics erased the other aspects of their selves.Composite characters also had the heightened risk of reinforcing stereotypes and minimizing the complexity of (re)socialization experiences.However, participants could choose to split their story into more than one character.For example, if sharing the gender, race, religion, and sexuality of one participant would make the character too identifiable but each of these components was integral to the participant experience, we could split the person into two different characters.In this way, the characters presented in the findings remained tied directly to specific individuals, but one person may have his or her story represented in more than one character.
Study participants were additionally among the first critics and audience for the written work, only after my graduate school advisors.I shared (and continue to share) works-inprogress with them to give them the opportunity to "speak about" and then "speak back" to (mis)representations (Mitchell et al., 2018: 4).This strategy is aligned with participatory arts-based research which invites participants to engage in continuous dialogue with written or visual work in order to shape, recreate, and speak back to the ways their voices are being represented (Mitchell et al., 2018).In reviewing the chapters together, I asked interlocutors: does your character speak to your experiences and does it feel like an authentic representation?Is there anything you think should be added or taken away from the stories?Rather than asking, "is it the truth?", as there is no absolute truth or reality in a postmodern approach, we considered together "in what ways is this true?" (Alexandra, 2015: 48).What does the story mean to them and what purpose does it serve in the book?We examined if their characters are appropriately altered to protect their identities.If they did not feel adequately anonymized, we worked together to brainstorm strategies on what else could be changed.As we contemplated these questions, I re-wrote and revised the manuscript, and then returned portions to them again, if they were interested in continuing the conversation.Although this process is time consuming, it creates space for restoration and repair when a participant feels misheard or misrepresented in the work. 19 Adopting a writing strategy that offers research participants opportunities to speak about and back to (mis)representations is particularly important when working with residents of all-encompassing institutions.As part of institutionalization, residents often have the external signs of their personal identities stripped away with a shift in the bodily and discursive ways they can represent themselves to others, especially if there are uniforms, strict standards for appearance, and protocols over behavior and language use as exist in the military (Centeno and Enriquez, 2016;Goffman, 1961).Individual voices can be lost (or intentionally silenced) in institutional attempts for uniformity and control.In writing, researchers make decisions that serve to either perpetuate institutional (mis) representations and non-consensual identity alterations, or support the agency of participants in shaping how they want to be seen in the world.

Conclusions
Ethnography within all-encompassing institutions contains specific challenges for researchers.All-encompassing institutions are settings that by definition have no separation between home, work, and play, maintain nearly constant surveillance, hold strict power differentials between authorities and residents, and subject residents to bodily and discursive control.These circumstances make it far more difficult for ethnographers to protect the identities of interlocutors who desire anonymity and confidentiality, as well as exacerbating power inequalities in research relationships.In addition, residents often experience heightened and cumulative physical, psychological, spiritual, and structural vulnerabilities, which introduce additional ethical questions into data collection and presentation.Delving into the research dilemmas that emerge in all-encompassing settings emphasize the extent to which context, space, and power matter in shaping the methodological tools that ethnographers choose to adopt.
The four strategies discussed in this article are designed to help ethnographers navigate the challenges inherent in all-encompassing institutional settings through adopting a feminist, equity-oriented approach to research.They encourage ethnographers to consider if and how participants' voices are centered in the research and writing plan, and the ways in which the study facilities partaker autonomy or further restricts their control over their lives and stories.Specifically, the strategies advocate for researchers to invite interlocutors to be co-collaborators in knowledge generation and presentation, and to pay close attention to how power dynamics between authorities, residents, and the investigator shape the research endeavor and possibilities.
The increased effort and time required to conduct research within all-encompassing institutions in a feminist and equity-oriented manner is worth the investment.The work has the potential to illuminate worlds and lives that are typically not visible to outsiders and replace dominant-and often stereotypical, reductive, and detrimentalrepresentations.Ethnographers' commitment to understanding the stories of those who have been historically separated, hidden, ignored or silenced has the potential to draw attention to social inequalities and pave the way toward policy change.In undergoing this sometimes difficult work, researchers may consider joining a group of peers and supervisors who are also committed to critical inquiry in order to have a sustaining and supportive community with which to reflect on complicated interpersonal and power dynamics that emerge in the field, process difficult emotional and social encounters, and problem-solve together (Dickson-Swift et al., 2008;Klostermann et al., 2020;Winfield, 2021).It is through collaboration between researchers and participants and support among colleagues that the conditions emerge for radically ethical scholarship both within in allencompassing institutions and across all ethnographic endeavors.7.During basic training, military recruits must rise, act, eat, and sleep according to the orders of their chain of command, are under nearly constant surveillance, and there exists a clear social boundary between the authorities-noncommissioned officers and officers-and new soldiers.8.For example, Svalastog and Eriksson (2010) discuss the merits of replacing "default anoymization" with "careful deliberation, together with research subjects, of how to handle the issues of identification and confidentiality" (105).9.As such, I will use also use the terms 'vulnerable,' 'disempowered,' and 'at-risk,' when referring to those experiencing physiological, psychological, spiritual, or structural vulnerabilities throughout this article.Physical vulnerabilities include chronic or acute health ailments, and being at high risk for physical harm (e.g., physical disability, injury, disease, drug/alcohol dependency, or being in dangerous areas like war zones).Those who are psychologically vulnerable may be experiencing high levels of stress or emotional distress, in response to specific events or underlying mental health challenges (e.g., grief, abuse, trauma, depression, anxiety, suicidal ideation).Spiritual vulnerabilities emerge for those who are questioning their faith, struggling to find a sense of meaning, experiencing gaps between existing beliefs and new realities, or are isolated from spiritual and religious communities.Individuals experience structural vulnerabilities when they are in situations where they are unable to act autonomously and make their own choices, marginalized due to race, gender, sexuality, or other characteristics, or lack the resources to meet their basic needs (e.g., legal status, housing/employment instability, poverty, food insecurity).10.I capitalize White in line with scholars who argue that capitalization is important to emphasize the nature of "Whiteness" as a social category, draw attention to its perceived neutrality, and refuse to "grant it the power to maintain its invisibility" (Ewing, 2020).11.Challenges related to negotiating access, relationships with institutional authorities, and self-care in these settings have their own specific dynamics and processes that require a separate set of strategies.For existing work on navigating access, research with more powerful participants, and ways to maintain physical, emotional, and spiritual well-being during immersive research, see Belur 2014;Ellis 2021b;Huckaby 2011;Winfield 2021. 12.When candidates came to me with sensitive questions or concerns about their lives, I would remind them that I was not a therapist and provide them with information on where they could receive free and confidential mental health support on post.However, many were concerned that going to a military therapist or chaplain could jeopardize their admission to USMA, even if sessions were confidential.In cases where I was most concerned about specific candidates, they did end up getting official support (often because a peer mentioned it to the chain of command or their company leaders, coaches, or teachers noticed they were struggling).Fortunately, there were no situations when a candidate came to me in acute crisis and I worried about harm to self or others.For more details on addressing the therapeutic relationship, see Winfield 2021.13.To identify candidates who were interested in partaking in the study, I first invited those who were in my squad and company to participate and later in the year expanded to include others in the battalion.14.We scheduled meetings based on their convenience to ensure that the research did not interfere with their institutional obligations or study periods.Often we did interviews during lunch and dinner time and I would order food for us from off-post restaurants.I was not permitted to directly compensate participants because of military research regulations; however, I could purchase them meals of their choice during our discussions and formal interviews.15.Again, I am coming from the perspective that anonymity and confidentiality are a positive outcome for research in all-encompassing institutions; for alternative approaches see Reyes (2018) and Svalastog and Eriksson (2010).16.Based on conversations with participants, I did not mask place.The historical and socio-cultural context of the location was integral to participants' experiences (Gieryn, 2000;Nespor, 2000;Reyes, 2018); for example, the prestige of being associated with West Point and history of exclusion and later attempts to integrate non-White, women, and LGBTQ+ soldiers contributed to the (re)socialization process.The only comparable programs are the Naval and Air Force Academy Preparatory Schools but their training programs, geographic locations, and proximity to their parent Academies are distinct.Anonymization would additionally deflate the possibility of internal and larger policy changes for the military as a result of the study, which was an outcome that was extremely important to the participants (Nespor, 2000).Moreover, masking place would not help with internal confidentiality because anyone who was at the preparatory school during my research period would immediately know the location.17.I use the term 'character' because published stories are inevitably incomplete representations of individuals' complexities and full live experiences.Instead, participants and I focused on developing a 'character' that was an authentic representation of this stage and piece of their lives (see strategy four).18.We discussed the benefit and challenges associated with using their own names and decided together to use pseudonyms.I did not have any participants request that I use their real names.My original IRB application with Princeton and USMA said that I would not record names or any personally identifiable details in research documents; therefore, if a participant had requested that I identify them in the published work I would have had to coordinate this with the IRBs.19.In order to acknowledge the time participants were investing in this process, I purchased them meals during our dialogue sessions.Unfortunately, regulations for research with active duty soldiers prohibited me from compensating them monetarily.