“Please. Do. Not. Share. Videos. Share. Cats.”: Counteracting Terrorist and Violent Extremist Content on Twitter during Terrorist Attacks

Abstract

Obtaining accurate information from social media during a crisis can be difficult, but should all information really be disseminated? Social media platforms actively filter out terrorist and violent extremist content (TVEC), but how are users themselves counteracting its spread? This paper aims to connect the research on media events with studies currently being conducted in information science and digital media research through a case study of tweets during the Vienna terror attack in late 2020. These tweets were manually coded in accordance with Braun and Clarke’s reflexive thematic analysis. This study shows that during the 2020 Vienna attack, GIFs shared on Twitter served three functions: amplification, personalisation and ethical practice. The paper ends with a discussion on the ways cats may function as a countermeasure against the prevalence of TVEC on social media during terrorist attacks and the implications of such countermeasures.

1. Introduction

While access to information is highly desirable during a terrorist attack, the ethical aspects of circulation have been debated. While relevant updates on the situation are clearly vital during an ongoing crisis, there are times when the sensitive nature of content (Mäenpää 2021)—especially for the families of victims—precludes publication in the media. On social media platforms, content moderation performs the function of controlling the sharing of information. User-generated, eye-witness content is incorporated into the media machinery as gripping news updates, something that presents a series of ethical dilemmas (cf. Mortensen 2011). Such content may be horrific, especially when it depicts a terrorist attack. In an ideal world, given that social media platforms are working actively to remove terrorist and violent extremist content (TVEC), this content would be filtered out in various ways. Twitter’s (2019) user guidelines include a “Sensitive media policy” that prohibits users from sharing graphic content depicting violent crimes or accidents and serious physical harm, including visible wounds. However, during the exponential dissemination of information in the wake of incidents such as terrorist attacks, it becomes difficult to moderate or even control such content.

Lately, there has been some pushback against the immediacy of information sharing in this fast-paced media environment from social media users themselves, centred around questions of what content is shared and how. This article explores how one common part of internet vernacular is being incorporated into this critique—namely, cats, especially when mediated via GIFs. Cats play a central role in online culture and social interaction; as Eppink (2016) puts it, “they rule the internet”. This paper examines how cat GIFs were used on Twitter during the terrorist attack in Vienna on 2 November 2020. Its purpose is to investigate these user practices in relation to TVEC in the wake of this specific terrorist attack and how the privacy of victims may be protected by strategically amplifying (Donovan and Boyd 2021) other types of content, such as cat GIFs. This work is dependent on a critical view of terrorism discourse, where the social construction of our view of what constitutes terrorism or not is politically grounded (Jackson et al. 2011; Homolar and Rodríguez-Merino 2019). The networked audiences of terrorist attacks might share this type of content as an ethical and strategic approach to tweeting during a terrorist attack. This has the potential to tell us more about the status of information flow in socially mediated spaces during ongoing disruptive events.

This paper is structured as follows. First, we present the research on information acceleration and the liveness of terrorist attacks in hybrid media and the prevalence and moderation of TVEC, ending on the idea of slowing down news streams and strategic amplification. Second, we describe the role of cats on the internet, specifically during terrorist attacks. This is followed by a description of the case, as well as the background to the 2020 Vienna attack, the data collection process and, finally, the ethical aspects of this research. This is followed by the analysis of three themes emerging from an inductive approach. The paper ends with a discussion of the ways cat GIFs may be used as a countermeasure against the prevalence of TVEC on social media during terrorist attacks and the real-life implications of doing so.

2. Previous Research

2.1. Information Acceleration: The Liveness of Terrorist Attacks

In the immediate aftermath of a terrorist attack, the information most likely to be sought by first responders, the news media and affected citizens may be lacking in real time. While such information gaps are filled jointly by organisations and the networked public (Liu et al. 2016), users may actively disseminate inaccurate or misleading information on social media platforms. Webb and Jirotka (2016) call this phenomenon a “digital wildfire” that once ignited, is difficult to fight and contain. Because of the rapid-fire tweeting which may develop during specific types of events, such as elections, Burgess and Baym (2020) have identified the growing political disproportionality of information flows due to the ease with which users can retweet content at the push of a button. They propose that during specific events, users should be forced to use the Quote Tweet function instead of Retweet, allowing them to add their own comments. This should suppress the overload of specific hashtags on Twitter concerning a given terrorist attack.

Terrorist attacks are mediated in a myriad of ways, not solely through traditional media outlets and social media. For example, 9/11 was included as a special episode on the television series The West Wing just three weeks after the attack and was framed as a pedagogical example meant to teach viewers how to understand the terrorist attacks (Holland 2011). As information technologies become faster, so information concerning disruptive events such as terrorist attacks becomes available in multiple spaces simultaneously. Because of this, media events are both decentralised and adapted to a myriad of actors and messages (Sumiala et al. 2016, p. 100). This immediacy also creates a stream of information from news organisations as they continuously break news in the aftermath of a terror event. These continuous updates create a sensation of “liveness” in ongoing reporting (Sumiala et al. 2019) and an imagined continuity of events in real time and by a number of actors. While information is highly prized in any crisis, the ethical aspects of sharing information is a central concern of journalistic practice in traditional media outlets. For example, visual news coverage of airplane crashes is filtered through various chains of ethical decision-making (Mäenpää 2021) that may well deem certain images too graphic to be published in the mass media, even if they might be informative for a certain readership. Specific controversies have especially shaped these ethical practices in news coverage. For example, wrongful descriptions printed by The Sun after the Hillsborough disaster quickly led to a boycott of the publication (Foos and Biscof 2022). Images from such disasters have also led to widespread critique. Such was the case of the man falling from the World Trade Center during 9/11, later referred to as “The Falling Man”. When printed in largescale publications such as the Chicago Tribune, Newsweek and Boston Globe, it received critique for being too explicit, and “given the reluctance in the press about showing human gore” (Zelizer 2004, p. 174), most of the images of victims were pulled from the news outlets.

There are also other risks associated with these accelerated news streams. According to Vosoughi et al. (2018), false news tends to be disseminated significantly farther, faster, deeper and more broadly than the truth. In the digital wildfire that may break out after a terrorist attack, networked publics themselves are resistant to the idea that more is better when it comes to information and may well adopt strategies to counter these forms of acceleration. Not only are social media platforms trying to eradicate terrorist and violent extremist content by improving their content moderation, but socially mediated ways of sharing information in violent crises may also be changing simultaneously.

2.2. Content Moderation of Terrorist and Violent Extremist Content (TVEC)

Most often, social media platforms actively filter out terrorist and violent extremist content (TVEC), not least because the platforms themselves are used by terrorists and violent extremists “for recruitment, dissemination of propaganda, communication and mobilisation” (OECD 2021, p. 6). This has put pressure on content sharing services to be transparent about their policies and reporting of such content. This was brought home after the Christchurch mosque shootings in 2019, when the gunman live-streamed his attacks against two mosques in Christchurch, New Zealand. This attack demonstrated the very real threat of TVEC being shared on platforms despite the measures taken to remove it. Two months after the shootings, heads of state and leaders from the tech sector were invited to The Christchurch Call to Action Summit (also called The Christchurch Call) by New Zealand’s Prime Minister Jacinda Ardern and French President Emmanuel Macron. The Christchurch Call was a joint commitment by governments and tech companies to eliminate TVEC from online platforms. In the call, it was stressed that:

All action on this issue must be consistent with principles of a free, open and secure internet, without compromising human rights and fundamental freedoms, including freedom of expression. It must also recognise the internet’s ability to act as a force for good, including by promoting innovation and economic development and fostering inclusive societies.

(Christchurch Call 2019, p. 1)

However, not all social media platforms have been performing these moderating practices, for example, Gab, Parler or MeWe. While in an ideal world harmful content would be automatically deleted from these platforms, there are several systematic problems to be overcome. The lack of transparency of commercial content moderation, the labour concerns of those working in the content moderation field, and the democratic issues involved are among the growing concerns expressed by many researchers (see Suzor et al. 2019; Roberts 2019; Gillespie 2018). Since the main focus of this study is on how terrorist and violent extremist content is shared, the following sections will deal with the specific situations surrounding terrorist attacks, and how such disruptive events create their own set of problems with regard to the governance of harmful content on social media platforms.

2.3. Slowing Down and Showing Up: (An Ethical) Strategic Amplification

While accelerated news streams and social media feeds have become the norm, there are incentives to slow these processes down in many different forms. Compared to the slow food movement, the slow media movement has not attempted to counteract the phenomenon itself but instead critiqued how they operate. In The Slow Media Manifesto, co-authored by Köhler et al. (2010), the authors specifically point to the need to use technology in mindful and deliberate ways. While they are critical concerning how tech companies run their businesses, their main focus is on users and how users interact with technology. They state that Slow Media should be progressive and not reactionary, stating that

it is because of the acceleration of multiple areas of life, that islands of deliberate slowness are made possible and essential for survival. Slow Media are not a contradiction to the speed and simultaneousness of Twitter, Blogs or Social Networks but are an attitude and a way of making use of them.

(Köhler et al. 2010, my emphasis)

There seems to be a need for these incentives. Mainstream media has an important role to play in a media landscape prone to the dissemination of misinformation, especially during a crisis such as a terrorist attack. As summarised by Luo (2019), editor of the New Yorker, “the pressures for speed and volume created by the digital age can’t be ignored—but they can be resisted”. Nonetheless, Vostal (2019) has suggested that these slow initiatives (slow food, slow travel and so on) focus heavily on slowness as a commodity, an individual attempt to counter a capitalistic form of acceleration. Although social media platforms are often blamed for the spread of disinformation and the accelerated speed of information, the mainstream media, too, are flawed in many ways. As news moves more quickly and profit is placed at the forefront of media infrastructure, what types of information are likely to be disseminated? In a blog post on the NiemanLab website, in her predictions for journalism in 2022, Zizi Papacharissi states that

the economics of journalism drive the algorithms. Let’s not get distracted and lose sight of the core problem. It’s not just the algorithms that need auditing. It’s also the profit-driven structure of news media that needs a do-over.

(Papacharissi 2021)

In the aftermath of a terrorist attack, there are political implications to an accelerated news stream. It is difficult to fully gauge the potential victims when news, information and photographs are disseminated on social media. Participating in information flows that can cause harm is problematic, even when information is merely shared and amplified, and these actions have potential consequences. In her report The Oxygen of Amplification: Better Practices for Reporting on Extremists, Antagonists, and Manipulators Online, Whitney Phillips (2018) strongly criticises the news media for amplifying hate group messages in particular, writing that “in the process, this coverage added not just oxygen, but rocket fuel to an already-smouldering fire” (Phillips 2018, p. 7). Additionally, as Donovan and Boyd (2021) recognise, the amplification of specific information is never a neutral act, no matter who is behind it. They call for a new editorial approach of “strategic amplification”, which recognises the ethical responsibilities of all parties, including social media users, involved in the flow of information. They state that the potential harm of spreading certain information should be weighed against the benefits and that “the more power a content creator or institution has within the ecosystem, the more obligation that actor has to be conscientious, ethical, and responsible—remaining prepared to be held accountable publicly” (Donovan and Boyd 2021, p. 346). This implies that not only should the news media expect to be held accountable but also social media users themselves.

3. The Internet Vernaculars of Cats

While cat memes gained in popularity online in the mid-2000s, the cultural significance of cats stretches back to at least the 18th century (Sewell and Keralis 2019). The anthropomorphic view of cats as conveyors of human feelings and behaviours makes them the “perfect memetic figure”, on par with the use of emoticons or emojis (Miltner 2014), and what Thibault and Marino (2018) call a “hyper-meme”, featuring “countless variations, iterations, and proliferations”. Although created by individuals, memes function as a viral social phenomenon continually developed by others. People reproduce memes as remixes of the original content in response to different sociocultural environments (Shifman 2013).

Memes in general are specifically grounded in contextualism, their interpretations dependent on cultural and subcultural understandings. Denisova (2019, p. 29) argues that the meaning of memes is derived from “their exploitation of context”. Memes are an ever-changing phenomenon and have developed over time to encompass both audio (see Abidin 2021) and coded language, such as the use of parentheses by specific groups (Tuters and Hagen 2022). GIFs are an animated form of meme, not technically videos but rather a sequence of images that, just as other types of meme, express emotion, humour or political statements.

Memes are not the only way context is collectively shaped and developed on the internet. How ordinary social media users interact and communicate on platforms depends on the social norms that are collectively created in digital spaces, what Gibbs et al. (2015, p. 257) call “platform vernaculars” that are “shared (but not static) conventions and grammars of communication, which emerge from the ongoing interactions between platforms and users”. While this communication may be built on memetic norms and forms, common internet vernacular is shaped by audio-visual and textual conventions that are shaped by different communities and vary over time.

Cats are only one part of the internet and, generally speaking, they constitute one of its better parts. “Animal and pet images play a key role in keeping Internet spaces habitable, light-hearted, and fun, but they are also not immune from broader Internet dynamics and problems, including negativity, toxicity, and so-called garbage”, says Jessica Maddox (2022, p. 3) in her book The Internet is for Cats: Attention, Affect, and Animals in Digital Sociality. In fact, the internet is jam-packed with garbage—including terrorist and extremist violent content. “We cannot talk about the cute, the wholesome, the relief, and the distractions”, Maddox continues, “without also talking about the bad, the ugly, the cumbersome, and what we need relief and distractions from” (Maddox 2022, p. 11). As such, cats cannot be studied in a vacuum, but must be related to the problem with, and prevalence of, both terrorist attacks themselves and terrorist and violent extremist content on social media.

Cats and Terrorism

Sharing cat memes, GIFs and images has become common whenever news of terrorism is shared in digital spaces, especially on Twitter. In their study of the motivations of digital publics for creating and sharing memes on Twitter during the Brussels security lockdown, Jensen et al. (2020) found that their reasons ranged from personal involvement in the subject matter, as an act of resistance, to the practice of creative self-realisation online. While, for some users, cats may represent a general political interpretation of the event itself (as one interviewee in their study stated, “Brussels will land on its feet like a cat”), they may also serve to tone down narratives of fear in times of terror. For example, Stockholm-based social media users turned to Twitter during the Stockholm lorry attack in 2017, mentioning cats, pizza and beer in their collective coordinating efforts to offer housing and lifts to people affected by the shutdown of public transport in the city (Eriksson Krutrök 2018). Such light-hearted responses to terrorism can be used as a means for digital publics to diverge from the dissemination of fearful narratives, often in mainstream media, such as after 9/11 (Altheide 2006). Memes can be a form of political activism against these narratives of fear, or what McCrow-Young and Mortensen (2021) have called “spectacles of fear”, explicitly as terrorists themselves produce these spectacles. Their study of Anonymous’ “Troll ISIS Day” shows how internet communities use humour to combat such spectacles, in what they call “counter-spectacles”. In this way, memes become potential political tools on social media, a countermeasure against terrorist propaganda. Similarly, in this study, cat GIFs are understood as a countermeasure against terrorist and extremist violent content on social media, especially in the interests of protecting victims from further harm.

4. Case Description and Data Collection

On 2 November 2020, a series of shootings occurred in Vienna, Austria. At around 8 PM, as people were enjoying time together in pubs and bars before the imposition of a new COVID-19 lockdown, a lone gunman opened fire in the busy city centre (Deutsche Welle 2020). The attacker was armed with several weapons, including a machete, and was wearing an explosive belt which turned out to be fake. By the time the police killed the attacker, just minutes after the attack had started, he had murdered 4 people and wounded 23 others, 7 of them critically, with both gunshot and stab wounds. The Jihadist group Islamic State (IS) subsequently claimed responsibility for the attack (Reuters 2020).

This study looks specifically at the platform Twitter as a case study of tweeting during terrorist attacks. The one-platform focus of this study does, however, come with its own limitations. Content travels across platforms, as has been studied in several previous studies (for example Baele et al. 2020; Macdonald et al. 2019), especially in relation to the origins and reach of internet memes (Zannettou et al. 2018). Because of this, the actual reach of the cat GIFs researched here is still unknown, and this study may only provide an initial inquiry into their real-life implications.

Before the data collection process could take place, I attempted to gain an awareness of how cats appeared in the Twitter feed1 after the attack. This was first and foremost a digital ethnographic process (Pink et al. 2016), sometimes referred to as netnography (Kozinets 2010) or internet ethnography (Haverinen 2015). In this process, I read, translated and followed comment threads on Twitter for several days. Primarily, this process was based on familiarising myself with the material by explicitly following the conversations where they went. This corresponds to how Sumiala and Tikka (2020) implemented the methodological rule by Marcus (1995) to “follow the thing” in their digital ethnography, where the researcher can “trace different types of actors and messages across a variety of online platforms” (Sumiala and Tikka 2020, p. 47). The “thing” that was followed in this study were the cats themselves, as they were posted on social media, shared and responded to. On an ephemeral level, this work included searching for specific words, hashtags or phrases and scrolling continuously to find instances with cats in these Twitter feeds. This allowed me to direct my attention to using specific hashtags and search words that were highly used at this specific point in time. These hashtags were specifically chosen based on the fieldwork stage of the analysis, where hashtags were identified based on their overall usage during this time. Additionally, in the journalistic reporting carried out in the aftermath of the attack, many of these hashtags were mentioned, adding even more emphasis to their significance during the attack.

While German is predominantly spoken in Austria, I found that much of the material was written in English and French. Since the data contained several languages, I was interested in the proportions of the use of each language in the data (see

Table 1. Proportions of the four different languages in the tweets.

Language	Count
French	504
English	450
German	13
Dutch	3

). Since the words for cats also corresponded with other languages, I found that three of the tweets were written in Dutch.

A total of 452,034 tweets using the hashtags #Wien2020, #Vienna, #Vienne, #ViennaAttack and #0211w2 were extracted from Twitter using Twitter’s v2.0 API for the whole month of November 2020. In this material, the words for cat and kitten in English, French (which was the most used languages found during the initial analysis) and German (since this is the main language spoken in Vienna) were abstracted into a separate dataset containing 970 tweets in which either of the search phrases for cats3 were included. Because of this, the tendency to share or discuss cats may be comparatively small compared to the rest of the tweeting behaviour. Indeed, this is a small proportion of the total tweets in the dataset. However, as this article focuses on understanding the role of this specific type of reaction on Twitter during terrorist attacks, the proportion is not the main reason for conducting this study. Rather, it contextualises the use of memes, GIFs and, especially, cats, during terrorist attacks.

The rationale behind using hashtags as a way of finding and collecting the data is due to the practice of communicating during crises on Twitter by incorporating hashtags as part of a “folksonomy”, a tagging system for information sharing (Vander 2007). In particular, users include hashtags in regard to breaking events, such as terrorist attacks. Hashtags fill a role for structuring the conversation regarding the unfolding of the event and connecting specific pieces of information with others. Hashtags have been commonly used for the purpose of data collection by researchers in a myriad of fields (Gevisa and Giovanni 2022), including crisis communication studies (cf. Starbird and Stamberger 2010) and studies of terrorism, specifically (cf. Sumiala et al. 2018; Eriksson Krutrök 2020).

After compiling tweets mentioning cats into a separate list, I read them one by one and coded them individually in accordance with Braun and Clarke’s (2022, pp. 35–36) reflexive thematic analysis. Thematic analysis was then completed in six phases: (i) familiarising oneself with the material, which in practice means reading (and re-reading) the dataset, and, in this study, translating and decoding the actual words; (ii) the researcher begins to code the material, which means creating and assigning specific code labels to the tweets; (iii) the many codes created are grouped into initial themes (here, “initial” is the key word, as these themes may well change and develop over time); (iv) the themes are developed and reviewed to identify and clarify each theme’s core organising principle; (v) the themes are demarcated and exhaustively analysed, usually by asking oneself “what story does this theme tell?”; and (vi) the writing process begins, during which specific data extracts are interwoven with an analytical narrative.

Ethical Considerations

The tweets analysed in this study have been qualitatively read and coded to find specific themes in the tweeting after the 2020 Vienna attack. While these tweets have been read and coded, the original tweets are not included in this paper. Even though these tweets are published in a public forum on Twitter, I have chosen not to quote them directly, except for tweets from police authorities, following Williams et al. (2017). Instead, I have paraphrased tweets in order to avoid them becoming “googleable”, a measure proposed by Markham (2012) as a way of avoiding readers being able to trace these tweets back to the source. This is especially important when researching political issues (Reilly and Trevisan 2016; Franzke et al. 2019).

5. Analysis

For the purposes of this paper, the sampled tweets have been qualitatively analysed using thematic analysis (Braun and Clarke 2006), with each tweet assigned a specific code depending on its content. At the second stage of analysis, these codes were clustered to form themes. To facilitate the analysis of different understandings of why cat GIFs were tweeted during the terrorist attack in Vienna, the tweets in the dataset were assigned codes corresponding to the general topic they related to or to their overall message, such as “help the police”, or “look at my cat”. After reading and re-reading tweets, these codes were grouped into three specific themes. These themes are amplification, personalisation and ethical practice. The codes and themes can be found in

Table 2. The figure shows the codes and themes of the analysis of the cat GIF tweets. The first column shows the codes created from the totality of the 970 tweets, and the second column shows how these codes are clustered to form three specific themes in this study.

Codes	Themes
help the police asked to do this please share/RT share no photos from police actions	Amplification
look at my cat here is a cute pic/video these are the names of my cats this is a cute non-cat	Personalisation
show respect to victims don’t help the terrorists thoughts and prayers to victims thinking of Vienna stay strong	Ethical practice

.

5.1. Amplification

Most tweets mentioning cats also included a plea for individuals to help the authorities in their efforts to stop the information from the attacks being circulated and possibly used by terrorists themselves. This was achieved by asking, for example, others to not “give the perpetrators the platform they seek”. A tweet sent by Polizei Wien (@LPDWien), the local police authority in Vienna, alluded to this plea in one tweet:

Shots fired in the Inner City district—there are persons injured—KEEP AWAY from all public places or public Transport—do not share any Videos or Fotos [sic]!

As one can see from this tweet from the local police, Twitter users were not urged to flood specific hashtags with other types of content, such as cat GIFs. In fact, people were asked not to share videos or photos. Cats have been prominent figures during other terrorist attacks, a phenomenon that may have its origins in the Brussels lockdown in 2015 (Jensen et al. 2020) and later repeated during the Barcelona attack in 2017 (Zorthian 2017). Many tweets referred to the cat GIF tweets as a way of supporting the authorities in their efforts, specifically by preventing videos already shared from the terrorist attack becoming highly visible on the platform by “bombing” or “drowning” Twitter with cat images, as in the following tweets:

Stop sharing pics as the #Austrian police ask. instead you can bomb twitter with cat GIFs with #vienna #wien tags to cover these images. [URL]

Drown the shocking videos instead, here are some cute kittens! [URL]

This rationale encourages people to create a distraction to drown out already present forms of harmful content. One of the clearest examples in which the logic behind this misdirection was fully explained was retweeted multiple times. In these tweets, this logic was explained in combination with a specific request to others to share their versions of this tweet. These tweets described to other users exactly which hashtags to use, combined with the cat images and GIFs, and the rationale. For example, these tweets read

#wienATTACK

#Vienna

#0211w

#ViennaAttack

#Vienne

Add: picture or a GIF of a cat, then: publish

It will flood searches about pictures/videos of the attack with cats! [URL]

While many of these were retweeted continuously, most of the tweets in the dataset were, in fact, retweets. Of the 970 tweets in the dataset, only 259 were original tweets; so, 711 of the tweets in the dataset were retweets. Retweets and encouragement to others to tweet in similar ways amplified the message of “drowning” or “flooding” the search terms on Twitter concerning the attack. In this way, the amplified message is an act of resistance against the ongoing information flows after the hybrid media event, in effect taking a stand against the fast-paced information concerning the terrorist attack. This form of misdirection would presumably counteract hybrid information spread as a strategic amplification (Donovan and Boyd 2021) of other content—in this case, cats (see Figure 1).

5.2. Personalisation

While many of the cat images shared on Twitter in response to the terrorist attack were standard GIFs used on a range of social media platforms, such as Twitter, Facebook or Instagram, there was also a range of more personalised cat images. For example, Twitter users shared images of their own cats, along with short movies of their cats doing specific things, such as climbing on tables or running in the backyard, together with tweeted texts such as

Instead of paying attention to these terrorists in #vienna this is a photo of our kitten Tequila. [URL]

So, these reactions did not solely pay homage to common internet vernaculars of cats but also included individual users’ personal lives and interpretations. In addition to showing their own cat images, users also mentioned the national origins of the cats, for example, in these two tweets:

A cat from Spain [URL].

Some cats from Antwerp to #Vienna victims ❤️ Please. Do. Not. Share. Videos. Share. Cats 😿 [URL]

While memes may be understood as creative input to a more extensive debate on terrorism and related tweeting practices, these personal images may be an additional format of such creative outputs and a form of personalised action frames (Bennett and Segerberg 2012). In my previous studies of post-terrorism tweeting, I have argued that mentioning beer, kittens and pizza in tweeting during the Stockholm lorry attack was a way of connecting to a sense of normalcy (Eriksson Krutrök 2018). Tweets such as these are thus intimately connected to a sense of social resilience, with social media users appearing to present a picture of nonchalance in the face of terrorism and eschewing fear-based reactions as a form of everyday political action (Highfield 2016).

5.3. Ethical Practice

One of the most prevalent themes in the tweets in the data sampled for this study was how sharing cat GIFs was a way of showing respect to the victims of the attack. In order to not share harmful imagery or information, social media users addressed how these cats were not only shared in support of the authorities themselves but also as an ethical practice. As one Twitter user wrote in response to a question from another user, the idea was explicitly not to offend victims:

@user it’s an attempt to flood Twitter with cat photos instead of graphic videos, it could mess things up for the authorities of offend victims. don’t let it dominate the news! it’s a kind of protest by ppl who are upset about this attack just like anyone would be. #Vienna #Wien

In addition to being sensitive to local victims, consideration was also given to the victims of previous acts of terrorism and how these individuals might be retraumatised by harmful images. The Charlie Hebdo terrorist attack in Paris in 2015 was specifically mentioned in this context:

Thinking about the victims of the 2015 attacks cause each one of these attacks awaken painful scars. And a kitten so they don’t come across violent images either. #0211w #wienATTACK #Vienna #Vienna #ViennaAttack [URL]

So, the victims of both the ongoing terrorist attack and previous acts of terror were directly referenced as the purpose of sharing cat GIFs. It therefore appears that these social media users created their own ethical position on what information should or should not be shared during terrorist attacks. By directing attention to something else, in this case cute cats, these publics could create a form of misdirection that can be likened to the “counter-spectacles” in the research by McCrow-Young and Mortensen (2021). These forms of spectacles were specifically directed towards the protection of victims, such as in the following tweet:

RT @user: My thoughts go out to Wien and the victims. All the love to you. In order to help out, put your cat on the feed. 🙏❤️ #viennaattack #vienna

In these tweets, cats were referred to as a way of helping out, specifically directed towards the potential victims. Because of this, the practice of using this form of common internet vernacular—cats—as a way of being deliberate in tweeting practices can actually be understood as a kind of ethical strategy aimed at producing a form of “noise” in times of terror that can drown out accelerated information streams in the hybrid media system. However, since this study is solely based on Twitter data, this finding cannot be generalised to the totality of platforms included in a hybrid media system. Instead, this may provide insights into how users on specific platforms make conscious choices in regard to information sharing during terrorist attacks.

6. Discussion: Counteracting TVEC with Cat GIFs

This paper has aimed to connect the research on media events with the ongoing studies in information science and digital media research through a case study of the terrorist attack in Vienna in late 2020. It shows that GIFs were shared on Twitter during the 2020 Vienna attack for three different reasons: amplification, personalisation and ethical practice. These themes explain how tweeting related to the different interpretations of why cats should be incorporated into tweets during terrorist attacks. Firstly, it was understood as a way of complying with a direct order from the authorities during the attack. These tweets often attempted to amplify cats and encouraged others to retweet the hashtags in combination with cats. Secondly, the tweets had personalised aspects such as “showing off” one’s pets or other cute pictures to share personally specific content. Thirdly, the wellbeing of the victims of the specific terrorist attack were explicitly referenced as the core motive for sharing cat images and GIFs.

Given that social media updates and content posted by bystanders play such a central role in the liveness of terrorist events (Sumiala et al. 2019), it is interesting to reflect upon what happens when social media users actively share cat GIFs as a form of distraction and when they actively discourage others from sharing updates and graphic content. While information is highly sought after, and accelerated information streams seemingly demand ever-more updates from the scene of the crime, this study shows how current internet vernaculars shape the way in which social media users act and interact on social media in response to terrorist attacks, specifically by amplifying other content. So, cat GIFs may function as a way of being evasive rather than exhaustive in the hybrid media climate of ongoing information streams during disruptive events.

This study demonstrates that the practice of sharing cat GIFs may stem from a politically motived desire to disrupt information flows on social media through amplification, as a personalised reaction and as an ethical approach to victims. In fact, no matter how frivolous they may seem, cats themselves should not be “dismissed as an object of serious scholarly inquiry” (Maddox 2022, p. 11). The findings of this paper show that this form of misdirection is explicitly linked to the desire to act ethically during terrorist attacks by not showing pictures of injured individuals, both for the sake of victims and witnesses to the attack. The ongoing media machine is thus disrupted. In this sense, strategic amplification (Donovan and Boyd 2021) may be a social practice built to function as a form of bottom-up content moderation. By drowning out the potentially harmful content associated with terrorist violence, these publics are diverting attention by not complying with common-place practices during terrorist attacks, such as live-tweeting or sharing images from the scene of the crime.

That said, while cats may well be clogging up the machinery, it is difficult to gauge the effectiveness and real-life consequences of such an approach. Since no baseline data have been acquired to show the “normal” conditions of Twitter or tweeting activities during other terrorist attacks, this study should not be generalised to include larger communication patterns on Twitter, or beyond. Nonetheless, this study has presented several analytical conclusions. Firstly, it might give more traction to the attack itself as individuals unable to find the information they desire on the feed in question may well feel inclined to actively search for further updates. Secondly, this is, in fact, still a minor reaction in relation to the majority of tweets in this sample (970 tweets as opposed to 452,034 tweets). However, this does not mean that these reactions should not be studied; in fact, these reactions can tell us something important about the way information is shared in hybrid forms in times of crisis, and about attempts to change the media system itself to become more deliberate and ethical. As proposed by Maddox (2022, p. 10), if we are interested in understanding the dark sides of the internet, such as terrorist and violent content on social media, or indeed terrorism itself, we need to be “examining this content as always already imbricated within the Internet’s broader sociocultural, technological dynamics and not separate and apart from it”. Future studies should focus on the role of vernaculars present in specific meme cultures and their role in information flows following terrorist attacks or other disruptive events.