Abstract
On 22 July 2011, a lone wolf terrorist attacked Norway. At the island of Utøya, he killed 69 people. This article asks how the youth at the camp on the island used social media (SoMe) in the emergency situation caused by the terrorist. Answers could give significant contribution to the growing research on SoMe and crisis, especially since there is a research gap when it comes to examine terror victims’ use of SoMe. This study is based on qualitative interviews with eight survivors. Based on the campers’experiences, how do they evaluate the opportunities and challenges of SoMe during a terrorist attack? What were the reasons for not using SoMe? What was the purpose of using SoMe during the attack? SoMe play an essential role in crisis communication strategies as well as being an increasingly important tool for the public. Although verification of SoMe content is difficult, SoMe have become important sources for journalists. This study offers best practices for journalists from victims. Moreover, it sheds light on how SoMe played a role for the victims in alerting, giving and receiving information, as well as building hope and resilience.
Introduction
On 22 July 2011, on the island of Utøya in Norway, 69 people were brutally killed. This happened during the traditional summer camp of the Labour Party’s youth organization, AUF. Previous research has shown that during the massacre, and in the hours that followed, campers at Utøya used their cell phones to connect to the outside world (NOU, 2012, pp. 14, 454).
Research highlights the need during emergencies for two-way information flow on social media (SoMe; Coombs, 2015; Wetzstein, Grubmüller-Régent, Götsch, & Rainer, 2014). Still, we do not have sufficient understanding of if and how directly targeted individuals deal with crisis through SoMe. This study is based on qualitative interviews with eight survivors. The aim is to shed light on some of the victims’ views on, and use of, SoMe during a terror attack. Furthermore, it will be emphasized whether SoMe contributed to resilience, which means the “ability to maintain a stable equilibrium” (Bananno, 2004, p. 20). My research questions are as follows:
Q1. Did the survivors from Utøya use SoMe?
Q2. What was the purpose of using SoMe?
Q3. What were the reasons for not using SoMe?
Q4. Did SoMe contribute to their resilience?
Q5. What were their opinions about the credibility of information on SoMe?
Q6. If there was any, how did they experience contact with journalists?
Q7. Did they have suggestions for improvement on the use of SoMe during a terror attack?
First follows a short outline of the terror attacks in Oslo and Utøya. Then, the next section reviews the research on the use of SoMe in crisis situations, SoMe on 22 July 2011, and research on resilience. There then follows a methodology chapter before the empirical findings are presented and discussed.
The 22 July 2011 Terrorist Attacks
On Friday 22 July 2011 at 15:25, Anders Behring Breivik exploded a 950-kg manure bomb in the government quarter in Oslo, Norway. Eight people were killed and 10 admitted to hospital, in addition to the injuries caused to several others. The explosion caused massive destructions in central Oslo (NOU, 2012, pp. 14, 17). The perpetrator then drove 38 km to the island of Utøya. The island, with an area of 0.12 km2, is owned by the youth organization, AUF. On 22 July, AUF held their traditional summer camp. A total of 564 people were present, when the terrorist, dressed to look like a policeman, started to shoot at 17:21 (NOU, 2012, pp. 14, 23-26). The police arrived at the island at 18:27 and the perpetrator was arrested at 18:34 (NOU, 2012, pp. 14, 30). Totally, 69 people were killed and 56 were admitted to hospital with severe injuries (Dyb et al., 2014, pp. 361). The survivors were taken to the rescue center at Sundvolden Hotel. In 2012, Commission on 22 July strongly criticized the authorities’ ability to protect the people at Utøya. The police should have acted earlier, the Commission said, and the possibility to stop the perpetrator before he reached Utøya was missed (NOU, 2012, pp. 14-15); this means that several of the killed campers could have been alive.
Litterature Review
SoMe, Verification, and Emergency Situations
Research suggests that SoMe play an essential role in crisis communication strategies as well as being an increasingly important tool for the public (Bunker, Ehnis, Seltsikas, & Levine, 2013). However, there is still a need to pay more attention to bottom-up approaches and ways to improve collaboration between authority response organizations and the public (Coombs, 2015; Wetzstein et al., 2014). As Bunker et al. (2013) write, If we had a better understanding of the use of Social Media technologies [. . . ] we are more likely to deal with information governance appropriately and effectively and especially in times of crisis. (p. 251)
During the Arab Spring of 2011, SoMe were used as means of communication and for organizing the uprising (Diakopoulos, De Chaudrhury, & Naaman, 2012). During and after the attacks at The Bataclan in Paris in 2015, SoMe turned out to be a powerful tool for coping, and Twitter’s #pourteouvert (open door) hashtag, for instance, directed people desperately trying to find a place to hide (Lee, 2015). However, also the terrorists monitor SoMe and traditional media to enhance their operation (Oh et al., 2011; Simon, Goldberg, Aharonson-Daniel, Leykin, & Adini, 2014).
Bruno (2011) claims SoMe are changing the way the media cover crisis events. We have seen that terror attacks brake on Twitter, that is, Mumbai attacks 2008 and Jakarta bombings 2009 (Cheong & Lee, 2011). On SoMe, a journalist may find “eyewitness experiences to be shared and official statements to be made” (Lee, 2015). However, Glad, Thoresen, Hafstad, and Dyb (2017) find that survivors from Utøya had negative experiences with journalists’ intrusive approach. In the terror attack in Mumbai, eyewitnesses’ accounts were circulated on SoMe and online news sites (Diakopoulos et al., 2012; Hermida, 2012). Studying how CNN, BBC, and The Guardian covered the Haiti earthquake of 2010, Bruno (2011) finds that these outlets initially used SoMe, but when reporters came to the island they did not need SoMe. Monitoring an unfolding crisis on SoMe may give an overview, but among eyewitnesses’ accounts and other SoMe content, there could be misinformation and rumors. Based on information from SoMe, traditional news media misidentified the gunman at the Connecticut school shootings in December 2012 (Zurawick in Schifferes & Newman, 2013). After the Boston Marathon bombing in April 2013, Reddit had to apologize for identifying innocent people as suspects (Chang, 2013). In the aftermath of the Paris attacks, there were tweets about a supposedly revenge attack at a refugee camp in Calais which turned out, in reality, to be due to an electrical fire (Kayali & O’Rourke-Potocki, 2015). Kovach and Rosenstiel (2007) define journalism as the discipline of verification. I see verification as the process of establishing whether information is true and accurate. Brandtzaeg, Lüders, Spangenberg, Rath-Wiggins, and Følstad (2015) claim that SoMe make verification even more important. Schifferes and Newman (2013) advocate the use of a verification tool to examine the three C’s; Contributors, Content, and Context. On the contrary, Kaufmann (2015) stresses that SoMe, at times, gain “more credibility than traditional media, precisely because their ‘reporting’ was considered more authentic” (p. 9).
SoMe and 22 July
In 2011, when the terror attacks on Norway took place, Facebook was the most popular social networking site with more than 500 million registered users worldwide (Facebook 2011 in Nadkarni & Hofman, 2011). As for Twitter, it had 275 million registered accounts in June 2011 and was responsible for 200 million daily tweets (Twitter Blog in Cozma & Chen, 2012, p. 14). People aged between 18 and 29 years use SoMe the most (Pew Research Center, 2012, p. 3). In 2012, Norway ranked second in the world for active SoMe users (Enjolras, Karlsen, Steen-Johnsen, & Wollebæk, 2013, p. 15).
The report from the 22 July Commission (NOU, 2012) describes how “Cell phones, social media and fast news dissemination spread the knowledge of the attacks quickly from the places of perpetration to large parts of the population” (pp. 14, 454). Andenæs (2012, p. 60) states that the use of SoMe exploded. Kaufmann (2015, p. 6) writes that the role of traditional media shifted and “social media were in fact attributed with more factual credibility than traditional media.” Actually, the news about the shooting at Utøya first broke on Twitter. Covering the bomb in Oslo, media could send journalists to the place of perpetration, but that was impossible at Utøya. A Norwegian small-scale study shows that Twitter was an important source for journalists during the shooting on the island (Kluge, 2012). Also, Konow-Lund and Olsson (2016) reveal that journalists could not ignore SoMe, although they initially tried to work more traditionally. Although “in the system for managing societal crisis, bottom-up oriented approaches from the public have begun to be considered more seriously” (Linnell, 2014. p. 70), several emergency organizations and authorities were not on SoMe. The Police (Hornmoen et al., 2018) and Norwegian Directorate for Civil Protection DSB (Brekke Jørgensen, 2012; Meling, 2012) were among the SoMe absentees.
When the shooting started at Utøya, the campers were trapped in a small area with the perpetrator. No one gained access until the police arrived more than an hour later. A master thesis studying the role of SoMe among the campers (Johnsen, 2012, p. 52) finds that Facebook was the main channel for communication and source for information. Twitter, which seven of the thesis’ interviewees normally used, was not important (Johnsen, 2012). This change in SoMe use Johansen links to the exceptional condition (Johnsen, 2012, pp. 45, 51). In addition, it may be explained with Facebook being the network to reach your loved ones, and that contact with them was the primary concern for the campers (Johnsen, 2012). This corresponds with Kaufmann’s (2015) findings that the general public use Facebook for personal statements and Twitter for analysis (p. 6).
Coping Strategies and Resilience
Kaufmann (2015) defines resilience as “a technique and mentality of self-governance during emergencies” (p. 13), linking resilience to behavior during a crisis. Some definitions focus on the aftermath and the multiple pathways to bounce back, regain one’s composure, and adapt well (Bananno, 2004; Newman & Nelson, 2012). I define resilience as “the ability to maintain a stable equilibrium” (Bananno, 2004, p. 20) during and after emergencies. In this article, I mainly look at coping strategies during the crisis. Even though they may seem to contradict one another, avoidance strategies and approaching strategies are both resilient strategies: For a person caught in the midst of a traumatic event, the two primary human responses are flight or fight (Newman & Nelson, 2012). Avoidance strategies may take the form of running, hiding, or diverting attention from the source of stress. Approaching strategies may include fighting back as well as “seeking information or closely monitoring the stressor” (Newman & Nelson, 2012, p. 19). Both flight and fight activities are active, resilient reactions. Passive reaction could be tonic immobility (Filkukova, Hafstad, & Jensen, 2016, p. 2). During a life-threatening catastrophe, our survival brain takes over and we act automatically in order to rescue ourselves. This “could have an impact on which impressions we save or supplant, and how we deal with the aftermath” (Frey, 2016, p. 174). This may influence how victims remember in the short — or long term— episodes could be blurry, odd details clear, and other things could be blotted out or forgotten. To heal again is to go through the process of integration; to integrate the memories, connect senses to emotions, and so forth (Tveito, 2011).
In emergency situations, people’s well-being and safety, as well as their need to know the whereabouts of family and friends, are pivotal. SoMe may be used for this purpose, as well as to alert about ongoing developments and allow people to adapt. Hermida (2010) considers Twitter to be an awareness system, “. . . enabling citizens to maintain a mental model of news and events around them” (p. 301). When a sudden crisis happens, we look for information, try to regain control, and seek meaning (Tveito, 2011). These are resilient strategies, as involvement provides a sense of control and increases a person’s ability to cope with the situation (Veil, Bucehner, & Palenchar, 2011). Kaufmann (2015) shows how . . . the network character of social media had a direct effect on how (self-)caring subjects came into existence, at the same time as these subjects translated the various affordances of the network into concrete resilience practices. (p. 14)
Methodology
When interviewing people about a life-threatening experience, one is asking them to go back to an extreme situation, which may trigger traumatic memories and reactions. Thus, careful consideration needed to be taken when recruiting and interviewing. In Norway, a coordinating group was established to overview the research on 22 July attacks. Its main object was “to safeguard the interests of those who were directly affected by the attacks when they participate in research” (Refsdal, 2014, p. 2). 1 The National Support Organization after 22 July Incidents 2 (for short National Support Group) was part of the coordinating group (Enebakk, Ingierd, & Refsdal, 2016, p. 17). Potentially, the participant pool size included all survivors from Utøya, a total of 495. In practice, the pool was reduced to a much smaller size. One explanation may be that the survivors were tired of being exposed and contributing their terror stories to researchers, the media, the police investigation, and as witnesses at the trial (Enebakk et al., 2016, pp. 15, 49). Another could be due to the careful recruitment process that declined the possibility to contact possible respondents directly. Then, I also excluded anyone under the age of 18. To rule out survivors with major health problems, I used the National Support Group to get in contact with interviewees. Twice, information about my research project and invitation to contact me was sent out to members of the National Support Group, 3 but only two survivors answered. Later, the National Support Group gave me direct contact information to survivors. I was in contact with 16 possible interviewee subjects. There were campers we contacted who did not want to be interviewed and others who had to withdraw for different reasons. The prolonged recruitment period was also due to the fact that I did not approach possible interviewees or conduct interviews on or close to memorial dates or at Christmas. Hence, the recruiting period and the process of conducting interviews were prolonged and I ended up with eight respondents.
All interviewees signed an informed consent. They could withdraw from the interview at any time. They were promised anonymity, and in this article they are referred to by a number. To make sure the interviewees felt in control, they received the interview guide in advance. Thus, they were able to prepare themselves for the questions. The interview guide was divided into three parts: addressing the pre-crisis with two questions, then the crisis, and post-crisis phase. Extra attention was paid to ensure the questions were open-ended, without trigger words, and that they were focused on SoMe and did not venture even more difficult experiences and memories from the terror. For instance, my main questions were, “Did you use social media during the attack? If you did, what was the purpose? If you did not use social media, did you think about using them? Why did you not use social media during the attack?” Other topics that I asked about were who they were in contact with and weaknesses and strengths of SoMe. As a precaution, I did not ask questions about resilience, but I asked whether they got information and a better overview of the situation through using SoMe. The interviews were done face-to-face in a safe and quiet environment. To secure a trustworthy atmosphere, the meetings started with repeating information about the research project, the interview, and the rights of the interviewee. The interviews ended with summing up, practically and emotionally.
All the interviewees were politically active and used SoMe on a regular basis. Several of them had local positions in AUF and most of them were candidates for the Labour Party at the local election in September 2011. It is likely to assume that the campers are people who understand societal processes, are up to date on communication on the newest platforms, and are conscious of how they use SoMe (Johnsen, 2012, p. 51). When the qualitative interviews were done in autumn 2015 to spring 2016, the interviewees ranged from 22 to 32 years of age. They comprised two females and six males. Since my sample is so distinct, it may influence the results.
The interviews were done more than 4 years and up until almost 5 years after 22 July 2011. The time that had passed could influence the memories of any interviewee. There is possible recall bias as survivors’ memories of traumatic events are subject to changes over time (Filkukova et al., 2016; Newman & Nelson, 2012). All my interviewees did at some points during the interview state that “this is a bit blurry” or “I don’t remember.” Some of the details were possible to verify, but others were not.
As described above, I did not have a question in the interview guide about resilience, but the interviewees talked about issues that led to reflections on staying safe and resilient behavior. This prompted me, when working with the transcribed interviews and analyzing them, to elaborate on resilience in the analysis. As seen in the theory part, there are several ways to act in order to gain resilience. Hence, resilience is revisited several times in different parts below.
Results and Discussion
Here I present and discuss my results. First, I write about use of SoMe. Then, I go into how SoMe were used to alert and say goodbye, discuss the reasons for not using SoMe, and how SoMe habits lead to different strategies. Next, I move on to how SoMe turned out to promote resilience. Then, the analysis turns to the problem of misinformation. Sections “Critical toward journalists” and “We are alive!” follow, before the interviewees suggest improvements.
Use of SoMe
All the eight interviewees used SoMe before 22 July 2011, everyone was on Facebook, four on Twitter, and some used Instagram and LinkedIn. After the bomb exploded in Oslo, SoMe and online news were used to find out what had happened as well as to locate family and friends in the capital. Approximately 2 hr later, the terrorist started shooting on the island. In Table 1, the column “During the attack at Utøya” needs some explanation: two people were in and out of Facebook, not using it actively. They both knew their status had been updated by others.
Social media use by the eight interviewees.
SoMe: social media.
Pre-crisis, four interviewees were especially dedicated to SoMe and had several accounts and large networks. Three of them used SoMe during the attack, but the fourth got his mobile wet. Four interviewees were only active on Facebook. During the attack, two of them passed through it and two did not use SoMe. The latter two are the ones expressing that they would not use SoMe in a new crisis situation. As seen in Table 1, all eight used SoMe once they were rescued and in the post-crisis period.
In accordance with Johnsen’s (2012) findings, my research shows that Facebook became more important than Twitter. This was due to the character of the networks, as the interviewees perceived Facebook as social and more personal and Twitter as more public and political (Johnsen, 2012). Also after the attack, Facebook was the prime choice both for personal contact and as a medium for information and organizing. This may be obvious since only four out of eight were on Twitter prior to the attack, but all of them were on Facebook. Moreover, the interviewees tell that their organization AUF used Facebook for contact, information, and organizing.
Alert and Last Farewell
The interviewees did alert by text message, phone, or SoMe, and seven of them tried to contact the police or got people outside the island to do it for them. At an early stage, one did alert on Twitter about the shooting. This interviewee had around a thousand followers at the time and a wider range of followers than on Facebook. Hence, Twitter was used because “it is a medium where news are diffused quickly and a lot of people get to know about the situation” (Interviewee 1). Before tweeting, the interviewee had rung the police, but they did not seem to grasp the seriousness of the situation. “My intentions were to break through the flood of information about the terror in Oslo and get attention” (Interviewee 1). The interviewee’s tweets were retweeted and informed the public about the shooting. An interviewee who alerted on Facebook says that “the reactions to one of my posts on Facebook showed that people outside the island, based on how it was presented on the news, had not understood the extent of the attack” (Interviewee 4). The first priority for seven of the eight interviewees was to tell people close to them that they loved them and say their last farewell. “In case this was the last thing I ever did, I told them on a text message that I loved them. Then, I could think about something else” (Interviewee 2) says one who used SoMe to relay what was happening, get information, and gain a better overview.
SoMe? The Main Thing Was to Survive (Interviewee 5)
There were practical reasons for not using SoMe — such as not having your mobile with you, that it was loading at the time when the terrorist started shooting, it was losing battery power, or it was destroyed in the water. As seen in previous research (for instance, Filkukova et al., 2016), people at Utøya threw away or turned off their cell phones, fearing that the shooter could trace them. Four interviewees point out that the terrorist could be using SoMe, which would be extremely dangerous for them. These four were not active on SoMe during the attack. They also stress the danger that noise or light from the mobiles could reveal their hiding place when the terrorist passed by. On the contrary, some of them think that if you cannot speak, written communication is a good alternative. The interviewees explain how the constant calls that came in put them in a dangerous situation and they whispered, texted, or wrote on SoMe that people were not to call them. It upset them most when journalists were calling at this life-threatening time, but also family and friends had to be stopped. One of them regrets texting something like, “Fuck, if you have seen the news, do not call me!” (Filkukova et al., 2016).
Some campers who did have their cell phones with them also chose not to use them. As one interviewee relates, “I was running, I could not even think about that” (Interviewee 8). Another makes the same point and adds that the main thing was trying to stay safe: “The will to survive kicked in” (Interviewee 5). Coping strategies such as hiding and keeping quiet, as well as closely monitoring the situation, are important flight reactions. As one interviewee says, It is dismal to watch your cell phone when you are on an island where someone is shooting and you don’t know where the shooter is. You have to be on guard all the time. If you are looking at the phone, you will not be observant enough to sound and movement around you. (Interviewee 5)
These interviewees went through the “. . . process of adaptation, of dealing with insecurity” (Kaufmann, 2013, p. 68), as their instincts and reactions were focused on staying alive.
Different Strategies in Similar Circumstances
Four of the interviewees were hiding throughout the attack, but regarding SoMe they acted differently. Two of them did not use SoMe actively. One reason is that both of them were preoccupied with the risk of being traced by the shooter. One says, “The terrorist was close. I did not know if I was safe, so there was no point in giving out information” (Interviewee 7). Contrary to them, the other two were active on SoMe while hiding. They had even more experience with SoMe and were active on several platforms prior to the attack. For them, SoMe were their link to the world outside the hiding place.
To save their lives, four interviewees swam and their cell phones were destroyed. Two of them did not consider using SoMe. Another was thinking about using SoMe, so he put the mobile in his mouth to keep it dry, but it was irretrievably damaged. The fourth tweeted before he started to swim. The latter two were the most experienced SoMe user of the four, which may be an explanation for having SoMe in mind in a stressful and dangerous situation.
A Life-Buoy and a Platform for Awareness
The active SoMe users gave and received information about what was going on — amounting to “resilient forms of emergency communication” (Interviewee 1, p. 9). One interviewee kept the cell phone in his hand the whole time. He had muted the sound and kept looking on the screen for updates. He used Facebook, Twitter, and online news media, outlets that supplied an awareness of, and an overview on, the situation. The mobile became a life-buoy and a way to control himself because “scanning for information and keeping the brain busy help you handle the situation” (Interviewee 4). Another interviewee states that even though the dangerous situation was prolonged, information on SoMe gave hope (Interviewee 1). The third interview subject who actively used SoMe says it was effective in “reaching a lot of people and the right people, those I knew, those I wanted to get hold of” (Interviewee 2). These examples indicate ways of actively securing one’s own safety and how resilient acts using 2.0 technologies promote a self-organized crisis response (Kaufmann, 2015).
The interviewee who tweeted early on in the shooting spree focused on alerting people outside Utøya, hoping that journalists would put pressure on the authorities to act. Still, to tweet was not an easy decision to make. When in a life-threatening situation, “you feel that everything matters” (Interviewee 1). This subject states that “you act on impulse or you have seconds to decide how to act” (Interviewee 1) and goes on to ask, Do you tweet when your friends are getting shot? Or should you use your energy on doing something else? It is a choice you have to take [. . . ] I felt responsible for taking care of the younger ones, so all together, there were many thoughts involved in how I chose to prioritize (Interviewee 1).
The three interviewees who used SoMe say SoMe served their purpose. Interviewee 1 did not get in contact with others on the island through Twitter, but following the tweet stream kept this subject aware of how people and the authorities were reacting and when help was going to come. The other two also believe they acquired a better awareness and understanding of the dangerous situation through SoMe. One says he learnt through SoMe that the perpetrator was dressed as a policeman, a fact that influenced how to act in an abnormal situation. Furthermore, from other campers’ use of SoMe, this interviewee confirmed that the decision to stay hidden was the right one. He says, “It was useful to learn what others in the same situation were thinking” (Interviewee 4). The other one reveals, I could not see much from where I was hiding. There are echoes on the island, so one shot sounded like five and I did not know where the shots came from. Towards the end, I understood on social media that there was only one terrorist. That was good to know. Then I learned on social media that he had been seized by the police. (Interviewee 2)
Not Necessarily True Information
The interviewees underline the advantages of fast and early communication in a crisis situation. Some of them mention that SoMe were particularly useful when asking people to take their boats out on the fjord to rescue people or when the health authorities and the blood bank utilized it (see also Ottosen & Steensen, forthcoming). Nevertheless, they are preoccupied with the difficulty of separating rumors and inaccurate details from facts — a particular problem amid the flood of information on SoMe as an emergency situation evolves. My interviewees were accustomed to filtering political information. Now, they emphasize, their safety could depend on filtering trustworthy, verified information in order to stay alive.
All the interviewees witnessed or heard the shooting. Several of them saw people getting shot or ran past dead people when they tried to escape. Others saw the terrorist. One interviewee was shot at, but the bullets missed. Speculations mounted on text messages, on SoMe, and by word of mouth: “It was confusing and contradictory information: It is one person, there are five terrorists, it’s war!” (Interviewee 8). As one interviewee says, “Spreading misinformation or rumours . . . it could be mortal if someone posts that it is safe when it is not” (Interviewee 3). Another subject realizes that There is no credibility. For instance, someone said that the police were on the island. Another said that now you are safe. We did not dare to trust it, and if you look at the time log you will see that we were right not to trust it. Anyone can post what they want on social media. There are no means to verify information. (Interviewee 2)
Those who used SoMe and even more so the ones who did not point to the enormous load of information out there and how difficult it was to handle many loose threads. SoMe . . . did not give an integral and complete picture of the situation, actually no one had that. There were single considerations from individuals. Limited information, but still useful. The information I got on the perpetrator was not necessarily trustworthy, still it helped me get a better picture of what others in the same situation experienced. (Interviewee 4)
On SoMe, there is a low threshold for posting, and anyone can uncritically post whatever they feel like. Furthermore, the information stream may contain facts but can, just as easily, spread fear, rumors, and conspiracy theories (Haataja, Hyvärinen, & Laajalahti, 2014). One from the island says, “There could be information on SoMe that will not benefit a person who is in a crisis situation” (Interviewee 7). “Twitter is a source of misinformation,” says another (Interviewee 5). After the police arrested the terrorist, AUF used a closed Facebook group to register everyone on the island, and one person was misleadingly reported to be alive (Interviewee 2). On the contrary, one interviewee stresses that on Twitter you can verify the accounts and then gauge the trustworthiness of a particular person (Interviewee 1). Another crucial aspect is having many friends and followers on SoMe —If you do not have a virtual network before a crisis situation, you cannot as easily pick up information (Interviewee 2). In that aspect, Utøya was unique: I knew a lot of people in the same situation and on the same place as me, people I knew and trusted. Their evaluation of their experiences and what they had of information was easier for me to consider since I knew them. (Interviewee 4)
Accordingly, the context and the contributors—two of Schifferes and Newman’s (2013) three C’s verification tools—contributed positively at Utøya. Although peer campers had credibility and their reports were authentic (Kaufmann, 2015), the young people found themselves in different situations around Utøya. They communicated their thoughts, fears, facts, and experiences from their particular positions and state of mind, which could be completely different from those of others on the island.
Critical Toward Journalists
Eyewitnesses’ accounts on SoMe were important to journalists on 22 July (Andenæs, 2012; Kluge, 2012; NOU, 2012, p. 14). It gave a direct link to the shooting and relayed authentic reports from the inaccessible island. My interviewees express dismay with the journalists who tried to contact them during the attack. This finding is in accordance with Glad et al. (2017) who report that when approaching the campers “reporters had lacked respect and compassion and that they were intrusive” (p. 6). Some of my interviewees were contacted on SoMe, and others received calls. One interviewee states, “the journalists kept calling me. The calls came in densely as in a hail-storm” (Interviewee 2). Another subject reveals that he could hear people whisper on the phone to journalists to stop calling them (Interviewee 3), since sound or light from the mobiles could reveal their hiding place. Furthermore, some of them posted on SoMe to prevent media contact. One relates how someone on SoMe gave her name to a British journalist. A call from the journalist put the interview subject in a very difficult position: “I was still on the island, and I was sort of in a shock, and I had not yet spoken to my parents” (Interviewee 7). Most of my subjects did give interviews later on, and their experiences at a later time accord with the views reported on by the 22 July Commission — the presentations of them in the media were done in a respectful way (NOU, 2012, pp. 14, 267). On the contrary, some feel that reporters and photographers acted intrusively at the rescue center at Sundvolden. Here, one interviewee tweeted about disrespectful journalists, hoping they would act in a more appropriate way (Interviewee 2). One interviewee also talks about the shift in the journalists’ use of SoMe: “There is a time before and after 22 July; the journalists understood that many people expressed themselves on social media, and that this could lead to news stories” (Interviewee 6). This interviewee agrees with research on how 22 July changed the Norwegian media’s use of SoMe (Konow-Lund, & Olsson, 2016), but he evaluates it in a more negative way: “They milked social media.” Several of the interviewees’ tweets were used by journalists. One of the campers recalls how a journalist phoned and asked for an interview to follow up a tweet she had posted from Sundvolden. She refused: Then the journalist asked if they could use my tweet in the newspaper, and I was sort of negative. Then the reporter said “But you know Twitter is public, so we are only being polite when we ask you.” I think that I just hung up. (Interviewee 2)
Generally speaking, my interviewees are positive toward SoMe’s influence on traditional media. It is good for public debate, one stresses, that voices that traditionally would not be heard can be quoted (Interviewee 4). In the acute phase, however, they argue for a change in journalists’ behavior. As I interpret what my interviewees say, these are their advice to journalists:
Renounce actions that might endanger the lives of directly attacked people: Do not call them or contact them on SoMe while they are trying to survive.
Quoting targeted people’s SoMe statements feel intrusive to them.
For most victims, media contact directly after they are saved is too early.
Later on, journalists may contact survivors, but in a respectful way.
I would add that people during and shortly after a terror attack are not necessarily credible sources for journalists. This is due to how the brain reacts during a life-threatening crisis and the implications on senses and memories (Frey, 2016; Tveito, 2011).
We Are Alive!
Once safe, all the interviewees posted on SoMe that they were alive: “Facebook for family and friends, Twitter for the public” (Interviewee 2). Since many had lost their mobiles, once rescued they asked to borrow one. Then, they called their loved ones before going on SoMe. When the survivors came to the rescue center, many of them stood in line for a computer in order to update that they had been rescued. Seven of them updated their status on Facebook, and two of them immediately tweeted. One asked a family member to update his status on Facebook since he was on the island and the battery on his mobile was low. He says, it saved me a lot of phone calls that I was safe, a lot of calls I was not ready to make at the time. It was a load that was nice to get rid off, because hundreds of phone calls . . . it would have been too much. (Interviewee 3)
Another subject puts it this way: “When you are physically worn out, you can’t stay on the phone for five hours calling everybody, so social media are a good way to spread the news” (Interviewee 6). The interviewees fed information into the closed Facebook group, and one of them started when on the island. This subject describes the situation when the police arrived: We had to stay away from the windows, because the police were not sure if there were more than one terrorist. First, we had to help someone with their shot wounds. Then we raided the kiosk for something to drink and eat. I felt I had to do something, and then I found the Facebook group, where you could share information on people you knew were safe. So, I accounted for the ones I knew that were alive and rescued. (Interviewee 4)
Although many survivors knew about people who died during the shooting, at this point that information was treated as unconfirmed. 4 The closed Facebook group and survivors’ own accounts on SoMe were important when they tried to find out who had survived or not. Relatives and friends of missing youth reached out to survivors on Facebook in order to get news about their loved ones.
Suggestions for Improved Use of SoMe
As can be seen in Table 1, crucially, six of eight interviewees say they would use SoMe in a crisis situation today. However, one points out that in a terror attack at a random place, SoMe would probably not be as useful as they were at Utøya, where he knew a lot of people (Interviewee 4). One of the two interview subjects reluctant to use SoMe during a terror attack says he believes SoMe could function better during slowly evolving emergencies such as a flood or a pandemic (Interviewee 5). All interviewees were asked whether they had any thoughts on how the use of SoMe in a time of crisis could be improved. First, they point to the fact that they use SoMe even more than they did in 2011. Now smart phones are in common use, which makes it easier to connect and alert. On the other hand, the smart phones’ batteries run out more quickly. Furthermore, several interviewees point out that Twitter and Facebook have improved tools and apps to use in a crisis. They also stress that the message function on Facebook was not as developed in 2011. At the time, people preferred to use text messages, whereas they today would use the Facebook message function or Snapchat. Maybe at some point, each of us will have a drone following us around, live streaming where we are and what we do (Interviewee 6).
The majority thinks SoMe can make a difference in crisis. Today, the crisis authorities are on SoMe, but the interviewees express the need for further developments in order for SoMe to function as a real two-way channel between crisis authorities and affected people. Sharing information orally or by written or visual communication with crisis authorities, without going public on SoMe with the same information, is a desire voiced by several interviewees. It should be possible for a victim to chat with the police, my interviewees think. They emphasize the importance of sending pictures and video. In this way, the police could see the perpetrator’s face or the crime scene and be prepared on arrival. On the contrary, there is research which indicates that fake photographs are a problem (Schifferes & Newman, 2013) and that videos and pictures are difficult and time-consuming to verify (Brandtzaeg et al., 2015).
Several of the interviewees’ peers and friends could have lived if the communication and alert system had been better and the police had arrived earlier at the island (NOU, 2012, p. 14). Knowing that, the interviewees are interested in improved alert routines and possibilities, including equipment, software, and gadgets. To verify facts on SoMe is difficult, and the interviewees like the idea of a verification tool. However, they agree that this tool should not be designed only for use in emergencies, since under pressure it is crucially important to use SoMe tools you rely on every day. Some of them assert that tagging should be introduced more widely, taught, and stressed to the public. To either use the emergency telephone or get in contact through for instance @nødetatennorge [@emergencydepartmentnorway] should be equally effective. “But it has to be organized differently than when we phoned in” (Interviewee 6). They are concerned about better routines and options in the police’s handling of information — they see a need for an improved system of receiving and using information and better ways of verifying information and people behind different accounts. “Perhaps one should log in to social media using one’s fingerprint?” one of them suggests (Interviewee 6). On the contrary, this interviewee is aware of the fine line between safety and surveillance: “Do we want social media that are open to everyone or do we want social media where you have to verify who you are?” (Interviewee 6). The problem of verification and misinformation on SoMe is a challenge they would like to see solved. To verify content and contributors at a random crisis scene probably will pose an even bigger problem than it was for them at Utøya. Several of the interviewees remark on the fact that they knew others on the island. One states that experiencing a terror attack at another location could have made it “extremely difficult to use social media as a source, since I would not have known anyone there” (Interviewee 4).
Analysis and Conclusion
The aim of this article was to shed light on victims’ use of SoMe during a terror attack. When studying SoMe use and its impact for directly targeted people, one has to keep in mind that victims’ first priority is to survive. Then, one needs to take into consideration that the situations victims find themselves in are different. Taking these considerations into account, my research shows that the interviewees say SoMe are efficient in reaching many people fast. I find that SoMe can help in alerting, monitoring the evolving crisis situation, getting and receiving information, and organizing. For SoMe to be useful, it is pivotal that the directly targeted people have massive experience with them. Under pressure, habits and tools you normally use are what help you when your survival brain kicks in and you act automatically (Frey, 2016). For instance, I find that the most active users automatically muted sound and turned off vibration on the mobiles. Also, my results show that under similar circumstances, the interviewees’ SoMe practices make an important factor: Although no targeted victim on an island will decline swimming in order to use SoMe, this study shows that the most active SoMe users had strategies to do both. In addition, in the context of hiding, the victims’ experiences with SoMe influence their attitudes toward SoMe and their actions. Furthermore, my interviewees point out that large virtual networks are important for contact outside and at the terror scene. In this sense, Utøya was unique since interviewees were able to connect with people in the same situation that they knew and trusted. When the pressure was reduced and they were safe, all interviewees think that SoMe were useful. During the attack, however, only the most experienced found it obvious that SoMe could be helpful.
Second, my study indicates — as Johnsen’s (2012) results confirm — that Facebook proved to be most useful. Since only half of my interviewees were on Twitter 22 July, this could be one explanation. Even so, I find that Facebook was the preferred channel to contact people, give, and receive information and organize.
My third main point concerns problems with SoMe use during crisis. The fear of the terrorist tracing them made an impact on the decision for the victims who renounced SoMe. This is a legitimate concern that targeted people need to take into consideration. That relates as well to worrying about sound and light from the mobile. Still, the most active SoMe users had strategies like muting the sound and turning off vibrations. Then, unreliable information on SoMe was another challenge for my interviewees, and they welcome a verification tool for content and contributors. However, the most skilled SoMe users found SoMe information to be useful and one points out that on Twitter it is possible to verify accounts. Still, all interviewees stress the need for better alert systems, a real two-way communication with the police as well as improved routines and options for handling information on SoMe during emergencies. Regarding these points, they make proposals that could have practical implications. Another problematic areal is how journalists approached victims. In light of the interviewees’ experiences with the media, this article proposes best practices for journalists that can make a difference to victims.
Empirically, I found that SoMe did contribute to the most active users’ resilience. To what extent SoMe were helpful in that regard would be a matter for further research. Especially so, since the ones who did not use SoMe during the attack also report on resilient behavior — and all results are in coherence with resilience theory. This small sampling cannot answer such a complex question, but my research indicates that if people have large networks and are active on SoMe, they will find that SoMe contribute in maintaining resilience.
In my opinion, the interviewees’ insights represent a significant and unique understanding of the advantages and difficulties of SoMe use in a terror situation. The number of respondents is too small for the findings to be representative, but I would argue that this study provides rich material about targeted people’s use of SoMe in emergencies. The sample of interviewees who are active SoMe users and politically engaged people from a country that ranked second highest in the world for active SoMe users (Enjolras et al., 2013, p. 15) may have contributed to a large extent to the results. It may be an asset for this research that my respondents are young, smart people who have thought a lot about what they experienced at Utøya. As such, my analysis may contribute to a better comprehension of the crucial decisions victims have to take. As SoMe and network technologies develop quickly, further research is needed on various aspects of SoMe use in crisis. Larger sampling and comparative studies on victims in different terror situations would be desirable to find representative results. I agree with my subjects that future terror actions will see more directly affected people using SoMe. Giving them, as well as crisis authorities, better tools and conditions for dealing and coping with terror situations is an argument for further research, especially since my results imply that SoMe can make a difference for terror victims.
Footnotes
Acknowledgements
Thanks to Harald Hornmoen, Steen Steensen, and Rune Ottosen for helpful comments.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is part of Researching Social Media and Collaborative Software Use in Emergency Situations (RESCUE), funded by The Research Council of Norway (grant no. 233975/H20).