Contesting #stopIslam: The dynamics of online counter-narratives against hate speech

By Eva Giraud, Elizabeth Poole and Ed de Quincey

On 22 March 2016 the hashtag #stopIslam began to trend on the social media platform Twitter, after terrorist attacks in Brussels for which the Islamic State claimed responsibility. As might be expected from an anti-Islamic hashtag, #stopIslam was originally designed to spread racialized and intensely negative narratives about Muslim communities. The dynamics of #stopIslam, however, were more complicated than they appeared initially, as the hashtag trended because people were appropriating it in order to condemn hate speech. Indeed what drew our attention to #stopIslam was the visibility it had afforded to narratives against hate, after the counter-narrative – rather than the original hashtag – was reported on by a number of national and international news outlets (including the BBC, CNN, Nigeria Newsdesk and Al Jazeera).

In 2016 we applied for a small grant from the British Academy/Leverhulme Trust, in order to explore a number of questions about #stopIslam that we felt were politically urgent. In light of the way that contestations of the hashtag had framed mainstream media narratives, we were initially interested in questions of voice. We wanted to know, for instance, who was participating in these counter-narratives and whether – despite all of the well-documented problems associated with social media – they can still offer a platform for perspectives that are often excluded from the mainstream media (or if it was dominated by well-intended allies in more problematic ways).

After purchasing Twitter data from the 30 days immediately after the bombings – about half a million tweets in total – we spent a year and a half analysing it using a combination of computational, quantitative and qualitative methods. Our findings ultimately shifted the project in a slightly different direction than we anticipated. Even though the hashtag was trending prior to the UK EU referendum and election of Donald Trump, what we found speaks directly to a political climate in which right-wing populism has gained heightened visibility. 

Our initial findings were relatively positive: The majority of the most re-tweeted tweets attempted to combat Islamophobia. A common meme was used by those contesting the hashtag, for instance, which stated “why are y’all tweeting #stopIslam when…” followed by images of white supremacist movements. We found similar cause for optimism when we looked at the most prominent twitter users who shared the tweets (in terms of the number of their followers), the majority of whom shared the counter-narrative. When it came to who was circulating counter-narratives we found a very geographically heterogeneous community of users, which was also diverse in terms of self-identified religious and ethnic identity. This heterogeneity, however, proved to be part of the problem.

In contrast, those sharing the hashtag with the most frequency were a very homogenous community of users who circulated it in line with its original meaning. Although uses of the hashtag to spread hate speech gained less visibility in terms of the sheer number of people who engaged with them on the day of the attack, they were circulated far more regularly (over time) and within a far more closely-knit community. #stopIslam was most frequently retweeted, for instance, with hashtags such as #islamistheproblem, #bansharia and #islamkills. Perhaps more worryingly, these messages were being tweeted in direct relation to US-based right-wing conservative groups on Twitter such as the Tea Party, CCOT (Conservatives on Twitter), TCOT (Top Conservatives on Twitter). #stopIslam was also routinely accompanied by hashtags such as #Trump2016 and #wakeupamerica. Links with the US right, and Trump in particular, were reinforced by the user bios of those sharing the hashtag, whose accounts were often adorned with patriotic imagery (such as flags, crucifixes and eagles).    

When looking at comments beneath tweets, a more worrying picture emerged still. Those who perpetuated the hashtag tended to be supported by those who agreed with their arguments. In contrast (with the exception of celebrities) those who used the hashtag to spread a more positive counter-narrative were routinely bombarded with anti-Islamic memes, statistics, and news reports originating from far right websites. Our findings suggest that the strength of this far right discourse is due to their tightly bound online communities, who persistently and aggressively attack any counter-narratives that emerge. These concerns seem borne out by the lack of longevity associated with the counter-narrative, which died a few days after it trended, in contrast to the ongoing use of the hashtag to spread hate speech.   

Although social media is often trivialised, our findings point to the need to take it seriously. It is possible that anti-racist activists feel there is nothing to gain by engaging in conflictual exchanges with the far right. This is perhaps especially the case when such exchanges are likely to be intimidating or emotionally exhausting. However, this project suggests that further work nonetheless needs to be done in understanding strategies engaged in by the far right – and how to respond to these strategies – in order to gain ground in the ‘information warfare’ in which we now seem to find ourselves.             

The team working on this project are: Elizabeth Poole (PI), Ed de Quincey (CO-I), Eva Giraud (CO-I), and RAs Mohammed Al-Janabi, Charis Gerosideris and Wallis Seaton.