ETHNOGRAPHY AND THEORY
Benjamin Shestakofsky - University of Pennsylvania
Upon being asked to contribute a reflection on how theory plays a role in my ethnographic research, I began to ponder whether there is any aspect of my work—and the work of ethnographers more generally—in which theory does not play a role.
Ethnographers dwell in the time and space of others. Using our bodies as research instruments, we participate in and observe social action, inscribing our interpretations to the page. Some researchers caution that ethnographers must be careful to undertake such activities without allowing theoretical presuppositions to color their observations. In practice, however, ethnographers can never escape theory, which may take the form of either explicit or tacit models of activity in the social world. Other scholars’ research and our own personal experiences inform our interest in the topics we choose to study. Once ethnographers have entered a social setting, we find that the field is an infinite manifold. It is theory that focuses our attention on particular aspects of the activities we observe. As we begin to gather data, we cannot help but draw it into dialogue with existing concepts as we develop insights and repeatedly return to the field to test our assumptions. Analysis of data is informed by reflections on how one’s positionality shapes one’s observations. And our writing requires us to make decisions about whose voices will be included or suppressed as we weigh considerations of audience and genre.
I take the latter approach of theory reconstruction in my own research, which intervenes in debates surrounding the future of work (Shestakofsky 2017). Discontinuity theorists predict that advances in artificial intelligence are poised to render workers across the economy obsolete. Continuity theorists counter that humans will continue to hold a comparative advantage over computers in fulfilling some tasks, and that new forms of labor will emerge in and around digital infrastructures.
In spite of the recent influx of interest in the future of work, we still know surprisingly little about the conditions under which software systems function autonomously, and when they rely on the assistance of complementary human workers, in real-world settings. Most contemporary research into the future of work operates at a high level of abstraction divorced from the concrete social contexts in which software algorithms and workers interact. Ethnographic researchers are thus uniquely positioned to shed new light on the relationship between software automation and human labor.
I draw on 19 months of participant-observation research at a high-tech startup company I call AllDone, spanning work sites in San Francisco, the Philippines, and Las Vegas. AllDone aimed to transform local service markets in the U.S. by using technology to more efficiently connect buyers and sellers of services ranging from house cleaning to wedding photography to tutoring and beyond. Each phase of the firm’s development revealed mismatches between humans and machines generated by the company’s shifting strategic imperatives. To address these problems, the company relied on two forms of complementary labor performed by a distributed, online workforce. When computers alone were unable to complete an operation, workers located across the Philippines provided computational labor, performing routine information-processing tasks to support or stand in for software algorithms. Workers in the Las Vegas area performed emotional labor aimed at helping users adapt to changing software systems. Instead of perfecting software algorithms that would progressively push people out of the production process, managers continually reconfigured assemblages of technology and human helpers, developing new forms of organization with a dynamic relation to technology. My study provides support for continuity theory while also revealing its limitations: Existing approaches fails to account for what I call the “discontinuity in continuity,” or how the texture of continuity is itself dynamic and discontinuous, the result of repeated transformations in human-machine configurations.
Recently I have begun to undertake an extension of this project in which I examine how members of each of AllDone’s three work teams experienced the firm’s rapid organizational flux. I was thrilled to have the opportunity to share an early draft of this work at the 2018 Chicago Ethnography Incubator. On March 8th and 9th, organizers Kimberly Kay Hoang, Kristen Schilt, and Forrest Stuart welcomed four faculty fellows and seven graduate student fellows to the University of Chicago. The first day featured a panel discussion on ethnographic practice featuring Tianna Paschel (UC Berkeley), Laurence Ralph (Harvard), and Iddo Tavory (NYU). Mary Gray’s (Microsoft Research) arrival was delayed by a blizzard, but she was able to join the group for the second day’s workshop. Before the workshop, the graduate student fellows shared one-page overviews of their dissertations, as well as 10-page empirical excerpts of their work in progress. Each graduate student’s writing was allotted a 45-minute session that included detailed commentary from three of the faculty in attendance, followed by a discussion with the full group. The workshop ended with a panel discussion on publishing featuring representatives from the University of Chicago Press and Stanford University Press.
In reflecting on the Incubator, I am particularly impressed by the depth of participants’ engagement with social theory—not to mention their generosity in engaging with with each other’s work. Each of these projects deserve far more space than I have been allotted to describe them:
- Paul Michael Atienza (UIUC) studies gay and bisexual Filipinos in Manila and Los Angeles who use social media, dating apps, and mobile messaging platforms to establish intimate relationships. Atienza asks how digital technologies shape these individuals’ notions of space and time, emotional attachments, self-presentation, and concepts of difference.
- Annie Hikido (UCSB) analyzes the experiences of women in black townships in South Africa who have turned their modest homes into guesthouses to host tourists. Hikido shows how these women trade on notions of “cultural authenticity” to make ends meet, aligning themselves with state-sponsored images of progress while also revealing the limitations of national economic development projects.
- Dana Kornberg (University of Michigan) worked alongside informal garbage collectors in Delhi. Kornberg’s study contributes to economic sociology by providing an in-depth account of the social and institutional underpinnings of an “informal” market, including a blend of patrimonial practices and bureaucratic bookkeeping.
- Jeffrey Omari (UCSC) examines the relationship between digital technologies, state, and society in the favelas of Rio de Janerio. Mobilizing fieldwork conducted among community organizers, activists, and residents, Omari reveals how state-sponsored projects ostensibly designed to foster democracy and digital inclusion are received—and, at times, resisted—by citizens.
- Melissa Osborne (University of Chicago) compares the experiences of college students from historically underrepresented populations across four institutions. Drawing from interviews with and observation of 150 undergraduates, Osborne uncovers how variations in campus institutions and resources provided by philanthropic foundations impact student trajectories in higher education.
- Ande Reisman’s (University of Washington) research centers on women who remain behind in Nepal when their male partners find migrant work abroad. Reisman examines how women’s use of remittances reconfigures the freedoms and responsibilities that they confront both at home and in their communities.
Throughout the workshop, faculty fellows emphasized the delicate balance that ethnographers aim to strike in mobilizing theory without allowing one’s data to be subsumed by it. Many of our conversations touched on the importance of focusing first on inscribing one’s (always partial) observations of concrete social practices, forestalling the work of explicitly linking one’s observations to existing concepts that might be used to describe them (e.g. “patriarchy,” “equality”). Constructing novel categories to explain what is going on in a field site allows one to enter into dialogue with existing theoretical distinctions, and ultimately to draw out their limitations in relation to one’s observations. Such an approach enables ethnographers to mobilize observations of everyday interaction to call into question—or, as Laurence Ralph put it, to “trouble”—taken-for-granted analytic categories. For example, Kornberg’s work complicates distinctions between “formal” and “informal” markets, while Omari’s research forces us to ask what we mean when we use phrases like “inclusion” or “democracy.” To this I would add that existing theory may be backgrounded during the initial writing process, but ethnographers ignore it at their own peril. Those who put prior work aside until later in the research process risk recreating existing theory de novo and failing to extend or advance research in their area of study.
Another notable feature of the Ethnography Incubator was that it brought together sociologists and anthropologists under the same roof. As Tianna Paschel pointed out, sociologists tend to prefer to engage with solving theoretical puzzles, while anthropologists are more apt to dwell in the “messiness” of the social world. At this workshop, however, there was little friction to be found between members of the two disciplines. Although theory was not an explicit focus of the Incubator, it was all around us. And that is just as it should be at a meeting of ethnographers.
- Burawoy, Michael. 1998. “The Extended Case Method.” Sociological Theory 16(1):4-33.
- Shestakofsky, Benjamin. 2017. “Working Algorithms: Software Automation and the Future of Work.” Work and Occupations 44(4):376-423.
- Snow, David A., Calvin Morrill, and Leon Anderson. 2003. “Elaborating Analytic Ethnography: Linking Fieldwork and Theory.” Ethnography 4(2):181-200.