Trolling democracy: anonymity doesn’t cause conflicts, bad site design does

‘Don’t feed the trolls’, people on social media are told. By forcing people to use their real names online, the managers of online forums hope to shame trolls into silence. But, Jennifer Forestal argues, this fails to tackle the root of the problem: a lack of the ‘close and direct intercourse and attachment’ which John Dewey believed was vital to democratic deliberation. We need to design online spaces where this can happen, rather than indiscriminately multiplying our connections and relying on top-down moderation. Some disruption is necessary when we talk about politics.


A drawing by Einar Bager (1887-1990) shows a youngster poking trolls with a twig. Source: Europeana via Malmo Museer and a CC-BY-NC-ND licence

With a number of high-profile incidents of online trolling—in which a New York Times editor was driven off Twitter by a barrage of anti-Semitism, for example, a troll intentionally tweeted a flashing strobe to a journalist with epilepsy, and women in all kinds of public positions are frequently threatened with assault, rape, and even death—the impact of internet trolls has gained renewed attention in recent months. Trolls are, it is argued, ‘ruining the internet’ and causing the downfall of democracy itself.

So how do we solve a problem like internet trolls? This question is more difficult than it first appears. Internet trolls highlight a tension at the heart of democratic communities, both on- and off-line. Trolls clearly intend to disrupt, even silence, discussions on a given site. But disruption is not always destructive.

Indeed, some disruptions are necessary in order to expand democratic politics beyond the constraints of formal deliberation, rational discourse, and traditional norms and expectations that can be limiting. This distinction, however, has not been clearly made in recent moves to identify and condemn ‘trolling’. The specifically democratic problem of trolls, then, is distinguishing them from these kinds of ‘activists,’ and remaining open to the latter while preventing the former.

Naming and shaming

In trying to negotiate this tension, social media platforms have largely focused on containing the negative effects of trolls rather than banning them outright: strategies vary widely, from prosecution, to moderation, to modifying the behaviour of trolls’ targets (this is the logic behind the omnipresent dictum ‘don’t feed the trolls’).

Often, sites turn to ‘real names’ or ‘verified identities’ as a way of compelling users’ good behaviour. People, so the argument goes, do not want their friends to know they are trolls. Naming policies work, in other words, ‘with a blend of incentivisation and good ol’ shaming.’ If users want their contributions to be taken seriously, they will conform to the norms of rational discourse rather than face the censure that comes from violating those expectations.

The problem is that shaming does not always work. What these tactics have in common is the idea that the threat of censure will somehow disincentivise trolls from unsavoury activities. Indeed, all of these strategies—and, most prominently, naming policies—are premised on the idea that users will be invested in maintaining the reputation of their online handle; the threats of banning, blocking, or shaming only carry weight if we assume that users care what other readers think and have some tie to their online identity.

By turning to these tactics, in other words, platforms have tried to install mechanisms of accountability without considering the social and spatial configurations that make it possible to hold individuals accountable in the first place.

Accountability needs ‘buy-in’

As I have argued elsewhere, an alternative approach to the problem of trolls is a turn to space, architecture, and design in order to build sites that can host activist democratic politics while protecting the community from the intrusion of trolls.

The role of space in structuring democratic interactions is often assumed, but not investigated. Software executives involved in the construction and management of online platforms, for example, often invoke the image of the ‘small town’ or local neighborhood as justification for imposing naming requirements on their sites.

Neighbours certainly can, and do, cooperate in productive ways; they often coordinate even without the threat of legal sanction. But neighbours do not interact in these ways because they simply know one another’s’ names. Instead, this willingness to ‘play nice’ is in large part due to spatial constraints.

Neighbours, by definition, share space with one another; because of this proximity, they run into one another frequently and come to understand how they are connected and how their actions affect those around them. Neighbours learn how to live together over time—and it is through these repeated interactions that incentives for cooperation, compromise, and civility become clear.

Trolls do not feel this same sense of connection to those they terrorise. Only in cases where trolls have been forced to confront their victims, and made to realise what they have in common, have they voiced regret or shame for their earlier behaviour. But if naming policies and other tactics are not the answer, how then do we build sites that reduce the likelihood of trolling by facilitating the kind of relationships marked by the ‘close and direct intercourse and attachment’ that democratic theorists have long argued is a crucial foundation for democratic politics?

Building digital democracies

Writing in 1902, the American philosopher John Dewey emphasised the importance of local spaces, like the school and neighbourhood, in facilitating democratic politics. Because of the smaller scale of these spaces, wrote Dewey, there was a greater possibility of “mixing people up with each other; bringing them together under wholesome influences, and under conditions which will promote their getting acquainted with the best side of each other.”

These kind of ‘social centres’ gathered a diverse set of citizens and brought them into close quarters. Through the repeated interactions that such a small space invited, individuals would not only come to see what they shared in common, but would also be incentivised to make adjustments to their habits, attitudes, and behaviours in the spirit of reciprocity and accommodation.

As platforms like Facebook work to achieve similar results in a digital environment—to ‘develop the social infrastructure’ that would ‘help people build supportive communities’—they should therefore resist the temptation to keep expanding the reach of social media platforms and relying solely on gatekeepers to maintain ‘civil’ discourse on the site. By indiscriminately multiplying our connections with one another, and using ‘top-down’ moderation to regulate our interactions, we can easily lose the kind of political intimacy that Dewey, and others, thought was so integral to the smooth functioning of democracy.

Instead of emphasising questions of access—of managing who can enter a space and under what conditions—I argue that we should also turn our focus to the design of the sites themselves. ‘The heart and final guarantee of democracy,’ Dewey tells us, ‘is in the free gatherings of neighbours on the street corner” and “in the free gatherings of friends in the living rooms of houses and apartments.’ As we consider the future of these digital platforms, then, we must therefore ensure that we build spaces in which these gatherings can take place.

This post represents the views of the author and not those of Democratic Audit.

jennifer forestalJennifer Forestal is Assistant Professor of Political Science (Theory) at Stockton University in New Jersey.  

Similar Posts