Svetlana Erpyleva, Public Sociology Laboratory
The essay explores the challenges of anonymizing ethnographic research in politically sensitive and repressive contexts, drawing on fieldwork in wartime Russia. It outlines a strategy for protecting research participants and locations while preserving the rich contextual detail essential to ethnographic analysis.
Introduction
In the summer of 2024, Public Sociology Laboratory published an adapted chapter from a forthcoming analytical report on the online platform Re-Russia (Roshal and Kappinen, 2024). The piece described how the Russo-Ukrainian war is experienced and talked about by residents of a small town in the Russian Urals, the town of Cheremushkin. A few days later, it was reprinted by one of the largest Russian “independent” outlets, Meduza, thus bringing even more readers to the piece. Within half a year, the article received a journalism prize for the best work in the “Analysis” category.
Part of the reaction to it was somewhat unexpected for us. Many people began trying to guess what the “real” Cheremushkin was—the name of the city was, of course, changed—and who the main character of the piece, whom we called “Tonya,” actually was. We were a little stressed and concerned: why were so many people putting effort into de-anonymizing locations and individuals from our research, as if it were some kind of amusing game with a big prize for the winner?
Nobody won, though—as far as we know. The “real” Cheremushkin and the “real” Tonya were never identified. This shows that we managed to successfully protect our research participants without sacrificing the analytical depth for which detailed contextual information was necessary. In this short essay, I will discuss the main principle of anonymization of ethnographic data that we relied on.
There is no need to discuss in detail why anonymization requires considerable thought and effort from researchers. It is one of the basic principles for ensuring the safety of research participants, which becomes especially crucial when studying politically sensitive topics in a repressive authoritarian context. At the same time, detailed descriptions of context are essential for ethnographic writing—providing generalized information about the entire “dataset,” as our quantitative colleagues often do, is not sufficient. How, then, can we comply with both of these requirements: protecting our interlocutors by not revealing any information that could allow them to be identified, while also providing “thick,” in-depth contextual descriptions of the field and our interactions within it? We tried to strike a balance by proceeding as follows.
Anonymization of Research Locations
First of all, which locations should be anonymized, and which can remain identifiable? In our ethnographic fieldwork, we did not anonymize locations that met two criteria: they were large cities, usually regional capitals, and the research there involved interactions with many people who were not connected to one another. In such cases, anonymizing individuals and the specific organizations or spaces where the research took place was sufficient; simply knowing the city did not make it possible to identify particular participants. This was the approach we followed, for example, in our research conducted in the regional capital of Kursk in the Kursk region, in Krasnodar, the capital of Krasnodar Krai, and in Ulan-Ude, the capital of the Republic of Buryatia.
The situation is quite different when data are collected in a small town with a dense community, where key members of the community know one another and the town itself has some distinctive characteristics. In such cases, revealing the town can immediately lead to the identification of central community members. For this reason, we anonymized such locations. In our fieldwork, these included, for example, the town of Cheremushkin in Sverdlovsk region, the settlement of Iglovka in the Kursk region, and the town of Udurgh in Buryatia.
To anonymize a town, it is not enough simply to invent a different name. Ethnographic writing relies heavily on context. The size of the town, its distance from the regional center, its economic life—key industries or enterprises, what it is known for—all of this can be crucial for the analysis. Yet at the same time, it can make the town identifiable even if its name has been changed.
For this reason, anonymization becomes not a technical task but an analytical one—it is inseparable from the process of analysis itself. The first step is to list all the key characteristics of the town or settlement that are important for understanding the context and for the analytical conclusions. Most of these characteristics should then be replaced with different but comparable ones, so that their analytical role remains the same.
For example, a town situated 50 kilometers north of the regional capital might become 60 kilometers south of it; transport connections consisting of three minibuses and one municipal bus per day might become two commuter trains and one municipal bus per day; a town-forming metallurgical plant might become a town-forming timber-processing enterprise, and so on.
When doing this work, it is important to remain attentive to the real political and economic geography of the region—that is, to make sure that even after these changes the town cannot be easily identified simply because all the other towns in the region look entirely different from the one being described.
For example, when we anonymized that same town of Cheremushkin, we first identified the key characteristics of the place that were important for our analysis. We then realized that roughly twenty towns in the region shared similar features. To protect anonymity, we blended together elements drawn from these different towns. As a result, any one of those twenty places could plausibly be the “real” Cheremushkin. In this sense, even with the help of AI one would not be able to identify the town—all twenty of them are equally plausible candidates.
We ended up having the following description of the town in our analytical report (Public Sociology Laboratory, 2024):
"The population of Cheremushkin is almost 12 thousand. A significant number of its residents are employed in the public sector and receive relatively modest salaries by Russian standards; the city is subsidized and directly depends on the distribution of regional and federal budgets. Working at a pulp mill is considered a relatively profitable employment option for Cheremushkin residents. After full-scale war broke out, according to the testimony of our researcher's interlocutors, business in this sector went into decline, since the products were mainly manufactured for export, which proved difficult under sanctions.
There are several penal colonies within an hour's drive from Cheremushkin. Some city residents either work there themselves or know many of the employees and prisoners personally. For this reason, residents of Cheremushkin know what is going on in the prisons, and the news about Wagner recruiting prisoners was common knowledge for many city residents.
According to Cheremushkin residents themselves, their city is relatively poor by regional standards. Many homes are not connected to a centralized water or gas supply. … Cheremushkin has the “classic” set of public spaces and organizations that you would expect to find in almost any Russian city of its size—a town square, cultural center, museums, churches, a few administrative buildings, schools and daycares. There are also several cafés, grocery and hardware stores, pharmacies and beauty salons.”
Anonymization of Individuals
The anonymization of people follows somewhat similar principles, though—except in the case of key informants and gatekeepers—it usually requires less effort.
First of all, there is no need to decide whom to anonymize and whom not to: everyone is anonymized.[1] Unlike locations, which always require detailed contextual description, many interlocutors in ethnographic writing appear only once or twice in the text. In such cases, it is usually sufficient to change their names and provide a few real social characteristics that would not allow them to be identified, even in a small town. For example, we typically indicate a person’s gender, age, and occupation—making sure to replace rare or highly specific occupations with something more general.
However, ethnographic texts almost always include a few “main characters”—people whom researchers return to again and again in the course of fieldwork. Conversations with them are particularly important for the analysis, which means we often quote them extensively and provide a range of biographical details that are analytically relevant. These people are often gatekeepers—individuals who play an important role in the community, are fully aware of the research, and help the researcher establish trust with other members of the community. In such cases, simply changing their names is not enough. If the rest of their biographical details remain unchanged, these details alone may make the person recognizable, especially in a small town.[2]
At this point, a principle similar to the anonymization of locations comes into play. We first assemble the set of biographical characteristics that define each of these key figures (there are usually only a few in any given text), and then replace most of them with similar but different attributes. The substituted characteristics should play the same social role—the one that is analytically significant—as the original ones. For example, a higher education degree in tourism obtained in Moscow may become a degree in management obtained in St. Petersburg; managing a well-known beauty salon in town may become managing a well-known restaurant, and so on.
This is, for example, how we described the gatekeeper “Tonya,” who played a key role in our ethnographic work in the town of Cheremushkin and who herself occupied an important position within the local community:
“Tonya is a young entrepreneur originally from Cheremushkin. In the city, she owns a beauty salon called ‘Style,’ a business which she has successfully managed for the past few years. In addition, she is the manager of the Ulybka Cafe.
Tonya is a model member of society with high social status and extensive connections. Various Cheremushkinites, from administrative workers and police officers to housewives and taxi drivers, all come to Ulybka for lunch or to celebrate birthdays, and sometimes just to chat and discuss local happenings. Since there aren't many jobs in the city, both the cafe and the beauty salon have an important social function. Their staff have a stable income, and many are able to master a craft from square one and gain experience that broadens their future life prospects. In addition, older city dwellers know Tonya's parents, who have a good reputation in the city. Due to all of the above, as well as Tonya’s personality—friendly, sensitive, sociable, simple and direct in her approach to a wide variety of people—Tonya enjoys universal respect throughout the city.
At the same time, Tonya has a ‘big city background.’ She received her degree in Moscow and has many friends in Ekaterinburg, St. Petersburg, and other big cities. Tonya keeps up-to-date on the top news stories and reads all the major independent media. In communication with others, Tonya generally does not hide her oppositionary and anti-war views, but at the same time does not impose them on anyone. She perfectly reads the unspoken conventions of communication, within which it is rather unacceptable to discuss ‘politics.’”
Conclusion
Anonymizing data in ethnographic research is not merely a “technical” task. Ethnographic writing depends on rich descriptions of context, yet such detail can also make places and people recognizable. To prevent identification while preserving the analytical value of contextual information, researchers cannot simply invent new names for individuals and research locations. Instead, they must identify the political, economic, and social characteristics of people and places that are important for the analysis and replace them with other, comparable features—ones that fulfil the same social role in the narrative. This is by no means an easy task. It is a highly creative process, but an essential one for responsible ethnographic writing.
About the author
Svetlana Erpyleva is a sociologist at the Public Sociology Laboratory. She received her PhD in Social Sciences from the University of Helsinki. Her past and present research focuses on collective action, political socialization, youth and children’s political participation, war perception in Russia, and various aspects of Russia’s wartime society.
Reading recommendations
Public Sociology Laboratory (2024). “We need to carry on:” Ethnography of Russian regions during wartime. Edited by S. Erpyleva and S. Kappinen. Analytical report. https://therussiaprogram.org/ps_lab_ethnography
Roshal, Ilya, and Sasha Kappinen. 2024. Parallel Cheremushkin: the absence and presence of ‘war’ in a provincial Russian city. Re-Russia, URL: https://re-russia.net/en/expertise/0165/
***
[1] In rare cases when some interlocutors do not want to be anonymized, our strategy has always been to explain that anonymization is a basic standard of academic writing in our discipline, and that their de-anonymization could potentially lead to the identification of other interlocutors from the same circle. In our experience, they have always agreed to be anonymized after this explanation.
[2] Of course, one could argue that the town has already been anonymized, and therefore biographical details would not reveal the person—after all, we do not know which town’s community is being described. However, in the interest of maximizing the safety of research participants, it is better to err on the side of caution: to anonymize individuals as if the locations themselves had not been anonymized.