Whereas there has been much research into the manufacture of ‘fake news’ to sow disunity within liberal democracies, little is known about how information disorders affect deeply divided societies. This paper addresses that gap in the literature by exploring how digital media are used to share misinformation and disinformation during contentious public demonstrations in Northern Ireland. It does so by reviewing the literature on social media information flows during acute crisis events, and qualitatively exploring the role of Twitter in spreading misinformation and disinformation during the 2014 and 2015 Ardoyne parade disputes. Results indicate that visual disinformation, presumably shared to inflame sectarian tensions during the parade, was quickly debunked in information flows co-curated by citizens and professional journalists. Online misinformation and disinformation appeared to have minimal impact on events on the ground, although there was some evidence of belief echoes among tweeters who distrusted the information provided by mainstream media.
Fake news or information disorders?
Digital media and counterpublics in ‘post-conflict’ Northern Ireland
Twitter and the Ardoyne parade dispute
Results and discussion
Tip of the iceberg?
The term ‘fake news’ has been used pejoratively by politicians in order to delegitimize media outlets that are critical of their conduct. Scholars have frequently challenged these narratives and argued that the ‘post-truth’ era is a crisis engulfing the entire information ecosystem, which in the case of the United States, can be attributed to the efforts by conservative media and right-wing political parties in the post-war period to undermine public trust in mainstream media (Lee, 2014). A related issue is that empirical work in this field has focused on how elections were affected by disinformation manufactured for financial gain in ‘fake news factories’, and amplified by social bots on social media in order to sow confusion and disunity (Ghosh and Scott, 2018; Shao, et al., 2018). This is despite the fact that such ‘pure’ forms of ‘fake news’ have been “rare or non-existent” in most countries to date .
In contrast, there have been few studies exploring how social media is used to circulate misinformation and disinformation in ethnically divided societies. In one such incident, anti-Muslim riots in Myanmar in July 2014, which were attributed to an unsubstantiated rumour spread on Facebook claiming that the owner of a Muslim tea shop had raped a Buddhist employee (Stanley, 2017). A similar story emerged in Sri Lanka in March 2018, when rioting was said to have been sparked by false claims circulated on the site that Muslims were planning to overthrow the Buddhist majority (Taub and Fisher, 2018). This paper builds on this work by presenting an in-depth analysis of ‘information disorders’ in the context of public demonstrations within ‘post-conflict’ Northern Ireland. It does so by providing an overview of the literature on the role of social media platforms in spreading misinformation and disinformation during acute events, and qualitatively exploring 7,388 tweets posted during disputed Orange Order parades in Ardoyne, North Belfast in July 2014 and 2015.
Fake news or information disorders?
‘Fake news’ is a ‘floating signifier’ deployed to project a certain image of how society should be structured . It is not only ‘linguistically defective’ but also a form of ‘propaganda’ closely connected with populist, reactionary ideologies, as well as former U.S. President Donald Trump’s attempts to undermine the credibility of mainstream media (Allcott and Gentzkow, 2017; Habgood-Coote, 2019). These are information flows that ‘pollute’ information ecosystems for either the financial or political gain of certain groups. In the case of the former, ‘clickbait’ news articles shared on social media were said to be generating income for Macedonian teenagers working in so-called ‘fake news factories’ in the run up to the 2016 U.S. presidential election (Wardle, 2017). In term of the latter, far-right groups wielded undue influence on the news agenda during the same campaign through their strategic use of clickbait and social bots to share content with partisan news outlets and mainstream media (Marwick and Lewis, 2017). Probably the most nuanced theoretical framework for exploring this ‘information crisis’ has been provided by Wardle (2017), who identified three different types of ‘information disorder’ affecting democratic states. These were:
- Misinformation: the sharing of false information without harmful intent;
- Disinformation: the intentional sharing of false information to cause harm to others; and
- Malinformation: the sharing of accurate information with the purpose of causing harm to others.
This paper focuses primarily on the first two of these information disorders. Contemporary definitions of disinformation are congruent with black and grey propaganda; the former referring to covert campaigns that use falsehoods to change the behaviour of a target audience, while the latter is characterized by uncertainty about the accuracy of the information as well as the validity of the source (Briant, 2015). These practices can be ‘turbocharged’ by social media platforms, which “host and organize access to public expression.”  The ‘virality’ of such content is said to be considerably higher when accompanied by a powerful narrative, a visual component, it provokes high-arousal emotions or comes from a reputable and trusted source (Wardle, 2017). However, disinformation designed to polarize opinion does not necessarily have universal effects on audiences. Citizens use heuristics, consult pre-existing social networks, and assess the source and content of political messaging before deciding on its authenticity (Tandoc, et al., 2017; Thorson, 2016). This feeds into the ongoing debate over whether social media create ‘filter bubbles’ or ‘echo chambers’ in which the algorithmic curation of newsfeeds, and the tendency of people to only engage in online groups populated by like-minded individuals, reinforces pre-existing opinions by limiting exposure to opposing views. The empirical investigation of these claims has been a primary focus of Internet studies to date. For example, one study found that people were exposed to ideologically diverse opinion on Facebook courtesy of the sharing of content produced by mainstream media, but they were still most likely to click on stories that were congruent with their opinions (Bakshy, et al., 2015). Yet, Dubois and Blank (2018) argue that news consumption is a multi-platform activity and thus people are unlikely to find themselves in hermetically sealed echo chambers. Bruns  characterizes online platforms as “powerful engines of context collapse” that expose their predominantly apolitical users to views that they might find undesirable. He asserts that ill-defined concepts such as echo chambers and filter bubbles distract from the root causes of political polarization, such as economic inequality, that predate the Internet. One should also be cautious about ascribing rationality to the information behaviors practised by politically engaged social media users. There is already some evidence to suggest that ‘civic-minded’ citizens share tabloid news on social media, irrespective of its accuracy, in order to spark debate with others (Chadwick, et al., 2018).
Online platforms such as Twitter provide fertile terrain for the dissemination of rumours and misinformation during acute events, those natural disasters and terrorist attacks which prompt short, intense bursts of activity from social media users (Burgess and Bruns, 2015). Much of this work has adapted the two-step flow model of communication proposed by Katz and Lazarsfeld (1955) to explore how information flows from ‘opinion leaders’ to citizens, with the former playing an important role in propagating false information and its correction (Pang and Ng, 2017). Twitter in particular was linked to the spread and debunking of misinformation during events such as the Queensland floods and English riots in 2011, and the 2013 Boston bombings (Bruns, et al., 2012; Gupta, et al., 2013; Procter, et al., 2013). While the number of tweets containing misinformation typically declined after corrective information was circulated by opinion leaders, this did not mean that such false information was dismissed by users. Research suggested that tweeters often struggled to distinguish between truthful messages and hoaxes during such incidents (Mendoza, et al., 2010; Starbird, et al., 2014). Overall, the evidence to date suggests that Twitter enables the rapid debunking of misinformation during acute events, but that tweeters are often unable to identify false information when they first encounter it on the site.
Digital media and counterpublics in ‘post-conflict’ Northern Ireland
While there has been much research into the weaponization of digital disinformation during democratic elections, there has been little exploration of how online platforms facilitate information disorders in societies deeply divided along sectarian lines. This paper uses a case study to address this gap in the literature by exploring how Twitter is used to share misinformation and disinformation during contentious episodes in Northern Ireland. These are moments in which existing levels of polarization within a society are ‘inflamed’ by “takeoff issues” . The 1998 Good Friday Agreement delivered a ‘negative peace’, defined by Galtung  as the “absence of organized collective violence,” that left the zero-sum perceptions of politics and territory held by the two main ethnic blocs relatively intact. Broadly, Catholics, nationalists and republicans continue to support reunification with the rest of Ireland, whereas Protestants, unionists and loyalists favour the continuation of the union with Great Britain (Edwards and McGrattan, 2010). There has been a long history of propaganda being deployed in support of these narratives, which has often involved efforts to delegitimize those of the ‘other’ community. The conflict (known colloquially as the ‘Troubles’) was an asymmetric ‘information war’ between the British state and the Republican movement, the umbrella term used to refer to the Provisional Irish Republican Army and its political wing Sinn Fein. The U.K. government actively promoted the view that journalistic norms of ‘impartiality’ should be sacrificed in favor of coverage that depicted these actors as ‘enemies’ of the state . In return, both Loyalist and Republican paramilitaries created their own media in order to counter representations framing them as ‘criminals’ who were unrepresentative of their respective communities (Curtis, 1996).
Dissident loyalists and republicans have had few opportunities in the news media to express their opposition to the Agreement since 1998. McLaughlin and Baker (2010) assert that this was due to the pervasiveness of the ‘propaganda of peace’, which mobilized support for a neo-liberal agenda conflating prosperity with peace and framed the Agreement as a ‘self-evident good’ beyond critique. This propaganda helped political elites manufacture consent for a consociationalist peace settlement that entrenched divisions between the two ethno-sectarian blocs, and failed to address the conflict-legacy issues experienced by both communities. In this context, subaltern counterpublics emerged online that challenged the political status quo. In the mid-2000s, anti-Agreement loyalists and republicans used their Web sites to outline their opposition to the peace process, with the former characterising Sinn Fein’s inclusion within the powersharing institutions as a perversion of democracy and the latter criticizing the political wing of the republican movement for abandoning the armed struggle and legitimising partition on the island of Ireland (Reilly, 2010). However, these Web sites functioned as ‘islands of political communication’ that were unlikely to provide a high degree of visibility for groups who received very little coverage in the traditional media.
The limited evidence to date suggests that social media has ushered in a new era of political partisanship, particularly among a small but vocal minority of (digital) citizens who use social media to share political information. Sinn Fein has been linked to orchestrated social media campaigns by so-called ‘Shinnerbots’ that seek to silence its critics. These anonymous Sinn Fein supporting accounts were responsible for the online harassment of Maíria Cahill in October 2014, in the wake of her allegations that senior PIRA figures had covered up her sexual assault by a senior republican (Emerson, 2014). Researchers such as Hoey have argued that Shinnerbots have ‘drowned out’ other voices during partisan political debates on social media over the past four years . Yet, with the exception of the U.K.-wide EU Referendum in 2016, there has been no evidence thus far suggesting that political disinformation shared on social media has influenced the voting decisions of Northern Irish citizens.
Twitter and the Ardoyne parade dispute
Previous research has shown how bloggers and journalists are more likely than citizens to be the sources of information flows during contentious episodes, such as the popular uprisings in the Middle East and North Africa in 2011 (see Lotan, et al., 2011, for example). These studies tend to focus on the role of social media in mobilising supporters rather than the spread of false information during such episodes. This paper adds to this literature and seeks to address that gap by exploring the role played by citizens and professional journalists in debunking misinformation and disinformation about the 2014 and 2015 Ardoyne parade disputes. This is one of the most contentious public demonstrations held in Northern Ireland each year and is viewed as a ‘bellwether’ for progress towards a shared society (or lack thereof). Both the nationalist Ardoyne residents and loyalist supporters of the Orange Order have competing claims in relation to the annual Twelfth July Orange Order parade. Many of the residents oppose the ‘sectarian’ parade and have called for it be rerouted away from the Ardoyne shopfronts on the Crumlin Road. Supporters of the Ligoniel Orange lodges accuse them of being intolerant of unionist and loyalist culture and claimed that they had a right to march on the arterial North Belfast route. These tensions illustrate the ‘ethnic poker’ that persist within highly segregated working-class districts in North Belfast, where zero-sum perceptions of space and politics have led rival interface communities to “unrealistically up the ante” against each other . Hayward and Komarova (2014) also argued that efforts to find a local solution to the contentious parade were hampered by the fixed ‘interpretative horizons’ of both sides.
The annual Twelfth parade in North Belfast has been a constant flashpoint in the post-Agreement era. Eight civilians and 105 police officers were injured in July 2005 as a result of the violence that flared after nationalist protesters were forcibly removed to allow the return leg of the parade to proceed along its traditional route towards the city centre. In July 2009, the Real IRA were blamed for a gun attack on PSNI officers and orchestrating a riot involving nationalist youths following the evening parade . Such incidents led the Northern Ireland Parades Commission, a quasi-judicial body created under the Public Processions (NI) Act 1998, to impose restrictions on the parade from 2012 onwards, such as a reduction in the number of participants. These failed to prevent violent clashes between nationalist youths and the PSNI following the return legs of the 2011 and 2012 parades, prompting the Commission to ban the return leg of the 2013 parade from passing by the Ardoyne shopfronts. A total of 70 PSNI officers were injured during four consecutive nights of “animalistic” violence by loyalist rioters in North Belfast that year .
The largely peaceful 2014 Ardoyne parade
This paper focuses on the 2014 and 2015 parades as they were the first in which there was significant social media activity linked to the dispute. The failure of political representatives to broker a solution to this impasse in July 2014, highlighted by the continued presence of a loyalist protest camp nearby, led many observers to fear that there would be a repeat of the 2013 violence. Tensions were further raised when leaders of the main unionist political parties, together with loyalist representatives and the Orange Order, announced a ‘graduated response’ to the Commission decision to alter the route of the return parade for the second year in a row. Further details on this were released on the eve of the Twelfth, including plans for a six-minute pause of all Orange marches the following day to represent the time it would take for the North Belfast march to return home via its traditional route . In response, the Greater Ardoyne Residents Collective (GARC) tried unsuccessfully to obtain a high court injunction preventing the outward leg of the parade from passing the Ardoyne shops .
Social media was used to both escalate and de-escalate tensions between the nationalist residents and the Orange Order in the run up to the Twelfth. This can be seen in the responses to an image shared on Twitter late on 11 July, allegedly depicting an image of Oscar Knox, a five-year-old who died from a rare form of cancer, being burnt on an ‘Eleventh night’ bonfire in County Antrim. These rumors were quickly and angrily refuted by loyalists, who ‘named and shamed’ two tweeters they believed responsible for starting them (Reilly, 2016). Evidence that the image had been digitally altered emerged later that evening. One tweet provided a link to the original image of the bonfire that had been taken by photographer Stephen Barnes in July 2013 . The image of Knox, that had caused so much anger among tweeters a few hours earlier, was clearly absent from the picture. It was noticeable that the number of tweets about the Knox incident sharply declined after visual counter-evidence was shared online. Moreover, this visual disinformation went unreported by the news media, which praised both sides for delivering one of the most peaceful Twelfths in recent memory.
The 2015 parade
In the absence of an agreement to resolve the impasse, the Parades Commission ruled that the 2015 parade would be subject to the same restrictions as in the previous year due to the “potential for public disorder” . They also placed limitations on a planned GARC protest against the outward leg of the parade which, as in previous years, had been permitted to follow its traditional route past the Ardoyne shops. GARC reacted angrily to their protest being restricted, urging local residents to ignore the ruling and come out in force during the morning parade . Loyalist groups such as the West Belfast Ulster Political Research Group (UPRG) claimed this was part of a republican strategy to provoke violent clashes between the residents and the Orangemen, thus leading to the outward leg of the parade being rerouted in future years.
The 2015 parade was marred by loyalist rioting, as protesters faced off against PSNI officers preventing marchers from returning home via their traditional route. A total of 24 police officers were hurt after loyalists threw missiles at their lines, with erroneous rumours circulating online that PSNI Chief Constable George Hamilton was amongst those injured . Baton rounds and water cannon were also deployed by officers in the nearby Woodvale Road as tensions flared shortly after the march arrived at the police barricade at 7.30 pm. Tensions were further raised when Orangeman John Aughey drove his car into a crowd of residents who had gathered on the Crumlin Road during the standoff between loyalists and the police. He was later charged with attempted murder for the attack, which resulted in 16-year-old Phoebe Clawson being seriously injured . Images of PSNI officers and residents lifting the car onto its side to free the teenager were shared by eyewitnesses and journalists on Twitter. Like in 2013, the violence was condemned from politicians from across the political divide, with Secretary of State Theresa Villiers calling the attacks on police ‘disgraceful’ .
This study focussed on how prevalent misinformation and disinformation was on Twitter in both years, and the role played by citizens and professional journalists in debunking these false stories.
There were two research questions that were investigated:
RQ1: How prevalent was misinformation and disinformation about the contentious parade on Twitter in 2014 and 2015?
RQ2: How did tweeters respond to this misinformation and disinformation?
These research questions were explored through a critical thematic analysis of tweets posted during two periods: 11–14 July 2014, and 12–15 2015. The two selected periods included key events such as the ‘Eleventh night’ bonfires and both legs of the annual Ardoyne Twelfth parade in North Belfast that caused such controversy. It should be noted that the 2015 Twelfth demonstrations were on 13 July because parades were not held on a Sunday for religious reasons. The text-mining software tool DiscoverText (www.discovertext.com) was used to collect these tweets.
A search was conducted using ‘Ardoyne’ in order to identify tweets referring to the North Belfast parade, including those that were not marked with the eponymous hashtag. A total of 7,388 tweets were identified, 1,842 in 2014 and 5,546 in 2015. The study found that ‘Ardoyne’ tweets peaked at 7 pm on the day of the Twelfth demonstrations in both years (see Figure 1). One explanation for this finding was that this spike in Twitter activity was the result of viewers of flagship television news programmes, such as BBC Newsline, turning to social media to follow what was happening during the return leg of the parade.
Figure 1: Tweets mentioning Ardoyne, 12 July 2014 and 13 July 2015.
Both corpora were dominated by retweeted content (64 percent in 2014; 55 percent in 2015), followed by original tweets and @replies to other users (11 percent in 2014; 6 percent in 2015). One interpretation of this was that there was very little conversation between tweeters in relation to the dispute. However, it should be acknowledged that responses to tweets that did not mention Ardoyne were excluded due to the keyword search used. It was also not possible to fully establish the representativeness of the tweets collected using a text-mining software such as DiscoverText.
A critical thematic analysis of these tweets was conducted in two phases: August 2014 and March 2016 . First, two coders read each comment to explore emergent themes from the data and then decide whether this content met the requirements of the study. As per previous studies of Twitter during acute events (Lotan, et al., 2011; Procter, et al., 2013), retweets and their original tweets were aggregated in order to create identifiable information flows. The six phases of TA proposed by Braun and Clarke (2013) were then implemented, from the reading of each tweet to the definition of themes that emerged from the entire corpus. Fieldnotes were used to capture relevant information, such as heavily retweeted content and angry back-and-forth exchanges between supporters and opponents of the Orange Order. The inductive construction of codes was completed through manual data analysis that focused on how tweeters interpreted key events and issues. Congruent with previous research using TA (Buetow, 2010; Braun and Clarke. 2013), numbers are not deployed to quantify the presence of themes in these data, but are used to identify key sample characteristics such as the number of retweets for certain posts. Public figures such as politicians and journalists are directly quoted due to the ethical stance adopted for this project. As per previous research into social media and contentious politics in Northern Ireland (Reilly and Trevisan, 2016), the focus here was on ‘what was said rather than who said it’. Therefore, de-identification and paraphrasing were deployed to protect unaware participants from any harm that might arise from their identification in academic publications.
A small minority of the tweets in both corpora were deemed non-relevant to the study. In 2014, there were 18 tweets (0.97 percent) excluded on this basis, including one from BBC Northern Ireland (BBC NI) reporter Mark Simpson on 27 June referencing a cross-community sleep-out in the North Belfast district was retweeted 13 times during the period of data collection . It was plausible that tweeters shared this image of children from both communities smiling and eating bacon sandwiches to counter the negative media coverage of the dispute in the run up to the Twelfth. However, it was excluded from the study on the basis that it did not directly address events on the ground during the contentious episode. In 2015, there were 42 non-relevant tweets (0.76 percent) included one referring to a mural in the district expressing support for Gaza retweeted eight times and another praising local footballer Joe Gormley.
Results and discussion
Misinformation and disinformation largely absent from information flows shaped by mainstream media
This study found little evidence of false information being circulated in relation to the Ardoyne dispute in both corpora. There was only one tweet propagating disinformation in the 2014 dataset which had been retweeted once during this period, with the photoshopped image of the Randalstown bonfire noticeably absent from these discussions of the contentious parade. There was much more rumoring on Twitter in July 2015, due in no small part to the incident in which John Aughey drove into a crowd of protesters and critically injured Phoebe Clawson. Four original tweets conveying minsinformation were identified, with one being retweeted 25 times and the others being shared fewer than four times during the period of data collection.
This misinformation and disinformation will be discussed in more detail below, but it is worth noting that the study provided further evidence of Twitter’s importance as an information network during acute events (Pang and Ng, 2017). Closer inspection of both corpora showed that the most heavily retweeted content was posted by professional journalists. In 2014, for instance, a picture posted by BBC NI reporter Mark Simpson, showing one of the Ligoniel lodges as they walked past a GARC observer, along with the text: “peaceful start to 12th July marches. No trouble as Orange Order parade passes Ardoyne shops” was shared 72 times . The integral role of mainstream media in these information flows was also illustrated by the high number of retweets of seemingly mundane updates posted by journalists, particularly during the 2014 parade. For example, a tongue-in-cheek tweet from BBC NI journalist Kevin Sharkey showing two dogs greeting one another (captioned ‘Trouble in #Ardoyne’), was one of the most retweeted that year (RT’d 19 times) . The real-time coverage provided by journalists showed images of crash barriers, police personnel and vehicles put in place to enforce the Parade Commission’s ruling relating to the return leg of the parade. Such content illustrated how professional journalists, presumably deployed in North Belfast in the expectation of a recurrence of the violence seen in 2013, had very little to report on.
While professional journalists were responsible for the most retweeted content in the 2014 corpus, the contentous nature of this public demonstration was reflected in the tweets posted by self-identified loyalists and nationalists. Both made claims about the legitimacy of the parade and related protests. Loyalist tweeters accused republicans of orchestrating the protests against the march and were highly critical of GARC for stirring up trouble. For example, a picture of a crowd of republicans gathered outside the shopfronts was shared 47 times, with the caption noting that the shops were shut and that there was no sign of the parade. In return, supporters of the nationalist residents shared malinformation that illustrated what they saw as the sectarianism of the marchers. Most notably, a picture purporting to show the names of ‘dead UVF terrorists’ on the drum of the Pride of Ardoyne band was retweeted 25 times during the period of data collection. Several tweeters even went so far as to share this image with local journalists including the BBC’s Mark Simpson, presumably in order to focus media attention on the links between loyalist paramilitaries and the Orange Order. However, there was a similar number of tweets (27) praising both sides for delivering the most peaceful Twelfth Ardoyne had seen in years rather than relitigating who was to blame for the impasse.
Further evidence of the role of journalists in structuring these information flows emerged during the 2015 parade. Reporters present at the scene were predominantly affiliated to local newspapers, such as the Belfast Telegraph and Irish News, and regional television news programmes produced by BBC NI and UTV. However, one of the most retweeted posts (120 retweets) in this corpus was posted by Sky News Northern Ireland correspondent David Blevins, whose tweet consisted of a picture showing protesters being soaked by the PSNI water canon along with the caption: “BREAKING: Water canon deployed after police officer injured during disorder at Ardoyne flashpoint, North Belfast” . Indeed, a recurring theme in tweets posted by journalists between 7 pm and 10 pm was the ferocity with which loyalists attacked police lines with bricks, bottles and other objects.
While it was anticipated that journalists would be retweeted more than citizens during the rerouted return leg of the parade, they also initiated most of the information flows during its outward leg. This included detailed descriptions of the heavy security presence in North Belfast, as well as reports on the arrest of two GARC protesters for assaulting a PSNI officer during the group’s counter-demonstration that morning. The latter story in particular highlighted the role of mainstream media in shaping information flows about the parade. News of these arrests was first shared by journalists such as Allison Morris (Irish News), who tweeted that apart from the arrests it had been a “fairly hassle free albeit wet morning” . Rebecca Black (Belfast Telegraph) provided subsequent updates confirming that two men had been arrested, rather than one as had originally been reported.
The media framing of these arrests was in sharp contrast to that of loyalist political group the West Belfast UPRG, who tweeted “PSNI moving in against DISSIDENT SCUM at Ardoyne !! Several arrests being made !!”, as well as sharing a picture of the police dispersing the counter-demonstration . The tweet, posted at 9.10 am, was shared 17 times by those supportive of the Ligoniel Orange lodges, but not by any news organization or professional journalist. The group was one of a small number of pro-loyalist Twitter accounts to accuse the PSNI and the news media of ‘not telling the full facts’ about the Aughey incident and having an agenda against unionist and loyalist culture. In return, tweeters expressed solidarity with the nationalist residents and praised them for not reacting to the provocation of the sectarian marchers. Overall, as in 2014, there was little evidence to suggest that either loyalists or nationalists were initiating information flows that travelled further than their respective constituencies.
Tweeters move quickly to debunk rumors and disinformation
Misinformation and disinformation appeared to have a very short life span in both corpora. This was particularly evident in the 2014 corpus, when the visual disinformation shared by one loyalist was quickly debunked by other tweeters. A picture of a protester holding up a mock-up road sign indicating that the Orange Order was not welcome in the area began to circulate on Twitter shortly after 7.30pm on 12 July. The tweet suggested that this illustrated the intolerance of republicans and highlighted the ‘unlawful’ nature of the GARC counter-demonstrations. Within a few minutes visual evidence was shared showing that the protester’s placard proclaimed “Love Thy Neighbour”, rather than the anti-Orange Order slogan in the doctored image. This was corroborated by an image of the same scene, taken and shared on Twitter by BBC NI journalist Kevin Sharkey a few hours earlier. The tweeter responsible for the photoshopped image was accused by several others of ‘spreading lies’ about what was happening on the ground at Ardoyne and mocked for their poor photoshopping skills. Whereas the original tweet had been shared only once, just one of these corrective tweets was retweeted six times during the period of data collection.
A similar theme emerged in relation to the Ardoyne parade and counter-demonstrations in July 2015. Most notably, images taken from many different vantage points were used to support claim and counter-claims about the Aughey incident. West Belfast UPRG was one of several accounts to share reports that the republicans had attacked the car, with no reference to the attempted murder of the residents in attendance . This was also the most retweeted piece of misinformation in both corpora (shared 25 times). It should be noted that another tweet alleging that the car had been attacked was posted by the same tweeter who had been responsible for the photoshopped image of the ‘Love Thy Neighbour’ placard the previous year. The small group of loyalist tweeters responsible for these information flows speculated about whether Phoebe Clawson had in fact been trapped under the vehicle, despite evidence to the contrary circulating on the site. Within a few minutes of the incident, first reported by journalists including the BBC’s Chris Buckler at 7.43 pm, the PSNI confirmed that the teenager had been injured and that Aughey had been arrested at the scene . It was a further two hours before local MLAs, such as Sinn Fein’s Gerry Kelly, and the PSNI Chief Constable George Hamilton confirmed her injuries were not life threatening.
The Chief Constable himself was the subject of a rumor later that evening. Several viewers of a Russia Today (RT) feed covering events in North Belfast claimed that they had seen a senior police officer, resembling Hamilton, being injured by an object thrown by one of the rioters at around 9.35 pm. By 10:50 pm, journalists such as the Impartial Observer’s Rodney Edwards tweeted to confirm that these reports were inaccurate . Hamilton later used his own Twitter account to thank members of the public for their best wishes and express his gratitude to the officers on duty in the interface area. Those viewing the RT feed and commenting on Twitter also speculated about whether one of the PSNI officers injured on camera had lost his ear during the violence. Hybrid media logics meant that many professional journalists used Twitter to source stories like the Hamilton rumor and correction, for publications such as the Belfast Telegraph. Indeed, this was one of the clearest examples in the study of how citizens and professional journalists co-curated news about the parade impasse.
There was one vignette illustrating how citizens used Twitter to prevent the spread of misinformation and disinformation that might incite violence during the North Belfast parade. One citizen tweeted a rumor that he had heard from an acquaintance suggesting that a deal had been put in place for the return leg of the parade to proceed via its traditional route past the Ardoyne shops. This tweet, posted a few days before the Twelfth, was only shared three times but clearly had the potential to raise tensions in Ardoyne. The same tweeter acknowledged that this information was untrue and apologized for spreading false information in a tweet posted on the morning of 13 July 2015. Clearly it was difficult to tell how many saw the original tweet, or how they responded to this false information. However, this corrective tweet showed how at least some tweeters were aware of the impact that such misinformation might have upon community relations.
Tip of the iceberg?
These were not the sophisticated disinformation campaigns seen during elections in 2016, nor the propaganda deployed by loyalist and republican paramilitaries to legitimize their ‘armed struggle’ during the Troubles. Rather, they were crude photoshopped images, misinformation, and unsubstantiated rumors shared by citizens on Twitter. While it was considered highly unlikely that these ‘agents’ were motivated by financial gain, it was plausible that they were seeking to inflame sectarian tensions surrounding these events. The study certainly provided further evidence to support Bruns’ (2019) contention that online platforms are ‘engines of context collapse’ that have the potential to expose citizens to oppositional views. Those using the microblogging site to search for information about the 2014 and 2015 parades would have seen both the competing claims of loyalists and nationalists, albeit that these were overshadowed by mainstream media coverage of the dispute. While this exposure was unlikely to facilitate the intercommunity dialogue needed to address the contentiousness of this parade, there was also no clear evidence to suggest that misinformation and disinformation on Twitter contributed to the violent clashes between loyalists and the PSNI during the 2015 dispute. Indeed, tweeters appeared to react to events as they unfolded rather than directly shape them.
There were two caveats that should be acknowledged. First, it is difficult to make too many inferences about the attitudes of ‘watchers’ (described pejoratively as ‘lurkers’) who viewed such content but did not comment on it, nor those who responded but did not include ‘Ardoyne’ in their tweets. Indeed, many may have continued to believe such misinformation and disinformation, even after it had been fact-checked. Research suggests that ‘debunks’ may even provide a degree of legitimacy to the original claim and deliver the publicity sought by those agents responsible for its creation (Shao, et al., 2018). Corrections to political misinformation are the least likely to be believed, even where such information is shared by users’ online social networks (Margolin, et al., 2017). These ‘belief echoes’ are caused by the affective ‘pull’ of the original piece of misinformation being much stronger than the correction, as well as the extent to which the false information corroborates pre-existing political views (Flynn, et al., 2017; Thorson, 2016). There has also been evidence to suggest that citizens only share fact-checking messages with their online social networks that are congruent with their political ideologies, while ignoring those that support oppositional views (Shin and Thorson, 2017). The failure of politicians to resolve the Ardoyne parade dispute meant that loyalists and nationalist residents were perhaps more likely to believe misinformation and disinformation that correlated with their respective views of each other.
Second, it is highly likely that false and misleading information about these events was being spread via other social media platforms. The research findings were based on the ‘low hanging fruit’ of publicly available tweets that included the keyword ‘Ardoyne’ rather than harder to reach data on the microblogging site and other social media (Burgess and Bruns, 2015). Only an estimated 18 percent of Northern Irish adults were said to be using Twitter during this period . It is inconceivable that misinformation about the Ardoyne dispute was not being shared on private Facebook pages, encrypted messaging apps such as WhatsApp, or indeed via face-to-face conversations. In contrast to the extensively documented propaganda of the Troubles, much of this content is ephemeral and therefore inaccessible to researchers. Hence, future research into the spread of misinformation and disinformation during contentious public demonstrations in divided societies should empirically investigate the multitude of channels, both on and off-line, through which it circulates.
The ebb and flow of misinformation and disinformation during contentious public demonstrations such as Ardoyne can ultimately be attributed to the failure of political elites to satisfactorily address their root causes, rather than platforms such as Twitter that publish such content online. Hayward and Komarova (2014) argued that efforts to resolve the Ardoyne dispute during this period were the product of a dysfunctional political system in which openness to alternative interpretations was limited by the actions of ethno-political elites in the Stormont Executive . Since 2007, the DUP and Sinn Fein’s electoral dominance has been based on their operationalisation of ethno-sectarian differences, which have left little room for compromise on issues such as how to deal with contentious parades. While online misinformation and disinformation is unlikely to generate the sectarian violence seen in Myanmar and Sri Lanka, it will remain a feature of these acute crisis events until such time as politicians from both sides develop collective narratives that address contentious public demonstrations.
This paper added to the emergent literature on information disorders by exploring how false information spread on social media during the Ardoyne parade dispute, a contentious public demonstration characterized as a bellweather for community relations in Northern Ireland. The study found that misinformation and disinformation constituted a very small proportion of the Twitter activity surrounding the 2014 and 2015 parades. Citizens directly challenged those responsible for sharing visual disinformation during this acute event, while journalists fact-checked unsubstantiated claims and refrained from amplifying misinformation in their coverage. However, the potential impact of social media activity upon events on the ground should not be overstated. There were no incidents of sectarian violence in these years directly attributed to false information shared online.
The scale of online misinformation and disinformation surrounding contentious parades in Northern Ireland remains difficult to ascertain. This case study revolved around the collection and analysis of tweets containing ‘Ardoyne’, ‘low-hanging fruit’ rather than harder to reach content circulated via private Facebook pages and Instant Messaging Apps. Future research should explore whether the apparently uncoordinated efforts to spread false information in this case study were replicated on other platforms.
Online misinformation and disinformation are likely to remain a feature of these parades for as long as they remain contentious. The Ardoyne impasse was symptomatic of the failure of political elites from the two main sectarian blocs to address issues such as controversial parades and protests. Although Northern Ireland is transitioning from conflict, the two mutually exclusive narratives on its constitutional status remain intact and zero-sum perceptions held by both main communities persist. Ethno-sectarian divisions in Northern Ireland have been consolidated rather than confronted by the democratically dysfunctional system of mandatory coalition between the unionist and nationalist parties, which was beset by scandals since its inception. Future research should explore how collective narratives on contentious public demonstrations, like Ardoyne, can be constructed in deeply divided societies transitioning from conflict.
About the author
Dr. Paul Reilly is Senior Lecturer in Social Media and Digital Society at the University of Sheffield.
E-mail: p [dot] j [dot] reilly [at] sheffield [dot] ac [dot] uk
Thanks to Sarah Reilly and First Monday’s reviewers for their feedback and support.
1. Chadwick, et al., 2018, p. 4.
2. Farkas and Schou, 2018, p. 299.
3. Gillespie, 2017, p. 1.
4. Bruns, 2019, p. 76.
5. Bode, et al., 2018, p. 217.
6. Galtung, 1967, p. 17.
7. Schlesinger, et al., 1983, p. 121.
8. For more on this, see https://www.thetimes.co.uk/edition/ireland/shinnerbots-drown-out-unionists-on-social-media-xsv02q9q8, accessed 10 September 2018.
9. Neill, 2003, p. 205.
10. Local Sinn Fein MLA Gerry Kelly was amongst those to criticise the Real IRA. More details can be found at http://web.archive.org/web/20190309100601/https://www.theguardian.com/uk/2009/jul/14/belfast-riots-real-ira-blame, accessed 10 August 2014.
11. H. McDonald, 2013, “Northern Ireland violence leaves 70 police officers injured,” Guardian (16 July).
12. A. Rutherford, 2014, “Implement Belfast Telegraph’s Ardoyne parade proposal, unionists urge,” Belfast Telegraph (11 July).
13. More information on these legal challenges can be found at http://web.archive.org/web/20190309100836/https://www.bbc.co.uk/news/uk-northern-ireland-28264670, accessed 10 August 2014.
14. The original image was available at http://www.demotix.com/news/2246569/protestants-prepare-11th-night-bonfires-northern-ireland#media-2246528, originally accessed 10 October 2014, no longer available.
15. For more on the decision, see http://web.archive.org/web/20190309103808/https://www.belfastlive.co.uk/news/belfast-news/north-belfast-parade-barred-again-9596250, accessed 10 August 2018.
16. This prompted an angry response from North Belfast MP Nigel Dodds. For more, see http://web.archive.org/web/20190309103926/https://www.newsletter.co.uk/news/ardoyne-residents-to-defy-parades-commission-restrictions-1-6845828, accessed 10 August 2018.
17. The Chief Constable denied these rumours and expressed his concerns for the injured PSNI officers. For more, see http://web.archive.org/web/20190309103808/https://www.belfastlive.co.uk/news/belfast-news/north-belfast-parade-barred-again-9596250, accessed 10 August 2018.
18. The 61-year-old was charged with two incidents of attempted murder. For more, see http://web.archive.org/web/20190309104146/https://www.theguardian.com/uk-news/2015/jul/14/attempted-murder-car-belfast-crash-marching-season, accessed 10 August 2018.
19. Villiers expressed her concern for the injured officers in the statement. For more, see http://web.archive.org/web/20190309104347/https://www.irishtimes.com/news/politics/orange-order-parades-all-sides-condemn-disgraceful-violence-1.2284398, accessed 10 August 2018.
20. Ethics approval was granted by the host institution prior to data collection.
21. Mark Simpson (@BBCMarkSimpson. “At a cross-community sleep-out in Ardoyne, Catholics & Protestants found what unites them — breakfast bacon.” 27 June 2014, 8.59 am. tweet.
22. Mark Simpson (@BBCMarkSimpson). “Peaceful start to 12th July marches. No trouble as Orange Order parade passes Ardoyne shops.” 12 July 2014, 7.41 am. tweet.
23. Kevin Sharkey (tv_kevinsharkey) “Trouble in #Ardoyne.” 12 July 2014, 8.53 pm. tweet.
24. David Blevins (@skydavidblevins).“BREAKING: Water canon deployed after police officer inured during disorder at Ardoyne flashpoint, North Belfast.” 13 July 2015, 9.14 pm. tweet.
25. Allison Morris (@AllisonMorris1) (“One man was arrested at shop fronts but apart from that fairly hassle free albeit wet morning #Ardoyne,” 13 July 2015, 8.42 am. tweet.
26. West Belfast UPRG (WestBelfastUPRG) “PSNI moving in against DISSIDENT SCUM at Ardoyne !! Several arrests being made !!” 13 July 2015. 9.10 am. tweet.
27. West Belfast UPRG (@WestBelfastUPRG) “Republicans have overturned a car at Ardoyne shopfronts !! http://t.co/AhfXLMSQZ6” 13 July 2015, 7.22 pm. tweet.
28. Victoria Stevely (@vstevely_utv) “Police clarify 16-year-old girl injured in Ardoyne crash — male driver of overturned car arrested at scene: http://t.co/9a2GhDSs9S,” 13 July 2015, 8.02 pm. tweet.
29. Rodney Edwards (@rodneyedwards) “Reports/tweets stating @ChiefConPSNI has been injured at Ardoyne NOT true. George Hamilton tells me he hasn’t been at Ardoyne since 7 am,” 13 July 2015, 10.50 pm. tweet.
30. Ofcom, 2017, p. 10.
31. A local accommodation was reached in July 2017, with the Orange Order agreeing not to return home via their traditional route and the residents not protesting on the outward leg. For more on this, see https://www.irishnews.com/news/northernirelandnews/2017/07/13/news/ardoyne-parade-passes-without-protest-for-first-time-in-almost-two-decades-1083148/, accessed 20 April 2019.
H. Allcott and M. Gentzkow, 2017. “Social media and fake news in the 2016 election,” Journal of Economic Perspectives, volume 31, number 2, pp. 211–236.
doi: https://doi.org/10.1257/jep.31.2.211, accessed 11 June 2021.
E. Bakshy, S. Messing, and L.A. Adamic, 2015. “Exposure to ideologically diverse news and opinion on Facebook,” Science, volume 348, number 6239 (5 June), pp. 1,130–1,132.
doi: https://doi.org/10.1126/science.aaa1160, accessed 11 June 2021.
V. Braun and V. Clarke, 2013. Successful qualitative research: A practical guide for beginners. London: Sage.
E.L. Briant, 2015. Propaganda and counter-terrorism: Strategies for global change. Manchester: Manchester University Press.
doi: https://doi.org/10.7765/9781847799630, accessed 11 June 2021.
A. Bruns, 2019. Are filter bubbles real? Cambridge: Polity Press.
A. Bruns, J. Burgess, K. Crawford and F. Shaw, 2012. “#qldfloods and @QPSMedia: Crisis communication on Twitter in the 2011 South East Queensland floods,” ARC Centre of Excellence for Creative Industries and Innovation, at https://www.oecd.org/governance/risk/Crisis%20Communication%20on%20Twitter%20in%20the%202011%20South%20East%20Queensland%20Floods.pdf, accessed 11 June 2021.
S. Buetow, 2010. “Thematic analysis and its reconceptualization as ‘saliency analysis’,” Journal of Health Services Research & Policy, volume 15, number 2, pp. 123–125.
doi: https://doi.org/10.1258/jhsrp.2009.009081, accessed 11 June 2021.
J. Burgess and A. Bruns, 2015. “Easy data, hard data: The politics and pragmatics of Twitter research after the computational turn,” In: G. Langlois, J. Redden and G. Elmer (editors). Compromised data: From social media to big data. London: Bloomsbury Academic, pp. 93–111.
doi: https://doi.org/10.5040/9781501306549.0010, accessed 11 June 2021.
A. Chadwick, C. Vaccari and B. O’Loughlin, 2018. “Do tabloids poison the well of social media? Explaining democratically dysfunctional news sharing,” New Media & Society, volume 20, number 11, pp. 4,255–4,274.
doi: https://doi.org/10.1177/1461444818769689, accessed 11 June 2021.
L. Curtis, 1996. “Republican publicity,” In: B. Rolston and D. Miller (editors). War and words: The Northern Ireland media reader. Belfast: Beyond the Pale Publications, pp. 197–207.
A. Edwards and C. McGrattan, 2010. The Northern Ireland conflict: A beginner’s guide. London: Oneworld Publications.
N. Emerson, 2014. “Just whose side are the Shinnerbots on?” Irish News (30 October), at http://www.irishnews.com/opinion/2014/10/30/news/just-whose-side-are-the-shinnerbots-on--106590/, accessed 10 September 2018.
J. Farkas and J. Schou, 2018. “Fake news as a floating signifier: Hegemony, antagonism and the politics of falsehood,” Javnost — The Public, volume 25, number 3, pp. 298–314.
doi: https://doi.org/10.1080/13183222.2018.1463047, accessed 11 June 2021.
D.J. Flynn, B. Nyhan, and J. Reifler, 2017. “The nature and origins of misperceptions: Understanding false and unsupported beliefs about politics,” Political Psychology, volume 38, number S1, pp. 127–150.
doi: https://doi.org/10.1111/pops.12394, accessed 11 June 2021.
J. Galtung, 1967. “Theories of peace: A synthetic approach to peace thinking,” International Peace Research Institute, at https://www.transcend.org/files/Galtung_Book_unpub_Theories_of_Peace_-_A_Synthetic_Approach_to_Peace_Thinking_1967.pdf, accessed 11 June 2021.
D. Ghosh and B. Scott, 2018. “#Digitaldeceit: The technologies behind precision propaganda on the Internet,” New America (23 January), at https://www.newamerica.org/pit/policy-papers/digitaldeceit/, accessed 11 June 2021.
T. Gillespie, 2017. “Governance of and by platforms,” In: J. Burgess, A. Marwick and T. Poell (editors). Sage handbook of social media. London: Sage, pp. 254–278.
doi: http://dx.doi.org/10.4135/9781473984066.n15, accessed 11 June 2021.
A. Gupta, H. Lamba, P. Kumaraguru and A. Joshi, 2013. “Faking Sandy: Characterizing and identifying fake images on Twitter during Hurricane Sandy,” WWW ’13 Companion: Proceedings of the 22nd International Conference on World Wide Web, pp. 729–736.
doi: https://doi.org/10.1145/2487788.2488033, accessed 11 June 2021.
J. Habgood-Coote, 2019. “Stop talking about fake news!” Inquiry, volume 62, numbers 9–10, pp. 1,033–1,065.
doi: https://doi.org/10.1080/0020174X.2018.1508363, accessed 11 June 2021.
K. Hayward and M. Komarova, 2014. “The limits of local accommodation: Why contentious events remain prone to conflict in Northern Ireland,” Studies in Conflict & Terrorism, volume 37, number 9, pp. 777-791.
doi: https://doi.org/10.1080/1057610X.2014.931214, accessed 11 June 2021.
E. Katz and P.F. Lazarsfeld, 1955. Personal influence: The part played by people in the flow of mass communications. Glencoe, Ill.: Free Press.
G. Lotan, E. Graeff, M. Ananny, D. Gaffney, I. Pearce and d. boyd, 2011. “The revolutions were tweeted: Information flows during the 2011 Tunisian and Egyptian revolutions,” International Journal of Communication, volume 5, pp. 1,375–1,405, and at https://ijoc.org/index.php/ijoc/article/view/1246/613, accessed 11 June 2021.
G. McLaughlin and S. Baker, 2010. The propaganda of peace: The role of the media and culture in the Northern Ireland peace process. Bristol: Intellect.
M. Mendoza, B. Poblete and C. Castillo, 2010. “Twitter under crisis: Can we trust what we RT?” SOMA ’10: Proceedings of the First Workshop on Social Media Analytics, pp. 71–79.
doi: https://doi.org/10.1145/1964858.1964869, accessed 11 June 2021.
W.J.V. Neill, 2003. Urban planning and cultural identity. London: Routledge.
doi: https://doi.org/10.4324/9780203402245, accessed 11 June 2021.
Ofcom, 2017. “Internet use and attitudes,” 2017 Metrics Bulletin (3 August), at https://www.ofcom.org.uk/__data/assets/pdf_file/0018/105507/internet-use-attitudes-bulletin-2017.pdf, accessed 11 June 2021.
N. Pang and J. Ng, 2017. “Misinformation in a riot: A two-step flow view,” Online Information Review, volume 41, number 4, pp. 438–453.
doi: https://doi.org/10.1108/OIR-09-2015-0297, accessed 11 June 2021.
R. Procter, J. Crump, S. Karstedt, A. Voss and M. Cantijoch, 2013. “Reading the riots: What were the police doing on Twitter?” Policing and Society, volume 23, number 4, pp. 413–436.
doi: https://doi.org/10.1080/10439463.2013.780223, accessed 11 June 2021.
P. Reilly, 2016. “Tweeting for peace? Twitter and the Ardoyne parade dispute, July 2014,” First Monday, volume 21, number 11, at https://firstmonday.org/article/view/6996/5649, accessed 11 June 2021.
doi: https://doi.org/10.5210/fm.v21i11.6996, accessed 11 June 2021.
P. Reilly, 2010. Framing the Troubles online: Northern Irish political groups and website strategy. Manchester: Manchester University Press.
P. Reilly and F. Trevisan, 2016. “Researching protest on Facebook: developing an ethical stance for the study of Northern Irish flag protest pages,” Information, Communication & Society, volume 19, number 3, pp. 419–435.
doi: https://doi.org/10.1080/1369118X.2015.1104373, accessed 11 June 2021.
P. Schlesinger, G. Murdock and P. Elliott, 1983. Televising “terrorism”: Political violence in popular culture. London: Comedia.
C. Shao, G.L. Ciampaglia, O. Varol, K-C. Yang, A. Flammini and F. Menczer, 2018. “The spread of low-credibility content by social bots,” Nature Communications, volume 9, article number 4787 (20 November).
doi: https://doi.org/10.1038/s41467-018-06930-7, accessed 11 June 2021.
J. Shin and K. Thorson, 2017. “Partisan selective sharing: The biased diffusion of fact‐checking messages on social media,” Journal of Communication, volume 67, number 2, pp. 233–255.
doi: https://doi.org/10.1111/jcom.12284, accessed 11 June 2021.
S. Stanley, 2017. “Misinformation and hate speech in Myanmar,” First Draft (16 May), at https://firstdraftnews.org/misinformation-myanmar/, accessed 11 June 2021.
K. Starbird, J. Maddock, M. Orand, P. Achterman and R.M. Mason, 2014. “Rumors, false flags, and digital vigilantes: Misinformation on Twitter after the 2013 Boston Marathon bombing,” iConference Proceedings, pp. 654–662.
doi: https://doi.org/10.9776/14308, accessed 11 June 2021.
E.C. Tandoc, Z.W. Lim and R. Ling, 2018. “Defining ‘fake news’: A typology of scholarly definitions,” Digital Journalism, volume 6, number 2, pp. 137–153.
doi: https://doi.org/10.1080/21670811.2017.1360143, accessed 11 June 2021.
A. Taub and M. Fisher, 2018. “Where countries are tinderboxes and Facebook is a match,” New York Times (21 April), at https://www.nytimes.com/2018/04/21/world/asia/facebook-sri-lanka-riots.html, accessed 11 June 2021.
E. Thorson, 2016. “Belief echoes: The persistent effects of corrected misinformation,” Political Communication, volume 33, number 3, pp. 460–480.
doi: https://doi.org/10.1080/10584609.2015.1102187, accessed 11 June 2021.
C. Wardle, 2017. “Fake news. It’s complicated,” First Draft (16 February), at https://firstdraftnews.com:443/fake-news-complicated/, accessed 11 June 2021.
Received 7 October 2019; revised 8 May 2020; accepted 29 June 2020.
Copyright © 2021, Paul Reilly. All Rights Reserved.
Countering misinformation and disinformation during contentious episodes in a divided society: Tweeting the 2014 and 2015 Ardoyne parade disputes
by Paul Reilly.
First Monday, Volume 26, Number 7 - 5 July 2021