Into the wild online: Learning from Internet trolls
First Monday

Into the wild online: Learning from Internet trolls by Andreas Birkbak

An Internet troll is a person who deliberately upsets users of online forums or social media. The term has been taken up widely in media discourses about democracy and the Web. Internet trolls and the act of ‘trolling’ thus speaks to a renewed significance of monsters and the monstrous in modernizing liberal democracies. Following ideas developed in science and technology studies (STS), monsters have the generative political capacity to teach us about heterogeneity and hybridity. Indeed, trolls have historically been understood not just as dangerous, but also as invitations to try to come to terms with ‘the other’, which cannot be ignored. In this paper, I explore this potential in relation to online trolling. I start by examining the rise of the troll metaphor in relation to online discourse and observe a shift towards an increasingly broad usage. I argue that Internet trolls are no longer only understood as acting for the sake of controversy itself. Today, the designation is also used to demarcate the boundaries of proper debate, i.e., by expanding the label of trolling to include things like information warfare, hate speech, and sometimes even political activism. Using the troll figure in this way invokes and reproduces ideals about deliberative democracy, where an ongoing public debate that meets certain standards of rationality and inclusiveness is understood as central to democratic societies. However, trolls per definition defy such terms, which means that their subversive political potential as monsters is contained rather than exploited in this frame. With the help of Belgian philosopher of science Isabelle Stengers’ use of the figure of the idiot, I suggest that to enter into a more interesting relationship with online trolls, we may have to open for the possibility that ‘there is something more important’, which is not articulated ‘seriously’, but is nevertheless crucial for the sort of issue-oriented take on democratic politics currently being developed in STS. More specifically, online trolling may be politically generative in the sense that trolls challenge the dichotomy between serious ‘public’ issues and ‘private’ jesting.


Introduction: The rise of the Internet troll
Living with trolls
Online trolling: A definition and some ambiguity
Fighting new trolls with old methods
Trolls and the depths of civil society
Monsters as productive others in STS
The problem of relevance in democratic politics
Discussion: The problematic relevance of trolls



Introduction: The rise of the Internet troll

The Internet is an extremely diverse phenomenon, but recently, one particular figure has taken a strong hold on our imagination of what might be found ‘out there’ on the Web. That figure is the troll. Since 2011, ‘troll’ has been a more popular term than ‘spam’. Google Trends, which lets users compare the relative popularity of Google search terms over time, at least suggest so:


Two popular Internet metaphors
Figure 1: Two popular Internet metaphors. Google Trends compares the relative volume of searches for ‘troll’ and ‘spam’ over the past decade. Spam is in decline from index 100 back in 2004, while ‘troll’ has risen, peaking in 2012.


Of course, a Google search for ‘troll’ does not automatically indicate an interest in online trolling. Accordingly, there could be multiple reasons for why ‘troll’ was able to dethrone ‘spam’ as the more searched-for word. One such reason could be that Peter Jackson’s blockbuster movie trilogy The Hobbit originally appeared in 2012, featuring famous trolls from the 1937 book by J.R.R. Tolkien. That said, the indication that search interest in trolls surpassed that of spam is suggestive. As is well known, spam is widely used shorthand for unwanted online advertising or bulk messaging. What started as a reference to a Monty Python sketch (according to the Oxford English Dictionary) has become institutionalized as a stable part of user interfaces of services such as Google’s Gmail, which now routinously offers a ‘spam folder’ for unsolicited e-mail messages.

Google Trends charts like Figure 1 can be useful to ‘think with’ (Birkbak, et al., 2015). In this case, the chart invites us to think about changes in Internet imaginaries. The comparison suggests that it might be time to think of spam as a somewhat outdated metaphor. It would seem that notions of trolls and trolling have taken over the crown when it comes to which metaphor best captures what is understood to be wrong with the Internet. The notions of Internet trolls and online ‘trolling’ have today become central to discussions about Web culture and online debate. While these notions are far from new, at least not in ‘Internet years’, trolls has moved away from forums like 4chan, into the center of popular notions about the Internet. Season 20 of the popular cartoon satire show South Park, which opened in September 2016, had online trolling as a core theme. The 18 August 2016 issue of Time featured a troll on the cover, with a related story entitled “How trolls are ruining the Internet” (Stein, 2016).


Time troll cover for 18 August 2016
Figure 2: Time troll cover for 18 August 2016.




Living with trolls

The use of a troll by Time and others draws on how trolls are not only understood to be abundant online. They are everywhere in fantasy novels, Hollywood movies, and before that, Scandinavian folklore and Norse mythology. This raises the question of what happens, more exactly, when we draw on these well-known monsters to understand Web culture, and specifically online debate and commentary? The question is worth spending time with, I suggest, because it might hold the key to imagining ways of coming to terms with online trolls in more productive ways in the future.

Answering this question requires a closer examination of how trolls have been imagined before the Internet, and an investigation of how the notion has been taken up in relation to online interaction more precisely. This matters because across the many appearances of trolls in popular fiction and folk tales there is no clear definition of a troll (Lindow, 2014; Arnold, 2005). The only thing that seems to be widely accepted is that trolls are mythical, monstrous creatures. The multiple meanings of trolls and trolling makes it worth while to pay attention to the choices we make when it comes to mobilizing trolls as a means to understand and discuss democracy in a digital age. Our use of the monster metaphor is related to the question of what kind of politics we want and expect to take place on (or rather with) the Internet.

I am going to suggest that today the notion of online trolling is used to reaffirm particular liberal narratives about civil society in Euro-American democracies. I will also propose, however, that the figure of the troll has more to offer. More specifically, the radical otherness of trolls may be used as a resource for an alternative kind of democratic politics that does not assume that a rational consensus is achievable. In making this argument, I draw on work in science and technology studies (STS) on monsters and the monstrous, and on recent thinking about the need to ‘slow down reasoning’ and stick to problematic issues as foundational of politics.

First, however, I will briefly define online trolling and at the same time observe that the meaning of the term seems to be shifting at the moment. I will then take this ambiguity and bring it with me into the discussion of how monsters have come to be seen as a resource for thinking in STS.



Online trolling: A definition and some ambiguity

The situation today suggests there are at least two understandings of online trolls in circulation. In the classic definition, trolling is the act of deliberately luring participants in online forums to go off topic in order to frustrate them, catch them unaware and stir up controversy:

“A troller is a[n Internet] user who constructs the identity of sincerely wishing to be part of the group in question, including professing, or conveying pseudo-sincere intentions, but whose real intention(s) is/are to cause disruption and/or to trigger or exacerbate conflict for the purposes of their own amusement.” [1]

In this understanding, a key feature of Internet trolling is to pretend to be sincere in order to more effectively disrupt online interactions. ‘Rick-rolling’ is probably one of the most iconic examples of trolling in this sense (Mantilla, 2015). What happens when you get rick-rolled is that you click on a link because another user (the troll) has given you reason to believe it will take you to something that is relevant to you, but instead of relevant content, you find yourself watching a YouTube video of the musician Rick Astley (hence the name ‘Rick-rolling’) performing his song Never gonna give you up.

Whether you like the music of Rick Astley or not, the trolling part of this experience lies in you having been lured off topic while believing the good intentions of the person who placed the link. Another technique is to write divisive comments and thus encourage heated argumentation. Here, the sport is to pretend to be a genuinely upset participant in the discussion, while only caring about achieving some sort of disruption. The reward would then be to make other people upset enough to escalate a conflict among forum users that might otherwise be sympathetic to each other.

Yet, in another more recent understanding statements may be considered trolling simply because they are straightforwardly angry and offensive. For example, Mantilla (2015) researches the very relevant and frightening phenomenon that misogyny is widespread in online commentary. She calls this a kind of trolling even though such comments are rarely about leading people off topic. To the contrary, hostile comments about women will often appear in relation to news articles where questions about gender are relevant, and those who post such comments are not just there to be amused.

What is going on in this second understanding of trolling is that the troll and the victim of trolling are no longer distanced from each other in terms of earnestness. Some people writing comments online are convinced that misogyny is a relevant position for them. The result of such comments can often be hard to distinguish from the results of ‘classic’ trolling: frustration, anger and heated arguments. But in the second definition, trolls can be equally frustrated. They do not remain distanced and in control, as in the first understanding, where trolls deliberately set up a trap or a scam.

The slide in the meaning of online trolling coincides with changing views on the origins of the term. Herring, et al. [2] suggested that ‘trolling’ is derived from the fishing technique of the same name, since the ambition to ‘lure’ people off topic can be compared to how a fishing boat drags a baited line behind it to see if any fish will bite. However, in its newer editions, the Oxford English Dictionary (2013) suggests that trolling in “the computing senses ... are probably influenced by trolls” in the folklore sense of “an ugly cave-dwelling creature depicted as either a giant or a dwarf.”

With the first and narrower definition of online trolling that focus on the deliberate attempt to lure people off topic, the metaphor of a fisher drawing a shiny object through water in order to tempt fish to bite is quite apt. But as the meaning of Internet troll slides away from amusement as the main object towards more straightforward versions of maliciousness or, more recently, ideologically motivated information warfare, the mythological figure of the troll seems to have become more relevant. Here, the online troll is no longer making a joke, which deprives the troll of a claim to some kind of innocence. In the second and broader understanding of online trolling, the trolls have bad intentions “all the way down” — and the mythological figure of the troll has become relevant.



Fighting new trolls with old methods

What happens when we understand and act on online debate with the mythological trolls as guides? Such comparisons are not just language games. Lakoff and Johnson (1980) famously pointed out that “the metaphors we live by” have implications for what actions are available to us. What sort of actions and what kind of politics follow along with the metaphor of trolls?

In Scandinavian folk tales collected during the nineteenth century — in an approach similar to that of the fairytale collection by Brothers Grimm — a troll is a “large, long-nosed, hairy and wild creature living in mountains and caves” [3]. The picture is familiar, but Jakobsson also argues that the meanings of troll found in earlier texts, notably the Icelandic sagas, are somewhat different [4]. Here, ‘troll’ was used as a “description for some worrying or abnormal characteristic of a human, and as a form of oath.” [5] This broader and more figurative sense of the word seems to be alive today through words such as the Danish “trolddom”, which roughly means magic or sorcery. As such, we have at least two quite different meanings of the word troll with us today: An early mythological meaning, where ‘troll’ is a magical quality that several entities can possess, and a later folklore meaning, where trolls are a group of monstrous creatures with specific physical attributes.

For now, I am going to focus on the folklore version of trolls. This also means leaving aside that fact that trolls can today be considered small and cute. Arnold (2005) distinguishes between large Icelandic trolls and smaller, ‘modern’ trolls in Denmark, where troll (trold) can indeed be used as a nickname for children. But the large, ugly trolls of Nordic folk tales are probably the most widespread version of trolls. In the stories by Tolkien (1937), for instance, trolls are large and dangerous, but stupid, and some live in mountain caves.

One of the features of the trolls that have been popularized through Tolkien and others is that trolls burst or turn to stone when exposed to sunlight (Kvideland and Sehmsdorf, 1991). This feature of trolls has been taken up in current media debates about online trolling. In Norway, a Google search for “when the Internet trolls burst” (“når nettrollene sprekker”) returns 1,820 results at the time of writing, including results from major news outlets.

Most of the top search results refer to efforts to track down and expose people that post hateful comments online. For example, one news article describes a (Swedish) TV show called Trolljegerne (The Troll Hunters), where a journalists “seeks out and confronts the Internet haters” (Ringheim, 2014). Such shows play on the idea that trolls will burst when they are exposed. In the case of Internet trolls, the show aims to reduce them to their much less imposing ‘off-line’ lives. The Swedish show has a Norwegian version too, even though some commentators argue that offering Internet trolls air time is not a good tactic, because they seek and thrive in the media spotlight (Sved, 2014).



Trolls and the depths of civil society

What is worth noticing about these shows is how a particular set of questions are raised about the relationship between a small group of individuals that can be tracked down and a general public debate that can suffer from their presence and decide to deal with trolls in various ways. The idea that Internet trolls can be found and exposed positions them as something that must be actively fended off and protected against, but also as something that can never be avoided completely since trolls are assumed to be naturally occurring somewhere ‘out there’. The mythological figure of the troll here contributes to a naturalization of Internet trolls as a threat that we simply have to find ways to live with. Such a way of understanding online trolling may be furnished by liberal and libertarian political philosophy:

“These practices [of trolling], while clearly problematic, are nonetheless widespread and often tolerated, due in part to the pervasiveness on the Internet of civil libertarian values that consider abusive speech a manifestation of individual freedom of expression.” [6]

Following a liberal understanding, the Internet is home to some kind of free-roaming civil sphere. ‘Abusive speech’ can in some instances be considered a sign that this civil sphere is healthy, since it suggests that people feel free to say whatever they want. The notion of online trolls as simply existing somewhere ‘out there’ due to the wild nature of the Internet fits this liberal narrative. Online trolls can then be understood as one of the necessary evils of individual freedom. The naturalization of Internet trolls becomes associated with a general naturalization of a civic or public sphere as some sort of natural and chaotic well-spring of freedom and independent thinking in modern liberal democracies [7].

Trolls fit into this scheme given that, according to Scandinavian folklore, they are “nature beings” [8]. Extending his knowledge of the troll figure in myths and folk tales to the recent surge of Internet trolling, Lindow observes how the Internet troll does a well-known job of keeping up social order by appearing as external to society:

“The original Internet troll(er)s behaved in a way that anyone familiar with Scandinavian folklore would instantly recognize: they threatened to punish improper behavior and thus upheld social norms, in this case the norms of a growing and quickly evolving medium. I am tempted to tie the arena in which the late twentieth-century electronic trolls operated with the rapidly changing late nineteenth and early twentieth-century rural society of Scandinavia, which gave us the classical trolls of fairy tale, legend and folk belief. Perhaps the faster things change, the more the norms are enforced not by the community but by less visible forces from outside it.” [9]

Lindow speaks of the norms of a new digital medium, the Internet, but this medium is no longer considered to be separate from the world around it, especially not politically (Rogers, 2013). Indeed, the fear of Internet trolls is today connected to a more general fear for the health of public debate in liberal democracies. For example, a column recently claimed that the rise of Donald Trump as a presidential candidate in the U.S. “shows how Internet trolls have ruined political debate” (Applebaum, 2015).

Here, the troll figure as one that upholds societal norms spreads beyond Internet communities and their ‘netiquette’ to general concerns with the health of civil society in a liberal democracy. As the ‘other’ of proper public debate, the Internet troll confirms the need for norms that regulate such debate to be rational and inclusive (Habermas, 1989). At the same time, Internet trolls naturalize civil society by feeding imaginations about its ‘true depth’ as a wild and free sphere. The idea of a naturally occurring civil force that underlies liberal ideas about public participation in democracy is combined with a reinforcement of the need to tame this civil force into a proper public sphere:

“The relative anonymity of the Internet releases some of the inhibitions of a civil society, resulting in flaming, harassment and hate speech online.” [10]

What follows from these liberal political narratives about civil society is that online trolls are handled through a dual tactic of naturalization and stigmatization. On the one hand, it is only natural and thus unavoidable that a free civil society will include a few monstrosities. This is a small price to pay for ‘the freedoms that we all enjoy’, according to the liberal narrative, including, perhaps most importantly for the discussion here, free speech. On the other hand, online trolls are stigmatized as exhibiting the opposite of what public debate in a free civil sphere should look like. There is no need to take seriously what these monsters write and do, because they exemplify that which it is important to be able to tolerate and avoid in a free liberal society.



Monsters as productive others in STS

In contrast to the liberal political narrative, which seems to dominate popular discourse today, researchers in the transdisciplinary field of science and technology studies (STS), especially in relation to feminist studies of technoscience and actor-network theory (ANT), have developed a positive account of the political role that can be played by monsters. In fact, it has long been claimed that monsters are of special interest. In 1991, Donna Haraway wrote of the significance of monsters:

“Inhabiting my writing are peculiar boundary creatures — simians, cyborgs, and women — all of which have had a destabilizing place in Western evolutionary, technological, and biological narratives. These boundary creatures are, literally, monsters, a word that shares more than its root with the verb to demonstrate. Monsters signify. We need to interrogate the multifaceted biopolitical, biotechnological, and feminist theoretical stories of situated knowledges by and about these promising and noninnocent monsters. The power-differentiated and highly contested modes of being of monsters may be signs of possible worlds — and they are surely signs of worlds for which ‘we’ are responsible.” [11]

Here, Haraway provides an optimistic view of the political potentials of monsters and the monstrous. We may draw on monsters, she suggests, in order to demonstrate overlaps, mixtures and entanglements where none were thought to exist. Indeed, this is a key ambition in the ANT-oriented strand of STS, as Law (1991) notes. Here, Law repeats the ANT argument that any ordering is the result of a carefully drawn together heterogeneous arrangement. Once we realize this, Law suggests, we are all monsters, all heterogeneous and without category.

Law’s further point is that what STS has been less insightful about are those monsters that do not maintain stabilizations that allow them to go unnoticed. In other words, even when STS in principle appreciates all beings as monsters, there has been a tendency to ignore those that are marginalized as monsters:

“Something like this seems to happen: first the dispossessed have no voice at all. Then, when they start to create a voice, they are derided. Then (I am not sure of the order), they are told that they are wrong, or they are told that this was something that everyone knew all along. Then they are told that they are a danger. Then finally, in a very partial form, it may be that their voices are heard and taken seriously. And it has been a struggle all the way.” [12]

Such claims about monsters, their marginalization and political potential may hold resources to think differently about Internet trolls. What, if anything, are the simultaneously stigmatized and naturalized Internet trolls “promising” of, to paraphrase Haraway? At first sight the question seems a hard nut to crack. What could possibly be promising about Never gonna give you up played over and over again on YouTube, not to mention manipulation and hate speech?

Haraway suggests that it is not enough to simply try and appreciate the “otherness” of monsters, it must also be appreciated how these others/monsters are not cohesive. One of Haraway’s sources of inspiration is the coyote in Navaho mythology:

“[C]oyote is not a very nice figure. It is a trickster figure, and, particularly in Navaho figurations, the coyote is often associated with quite distressing kinds of trickster work. Coyote is about the world as a place that is active in terms that are not particularly under human control, but it is not about the human, on the one side, and the natural, on the other. There is a communication between what we would call ‘nature’ and ‘culture’, but in a world where ‘coyote’ is a relevant category, ‘nature’ and ‘culture’ are not the relevant categories. Coyote disturbs nature/culture ontologies.” [13]

Similarly to Haraway’s coyote, the Internet troll is not a very nice figure. This much has already been established. In fact the notion of a trickster may be quite apt for thinking about what an Internet troll does. This also means, however, that Haraway’s argument about monsters and tricksters, as challenging of categories we tend to take for granted, may apply to online trolls as well. Existing research on trolls in mythology and folklore, it turns out, is not far removed from Haraway’s argument about the abilities of monsters to undermine taken-for-granted dichotomies. As Lindow notes:

“Trolls who choose to be trolls, who choose to cut themselves off from normal human society, raise basic questions about what it means to live in society, and what society means. Trolls have always raised this question . . . .” [14]

In this passage, after a tour de force of how trolls have appeared and disappeared through the centuries, Lindow discusses the recent tendency in the U.S. to designate homeless people as trolls. Without endorsing this tendency, he observes that one thing homeless perhaps have in common with trolls is that “many more people have opinions about homeless people than have actually had anything other than very casual encounters with them” [15]. The same is true for trolls in Scandinavian folklore, Lindow claims. Trolls were abundant, especially as an unseen but widely imagined threat. As such, the use of the notion of trolls may say more about people using it than about trolls themselves. Realizing this might be one way in which trolls can help us raise questions about the societies in which we live. The marginalization of certain creatures may say more about the community that marginalizes than about those being marginalized.



The problem of relevance in democratic politics

What dichotomies could online trolling help us see and handle differently? Recent work in the ANT-inspired line of STS research suggests that we may have to think differently about democracy. My proposition is that there might be a more interesting role for trolls to play here. The central idea is that public participation in politics does not turn around a civil sphere of rational deliberation, but rather around specific issues that engage people by implicating them in ways that are beyond their control (Marres, 2012; 2007).

Such issues cannot be handled by public debate and parliamentary politics, because issues, as the word is used here, are exactly characterized by breaking with the boundaries of existing categories and institutions. In fact, this is what makes them engaging for people and thus relevant for democratic politics in so far as democracy is about public participation in politics. Marres (2012) describes this sort of public engagement as a sparked by the ‘ontological trouble’ that arises when one’s habitual way of life is disrupted by consequences of actions taken elsewhere.

This understanding of publics as those implicated by indirect consequences of actions is inspired by the pragmatist political philosophy of John Dewey (1927). However, while Dewey never really gave up the idea that some kind of social community should be equipped to deal with issues through an ongoing process of collective inquiry, Marres (2012) emphasizes how indirect consequences of actions produce a situation marked by a ‘problem of relevance’ in relation to who and what makes up this community. It is not just a question of identifying ‘those implicated’ by an issue, because who and what those entities are depends on how one is implicated. This is the sort of ontological trouble that removes deliberation from the center of democratic life, since public debate assumes that some kind of public space can be delineated and a set of legitimate speakers identified.

The alternative to deliberative democracy does not have to be a reduction of democratic politics to simple power games. Drawing on pragmatist philosophy and work in STS, Stengers (2005) has proposed the notion of ‘cosmopolitics’ as an alternative to the Habermasian idea of rational argument in a public sphere. The challenge of cosmopolitics, following Stengers, is to “slow down reasoning and create an opportunity to arouse a slightly different awareness of the problems and situations mobilizing us” [16]. This ‘slowing down’ is relevant to situations where publics are sparked into being by issues where there are significant problems of relevance (Marres, 2007). The challenge then becomes one of not rushing to establish what is the matter and how to deal with it, but one of finding ways to keep the situation open.

Stengers introduces the figure of the idiot as an example of a character that can assist in slowing down reasoning because the idiot “resists the consensual way in which the situation is presented and in which emergencies mobilize thought or action” [17]. The way the idiot does this is by insisting that “there is something more important” while refusing to reason about what this is [18].

We may notice some similarities and differences here between the figure of the idiot and the figure of the troll. Similarly to the idiot, the troll refuses to engage in rational argument. Instead the troll presents us with something beyond explanation, whether it is being insincere or simply full of angry opposition. Contrary to the figure of the idiot, it is not clear whether the troll believes there to be ‘something more important’. But the potential of the troll to disturb consensus nevertheless remains.

Making something of this potential, however, depends on escaping liberal narratives where the troll is ‘the other’ in a sense that reproduces a distinction between a wild or untamed way of behaving in online forums and a civilized way of behaving. Coming to more interesting terms with the existence of online trolls seems to require a different kind of politics that puts a price on being able to question ideas about civilized behavior and proper understandings. Such an understanding of politics is available in the abovementioned mentioned work in STS, which puts the indeterminacy of problematic situations (or issues) at the center of democratic politics.



Discussion: The problematic relevance of trolls

Having the figure of the troll make a difference here might benefit from thinking more carefully about what role trolls have played in other settings than online forums. To come back to Lakoff and Johnson’s work on metaphors, this does not have to be a question of abandoning the metaphor of the troll. It can be a question of exploring further the “reverberations” that metaphor creates [19]. Exploring trolls will allow for not just thinking differently about something but also open for other kinds of actions where metaphors are understood as performative.

Lakoff and Johnson (1980) offer an anecdote about how an Iranian student understood the phrase the ‘solution of our problems’ as referring to not a means of solving problems, but a boiled-down and concentrated version of the problems in question. ‘Solution’ was understood as a chemical metaphor, referring to a liquid mixture, rather than in the everyday sense. This metaphorical accident is suggestive of a way to think about problems not as things to be overcome, but things that can be ‘catalyzed’ or not (to stay with the chemical metaphor). Pursuing this idea, Lakoff and Johnson propose that “the reappearance of a problem is viewed as a natural occurrence rather than a failure on your part to find ‘the right way to solve it’.” [20].

Such an understanding of the role of problems would fit a more issue-oriented version of democratic politics, and it would arguably also shift the role that trolls can play. If trolls can no longer be understood to simply be ‘in the way’ of addressing problems ‘seriously’, trolling might be appreciated as forcing us to slow down reasoning. Such ‘slowing down’ will have to include avoiding hasty conclusions about what Internet trolls are and what motivates them. The resources for doing so can both lie in the richness of thinking about trolls prior to the Internet and in more careful studies of those beings that we refer to as Internet trolls. As an online community manager said after dealing with trolling for a period of time:

“The idea of the basement dweller drinking Mountain Dew and eating Doritos isn’t accurate (...) They would be a doctor, a lawyer, an inspirational speaker, a kindergarten teacher. (...) It’s more complex than just being good or bad.” [21]

Moving beyond simplified versions of good and bad is the first step if we want to live more productively with Internet trolls. Another step is to appreciate how trolls, and monsters more generally, have been generative in the sense of forcing us to slow down reasoning at several points in our history, including questioning what counts as natural and what does not. Such realizations about historical interactions with trolls might inform our current dealings with their online relatives. As Lindow [22] puts it: “The details have changed; these trolls are not large or small, shaggy or misshapen, changers of shapes or perceptions, but they still blur categories, and they still disrupt.”

The recent rise in the use of trolling in the context of political campaigns provides perhaps the most clear example of how particular instantiations of trolling can be perceived as tolerable or even desirable. When U.S. politician Rick Perry’s positions on abortion sparked politically motivated trolling behaviour on his Facebook page, this activity was understood by respondents as “trolls with whom the public can sympathize and even empathize” [23]. Such dynamics points to a newer tendency to recognize certain kinds of trolling behaviour as ideologically motivated and perhaps best understood within an ethics of political activism (Fichman and Sanfilippo, 2016). The notion of ideological trolling is thus useful for capturing one way in which trolling can be politically productive, but it is somewhat at odds with the potential for disruption put forward in this paper, in so far as ideological trolling often reproduces rather than disturbs existing political categories.

The use of trolling as a form of “soft influence” in international politics, such as the case of the “Kremlin trolls,” is perhaps the most emblematic example of ideological trolling (Zelenkauskaite and Balduccini, 2017). While the systematic use of trolling to shift public opinion (astroturfing) is a significant phenomenon worth more scholarly attention, such dynamics also position trolling as counter to a public sphere of free and sincere public opinion-forming, which might lead away from appreciating trolls as the radical other that challenges existing norms for proper action. Following the framework proposed in this paper, it seems key to insist on the uncertainty as to whether, for instance, Kremlin trolls are in fact ‘hired guns’ or rather tricksters piggybagging on international tension to achieve distruption (Zelenkauskaite and Niezgoda, 2017). Such uncertainty challenges us to think twice about the ideological maps we use to navigate political conflicts and consider what else may be of relevance.




In so far as they still distrupt, Internet trolls may become uncanny allies in the politically productive blurring of categories — if we stick with the ambivalences inherent to the figure of the troll. The ongoing slide in the meaning of online trolling noted at the outset of this paper is useful in this perspective. As a consequence of the proliferation of the use of the troll figure in relation to online debates, we can no longer be sure if trolls are sincere or not when they make their antagonizing comments. This uncertainty is interesting because it brings the problem of relevance to the fore. Internet trolls suggest that ‘there is something more important’ without explaining what this is and without justifying whether it is relevant or not. Instead, victims of trolling are left to wonder what happened and how it all makes sense. Such wonderment and sense-making has the capacity to force us to question our existing understandings of a situation. As such, Internet trolls specifically help us question the dichotomy between statements relevant to public debate and statements that are merely private. No doubt, the affordances of online technologies contribute to this by making informal statements visible and traceable in new ways. But the potential is misunderstood, or at least ignored, if we stick to dominant liberal narratives about online debate, where trolls are natural inconveniences to be contained as best possible. Trolls become much more interesting if they are not conceived as either natural or cultural, but as what they truly are: Monstrous figures that force us to reckon with things currently outside our radar of relevance. End of article


About the author

Andreas Birkbak is an Assistant Professor in the Techno-Anthropology Research Group at Aalborg University (AAU) in Copenhagen. His research is on digital methods and democracy, drawing on science and technology studies. Andreas holds a Ph.D. in Techno-Anthropology (AAU), an M.Sc. in Social Science of the Internet (Oxford), and an M.S.c+B.Sc. in Sociology (University of Copenhagen).
Twitter: @communaut
E-mail: ab [at] learning [dot] aau [dot] dk



The author thanks Steffi Steinhart, Diego Compagna, and First Monday’s reviewers and editors.



1. Hardaker, 2010, p. 237.

2. Herring, et al., 2002, p. 372.

3. Jakobsson, 2008, p. 41.

4. Ibid.

5. See also Arnold, 2005, p. 112.

6. Herring, et al., 2002, p. 372, drawing on Pfaffenberger, 1996.

7. For an excellent critique of the naturalization of civil society, see Somers, 1995a; 1995b.

8. Lindow, 2014, p. 9.

9. Lindow, 2014, p. 140.

10. Herring, et al., 2002, p. 371.

11. Haraway, 1991, pp. 21–22.

12. Law, 1991, p. 2.

13. Donna Haraway interviewed in Markussen, et al., 2000, p. 12.

14. Lindow, 2014, p. 138.

15. Ibid.

16. Stengers, 2005, p. 994.

17. Stengers, 2005, p. 994; see also Michael, 2012.

18. Stengers, 2005, p. 994.

19. Lakoff and Johnson, 1980, p. 129.

20. Lakoff and Johnson, 1980, p. 131.

21. Jessica Moreno quoted in Stein, 2016.

22. Lindow, 2014, p. 141.

23. Sanfilippo, et al., 2017, p. 1,807.



A. Applebaum, 2015. “Donald Trump, voice of the bottom-feeders,” Slate (21 August), at, accessed 22 February 2018.

M. Arnold, 2005. “Hvat er tröll nema Þat? The cultural history of the troll,” In: T. Shippey (editor). The shadow-walkers. Jacob Grimm’s mythology of the monstrous. Tempe, Ariz.: Arizona Center for Medieval and Renaissance Studies, pp. 111–155.

A. Birkbak, M. Petersen and T. Elgaard Jensen, 2015. “Critical proximity as a methodological move in techno-anthropology,” Techné: Research in Philosophy and Technology, volume 19, number 2, pp. 266–290.
doi:, accessed 11 April 2018.

J. Dewey, 1927. The public and its problems. New York: H. Holt.

P. Fichman and M. Sanfilippo, 2016. Online trolling and its perpetrators: Under the cyberbridge. Lanham, Md.: Rowman & Littlefield.

J. Habermas, 1989. The structural transformation of the public sphere: An inquiry into a category of bourgeois society. Translated by T. Burger with the assistance of F. Lawrence. Cambridge: Polity.

D. Haraway, 1991. “The actors are cyborgs, nature is coyote and the geography is elsewhere: Postscript to ‘ Cyborgs at large’ ,” In: C. Penley and A. Ross (editors). Technoculture. Minneapolis: University of Minnesota Press, pp. 21–26.

C. Hardaker, 2010. “Trolling in asynchronous computer-mediated communication: From user discussions to academic definitions,” Journal of Politeness Research, volume 6, number 2, pp. 215–242.
doi:, accessed 11 April 2018.

S. Herring, K. Job-Sluder, R. Scheckler and S. Barab, 2002. “Searching for safety online: Managing ’trolling’ in a feminist forum,” Information Society, volume 18, number 5, pp. 371–384.
doi:, accessed 11 April 2018.

Á. Jakobsson, 2008. “The trollish acts of Þorgrímr the Witch: The meanings of troll and ergi in medieval Iceland,” Saga-book of the Viking Society for Northern Research, volume 32, pp. 39–68.

R. Kvideland and H. Sehmsdorf (editors), 1991. Scandinavian folk belief and legend. Minneapolis: University of Minnesota Press.

G. Lakoff and M. Johnson, 1980. Metaphors we live by. Chicago: University of Chicago Press.

J. Law (editor), 1991. A sociology of monsters: Essays on power, technology and domination. London: Routledge.

J. Lindow, 2014. Trolls: An unnatural history. London: Reaktion Books.

K. Mantilla, 2015. Gendertrolling: How misogyny went viral. Santa Barbara, Calif.: Praeger.

R. Markussen, F. Olesen and N. Lykke, 2000. “Cyborgs, coyotes and dogs. A kinship of feminist figuraions — Interview med Donna Haraway,” Kvinder, køn & forskning, number 4 (“Overskridende metoder”), pp. 6–15.

N. Marres, 2012. Material participation: Technology, the environment and everyday publics. Basingstoke: Palgrave Macmillan.

N. Marres, 2007. “The issues deserve more credit: Pragmatist contributions to the study of public involvement in controversy,” Social Studies of Science, volume 37, number 5, pp. 759–780.
doi:, accessed 11 April 2018.

M. Michael, 2012. “‘What are we busy doing?’ Engaging the idiot,” Science, Technology, & Human Values, volume 37, number 5, pp. 528–554.
doi:, accessed 11 April 2018.

Oxford English Dictionary, 2013. “troll,” at, accessed 11 April 2018.

B. Pfaffenberger, 1996. “‘If I want it, it’s OK’: Usenet and the (outer) limits of free speech,” Information Society, volume 12, number 4, pp. 365–386.
doi:, accessed 11 April 2018.

T. Ringheim, 2014. “Oppsøker og konfronterer netthatere – på TV” (9 June), at, accessed 22 February 2018.

R. Rogers, 2013. Digital methods. Cambridge, Mass.: MIT Press.

M. Sanfilippo, S. Yang and P. Fichman, 2017. “Managing online trolling: From deviant to social and political trolls,” Proceedings of the 50th Hawaii International Conference on System Sciences. Hawaii, at, accessed 11 April 2018.

M. Somers, 1995a. “Narrating and naturalizing civil society and citizenship theory: the place of political culture and the public sphere,” Sociological Theory, volume 13, number 3, pp. 229–274.
doi:, accessed 11 April 2018.

M. Somers, 1995b. “What’s political or cultural about political culture and the public sphere? Toward an historical sociology of concept formation,” Sociological Theory, volume 13, number 2, pp. 113–144.
doi:, accessed 11 April 2018.

J. Stein, 2016. “How trolls are ruining the Internet,” Time (18 August), at, accessed 22 February 2018.

I. Stengers, 2005. “The cosmopolitical proposal,” In: B. Latour and P. Weibel (editors). Making things public: Atmospheres of democracy. Cambridge, Mass.: MIT Press, pp. 994–1,003.

R. Sved, 2014. “Nettroll kveler debatten” Kommunal Rapport (29 August), at, accessed 23 August 2016.

J.R.R. Tolkien, 1937. The hobbit, or, There and back again. London: George Allen & Unwin.

A. Zelenkauskaite and M. Balduccini, 2017. “‘Information warfare’ and online news commenting: Analyzing forces of social influence through location-based commenting user typology,’ Social Media + Society (17 July).
doi:, accessed 11 April 2018.

A. Zelenkauskaite and B. Niezgoda, 2017. “‘Stop Kremlin trolls:’ Ideological trolling as calling out, rebuttal, and reactions on online news portal commenting,” First Monday, volume 22, number 5, at, accessed 11 April 2018.
doi:, accessed 11 April 2018.


Editorial history

Received 22 February 2018; revised 6 March 2018; accepted 11 April 2018.

Creative Commons License
This paper is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Into the wild online: Learning from Internet trolls
by Andreas Birkbak.
First Monday, Volume 23, Number 5 - 7 May 2018

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.