Click here to read GNET's latest report ‘Fogging’ and ‘Flooding’: Countering Extremist Mis/Disinformation After Terror Attacks

Censoring Extremism: Impact of Takedowns on Islamic State Visuals

Censoring Extremism: Impact of Takedowns on Islamic State Visuals
26th November 2021 Kayla McMinimy
In Insights

Non-state extremist groups like Islamic State (IS) and al-Qaeda utilise various online channels to disseminate strategic information to members, sympathisers, and potential recruits. Extremist groups disseminate official and user-generated content online to propagate their ideological goals, recruit, radicalise, and reinforce narratives. Scholars assessing the reactions to online account disruptions recognise that groups targeted for takedowns respond by migrating to more secure platforms such as Telegram emerging as a vital communication channel for IS after deplatforming from more popular sites like Twitter and Facebook. More recently, Telegram undertook an extensive effort to censor IS-related channels and accounts that slowed the online activity of the group considerably. Yet, as the group’s recent shift to alternate networks, IS is still working to rebuild its Telegram networks, albeit in a less robust form.

While researchers have carefully examined online migration and access patterns of extremist groups after deplatforming campaigns, any corresponding changes in the messaging content appearing in their official publications remain unexplored. To examine the content of official IS media messaging in terms of priming, agenda-setting, and framing at a time of severe channel disruption we asked the following: If the number of visual images in al-Naba will decline after a major suspension of IS’s Telegram networks; and if the media content of visual images in al-Naba will change after major suspensions of IS’s online networks.

Our study’s timeframe spanned the 6-month period before and after the Telegram takedown to account for the ideological war over time that IS reinforces through its published visual messaging. Our image dataset includes 100 percent of the visual images displayed in al-Naba IS’s only consistently produced newsletter or magazine. Visual messaging serves as a key contributor to agenda-setting, priming, and framing functions of the media. To list some major characteristics of visual images: visual images overcome language and literacy barriers, increase audience attention, recall, and emotional responses, to identity formations and group-based ideologies, elevate the impact of violent actions, and serve as condensation symbols of the group’s key narratives. We analysed 58 issues in total: 29 issues before the takedown (Issues 181–209), and 29 issues following the takedown (Issues 210–238). In order to examine whether shifts in the visual messaging of al-Naba occurred after Europol’s disruptions of IS’s online message distribution, we utilised the codebook previously used in studies looking at visuals in different IS magazines.

Our coding categories fell into three overarching thematic clusters: the first related to militancy and its outcomes; non-militancy activities; and, presentational elements. We conducted a statistical analysis, Loess regression over the cluster groups to minimise any alternative explanations in the longitudinal trends. To examine any shifts in IS’s visual strategy within the three clusters after the takedown of the group’s accounts and channels, we conducted a chi-square analysis of all coding categories at the 0.01 level of significance and we used Fisher’s Exact test for cells with less than 5 (please refer study for counts and residuals of each category).

The answer to our first question was no. There were 260 images in the 28 issues before the takedown and 290 images in the 28 issues following the takedown. This showed that the number of images in al-Naba increased after Telegram’s takedown of IS networks by 12%. Our second question was yes in all categories in the militancy cluster (military role, military outcome, cause of death, and about to die), one non-militancy cluster (infrastructure), and one presentational cluster (viewer distance). To further understand these changes, we went back to the priming and agenda-setting scholarship.

Priming: al-Naba audiences have been primed to expect a weekly newsletter and though faced with aggressive challenges to their communication networks, al-Naba continued production and distribution while increasing the quantity of images published. Thus, the fact that the magazine continued uninterrupted after a major takedown by the preferred platform of communication of the group reinforced the expectations of the audience.

Agenda-setting: al-Naba also altered the agenda-setting priorities incorporated into its visual messaging as the newsletter placed greater emphasis on the perspective that coalition forces posed an ongoing threat to the region. The group more than quadrupled its displays of non-IS militants, indicating a focus on the enemy and their continued threat. Regardless of the actual level of active coalition forces deployed in the region, repeated visual reminders of force presence heighten audience perceptions of the dangers posed by enemies.

Framing: Additionally, the publication more than doubled its number of photographs displaying death carried out by IS fighters. In doing so, IS shifted the burden of personal sacrifice in their virtual elimination of photographs of its own martyrs while increasing the number of images of dead bodies. Presumed deaths fell by 17 percent and in conjunction with the increased number of dead bodies visually document successful outcomes against enemies rather than leaving outcomes to the audience’s imagination. When enemies were photographed, the visual components of the images worked to dehumanise enemies as indicated by the higher than expected number of images showing militants at a social and public distance rather than closer shots which focus on unique characteristics of individual people.

Presentational display of visual credibility and dominance indicators did not change demonstrating continued implicit assurances the group is resilient and powerful. Images containing leaders and state-building functions also remained consistent before and after the take-down. The group highlights the ongoing need for IS members to remain alert regarding enemy leaders even during contested message distribution networks. Similarly, the reliability of state-building visuals reinforces their long-term goal of building a caliphate that will not be communicatively detoured based on temporary online attacks.

Take-Aways

The study at large illustrated that in the aftermath of the major Telegram takedown of its significant accounts, and thus channels of information distribution to the group’s global audience, IS used their visuals to reinforce the group’s continued viability, emphasize its public agenda, and frame the group’s enemies all to ensure their continuity. The vehicle of changing certain visual media strategies while holding others constant underscores the nuanced intersections of channel and message in the mediated communications of extremist groups whether conscious and unconsciously conceived.

The consistent adherence to its standard publication and distribution schedule of al-Naba reinforced the primed expectation of the group’s ongoing online and offline media presence and belied any premature assumptions of the group’s total defeat. The increased displays of dead enemies work to project on the ground physical strength and resilience in the face of online existential attacks. The stable frequency of images highlighting enemy leaders in the newsletter reinforced the ongoing need for IS members to remain alert even during periods of disruption to its online communications.

The findings in the study also help to further understand how groups like IS visually react to online and offline challenges to their continued existence. Certain shifts in visual strategy reappear regardless of whether the challenge occurs online or offline. In trying to understand the effects of censorship on such groups it is crucial to examine their publication content post and prior.

We would like to thank our co-authors Carol K. Winkler and Monerah Almahmoud, a women-led research team.

Kayla McMinimy is currently a Craigie Fellow and formerly a Transcultural Violence and Conflict Fellow at Georgia State University. Her work focuses on digital, visual, and multimodal communication with attention to media, politics, and mis/disinformation. She has published in Behavioral Sciences of Terrorism and Political Aggression and Terrorism and Political Violence.

Ayse D. Lokmanoglu is a Post-Doctoral Fellow at the Center for Communication & Public Policy at Northwestern University. Her work focuses on malign digital campaigns (hate speech, extremism, disinformation) and utilizes computational, quantitative, and qualitative approaches to examine harmful narratives and digital messaging. She has published in journals including Studies in Conflict and Terrorism, Terrorism and Political Violence, Cambridge Review of International Affairs, Journal of Health Communication, and in edited volumes including Islamist Approach to Governance, Radicalization and Counter-Radicalization, and others.