“Trump Shit Goes into Overdrive”: Tracing Trump on 4chan/pol/
Sal Hagen · M/C Journal · 2020
Content warning: although it was kept to a minimum, this text displays instances of (anti-Semitic) hate speech. During the 2016 U.S. election and its aftermath, multiple journalistic accounts reported on “alt-right trolls” emanating from anonymous online spaces like the imageboard 4chan (e.g. Abramson; Ellis). Having gained infamy for its nihilist trolling subcultures (Phillips, This Is Why) and the loose hacktivist movement Anonymous (Coleman), 4chan now drew headlines because of the alt-right’s “genuinely new” concoction of white supremacy, ironic Internet humour, and a lack of clear leadership (Hawley 50). The alt-right “anons”, as imageboard users call themselves, were said to primarily manifest on the “Politically Incorrect” subforum of 4chan: /pol/. Gradually, a sentiment arose in the titles of several news articles that the pro-Trump “alt-right trolls” had successfully won the metapolitical battle intertwined with the elections (Phillips, Oxygen 5). For instance, articles titled that “trolls” were “The Only True Winners of this Election” (Dewey) or even “Plotting a GOP Takeover” (Stuart).The headlines were as enticing as questionable. As trolling-expert Whitney Phillips headlined herself, the alt-right did not attain political gravity solely through its own efforts but rather was “Conjured Out of Pearl Clutching and Media Attention” (“The Alt-Right”), with news outlets being provoked to criticise, debunk, or sensationalise its trolling activities (Faris et al. 131; Phillips, “Oxygen” 5-6). Even with the right intentions, attempts at denouncement through using vague, structuralist notions–from “alt-right” and “trolls” to “the basket of deplorables” (Robertson) – arguably only strengthened the coherence of those it was meant to disavow (Phillips, Oxygen; Phillips et al.; Marantz). Phillips et al. therefore lamented such generalisations, arguing attributing Trump’s win to vague notions of “4chan”, “alt-right”, or “trolls” actually bestowed an “atemporal, almost godlike power” to what was actually an “ever-reactive anonymous online collective”. Therefore, they called to refrain from making claims about opaque spaces like 4chan without first “plotting the landscape” and “safeguarding the actual record”. Indeed, “when it comes to 4chan and Anonymous”, Phillips et al. warned, “nobody steps in the same river twice”.This text answers the call to map anonymous online groups by engaging with the complexity of testing the muddy waters of the ever-changing and dissimulative 4chan-current. It first argues how anti-structuralist research outlooks can answer to many of the pitfalls arising from this complex task. Afterwards, it traces the word trump as it was used on 4chan/pol/ to problematise some of the above-mentioned media narratives. How did anons consider Trump, and how did the /pol/-current change during the build-up of the 2016 U.S. elections and afterwards?On Researching Masked and Dissimulative ExtremistsWhile potentially playing into the self-imagination of malicious actors (Phillips et al.), the frequent appearance of overblown narratives on 4chan is unsurprising considering the peculiar affordances of imageboards. Imageboards are anonymous – no user account is required to post – and ephemeral – posts are deleted after a certain amount of activity, sometimes after days, sometimes after minutes (Bernstein et al.; Hagen). These affordances complicate studying collectives on imageboards, with the primary reasons being that 1) they prevent insights into user demographics, 2) they afford particularly dissimulative, playful discourse that can rarely be taken at face value (Auerbach; de Zeeuw and Tuters), and 3) the sheer volume of auto-deleted activity means one has to stay up-to-date with a rapid waterfall of subcultural ephemera. Additionally, the person stepping into the muddy waters of the chan-river also changes their gaze over time. For instance, Phillips bravely narrates how she once saw parts of the 4chan-stream as “fun” to only later realise the blatantly racist elements present from the start (“It Wasn’t Just”).To help render legible the changing currents of imageboard activity without relying on vague understandings of the “alt-right”, “trolls”, or “Anonymous”, anti-structuralist research outlooks form a possible answer. Around 1900, sociologists like Gabriel Tarde already argued to refrain from departing from structuralist notions of society and instead let social compositions arise through iterative tracing of minute imitations (11). As described in Bruno Latour’s Reassembling the Social, actor-network theory (ANT) revitalises the Tardean outlook by similarly criticising the notion of the “social” and “society” as distinct, sui-generis entities. Instead, ANT advocates tracing “flat” networks of agency made up of both human and non-human actors (165-72). By tracing actors and describing the emerging network of heterogeneous mediators and intermediaries (105), one can slowly but surely get a sense of collective life. ANT thus takes a page from ethnomethodology, which advocates a similar mapping of how participants of a group produce themselves as such (Garfinkel).For multiple reasons, anti-structuralist approaches like ANT can be useful in tracing elusive anonymous online groups and their changing compositions. First, instead of grasping collectives on imageboards from the outset through structuralist notions, as networked individuals, or as “amorphous and formless entities” (see e.g. Coleman 113-5), it only derives its composition after following where its actors lead. This can result in an empirical and literally objective mapping of their collectivity while refraining from mystifications and non-existent connections–so often present in popular narratives about “trolls” and the “alt-right”. At the same time, it allows prominent self-imaginations and mythologizations – or, in ANT-parlance, “localisations of the global” (Latour 173-190) – rise to the surface whenever they form important actors, which, as we will see, tends to happen on 4chan.Second, ANT offers a useful lens with which to consider how non-human actors can uphold a sense of collectivity within anonymous imageboards. This can include digital objects as part of the infrastructure–e.g. the automatically assigned post numbers having mythical value on 4chan (Beran, It Came From 69)–but also cultural objects like words or memes. Considering 4chan’s anonymity, this focus on objects instead of individuals is partly a necessity: one cannot know the exact amount and flow of users. Still, as this text seeks to show, non-human actors like words or memes can form suitable actors to map the changing collectivity of anonymous imageboard users in the absence of demographic insights.There are a few pitfalls worth noting when conducting ANT-informed research into extremist spaces like 4chan/pol/. The aforementioned ironic and dissimulative rhetoric of anonymous forum culture (de Zeeuw and Tuters) means tracing is complicated by implicit (yet omnipresent) intertextual references undecipherable to the untrained eye. Even worse, when misread or exaggerated, such tracing efforts can play into trolling tactics. This can in turn risk what Phillips calls “giving oxygen” to bigoted narratives by amplifying their presence (“Oxygen”). Since ANT does not prescribe what sort of description is needed (Latour 149), this exposure can be limited and/or critically engaged with by the researcher. Still, it is inevitable that research on extremist collectives adds at least some garbage to already polluted information ecologies (Phillips and Milner 2020), even when “just” letting the actors speak (Venturini). Indeed, this text will unfortunately also show hate speech terms below.These complications of irony and amplification can be somewhat mitigated by mixing ethnographic involvement with computational methods. Together, they can render implicit references explicit while also mapping broad patterns in imitation and preventing singular (misleading) actors from over-dominating the description. When done well, such descriptions do not only have to amplify but can also marginalise and trivialise. An accurate mapping can thereby counter sensationalist media narratives, as long as that is where the actors lead. It because of this potentiality that anti-structuralist tracing of extremist, dissimulative online groups should not be discarded outright.Stopping Momentarily to Test the WatersTo put the above into practice, what follows is a brief case study on the term trump on 4chan/pol/. Instead of following users, here the actor trump is taken an entry point for tracing various assemblages: not only referring to Donald J. Trump as an individual and his actions, but also to how /pol/-anons imagine themselves in relation to Trump. In this way, the actor trump is a fluid one: each of its iterations contains different boundaries and variants of its environment (de Laet and Mol 252). By following these environments, can we make sense of how the delirious 2016 U.S. election cycle played out on /pol/, a space described as the “skeleton key to the rise of Trump” (Beran, 4chan)?To trace trump, I use the 4plebs.com archive, containing almost all posts made on /pol/ between late-2013 and early 2018 (the time of research). I subsequently use two text mining methods to trace various connections between trump and other actors and use this to highlight specific posts. As Latour et al. note, computational methods allow “navigations” (593) of different data points to ensure diverse empirical perspectives, preventing both structuralist “zoomed-out” views and local contexts from over-dominating. Instead of moving between micro and macro views, such a navigation should therefore be understood as a “circulation” around the data, deploying various perspectives that each assemble the actors in a different way. In following this, the case study aims to demonstrate how, instead of a lengthy ethnogr