It has been standard practice in anthropology to change the names of the people and places we analyze, but recently scholars have been questioning the necessity and even possibility of keeping participants anonymous, especially when they already have a social media presence. In this post, I share what I did to anonymize my research site and participants, and I do my best to start a discussion about the broader issue of anonymization now that detective work can be as simple as plugging a few search terms into Google.
When anthropologist Cathy Small enrolled as an undergraduate in her own university ten years ago to do the fieldwork that resulted in My Freshman Year: What a Professor Learned by Becoming a Student (2005), she knew that she wanted to protect the identities of her participants and institution by referring to them using pseudonyms. She called herself “Rebekah Nathan” (an excellent choice of pseudonym if you ask me) and Northern Arizona University “AnyU” (a play on its initials, NAU).
Small explained that she considered anonymity a standard anthropological practice and that she underestimated the interest her hidden identity would generate. Jacob Gershman, a reporter for the New York Sun, drew from the rich details she included in the book to unmask her and her university before it had even been published. Gershman wrote that, “whether by choice or accident, [she] also planted in her ethnographic study many clues about her identity[:] her university is located near Las Vegas, is surrounded by mountains, and has a hotel and restaurant management school.” The risk of my readers playing detective stuck in my mind as I started working through my own ethical commitments to my participants.
The organization where I did most of my fieldwork — a regional women’s weaving cooperative in Quetzaltenango, Guatemala that exports textiles to the U.S. and Europe — has a strong online presence: in addition to a website, it has accounts on Facebook, Twitter, Instagram, and Pinterest. It has also hosted a large number of tourists, who have recorded their experiences in travel blogs, social media posts, and comments on travel review sites. These materials are typically available for public consumption (with the exception of social media sites that may permit varying degrees of access to their contents, like Facebook, which is an ethical discussion for another day) and present researchers with a tantalizingly rich source of self-presented commentary. However, this easy accessibility also makes it difficult to maintain pseudonyms for the organization and its leaders.
Why was it important for the cooperative federation to have its identity disguised? The board of directors had concerns on a few levels, the most prominent being concern about potential retaliation from members of the military for speaking out about the recent genocidal civil war in Guatemala (1960–1996), which is still highly contested. While the organization’s members spoke freely about their experiences in the war to visiting tourists, they wanted to control the audience and context of their testimonies.
The organization already makes its own decisions about how much information to share and withhold online, in negotiation with its international volunteers. US volunteer “Rosalyn” wanted to share some members’ personal stories of their suffering during the war on the website: “When I came in for a talk they were quite open about their own experiences. The one woman was telling us about how they burned her family alive in their house.” She interviewed the president of the cooperative, who spoke at length about the war and then said that she did not want that material to go on the Internet. Rosalyn discussed a couple of possibilities with the president, including focusing on the parts about her childhood and leaving the war out of the finished interview, or publishing the interview anonymously, without her name or town, and she opted for the former. In talking to tourists and allowing the volunteers to post certain stories and images, the organization is constantly weighing how much risk they are willing to take on, so I viewed my responsibility as mitigating any possible impacts of the stories I shared and using basic techniques to make it so that a casual reader would be less likely to be able to associate the name of the organization with my work.
Practically speaking, is there anything we can do to take advantage of the wealth of data available on blogs and websites without compromising our commitment to our participants? Asking a colleague about this issue, I realized that we had independently come to the same solution to incorporate some blog material without inviting the revelation of institutions’ actual names. We presented a few of the most enticing quotes from blogs, the ones we simply couldn’t let go, as interview material, to discourage readers from plugging the quotes into search engines and finding the original blogs. As a researcher, this is a somewhat unsatisfying compromise, because statements made in a blog, intended to be broadcast to an audience, have a different quality than statements solicited in an interview, and should be analyzed differently. However, our guiding principle has to be to restrict harm above all else.
At other times, I paraphrased what someone had said or deliberately cut out words to make it more difficult to search for the quote online. I also tried to use non-online sources such as flyers rather than web sources when talking about how the cooperative represents itself, and deliberately kept the web sources I used out of the final bibliography.
Once, when I was reading a dissertation about Quetzaltenango, I thought I recognized the organization where I was doing fieldwork, referred to with a pseudonym. The anthropologist cited a substantial amount of online material, and checking the bibliography revealed the organization’s actual name associated with its website! It seems likely that this researcher assigned pseudonyms as a matter of standard anthropological practice, counting on the relative obscurity of his research and site to limit the risk of discovery to the organization. This is the kind of pitfall I hope to avoid by using the strategies described above (and any other approaches you can suggest in the comments).
Scholars have pointed out that many groups and individuals may actually want to be identified by name, to gain recognition or aid in their struggles. However, we don’t always get to take the more straightforward approach of naming names. My goal here is to start a conversation about how to handle situations where our participants have explicitly asked to have their identities masked for reasons of reputation management, legal protection, or personal safety.
5 thoughts on “Pseudonyms 2.0: How Can We Hide Participants’ Identities When They’re on Pinterest?”
Before we get to how to conceal identities, we should, perhaps, first ask those most directly concerned about their preferences. In my case, the advertising creatives I study are eager to have their names associated with their work. As a matter of professional courtesy I show them what I have written and ask for feedback before publication.
This is a pressing issue, and it was interesting to read about your obfuscation strategies. However, I’m not sure how safe or effective they are. For one thing, I suppose it depends on amount the attention the study gets (cf. the Alice Goffman case).
Another thing to consider, to use an analogy from cryptography, is how future-proof these methods are. Of course we won’t know the kinds of tools available in the future – who would’ve thought 10 or 20 years ago that Google’s Search by Image or Street view are available for free, for anyone, today? Still, it seems safe to assume that simply decontextualizing your quotes is not sufficient defense against a script that goes through the publication and tries to find text or image matches in Google’s databases, for example.
I’m currently writing a PhD dissertation and grappling with these kinds of questions too.
I completely agree that the measures you take will depend on the level of scrutiny your study is expected to go through (which may be difficult to anticipate) as well as the anticipated level of risk to your participants should their identities be revealed. For instance, had my participants been adamant about completely preventing material about their experiences in the war from going on the Internet, I would have made different decisions about including text drawn directly from the Internet that could ultimately be used to trace their statements back to them. There was material that I chose to completely omit from my dissertation on the basis that it would have been harmful to include it, and it may be that those considerations will expand as technology develops. When using blog material I did make what I thought of as “silent changes” or synonymous substitutions to some of the words in addition to decontextualizing them, but as you note, text algorithms are powerful and it’s difficult to anticipate how much more powerful search technology will become in a relatively short period. Do you have any thoughts on potential solutions, aside from leaving material out of the dissertation, which can be quite painful?
Of course the first step is to consult with your participants and figure out if they want to be directly identified, which makes everything easier and more straightforward. My goal with this post is to look at situations in which you’ve consulted with your participants and they have reasons for not wanting to be identified by name… how do you proceed with your research then? In my research site, they had concerns about reputation management, which I didn’t go into in the post, as well as the ubiquitous anxiousness in Guatemala about the potential for violence to erupt again at any moment. I agree that the most important thing here is for everything to be done in consultation with your participants.
Thanks for clarifying your situation. I am reminded of Marcus and Fischer’s remark in Anthropology as Cultural Critique that,
“Anthropology no longer operates under the ideal of discovering new worlds like explorers of the fifteenth century. Rather we step into a stream of already existing representations produced by journalists, prior anthropologists, historians, creative writers, and of course the subjects of study themselves.”
Now this situation is further complicated by the increasing likelihood that both our collaborators and those with other interests in monitoring their behavior have access to social and other digital media.
Comments are closed.