Who is “I”?: Subjectivity and Ethnography in HCI

2.1 Ethnography – A Brief Context

Ethnography is anthropological practice that emerged in the early 20^th century, primarily characterized by the researcher's immersion in the “field” [4,18,25]. The ethnographer conducts activities including “making anecdotal observations, doing discourse analysis, analyzing the use of space, conducting extended case studies, and other practices” [18] by participating in the culture under study for extended periods, and then reporting her research findings back to the research community. In anthropology, theoretical and methodological resources offer best practices while articulating norms of rigor in reporting, such as motivating the chosen data collection methodology, including field notes and verbatim quotes, and using triangulation of the data to support claims (e.g., [18,84]). Researchers, using these norms, must legitimize their findings and research outcomes to the epistemic community.

Yet, many published accounts of researcher's experiences [3,15,34,65] indicate that building one's ethnographic research repertoire is often a personal endeavor, cultivated over time by actually doing fieldwork [68]. “There is a tradition in cultural anthropology that one cannot be told how to do fieldwork” writes Agar, [68:2] implying that the only way to learn to do fieldwork is to be “thrown” into a community and emerge on the other side as a (hopefully) successful ethnographer. While Agar is likely trying to be provocative, he brings attention to two concerns with regards to ethnographic practice. First, while books and best practices offer guidance, what happens in real time is guided by the ethnographer's personal and subjective way of conducting oneself in the field [68:2–4]. As a result, the ethnographer often must also have a degree of self- and social-awareness and develop it to improve their craft [84:8–13]. Second, because ethnographers are often a single point contact between the culture and the epistemic community, they also must overcome the skepticism with regards to the scientific legitimacy of their work through their reporting [14]. To obtain this scientific legitimacy, early ethnographic researchers often retrofitted their findings to the rigor expectations of traditional scientific paradigms, adopting a scientific, dispassionate, and authoritative voice [64].

However, ethnography today, while still committed to rigor in evidence gathering and analysis, is also recognized as subjective and interpretivist in nature [57,64,68]. By “interpretivist,” ethnographers often stress that the work must offer an account of the emergent interactions among “1) the observed 2)[…] the observer 3) the representational style selected to join the observer and the observed 4) the role of the reader engaged in the active construction of the tale (the audience)” [64:xv] – a description that aligns with the ethnographic practice of the authors of the present work. In recent years, “confessional accounts” (e.g.: [1,2,15,58,65]) autobiographical monographs (e.g.: [15,64,68]) and experiential ethnographies that show-case first person researcher's experiences (e.g.: [3,46]) has become an important ethnographic subgenre. The goals of such ethnographic writing are varied. One is to explicate what actually happens in the field, while acknowledging the partial perspective of the researcher providing the explication, thereby furthering the research in the domain [16,34,70,76]. This researcher-focused genre also makes space for researchers to talk about their emotional labor, share their experience, and consequently contribute towards a repertoire of community practice. This epistemic culture in anthropology informs an aspirational goal for the present work: How might we contribute towards building such an epistemic culture?

HCI researchers originally adopted ethnographic methods to overcome the limitations of the earlier HCI methods to capture the context of the “end-user” [4].  A large body of work in HCI applies ethnographic methods as a part of research for design, notably including ethnomethodology, which is particularly well positioned to account for interactions at a level of granularity that has proven useful for informing concrete design decisions. Since the 1990s, HCI practitioners have expanded the concept of “user-context” from a “dialogue” between a single user and a system to factor in macrostructures such as socio-technical systems, polities, cities, and related cultures [13,21,22,40]. In these broader contexts ethnography is used to present an interpretivist understanding of the people, their cultures, and possibilities of technology use rather than specific design outcomes. Dourish [37] describes interpretivist view of ethnography as an act of interpretive theorization developed by the ethnographic researcher to understand and help others witness what the ethnographer experienced and witnessed during fieldwork. The objective is to expand our understanding of the design space rather than converge at specific design outcomes.

A salient feature of an interpretivist view of ethnography is the perspectival (i.e., subjective) nature of the practice – – “ethnography is always, inherently, a perspectival view, and that this perspectival quality is critical to what ethnography is” Dourish [37:544]. Consequently, researchers’ subjective experience has become a central theme in some recent works. Some examples include somaesthetic interaction design [50,56,85], experience design [39,60,66,91], humanistic HCI [7], sensory ethnography [75], and autoethnographic research (e.g., [20,31,79]). Collectively, these and similar works are helping to shape a research trend in which the interpretivist stance is foregrounded, though in many cases, implicitly rather than explicitly. This trend both provides foundations for the present work and opportunities to help advance them.

2.2 Reflexivity in HCI

Early ethnographic projects involved the study and documentation of remote, inaccessible cultures such as tribes in remote villages, places in global south, and other locations that were outside of, and not understood by, Western researchers [74]. The ethnographic researcher became the single authoritative source able to represent these cultures, often by essentializing them. Subsequent social science research has critiqued early ethnography as a power-imbalanced, othering practice situated within a colonized, western, and privileged undertaking [16,26,41,52,55,70,74]. Reflexive ethnographic research practice provides ways to diffuse cultural essentialization and humanize the otherwise authoritative “god voice” [47] behind dispassionately written objective-seeming ethnographies [64,83]. By encouraging ethnographers to examine their role and impact on the research and participants, reflexivity involves turning the lens on oneself and including that positionality and perspective as part of the contents of ethnography. With its roots in feminist and emancipatory philosophies, reflexive research practice entails recognizing and disclosing one's subjectivity and its influence on our research [16,76].

Within HCI, the rise of reflexive ethnographic research is relatively recent, though reflexivity can be found in related areas. Feminist HCI encourages interaction designers to take responsibility of their agenda-setting role so as to develop designs that embody social justice and emancipation [9,12,82], participatory design includes a reflexive, participant driven emancipatory practice [11,22,48,86]. Some recent works explicitly leverage reflexivity to study the researcher-self, such as autoethnographic design projects that have the self as the participant and object of study [20,31,78,79]. Similarly Garcia and Cifor [44:190:6] reflexively show that “individual and collective feelings, experiences, or perspectives as a vital component of the research process” in their duo ethnography, and Howell et al. [51] extend that to a trioethnography. Balaam et al. [6] center the qualitative researcher's emotional labor through stories of personal experiences of researchers. Devendrof et al. [32], raise a methodological challenge that highlights the struggles of collecting emotional data and sharing these emotional experiences while also maintaining the researcher role.

This research foregrounds the ethnographer's subjectivity and explicates the methodologies they used - writing notes privately and sharing with each other, engaging in a dialogical understanding of each other and themselves [6,32,43,51]. However, at an individual researcher level, the concept of self and how it was arrived at individually, the challenges a researcher faces internally in doing so and how she overcomes that, remain largely tacit. Garrett et al. [45:1] contribute towards this issue by foregrounding the investigation of the self through “the felt self, inter corporeal self, socio-cultural and political self, and entangled self” as a method to aid researchers in developing their ethical sensibility.

The present work is thus situated within reflexive ethnographic research in HCI that foregrounds the “I” in the ethnographic accounts. We propose an approach that empowers ethnographic practitioners in HCI to learn about and cultivate their subjective ethnographic “I”, by using Doucet's notion of Gossamer Walls.

2.3 Gossamer Walls and Ethnographer's Subjectivity

We draw upon sociologist Andrea Doucet's conceptualization of relational ways of knowing to frame our reflections and reflexive process. In talking about the use of her metaphor “gossamer walls” she writes:

The metaphor of gossamer walls, combining the sheerness of gossamer and the solidity of walls, provides for creative ways of conceptualizing reflexivity in temporal and spatial terms as well as to consider the constantly shifting degrees of transparency and obscurity, connection and separation that recur in the multiple relations that constitute reflexive research and knowing [69]

Borrowing from Anne Michael's novel Fugitive Pieces, Doucet further claims that gossamer walls mediate mediates three different sets of relationships:

3.1 Interpretive Reflexive Approach and Impressionist Writing

In this section, we first present the broader research context within which the present work is situated and explain our research methodology. We follow that with a detailed review of how we reflexively engaged with our fieldwork experience. We also explicate our use of Van Maanen's “impressionist tales” – ethnographies that are written as stories that show the readers the researcher's experience and allow us to “jointly examine” the culture which we are studying and the ethnographer's ways of knowing – as our narration style [64:101-124]

3.2 Research Context and Data

The present study is part of a broader ongoing collaborative research program that focuses on bottom-up innovation in small-scale urban farmer's practices in the context of sustainable food production. It draws upon our ethnographic fieldwork in small-scale urban and peri-urban farmers around Indianapolis, Indiana at the beginning of 2020 and during the pandemic. Our ethnographic practice is predominantly based in observation, which includes going to a field site to observe practices first-hand and in person; using walking probes [29]; interviewing participants; producing ethnographic jottings in real-time; creating diverse visual representations including photographs, sketches, and diagrams; and so forth. These methodologically follow [84:47-82] in structure and process. In the field, the lead author made jottings and mental notes [84:50], and then subsequently wrote them up as full-field notes [84:50–54].

Subsequent analysis is interpretivist in nature. Interpretivist analysis has been characterized as an “effort after meaning,” which is to say that significances reveal themselves over time through a kind of labor [23]. That labor is primarily constituted by an iterative back-and-forth movement between concrete particulars within the ethnographic experience and our representations of it (i.e., in jottings, sketches, transcripts, etc.), and more holistic themes and design-related issues of focus. This motion is back-and-forth, because both the concrete particulars within our evidence and broader themes direct the analyst's attention back to the other, in an iterative and mutually informing way, as their meanings gradually come into awareness. We have previously published empirical findings from that research [19].

Here, we focus specifically on three farm visits from that ethnographic undertaking to stage a reflexive dialogue between ourselves and our fieldwork. We conducted research during the later part of 2021 and early 2022 as a retrospective analysis of our research practice. We seek to develop an explicit self-awareness as part of both the study design and the analysis of the ethnographic experience. To develop this explicit self-awareness, we use Doucet's three-part relationship we describe section 2.3: with herself, her research participants, and her epistemic community. To foreground these relationships and understand how they shaped the subjectivity and thus the research practice of the first author, we iteratively did the following.

First, the first author developed reflexive notes emphasizing her personal experience. She re-engaged with the material outcomes of the 2020 field visits in the form of transcripts and voice recordings, images, field-notes and initial interpretations and analyses done in the past. Reengaging with these materials enabled the first author to re-familiarize herself with her in-the-field experiences and the perspective from which she had interacted with the farmers. She then used “confessional tales” [64:73-100] as a medium to express her subjectivity. Confessional tales showcase the researcher's point of view with as much naturalness/veracity as possible in describing how the research work came to be [64:73-100]. This makes it a suitable methodological writing tool to surface one's experiences by forcing one to put oneself at the front and center of the conversation. As the name suggests, these are confessions; thus, they also require the intentionality of searching for and exposing one's personal thoughts and experiences and are not just performative writings.

Informed by humanistic interpretation [7], we then tasked ourselves with identifying connections between the first author's reflexive notes and fieldnotes from the past. Specifically, we leveraged explication de texte [72] or close reading, from the humanities. Close reading leverages the hermeneutic aspects of interpretation and facilitates subjective engagement with the materials. Our interpretive process was theoretically informed by Doucet's reflexivity framework and John Van Maanen's concept of impressionist tales [64:119]. By iteratively going through this process, we connected the first author's subjectivity – her perspectival view - to her 2020 interpretation of the data, specifically identifying and demarking “data” closely associated as an outcome of the three ethnographic relationships Doucet proposes.

We then iteratively discussed and refined these connections to formulate them as impressionist tales by leveraging the structural conventions presented in Maanen's work [64:101-124]. These structural conventions include playing attention to the textual identity, narrative, novelistic event-by-event fragmented style, dramatic undertones and characterized portrayals of the actors (researchers, participants, etc.) [64:103-105]. The stylistic goal of impressionist tales is to be able to present the “doing of fieldwork.”

The story itself, the impressionist's tale, is a representational means of cracking open the culture and the fieldworker's way of knowing it so that both can be jointly examined.” Consequently “the epistemological aim is then to braid the knower with the known [64:102]

We thus used impressionistically written ethnographies to craft narratives that allowed us to showcase the “doing of the fieldwork” not just through the first author's tangible actions, but rather through her feelings, emotions, and aesthetic experiences.

3.3 Who is “I”?

We now present an autobiographical essay – “a retrospective confession” by the first author, with the hopes that the readers can witness, in part, the making of the subjectivity of the “I” behind the reflexive accounts in section four.

My early career training was in Informatics Engineering, which upholds values of scientific rigor, precision, accuracy, generalizability, mathematics etc to produce accurate and reproducible program codes. Yet, I find humanistic interpretation closer to my disposition than the logical flow of writing a code. I often coded intuitively, resulting in definite functional outcomes but really messy programs. In programs, messiness is undesirable, and yet the messiness of my field notes often surprises me at my own tacit sensitivity to social, communicative, and embodied phenomena. There is a messiness to this process that my engineering training taught me to reject, and yet when it happens, it makes me feel alive as a researcher. How could I reject as lacking rigor the epistemic conditions that were most fruitful to my research?

Embracing this apparent “messiness” as a postmodern ethnographer, I posit that my research offers one perspective on research participants and their context and cultures. Philosopher Robert Pool [77] shows that the key idea behind being a postmodernist researcher means acknowledging that “reality” is not an unmediated, objective phenomenon, but rather a partial perspective. His view entails ongoing self-directed doubt about how we see the world and define “knowledge”. My subjectivity thus has evolved from pursuing “The Truth” worldview of scientific engineering, to acknowledging the perspectival nature of ethnographic accounts.

I started the research project on innovative practices of small-scale farmers in the Spring of 2020. I have never done any sort of garden work, farming, or even lived close to a farm. I am an Indian and I grew up in the city of Pune, and all I ever knew was traffic, noisy streets, and pollution. When we as a team visited farms, I was enraptured by how content I felt as I listened to the participants share their stories, as I pet chickens, ate beef stew and that feeling often reiterated the question I asked above: Who am I here but an outsider witnessing a life I know nothing about? An imposter almost.

This feeling is not helped by the endless hours I have spent trying to make sense of all the data I have captured during the 2020 field visits. My field notes sometimes shock and surprise me but also make me wonder “Why did I care about that strange thing?”. I might have spent 100s of hours making sense of my own mind through application of theory, coding, and rigorous analysis over the last two years, only to decide that maybe I want to scrap the idea of writing up that fieldwork. It was on this precipice of giving up that I stumbled across Doucet's work and felt the same excitement I feel when I find interesting connections in my field notes. By familiarizing myself with different ways of writing ethnographies, I can finally begin to pacify the imposter and embrace different ways of knowing. The present work is me following that intuition and cherishing the messiness of the process. In this work, I want to explore the boundaries of subjectivity in ethnographic fieldwork in HCI, maybe attempt to push them a bit. In that process, I hope to learn as much about the people I research as I do about myself and ethnographic research practice within my epistemic community.

With this autobiographical essay, we set the background for our presentation of our impressions as findings.

4.1 The Researcher's Relationship with Herself

The researcher's relationship with herself represents different versions of herself, spatially and temporally located across her lifespan and experiences. Doucet thus describes the first gossamer wall - her relationship with herself as “ghosts” across time and space:

on the other side of a first gossamer wall are relations with our many selves as well as with ‘ghosts,’ deeply buried across time and space, that may come back to haunt us when we are physically and emotionally invested in our research [34:73]

Doucet implies that, while doing research, we recall fragments of those past selves “that come back to haunt us.” Personal and emotional investment accentuates the sheerness of the gossamer, enabling a dialogical conversation with these past versions and see how they have shaped us, and our research practice.

The first author experienced the “sheerness” more than a year later, and after iteratively going over the transcripts and fieldnotes for an extended period. The moments of failures, the frustration of not being able to capture and unpack her understanding of the farms, and yet the personally motivated drive to explicate her perspective led her to find traces of her own self in the data that filled her field and analytic notes. Doucet's conceptualization of past “ghosts” empowered the first author to embrace vague childhood memories and experiences and set up a dialogue between herself and these past versions. Thus, here, we present an impressionist tale, titled “Abandoning the researcher.”

4.1.1 Impressionist Tale: Abandoning the Researcher

Ethnography in HCI and anthropology relies on jottings, field notes, interview transcripts, photographs, videos, etc. to capture and store research data. In addition, the researcher must maintain a regular dialogue with her research questions to observe and note anything that responds to them in interesting ways. That is, she must always maintain a boundary between herself as the “researcher” and herself as an individual. But as extensive research in social sciences has shown, a strict separation of the human-being and the researcher is virtually impossible. This impressionist tale thus narrates the difficulties of maintaining the separation between one's person and the role of the “researcher.” It presents a situation, when at the spur of the moment, the researcher metaphorically “abandons” the researcher role and focuses on experiencing her surroundings and field visit.

This tale is about a field visit to Wilma's peri-urban farm [anonymized] around Indianapolis. She raises lambs for meat, chickens for eggs, and has a small vegetable garden. During the initial analysis phase in 2020, in the first author's imagination, Wilma's farm represented ideal pastoral living, away from technology and the city. This author even questioned why anyone might introduce technology into the seemingly “ideal” pastoral place and lifestyle of the small farm. Yet, when it came down to articulating and defining this “pastoral living”, the first author could only describe how she felt and was convinced that that intangible feeling meant cozy pastoral living. For her, ideal pastoral living was not a clearly articulated construct, but rather a feeling. Staying with this feeling, the first author immersed herself in her field notes and listened to the audio recordings. The recordings didn't just replay what they talked about, but also the sounds of chicken and sheep and the biting cold February wind. This multi-sensorial (re-)immersion by the first author, writing, and rewriting reflexive confessional dialogues with herself, coupled with iterative discussions between the authors, helped us to braid the knower with the known as presented below.

Raised as city girl, farms meant a school field trip where we drank frothy buffalo milk, petted calves, and enjoyed the fresh countryside air. It was also a place of escape, away from the city. These distant memories of school field trips did not resurface as distinct, quotable memories, ready to be documented in my researcher positionality. Instead, I felt them. Standing at the doorstep of Wilma's farm for the first time one cold February morning, the crispness of the peri-urban air in my lungs felt strangely nostalgic. I felt alive, and from that moment on I left the researcher outside in the car. Or did I? 

Wilma guided us through the hallway, and we placed our coats in a closet opposite to a cupboard with denim overalls, boots, and the works. We went straight to the kitchen – where she was making a beef stew for us for lunch. We seated ourselves at the kitchen table. As we discussed her farm's history, I was mesmerized by our surroundings. There was fresh bread waiting for us, plates, and bowls ready to be filled, and the rich aroma of steaming beef stew on the stove top. At noon, she urged us to help ourselves to stew, coleslaw, and a serving of bread with butter. The hot, thick beef stew was perfect for a cold winter lunch and the coleslaw crunch complemented and lightened up the rich, spicy flavors of the stew.

We sat at the dining table, behind which there was a stack of eggs and egg cartons, ready to be packed and sold. The chickens could be seen around their pen, right outside the window near the dining table. All of this was in alignment with the feeling Wendell Berry - a farmer and writer [17] had made me feel through his writings on agrarian American agriculture - Wilma's farm became the infatuation I have today with the notion of pastoral living. Amidst all this, I could not exactly locate or find the researcher within me. I was just a person, escaping the fast and loud city life, enjoying the beef stew Wilma made for us. In a way, it felt like a school field trip. I pet the chickens, was smitten by the adorable sheep and consumed by the smell of hay and fodder in the barn. I felt content. I did not hold onto my notebook either and did not take any in-the field notes. I am so glad I did not, because then I could hold three eggs in my hand alongside Wilma as she showed us around.

Figure 1:

View Figure

But in another sense, I did take notes. I took them in the palm of my hands as a permanent sensation of picking up and holding three eggs delicately. I will probably never forget the sense of pride and accomplishment I felt in placing the eggs safely in the collection basket. I took the notes on my fingertips as I felt the waxy, leathery feathers of a chicken I pet. I wrote them in the memories of the strange sounds the chickens made and in the silly expression the chickens had when they looked at us. These sensations resurfaced long-forgotten memories of school time farm visits. Of farm being about an escape from the city life and maybe, just maybe, the following analysis for Wilma's farm was associated with this feeling rather than anything that was precisely said or documented.

Wilma's farm was representative of the romanticized vision of farming city dwellers had come to imagine—a remote, cozy home surrounded by pastures and grazing animals, far removed from bustling city life. (First author's analytical notes during the interpretation phase between February 2020-Feb 2021).

This single sentence interpretation, however, does not have “data” grounding in the traditional scientific sense of the word. Sure, photographs we took on the farm coupled with some choice quotes from Wilma might help us claim that the lifestyle was the most attractive aspect of Wilma's farm. However, we argue that the interpretation presented here is not a summative product of all the tangible information we collected. Instead, through the first author's account, we showcase how the first author's subjectivity, and latent, sensorial experiences, shaped her feelings during the field visit and how that led to consequent interpretation of the farm. These feelings emerged on Wilma's farm because the experience intersected with the first author's lived experience and elusive “ghosts” from the past, and her leisurely reading of Wendell Berry's essays on American agriculture [17], amongst other things that might inform her subjectivity.

As others have before us, we found that that which can be merely factually observed has only potential meaning until it is actively noticed and interpreted by a sense-making ethnographer. For example, the first author noticed the layout of the house and the farm but saw it as a feeling and ambiance; as an experience that was unique to her. A layout map of the farm would accurately provide the location of different elements of the farm down to the last inch, but what would it tell us about Wilma's farm but as limited a perspective, much like the researchers initially had of the women's movement in the opening sequence of Kitchen Stories [92,94,95] ¹*.

This notion of perspectival view, of theorizing and interpreting the data is central to the interpretivist reading of ethnography [30,37] . However, in doing the above exercise, we also discovered that teasing out the subjective elements of how we interpret our field visits is a laborious task, on account of our naturally limited awareness of our internalized subjectivity. Doucet contends that she made these connections only when she entered “the stage of physical and emotional exhaustion” and when the words of her research participants “filled [her] waking and sleeping hours and rolled through [her] conscious and unconscious mind” [34:75]. In the case of the first author, it was over almost two years later, having heard the recordings and read the jottings and written narratives uncountable times, before they started to “roll through” her mind. Both Doucet's theory and the first author's direct experiences concur in appreciating these connections as hard-won and oftentimes incomplete, especially in light of new experiences and repertoires – an appreciation that is always unfinalized [66:69]. Coupled with Van Maanen's confessional tales as a tool to write and express this dialogue, we could create the emotional immersion needed for this account—a tactic that we share in hopes of supporting fellow ethnographers in HCI.

This discussion so far has helped start building a toolbox for our first objective – provide an actionable approach to unpacking one's own subjectivity. But we also must consider its possible contribution to the broader HCI ethnographic research field. We posit that, to fully leverage the potential of reflexive approaches such as the one we propose and others (e.g., [43,51,80,83]) have advocated for, there needs to be an epistemic incentive to do so. But we save that for a later discussion and for now, we turn to how the researcher's relationship with the participants might play a role in how we do research.

4.2 The Researcher's Relationship with the Participants

The researcher's relationship with the participants examines intersubjective knowing, mediated through a dialogical interaction between the researcher and the participants.

Doucet writes of these relations:

the multi-layered relations between researchers and research respondents, relationships that can involve oral, audible, physical, emotional, textual, embodied, as well as shifting theoretical and epistemological dimensions. [34:73]

This relational knowing is mediated through an ever-changing, elusive gossamer wall, which comprises our embodied understanding of emotions and body language, empathic understanding of one another and the respondents’ and researcher's living personal histories and subjectivities. It showcases how both the researcher and the researched influence each other during an ethnographic encounter. This form of exchange contrasts with the idea of empirical, objective “data” because it violates the principle of minimizing “reactivity” – that is, minimizing the influence of the researcher and the participant on one another.

4.3 The Researcher's Relationship with Her Epistemic Communities

The researcher's relationship with her epistemic communities represents how epistemic affiliations and training interact with the researcher's subjectivity. This relationship, the third gossamer wall Doucet writes about is to “recognize the theoretical and epistemological, or epistemic communities that influence our work in subtle and explicit ways.” [34:73]. This relationship reframes how we understand the process of producing knowledge. Doucet writes that when we recognize the influence of our training and epistemic commitments, we can see that “producing knowledge is less a matter of face-to-face confrontation with the data than of negotiation within an epistemic community” [34:81]. This negotiation is an implicit and sometimes even explicit activity we engage in when we collect data or analyze it, when we write up our research to a certain audience and so forth. Consequently, the researcher's relationship with her epistemic community enables, limits, and shapes what it is possible for us to think or see.

In the following impressionist tale, the not-so subtle pressures of the first author's epistemic commitment to sustainable HCI research manifested as an expectation for the farmers she met to demonstrate exemplary commitment to nature, earth, and soil (e.g.: [59,62,71]). This expectation was also shaped by imageries of the caring, living-off-the-land, frugal farmer for sustainable agriculture from Wendell Berry's essays on American agriculture [17]. Thus, when faced with an ethnographic encounter that shook this understanding of sustainable farming, the first author, as we shall see, struggled to reconcile her epistemic position and what “data” was saying.

5.1 Who is “I”? In Pursuit of Cultivating the Ethnographer's Subjectivity

Throughout this paper, we use empowerment to describe our experience of coming to terms with our subjectivity. Our methodology - Doucet's reflexivity approach (i.e., Gossamer Walls) coupled with Van Maanen's confessional and impressionist tales is empowering as a discursive mode because it normalizes the emotional, ever-shifting human experience of being an ethnographic researcher. Extending reflexive ethnographic research in HCI (e.g.,[43,51,80,81,83]) thus, we contribute an actionable approach that provides ethnographic researchers resources to unapologetically engage with and express their subjectivity as a constitutive dimension of their intellectual contributions. Doing so helps to address some of the conflicts raised in HCI already discussed in this paper [6,32]. We argue that ethnographic researchers in HCI must often negotiate with their own aesthetic and emotional experience to write ethnographies that are acceptable and legitimized. This negotiation can be internal, potentially resulting in a truncated or sometimes even discarded datapoints or project. Or it can be external – by providing extensive justification for the perspective presented using inadequate epistemological tools to accommodate it. Consequently, researchers struggle in knowing their subjectivity, presenting it, truncating it, or sitting with the insecurities and questions it raises.

The present work addresses these struggles as follows. First, by presenting detailed ethnographic accounts that “expose” the first author's experiences being an ethnographic researcher, we hope to emphasize the similarities across experiences such as those of Balaam et al. [6], Devendrof et al. [32] and ours. Second, Doucet's reflexivity approach granularizes reflexivity down to the lived experiences of being human. That is, being a reflexive, self-aware practitioner is not an exclusively a deliberate cognitive act, but also involves elusive, sometimes surprising and at other times frustrating dialogical and multi-sensorial process of self-discovery. “Gossamer walls” as an oxymoronic metaphor and the three-part relationship of the researcher with herself, the research participants, and the epistemic communities helps one visualize and unpack this dialogical process. More importantly perhaps the metaphor shows that completely separating the “I” from the outcomes of ethnographic undertakings to represent “the objective truth” is impossible. Finally, Van Maanen's stylistic descriptions of confessional and impressionist tales provide the necessary rhetorical tools for rigor to write and present ethnographic experiences without fitting into notions of “objective truth”. Putting all this together, we hope to encourage ethnographic researchers in HCI to be curious about, rather than suspicious of, their own relationships to the external world and their sense-making of it.

This approach also has implications for how the work should be used—both in peer reviewing and especially once it is published. We have talked about the role of the ethnographer in all of this, but all this also suggests what the configuration of the ideal reader of such ethnographic work might do, that is, to reimagine how reflexive, candid ethnographies can be read.

5.2 What Does “I” tell us? Legitimizing Perspectival Writing and Partial Knowledge

I say ‘we,’ but I did not ask a single question. I sat diagonally to her, nodding, making eye contact, listening, and taking shabby notes, and holding the recorder. Maybe I embodied the proverbial fly on the wall, but it was not intentional, rather an outcome of my own, generally well-masked social anxiety. – First author's impressionist account of meeting Ava, section 4.1

The “I” in this account is self-conscious, and hesitant in accepting the legitimacy of her perspective. The “I” is also curious about how she might look like, from outside, as a non-judgmental self-reflexive observation. What does this “I” tell us about the ethnographic encounter itself? Or the “data” that might follow such a statement? One thought, reflecting contemporary post-positivist empiricism, is that the data that follows is not reliable. Yet, an ontological position that draws from idealism and/or relativism posits that reality is mediated through perception and interpretation by the human mind, and through co-constructed social meanings might facilitate a different reading of this. The “nodding, making eye contact, listening, and taking shabby notes” might hint towards the speed and length at which the speaker might have spoken. Depending on how one looks at it, such nuanced and raw presentation of oneself might make the “I” an unreliable narrator, or might induce feelings of sympathy, and trust on account of the vulnerability shared. Interpretivist, reflexive, ethnographic research posits that the “I” who writes vulnerably is in fact a more reliable narrator than the one who only claims “Ava dominated the whole conversation” as though this were an unmediated fact, rather than a social perception. Foregrounding the “I” that witnessed Ava dominate the conversation adds precision to what is reported as research findings.

The present work provides an approach that can empower ethnographic researchers to engage with and express their subjectivity, thereby foregrounding the vulnerable “I” in their research findings. However, for ethnographic researchers in HCI to undertake an intensive reflexive practice such as this, there is limited epistemic incentive. For example, [83] and more recently [6,80] suggest that HCI as a field favors the realist – objectively, authoritatively written ethnographies; that is the implicit expectation of the reader due to post-positivist roots of HCI. Thus, we argue that if we are to empower ethnographic researchers to discover and cultivate their subjective ethnographic perspective and practice, the role of the reader in engaging with interpretivist ethnographic writings must also be re-examined. Dourish [37] suggests that ethnographic writings can be read as the theorization of what the researcher gathered during the ethnographic work, rather than just empirical findings [37]. The readerly goal is to help open the design space and to facilitate alternative ways of imagining technology for different socio-cultural contexts. Another, similar, approach is to read them as “humanistic essays” [7:69]. Humanistic essays provide the space to express critical thinking by focusing on the notion of “expert subject” that crafts arguments and investigates the subject under study. [7] argue that the structure of such writing is crafted to “reflect doubts, dialogism, dead ends, and other signs of ‘enacting the struggle for truth.’” The purpose of reading an essay then is to “sit alongside [the researcher] and ponder the truth with [the researcher]” [7:69].

Both these forms of reading emphasize the active role of the readers of ethnographic writings in this form of epistemology [42,64:xv]. For example, another HCI researcher, reading our doubts and reservations about Ava, could have been inspired to investigate social justice oriented HCI's role in unpacking the dynamics at youth support groups. Amelia's farm – exemplifying the tensions between idealistic sustainable farming and financial profit could spark a debate on the construct “sustainable farming” in HCI. In short, reading ethnographies as dialogues could reproduce salient features of other modes of intellectual debate, including seminars, brainstorming sessions, and so on.

For this dialogical relationship to flourish as a part of knowledge production a larger epistemic concern also benefits from ongoing reexamination. We posit that the post-positivist scientific ideals of perspective-free truth often show up as “ghosts” from the past in how research rigor is defined, produced and published in HCI, other times as expectations of implications for design [37]. That is, the “subtle epistemic pressures” also shape and define our epistemology. For example, in the case of Ava, a skeptical question we might ask is “Can we know more and be more certain of our research findings if the first author had shared her findings with her participants, making them an object of focus and discussion, generating new data and insights collaboratively?” While COVID limited our ability to do so, one might also question the implicit understanding that “saturation and triangulation” arrive at perspective free truth. In the work of Balaam et al. [6] for example, the ethnographic researcher was hesitant, despite of using the best practice of verifying and clarifying with the participants: “Once they had finished, I often tried to ask gentle questions to clarify parts that were muddled, but this changed the way details were re-told. I started to question whether I was putting words in their mouths.” What Doucet's relational knowing highlights is that “saturation” i.e., knowing the whole truth is also often formulated by the researcher's situated judgement. Braun and Clarke [24] similarly highlight the subjectivity of “saturation”. By thus recognizing interpretivist ethnography as a subjective practice, we challenge the idea that methodological “proofs” alone can unequivocally prove that we have presented is the “objective truth.” We argue that in many ways, the pursuit of objective truth to “prove” the validity of data and reach “saturation point” legitimizes the authoritative and omnipotent “God voice” [47], thereby essentializing the participants we study. While HCI has done extensive work to emphasize emancipation of the underserved and subversion of power dynamics between designers and “users” [9,22], there needs to be more work to cultivate epistemological openness to partiality of knowledge.

Becker, in his article on epistemology of qualitative research critiques the discipline of epistemology: “Epistemology has been […] a negative discipline, mostly devoted to saying what you shouldn't do if you want your activity to merit the title of science, and to keeping unworthy pretenders from successfully appropriating it” [14:2] – a sort of a gatekeeper. We argue that when faced with balancing the tensions between post positivist epistemological “should” and the perspectival nature of interpretivist ethnography, epistemology does act a negative, because it portrays the decision as a black and white choice. One way forward would be then is to find alternative epistemologies that might facilitate a collaboration between design-oriented goals and interpretivist ethnographies. Khovanskaya et al. [54] for example show that by maintaining and even accentuating the gap between design and ethnography, generative outcomes can be produced. Their work implicitly highlights how epistemological underpinnings define our perspective and how we see research sites and it is in-fact generative rather than limiting to knowledge production. Their work exemplifies a situation where the reflexive methodology and ethnographic writings we propose in the present work might further advance design research.

The approach we propose thus further extends issues raised by [37,80,83,89] by advocating for explicit articulation of the ontological and epistemic background of interpretivist ethnographic practice. Certainly, there is room for other responses to this than our own, but we do believe that ethnographic accounts that include the doubts, hesitant assertions about the participants and member meanings, and yet offer evidence of people and practices associated with computing and interaction, can contribute towards scientific knowledge of and design-oriented interventions in the real world.