Editor’s Note: Reddit. Facebook. YouTube. Twitter. These days it’s difficult to go anywhere online without encountering an anonymous troll (or ten). Debates about trolling, which is best described as deliberately antagonistic or otherwise provocative online speech and behavior, have even seeped into congressional hearings. In research conducted between 2008 and 2012, Whitney Phillips @wphillips49 –who received her PhD in English with a Digital Culture/Folklore emphasis from the University of Oregon– investigated the origins and subcultural contours of online trolling. Using a combination of cultural studies, new media studies and ethnography, Whitney postulated that there is more to trolls and trolling behaviors than detractors might initially think.
The subject of Whitney’s research leads us to ask, how does one conduct ethnographic research on an anonymous, and at times malicious, online population? In the first post of her three-part guest series, Whitney shares with us how she tackled the ethical pitfalls of her groundbreaking research. She also discusses how these pitfalls allowed her to make larger claims about trolling, with particular focus on the striking overlap between trolling and mainstream behaviors. We look forward to her next post on her participant observation expeditions with trolls.
Check out past posts from guest bloggers.
My name is Whitney Phillips, and I study trolls. Well, not just trolls. I’ve also written about meme culture, so-bad-it’s-good fan engagement (my essay on the kuso aesthetic in Troll 2 is forthcoming in Transformative Works and Cultures), and online shaming. But for the better part of five-ish years, my life has revolved around trolls and the trolls who troll them. The title of my dissertation—THIS IS WHY WE CAN’T HAVE NICE THINGS: The Origins, Evolution and Cultural Embeddedness of Online Trolling—pretty much says it all.
As I will discuss in this and several subsequent guest posts, my research experiences have been something of a mixed bag. Writing about trolls (to say nothing about working with trolls) has certainly been engaging, but has also proven to be the most consistently frustrating, challenging, and at times downright infuriating endeavor I have ever attempted. Which is one of the main reasons it has been so engaging, go figure.
Because in the end, it was the complications—the incomplete data sets, the trolls’ endless prevarications, the incessant march of subcultural change—that gave rise to my basic argument, the nutshell version of which can be found in my response to the Violentacrez controversy. As I argue, trolls are agents of cultural digestion; they scavenge and repurpose mainstream content, allowing one to extrapolate what’s going on in the dominant culture by examining what’s going on in the troll space. I could not have written my way into this argument if things had gone according to plan. I needed those roadblocks, even if at the time they made me want to rip out my hair.
The first and most initially intimidating of these roadblocks was the fact that my research subjects were anonymous. And not just anonymous (i.e. nameless), but in the case of trolls on 4chan’s /b/ board, Anonymous (i.e. the Internet Hate Machine, at least from the mid/late 2000s-around 2010).* Although this fact certainly sped up my IRB request (research conducted with anonymous subjects is designated exempt, so while I was required to submit a research protocol it was quickly and painlessly approved), it complicated everything else.
For one thing, and most obviously, I had no way of knowing who exactly I was dealing with. Who the trolls “really” were in real life—in terms of race, class, gender, sexual orientation, employment, level of education, and any other relevant information—was simply unverifiable. There is no trolling census; there is no stat box to click whenever you encounter some anon on 4chan. Sometimes the trolls would self-identify as this or that, but even these details were suspect. They were trolls, after all. Even during my research on Facebook, which was more directly participatory (on 4chan I was primarily an observer, while on Facebook I was able to embed myself within a group of RIP trolls), trolls would often joke about telling me nothing but lies in between telling me what they swore was the truth (Gabriella Coleman discusses a similar dynamic in her paper “The Ethnographers Cunning: The Return of the Arm Chair and Keyboard Anthropologist,” which she presented at the annual meeting of the American Anthropological Association in 2011). I can’t count the number of times a troll—even trolls I “knew” “well” from months and months of online interaction—would recount some involved personal trauma and then turn around and “lol jk” me. This was always much funnier to them than it was to me.
Trolls’ anonymity didn’t just hinder my ability to establish concrete demographics. It also made it almost impossible to reach out to potential research subjects, particularly while researching 4chan’s /b/ board, which is neither searchable nor archived (at least not on-site, though there are some third party archiving services like chanarchive that save backups of more popular threads). Questions of anonymity were thus compounded by questions of ephemerality. Not only would threads shuffle off the server coil within a half an hour, sometimes even faster, the identities of participants on each thread were, and could only be, a mystery. Though trolls on Facebook were somewhat easier to track down (I was sometimes able to get in touch with particular trolls with the help of existing collaborators), these trolls often operated behind a half dozen alt accounts, many of which were throwaway (i.e. created for one specific raid) and therefore not easily traceable.
One implication of having limited access to my research subjects was that questions of motivation were, and had to be, taken off the table. Not because I wasn’t interested in why trolls did what they did. Of course I was interested. But more often than not, I had no way of interacting with the troll(s) in question, and therefore no way of asking. Even when I could ask, there was no way to know if the trolls’ answers were genuine or merely part of the performance. Or maybe a little bit of both, as I suspect was usually the case.
Instead of trying to mine information that simply wasn’t available to me, I decided to focus on the information that was. My primary line of inquiry thus shifted from “what do trolls feel about what they do” to the much simpler “what do trolls do.” And what trolls do is engage in behaviors that are gendered male, raced as white, and marked by privilege. This demographic might not be literal, but it is symbolic—and more importantly, it is verifiable. Also verifiable are the ways in which trolls’ behaviors gesture towards, and in some cases directly parrot, ostensibly “normal” mainstream attitudes and behaviors. For example, trolls’ rhetorical and behavioral tactics—particularly in response to mass-mediated tragedy—echo precisely the sensationalism, spectacle, and emotional exploitation routinely deployed by corporate media outlets. Furthermore, their grotesque pantomime of masculine domination and white privilege call direct attention to remaining strongholds of institutionalized sexism and racism. This I could see, this I could confirm, and so this is what I chose to focus on — from which emerged my theory of cultural digestion, which is comparable to the process by which a scientist might infer an animal’s diet based on its –shall we say– “output” (just let the metaphor sink in; you’re welcome). In the process of grappling with what I couldn’t know, in other words, I stumbled upon a thesis.
Anonymity wasn’t the only solution wrapped in a roadblock wrapped in weird metaphors and a migraine. Even more difficult to navigate was the fact that the troll space kept changing. Not just in terms of the seemingly endless stream of memes either created or amplified by trolls, although those were slippery fishes unto themselves (for the basic and practical reason that by the time I’d written an explanation of a specific meme, the meme would have undergone further iterations). Additionally, and much more problematically, the subculture itself changed. Originally, I was writing about present-progressive behaviors: trolls on 4chan’s /b/ board and on Facebook were still doing the things I was discussing. By 2010, I found myself using the past tense more and more frequently (for an example, see my contribution to Henry Jenkins, Sam Ford and Joshua Green’s Spreadable Media). By mid-2011, past tense had become the default, particularly after Facebook began pushing back (and pushing back hard) against on-site trolling. At first I was reluctant to acknowledge this shift, because—well because I’d been working in the vein of present tense for so long. I couldn’t, and frankly wasn’t willing, to see past my original blueprint. But as much as I may have wanted to, I couldn’t ignore my own findings.
And so, as always, I improvised. The troll space was changing; trolling subculture was changing (but not disappearing, not exactly; trolling behaviors are as prevalent now as they were then. But a great deal of what emerged from the subcultural ooze in the early/mid 2000s, including a very specific understanding of the term “troll,” has since been folded into more mainstream internet culture, which is a discussion unto itself). Instead of fighting these changes, I decided to place them front and center. Though it had started as an examination of emergent online behavior, the dissertation had grown into a study of the subcultural lifecycle—hence the subtitle “The Origins, Evolution and Cultural Embeddedness of Online Trolling.”
However frustrating this decision might have been in the moment (again, sometimes a gif is worth a thousand words), this was the best thing that could have happened to me and to my project. For one thing, it forced me to consider the academic shelf life of my arguments. I had already encountered some resistance to my project on the grounds that it was a “novelty” topic; I needed to make sure the dissertation remained readable and relevant even in the face of profound subcultural change. For another, the painful realization that I was dealing with a moving target forced me to think carefully about the relationship between trolling subculture and the dominant culture, particularly in the United States. This lead me back to my theory of cultural digestion. If this theory were true, if trolls really were cultural scavengers, it would stand to reason that as mainstream culture changed, so too would trolling subculture–a theory born out by the changes I was (nervously) watching unfold, and which I decided to address in my seventh chapter, titled “The Lulz are Dead, Long Live the Lulz: From Subculture to Mainstream.” The connection between trolls and the culture that spawned them was already there, it’s just that I hadn’t yet realized what questions to ask, and therefore which dots to connect. I needed a catalyst, and it came in the form of a methodological crisis.
Needless to say, this wasn’t a pretty process. (A thousand blessings upon all those who tolerated my…let’s say fluctuating moods.) But by dismantling my project and rearranging all the pieces, I was able to build an infinitely stronger argument, one that ended up being as much a cultural postmortem as it was a subcultural case study. I could not have done this if I had not had ALL OF THE RUGS continuously pulled out from under me, and for that I am very grateful (if still somewhat bruised).
I’ll be expanding on that theme in my next post, which will focus on my role as participant observer within the troll space. SPOILER ALERT: it was weird. Till then, then…
*In recent years –I’d place the cutoff around late 2010– the already-nebulous mass noun “Anonymous” has become even trickier to characterize. What once was nearly synonymous with chaos and trollish disruption has taken on increasingly activist connotations, particularly post-Occupy. For more on this (still unfolding) transition, see Gabriella Coleman’s outstanding work on the political arm(s) of Anonymous.
(all gifs courtesy of PhD Stress, a very funny single-topic Tumblr)