There was a recent escalation of unrest in Leicester between Indian and Pakistani fans following the Asia Cup cricket match on August 28th, 2022, which turned violent. The current events suggest an unprecedented escalation – the rivalry between India and Pakistan was portrayed as a Hindu aggression towards the Muslim community, driven by malicious online narratives and social media activities. Interestingly, according to the 2011 census, Leicester is 13% Muslim, 12.3% Hindu, 22.3% of Indian origin, and 1.9% of Pakistani origin. This is important, as Pakistani – origin population is not a majority. Still, this malicious narrative gathered so much momentum as it got portrayed as a communal aggression instead of a nationalistic rivalry. This article chronicles the events from August 27th to September 19th. It analyzes cyber social activities through social media platforms such as Twitter, YouTube, TikTok, Instagram, etc., to understand the connection between online and real-world activities.
On August 28th, Indian cricket fans gathered in Belgrave, Leicester, to celebrate India’s win over Pakistan in the Asia Cup Series. On August 29th, Twitter user @tragicbud’ tweeted one of the earliest malicious narratives that had a high engagement, including a short video that went viral. Since deleted, the narrative tied violence to nefarious ‘Nazi’ and ‘Hindutva’ ideology and nationalist objectives, as opposed to a nationalistic rivalry. The video shows Indian fans carrying the Indian flag and shouting ‘Pakistan Murdabad’ accompanied by brawls between Indian and Pakistani fans. Organizations of repute were found in the reply network of the tweet, such as the Islamophobia Response Unit, and MAMA UK. These responses suggest a prima facie assumption that Islamophobia was the catalyst for these fistfights. Majid Freeman (20.8k+ followers) quote tweet on August 29th, which was further amplified on Twitter, claiming that ‘Pakistan Murdabad’ chants were religiously motivated, seems to give a significant lift to this narrative. Majid Freeman is a British Muslim former aid worker who has voiced support Al-Qaeda, and ISIS on his social media. British Pakistani networks used Cyber swarming – tactically alarm or recruit followers to deploy en masse in both cyber simultaneously and, potentially, real-world domains – to call for a response. These were followed by tweets exemplifying these calls and portraying the Hindu community as a dangerous threat, which has since been deleted. These tweets consistently ranked in the most engaged tweets between August 30th and September 4th.
Post-September 4th marks a period of communal clashes and rampant misinformation amplification, characterizing British Hindus as an aggressive, hyper-nationalist, and fascist threat. A summary of the critical events, correlating social media activity and the violence that followed, is outlined in the section (Timeline Visualization of Key Events). Exhorted the users to violent actions against the Hindu community through these conspiracies. This misinformation was extremely penetrative, amplified by Al-Jazeera and Indian Outlet ‘The Wire’. There was also misinformation targeting the British Muslim community, calling for reciprocal violence in self-defense.
Anti-Muslim and anti-Hindu disinformation are distinguished in two critical ways. First, anti-Muslim disinformation primarily focused on blame attribution, while anti-Hindu disinformation was highly specific, identifying actors, alleged crimes, and mobilizing targeted actions. Secondly, spread anti-Muslim disinformation almost exclusively through and within a robust, dense network of users in India. Anti-Hindu disinformation spread through a more diffuse Islamist network, penetrating more diverse audiences, including several media outlets.
Analyzed online social media activities between the period of August 27th to September 19th in the following ways:
1. Twitter chatter analysis:
Time series analysis: Strong reactivity on the platform is observed. Mentions of ‘Hindus’ outstripped comments of ‘Muslims’ by nearly 40% during peak online activity.
Natural Language Processing: Purpose: Analyze semantic themes (topic networks) associated with the Hindu and Muslim communities using multiple NLP tools. ‘Muslim’ word usage mainly was around themes of protest, confrontation, and outrage and did not support the notion that Muslims were depicted with ideological or religious threads. The network for ‘Hindus’ appears to describe terror, threat, and dominance as critical themes. Terms such as “abuse,” “Rampage,” “terrorize/terrorize,” and “unleash” cluster closely with Hindus. In addition, unlike the Muslim topic network, ethnic slurs appear with accusations of “genocide”.
Analysis of Hindu/Muslim Twitter spheres: Purpose: Map connections between social media users within Hindu and Muslim Twitter spheres to understand the interaction patterns within each community using a ‘twarc network’. While Hindu users engaged the most with police among all social media accounts, Muslim networks revolve mostly around Majid Freeman and 5PillarsUK (online Muslim community).
Analysis of religious slurs: NCRI found a sharp rise in the use of Hindu phobic slurs, particularly on September 20th, as an Islamist mob surrounded the Durga temple in Birmingham. There was a marginal increase in Islamophobic retorts during this period.
Tweet Sentiment analysis: Purpose: Analyze Toxicity, Insult, Identity attack (blame), and threat in tweets through time series analysis. Identity attacks were most prominent. A study of the top 500 most engaged tweets with an identity attack found that tweets originating from the U.K. depict both Hindus and Muslims as aggressors of the violence, with slightly more blame cast on Hindus. However, tweets originating from India disproportionately blame Muslims, suggesting a sphere of reactivity and amplification of responsibility by a network of self-identified Indian accounts. While topic network data suggests the portrayal of Hindus as aggressors, this data means a highly orchestrated echo chamber from India amplifying tweets solely blaming Muslims, with evidence of bot-like inauthentic activities. It suggested these events fed into ethnic reactivity and fueled hyper-sensational nationalism.
Analysis of Bot Activities: Evidence of bot-like accounts that disseminated anti-Hindu and anti-Muslim messaging was found. These tropes reinforce ancient ethnic hatreds and undermine inter-community trust. Focused Anti-Muslim bot accounts on socio-economic indicators to differentiate Hindus and Muslims. The bots seek to project Hindus as a model minority in contrast to Muslims. In addition, the tweets suggest that Muslims are part of a global Al Qaeda and ISIS conspiracy, generalizing the entire Muslim community as bloodthirsty and criminal. A hallmark of anti-Hindu bot-like accounts was the creation of a binary between “Hindus” and “Hindutva extremists,” where the latter is alleged to have inflicted violence on the Leicester community. They also emphasize that the Hindu and Muslim communities had lived peacefully before the arrival of “extremist Hindus”, alleging that the Hindus involved in the Leicester violence have extremist political views, described as Nazist.
Hate Speech Analysis: Purpose: Quantify incitement to violence online using hate speech detection AI models. 70% of violence incited on Twitter was against Hindus, and 30% against Muslims.
These findings suggest that dominance and conspiracy narratives circulated by British-Pakistani networks are mainstreamed as primary depictions of Hindus on Twitter. Online histories mischaracterized Hindus as instigators and Muslims as responding to aggression. NCRI assesses that these narratives were deployed as a pretext for violence.
2. YouTube and TikTok Data Analysis:
Classification Analysis: Purpose: Classify which community was portrayed as the aggressor and the victim. From the 24 most viral (views > 20,000) TikTok videos, 17 portray Hindus as the aggressor, 3 Muslims as the aggressors and the remaining four do not specify. Similarly, out of the 14 most viral (views >10,000) videos on YouTube before the violence, 5 portray Hindus as the aggressors, 4 Muslims as aggressors, and the remaining five do not assign blame.
3. Other Online Activities:
OSINT reveals that both anti-Hindu and anti-Muslim memes, caricatures, and messaging have also been rising. Anti-Muslim caricatures actively dehumanize Muslims and fear monger about the potential for an Islamic invasion. At the same time, anti-Hindu memes suggest Hindus were “deserving” of the violence inflicted on them based on their inherently “fascist” traits.
To conclude, sustained violence in Leicester did not persist through organic outrage. Disinformation about Hindus portrayed them as bloodthirsty and genocidal. Mainstream media platforms, including the BBC, The Guardian, and The New York Times, failed to perform due diligence on Majid Freeman and amplified the voice of a conspiracy theorist and extremist sympathizer, a central agitator in the events of Leicester. His malicious narrative was disseminated by a network of groups mobilizing on-the-ground activities. The findings portray a worrying system of organized violence that allows for runaway escalation outside the reach of law enforcement, lawmakers, and mainstream media. Violence engendered through social media platforms regularly reaches episodic spillovers of inter-group conflict that now threaten vulnerable communities worldwide. Reacting collaboratively to false narratives, allegations, rumors, or biases propagated by social media may prevent the next outbreak of sectarian violence and or conflict instigated by outside agitators.
References: Cyber Social Swarming Precedes Real World Riots in Leicester: How Social Media Became Weapon for Violence, Network Contagion Research Research Institution, Rutgers University