An Ethics of (not) Showing: Citizen Witnessing, Journalism and Visualizations of a Terror Attack

ABSTRACT

This paper explores how news organizations visualize crises in a digital media ecology shaped by citizen witnessing and widespread image circulation on social media. Empirically, the paper draws on the events of 7 April 2017 in Stockholm when a hijacked truck plowed into crowds, killing five and injuring several others. The study is informed by theoretical perspectives and research on citizen photojournalism and witnessing, and examines the function of visual citizen contributions, degrees of explicitness in visual coverage and the impact of proximity on visualizations of crises. Qualitative analyses of visual content and text in digital and print editions of four Swedish newspapers showed both enhancing and featured positioning of visual eyewitness contributions from the public, few examples of explicit imagery, and journalistic commentary foregrounding ethical dimensions of showing and seeing. Based on the findings, the author considers journalism’s civic response as a strategy for staking a claim to credibility and ethics at a moment when the journalistic gatekeeping position is called into question.

KEYWORDS:

• Ethics
• gatekeeping
• newspapers
• photojournalism
• Sweden
• witnessing
• terrorism

Introduction

In the afternoon of 7 April 2017, a delivery truck careened down a busy pedestrian mall in the center of Stockholm, mowing down people in its wake before crashing into a department store after about 500 meters. Five people including an 11-year-old were to lose their lives and several more were injured. Later that afternoon, the Swedish prime minister, Stefan Löfvén, in a televised address, referred to the event as an act of terrorism, stating, “Sweden is under attack.” The police apprehended the fleeing driver the same evening, assisted by tips from commuters who had spotted the person’s likeness to public-transportation system CCTV footage released to the news media. While the identity of the driver was not revealed at first, it became public once he claimed responsibility for the attack on behalf of the Islamic State-a claim that was later discounted as it became clear he was acting alone; an Uzbek citizen whose asylum application had been denied by Swedish immigration authorities, he was later tried and convicted on several counts of terrorism.

The events in Stockholm are reminiscent of several other attacks in cities across the world where, in a familiar pattern, vehicles are weaponized with the purpose of striking members of the public. While professional photographers were on the scene in Stockholm during or shortly afterward the attack, several members of the public who had escaped unharmed or witnessed the attack photographed or recorded the scene with their smartphones. Parallel to the breaking-news coverage, explicit imagery of casualties went viral on social media, bypassing the boundaries of showing set by the news media. Research has shown that visual contributions from the public, in particular in crisis coverage, present journalism with opportunities as well as challenges. Imagery made by witnesses or others affected by an event may offer alternative and even more credible perspectives to news coverage (Allan 2014; Andén-Papadopoulos and Pantti 2011; Puustinen and Seppänen 2011). Yet such content may also pose challenges to news organizations through the explicit nature of the imagery (Allan 2014; Andén-Papadopoulos 2013). Thus, at a time when visual content frequently circulates on social media first, the gatekeeping position of news organizations is being weakened by a fast-paced image culture, placing questions of ethics and journalistic power at the forefront of the production of news.

The aim of this paper is to explore journalistic strategies for visually covering crises in a convergent media ecology (Mortensen, Allan, and Peters 2017) shaped by citizen witnessing and widespread image circulation on social media. It does so through the empirical example of the events of 7 April2017 in Stockholm. Drawing on theoretical perspectives and research on citizen photojournalism and witnessing, the study considers degrees of explicitness in visual coverage (Hanusch 2012). A dual aim is to explore how proximity impacts visualizations in the news media (Hanusch 2012; Zelizer 2002). The attack on 7 April was a crisis of major national impact as the first suicide bombing causing casualties among the public in Sweden. While Swedish scholars have examined the coverage of 7 April from various perspectives, including media framing (Johansson and Truedson 2018; Nord, Färm, and Jendel 2018), this study contributes a hitherto overlooked area, namely how the news media visualized the event.

There are various terms for visual contributions from the public, each referring to different kinds of practices, purposes and visual content published by the news media (see Andén-Papadopoulos and Pantti 2011). The type of contribution in focus for this study is most closely aligned with “citizen witnessing” and “accidental photojournalism,” (see Allan 2014) that is, an unplanned recording of events as they unfold. While photojournalists also are eyewitnesses, the terms “eyewitness,” “eyewitness contribution” and “citizen witnessing” will nevertheless be used henceforth in this article in reference to members of the public who visually record an event and the visual content they contribute to the news media.

Literature Review

Citizen Witnessing, Journalism and Visualizations of Crises

Eyewitnessing as a “journalistic key word,” according to Barbie Zelizer (2007, 424–425), has remained an important part of journalism’s truth claim throughout its modern history. Photography is the primary instrument for fulfilling journalism’s claim to witness and provide evidence, while simultaneously engaging audiences through the perceived emotional impact of photographs (Becker 1992; Linfield 2010). Though not a new phenomenon, the public’s witnessing through photography and video may be more ubiquitous than ever before, according to a growing literature on the various aspects of citizen witnessing or citizen photojournalism (Andén-Papadopoulos and Pantti 2013; Patrick and Allan 2013). Writing a decade ago as one of the first scholars to address this phenomenon, Zelizer (2007) argues that journalism has outsourced and lost its claim on eyewitnessing to non-professionals in a reliance on their images and video footage from global crises. Stuart Allan, in turn, refers to the manifold perspectives and circumstances that might make people “accidental photojournalists” finding themselves at the scene of an unfolding event: “victims, bystanders, first responders, officials, law enforcement, combatants, activists or the like” (Allan 2014, 3).

Eyewitness imagery, easily recorded, transmitted and shared with digital technology, is a potential source of live material when journalists are not on the scene (Allan 2014; Andén-Papadopoulos and Pantti 2011; Caple 2014; Mortensen 2015; Mortensen and Gade 2018; Pantti 2013). Furthermore, the first-person perspective and a close-up “amateur” perspective, especially in video content, has been perceived by journalists as well as some audiences as emotionally authentic and in certain contexts more credible than professionally made content (Pogliano 2015; Puustinen and Seppänen 2011; Williams, Wahl-Jorgensen, and Wardle 2011). Yet, citizen contributions have also been found to pose challenges to the journalistic process of authentication (Andén-Papadopoulos and Pantti 2013; Niekamp 2011; Pantti and Andén-Papadopoulos 2011; Pantti and Sirén 2015) and to journalistic ideology (Mortensen 2011, 2014; Sjøvaag 2011; Williams, Wardle, and Wahl-Jorgensen 2011). Studies examining the regular news flow have found low rates of usage for citizen-produced content, attributed to journalistic ideology (Brennen and Brennen 2015; Pantti and Bakker 2009), low image quality (Mortensen and Gade 2018; Nilsson and Wadbring 2015), and a lack of newsroom resources to verify and process contributions (ibid).

The difficulty in determining the impact of citizen witnessing on journalism may lie at least partly in the manifold motivations for recording an event, an indication that more empirical research is needed. Practices and motivations of citizen image-makers (Grayson 2015), the identification as professional photographer or citizen, as well as technical skills and resources have been found to have an impact on the imagery produced and whether the image-maker offers it to a news organization (Bock 2008; Yaschur 2012). Credibility and truth-claim, key journalistic concepts, have been addressed by scholars seeking to move beyond the professional-amateur binary (Allan and Peters 2015; Pogliano 2015). Citizen contributions have also been considered as innovative participatory practices (Aubert and Nicey 2015; Gorin 2015) and as contributions challenging boundaries between lay persons and experts (Mortensen 2015). Furthermore, citizen contributions have been found to make the journalistic gatekeeping process more transparent by challenging journalism’s “power to define what terror looks like” (Allan 2014, 146).

The moral claim embedded in looking at and showing the suffering of others has been problematized in an extensive critical literature summarized by Hanusch (2012) as either being concerned with the extent and proliferation of explicit imagery (see also Moeller 1989; Sontag 1977, 2003), or with how violence and the impact of violence are represented (see also Lutz and Collins 1993; Parry 2011). A common theme in the literature concerns the gap between the journalistic drive to show and the impact of images. Susie Linfield reflects over a much-debated conundrum: “ … Seeing does not necessarily translate into believing, caring, or acting. That is the dialectic, and the failure, at the heart of the photograph of suffering” (Linfield 2010, 33).

Proposing an “ethics of showing” on the part of the image-maker, Linfield suggests that it is the viewer who must be responsible for an “ethics of seeing” (60) and for acting on the circumstances visualized by photography. Similarly, other writings on eyewitnessing and “media witnessing” (Frosh and Pinchevski 2014) consider witnessing broadly to implicate both seeing and saying (Peters 2001). According to this perspective, witnessing is discursive and carries a civic responsibility that applies to the news media and to members of the public who may see and experience an event, photograph it and decide to share the imagery.

Yet visualizations of terror pose different concerns than those aimed at evoking empathy in a humanitarian crisis. Mitchell (2011) uses the term “The War of Images” to describe a digital context emerging after 11 September where dueling visualizations are digitally played out in the so-called “War on Terror.” An act of terror on domestic soil presents particular challenges to local news organizations. Terrorism is highly mediated, as perpetrators create their own visualizations and draw on media coverage for propaganda purposes (Mogensen 2013). Yet, news organizations are charged with informing the public in a crisis situation (Nord, Färm, and Jendel 2018) while journalists themselves are also experiencing the traumatic event (Simonsen 2015; Zelizer 2002). Thus, when visual materials from the scene are available, what news organizations choose to publish has bearing on how the public interprets the event, its meaning and impact on society.

Explicit visuals might inflame and incite hatred and stir a sense of chaos, whereas an absence of images might suggest the event had little impact, potentially raising criticism of censorship (Allan 2014; Hanusch 2012). However, research has shown that journalistic visualizations in “western” crisis reporting are generally quite restrained (Hanusch 2012; Zelizer 2005), perhaps a shift from an earlier journalistic ethos of “if it bleads, if leads.” Yet, images are also political. They tend to be more explicit with greater geographical and cultural distance, according to the literature. Western news media have been found to represent “our dead” more respectfully than the death of a “distant other” who may be represented as the enemy in conflict coverage (Fahmy and Kim 2008; Griffin and Lee 1995; Parry 2011) or as an objectified, inferior “other” victim of disasters (Chouliaraki and Stolic 2017; Lutz and Collins 1993).

News coverage frequently shows the effect rather than the moment of impact (Caple 2013; Hanusch 2012). A strategy used when reporters are not on the scene, or when the event may be unfolding or is not easily photographed, metonyms also work as powerful symbols in journalistic coverage. Indirect visualizations are tools used for not showing the impact of violence and war, employed for ethical concerns, out of concern with audience reactions or for other reasons (Hanusch 2012; Zelizer 2005). In an analysis of “how news images move the public,” Zelizer (2010) found that journalism, faced with the choice of representing the world as is or as if, chooses as if. The subjunctive mode, in this context, is characterized by open-ended images leaving room for viewers to imagine the outcome. Zelizer’s concern is how cultural forces shape the public’s understanding of political processes through such visualizations.

In conclusion, the virality of image flows poses challenges to journalistic witnessing and editorial deliberations concerning images. Photographs and video tend to circulate on social media first, potentially impacting newsroom selection processes. As Mortensen, Allan, and Peters (2017) note, in a study of the photographs of the Syrian boy Alan Kurdi, “Longstanding conventions tend to unravel in digital contexts” (75). Specifically, these authors argue, the question for news organizations was not whether to publish the photographs of the drowned three-year-old boy since they were already available on social media. Rather, the issue was “how to present them so as to direct the symbolic impact of their power” (ibid).

Study Design and Method

This study explores challenges facing contemporary journalism to visualize crises. It does so through an analysis of the events of 7 April 2017 in Stockholm. Specifically, it examines how images and video made by members of the public contributed to the coverage, and how the attack and its impact were visualized by the news media, especially in terms of its human impact. The study sets out to answer three research questions through analyses of visual and written content published in four Swedish newspapers. The first question addresses an anticipated contribution from citizen eyewitnesses by focusing on the function of images. The aim is to understand how citizen visuals contributed and the journalistic strategies for managing the content. The second research question focuses on degrees of explicitness and journalistic techniques for managing and opening spaces for showing, seeing and withholding. Based on the literature and the photo policy of Swedish news organizations that protects the privacy of victims, few explicit images of casualties were expected to emerge. However, as Hanusch (2012), notes, the interest is not quantity, rather how the impact of crisis is visualized in the public sphere. The second question also explores how the response of society was visualized by the newspapers. The third research question examines written commentary and articles from the empirical materials addressing editorial deliberations and other visualizations of the event.

Research Questions

1. How did the newspapers use contributions from the public in their visual coverage of the attack and its aftermath?

2. How did the newspapers visualize the impact and response to the attack?

3. How did the newspapers address ethical aspects of visualizing the impact of the attack?

A majority of the Swedish public received information about the attack from news sites (Nord, Färm, and Jendel 2018), a motivating factor for the choice of newspapers as empirical material. Four national-circulation newspapers with a combined wide reach were selected: Dagens Nyheter, the leading-circulation legacy newspaper; Svenska Dagbladet the second-largest legacy daily in Stockholm; and the tabloid-content dailies Aftonbladet and Expressen. Dagens Nyheter and Expressen are owned by the Swedish media company Bonnier AB, while the Schibstedt Media Group of Norway owns Svenska Dagbladet and Aftonbladet. All four newspapers have the resources to cover national stories and to produce their own visual content, another important criterion. Aftonbladet and Expressen, each other’s main competitor, are known for their breaking-news emphasis and an extensive use of visuals, and both newspapers regularly invite contributions from the public (Nilsson and Wadbring 2015). All newspapers are based in Stockholm, and had reporters and photographers on location during or shortly after the attack. The use of images and judgments of news value and privacy were expected to differ between the legacy titles, expected to be more restrictive, and the tabloid-content titles, expected to be more explicit. However, while the analysis compared the publications, the main purpose was to learn how the newspapers, taken together, handled eyewitness contributions and boundaries for visualizing the attack and its human impact.

While research has shown that citizen contributions are most likely to appear on digital platforms (ibid), the print edition was included to allow for an analysis across platforms. All images relating to the attack and its aftermath published in the four newspapers’ digital editions on 7–9 April and in the print editions on 8–10 April were included. The rationale for the chosen time frame was the study’s focus on the attack and its immediate impact. Furthermore, ethical questions regarding the publication of content were deemed to be most likely to emerge within this time frame, which is also when any contributions from the public was deemed most likely to appear, following the literature on eyewitness contributions.¹ A total of 694 visual entries were identified in the digital editions for the four newspapers combined, and a total of 401 images were identified in the print editions on the selected dates.² This content formed the corpus for the visual analyses.³

In order to address the first research question, the 694 visual entries identified in the digital editions and the 401 images identified in the print editions were examined for eyewitness content. Since the study focuses on breaking-news coverage and coverage of the immediate impact of the attack, it was deemed plausible that eyewitness contributions primarily would relate to casualties or to the experience of being caught in the middle of the attack or its aftermath. Research has found that visual contributions from the public tend to fall in three categories: eyewitness content from breaking news, and two categories of little relevance to this study and therefore not included in the analysis: private imagery—such as family portraits—and social-media content that have little relation to journalism (Brennen and Brennen 2015; Nilsson and Wadbring 2015; Pantti and Bakker 2009).

A simplified content analysis was conducted in order to chart the extent and type of published eyewitness content and to provide a basis and structure for the qualitative analysis.⁴ The quantitative analysis noted date, source,⁵ topic, single or repeated publication, cross-format and title publication, image size, page for print edition and placement—such as whether a stand-alone image or video, or inclusion in slideshow or other form of presentation.⁶ The results were not analyzed quantitatively and were not presented in tables or graphs since a quantitative analysis was not deemed suitable for the study’s focus on how eyewitness content was used in the coverage. Thus, the materials were analyzed qualitatively through a close reading of images in published context.⁷ In order to explore a potential variety of uses of eyewitness imagery, the analysis focused on salience (Becker 2000), information value (Caple 2013), composition and technique (Langton 2009) and distance (Caple 2013; Kress and Van Leeuwen 2006). Salience identified the most prominent uses of the published eyewitness visuals and provided a focus for the presentation of the findings. Information value was applied in the analysis to visual information about a news event, including time, place, and moment. While highly informative imagery may construe high news value, less informative imagery can also present high news value as unique or as the only visualization available. Composition and technique, while part of the information value of news photographs, were considered for the purpose of this analysis as communicative tools including framing, angle, perspective and other technical and aesthetic means. One characteristic of eyewitness content, according to the literature is an “amateur aesthetic” with imperfect framing and out-of-focus shots. Distance was applied to distance between the camera and the photographic subject, ranging from intimate (close-up) to personal and social with increased camera distance. While a range of camera distances are used in news photography, an intimate camera distance has been associated with citizen contributions. Two descriptive terms were applied to the use of eyewitness imagery more broadly: featured for imagery foregrounded in the story, and enhancing for content supporting an aspect of the coverage; the terms were deemed useful for analyzing what made citizen contributions stand out in published context.

In order to answer the second research question, a series of qualitative analyses were conducted on photographs and video clips published on the selected dates in digital and print editions. The corpus included imagery made by the newspapers’ staff, freelancers and agency photographers as well as eyewitnesses. The analysis applied Hanusch’s (2012) five categories of visibility of death expanded here to include injuries: Implied death or injury, where the impact is “indicated” but not shown, for instance by showing covered bodies, Fractional death or injury, where only a part of a body is shown, or casualties are shown blurred or at a distance; Masked death or injury, showing bodies but with faces covered or not shown; Unveiled death or injury, showing casualties uncovered; Contorted death or injury, the most graphic, showing bodies, often being handled (662). A second part of this analysis focused on the response to the attack by examining the relationships within the photographic frame, specifically whether the deceased or injured person is alone or accompanied and tended to as well as the expressions and actions of those individuals. The purpose was to analyze the visualization of the public response as part of the immediate impact and the visualization of the victims. The analysis was further informed by newspaper design and framing techniques for images in the editing and production process, which were noted—including cropping, digitally pixelating, image size and placement—such as on page one or on inside pages of the print edition.⁸

The third research question was addressed through a qualitative analysis of articles addressing the selection process for images, specifically deliberations about whether to publish or withhold explicit imagery from the event and eyewitness content more broadly. The purpose was to ascertain if and if so how journalists articulated their role and responsibility as gatekeepers to readers. This analysis draws on the work of Mortensen, Allan, and Peters (2017), who in turn apply Erwin Goffman’s notion of the “front stage” to journalism as a performative process. The focus in this study is journalists’ discourse about the editorial process and about claims of journalistic credibility and ethics. The empirical material consists of relevant articles published in print editions on the selected dates. The time frame was chosen to coincide with the attack and its aftermath, and with the controversy surrounding the virality of social media images in the days following the attack that caused widespread rumors and discussions about the ethical boundaries of social media.⁹

The findings are presented in the following section, opening with eyewitness contributions (research questions 1) followed by the newspapers’ visualizations of the attack and its impact (research question 2), and concluding with how the newspapers in published articles addressed visualizations of the event (research question 3).

Findings

The Feeling of Being There

All four newspapers published eyewitness imagery in digital and print editions, most prominently on the day of the attack and the following day with similarities as well as differences in uses of the content. The most extensive use of eyewitness visuals was as expected found in the digital editions of the tabloids. Among the visuals identified as news photos of the attack and its aftermath in the digital editions of Aftonbladet and Expressen combined, 60 unique visual entries were identified as eyewitness news photographs or video. Dagens Nyheter and Svenska Dagbladet ran a combined 11 unique eyewitness visuals in their digital editions. A smaller number of visuals published first in the digital edition ran the following day in the respective newspaper’s print edition: 24 in total for all four newspapers, of which 13 were published in Aftonbladet, 15 in Expressen, five in Dagens Nyheter and one in Svenska Dagbladet. The relative prominence of this content in the tabloids was expected given that both newspapers solicit eyewitness visuals for their regular coverage (Nilsson and Wadbring 2015). Eyewitness contributions included video of the impact, footage and stills of people fleeing the scene, and to a lesser extent video and stills of injured people, first responders and eyewitnesses. The most salient aspects of eyewitness visuals in terms of their use in the coverage were: video clips as part of the breaking-news coverage, photographs positioned either as lead images or as part of visual coverage in digital and print-editions, and photographs included in slide shows in the digital editions.

Eyewitness video was featured in breaking-news stories uploaded shortly after the attack and posted as screen shots. The content was featured through repetition, prominent placement and presentation. Video clips also filled an enhancing function, either as teaser on a digital site or on the cover of a print edition the following day. The clips, characterized by a blurred image quality due to camera movement or lack of focus, conveyed the feeling of being on the scene (Andén-Papadopoulos and Pantti 2011; Puustinen and Seppänen 2011). Some clips contained little information about time, place and moment yet provided a unique perspective. This was found for example in the early live TV coverage of Expressen.se, where a TV-camera zoomed in on a smartphone playing a clip showing crowds running from the scene as the reporter interviewed an eyewitness.¹⁰ Thus, the use of eyewitness visuals in the early coverage corresponds with findings in the literature in its prominent use of video (Niekamp 2011; Zelizer 2007).

High information value in reference to time, place and moment and with foregrounded aspects of composition was found in some photographs featured through placement, image size, headline or accompanying story. One of the most circulated eyewitness contributions showed people fleeing the scene with the burning truck in the background. The image’s provenance emerged on the second day of its publication in Dagens Nyheter, here accompanying an interview with the image-maker, a survivor of the attack. Another example of eyewitness imagery with high information value was a birds-eye view of the hijacked truck crashed into the department store Åhléns. Published as the lead cover image of the Dagens Nyheter 8 April early print edition, the image was made from an office building across the street. It provided a panoramic overview of the impact of the attack, a function filled by eyewitness imagery elsewhere in the coverage. ¹¹ While few photographs with foregrounded information value and composition were found in the examined materials, they challenge the notion that citizen contributions are of lesser quality (Mortensen and Gade 2018; Nilsson and Wadbring 2015).

Another function of eyewitness imagery in digital editions, in particular in aftonbladet.se, was to integrate still images in slideshows augmented during the afternoon of 7 April and the following days, a form regularly used by the newspaper and a common presentation of visual content on digital news sites (Caple and Knox 2012). Eyewitness shots were inserted into slideshows along with staff or freelance images, contributing different perspectives of reactions to the attack or of first responders attending to casualties. In this published context, eyewitness imagery had an enhancing function, as did the professional images included in the slide shows (ibid). In the examined sequences, some eyewitness images contained a high level of information while others, though less informative, contributed a unique perspective, such as of an arrest of a possible suspect.¹²

In contrast to findings in previous studies, few examples of an intimate close-up perspective were found among the 7 April eyewitness contributions (Pogliano 2015; Puustinen and Seppänen 2011). A number of eyewitness images and videos were made at a social or long distance, either providing a birds-eye view or a panning perspective of a chaotic scene. This may be the result of a lack of access or of a desire to flee the scene. Photojournalists, in contrast, are trained to walk towards the scene (Yaschur 2012). Furthermore, technical limitations and choice of camera mode may have an impact on the suitability of the content (Nilsson and Wadbring 2015). There were few still images among the published eyewitness content over all. In contrast to video, photography forces the image-maker to focus, frame and interact with the situation, which may not have been a possibility for eyewitnesses caught up in the event (Yaschur 2012).

An image-maker’s motivation and resources and the circumstances of the event may shape the content produced as well as its journalistic usefulness (Bock 2008; Grayson 2015; Yaschur 2012). Given the proximity of the newsrooms and the central location of the scene of the attack on 7 April, reporters and photographers arrived soon after impact. This is in contrast to situations where eyewitness content provides the only visual coverage. The circumstances on the site of the attack may also have contributed to eyewitnesses’ experiences soon after impact as the area was cordoned off as a crime scene and triage area. Furthermore, a high threat level remained in place throughout the afternoon and early evening as the downtown area was virtually evacuated by the police. Yet another possible factor is a lack of interest among the public to offer imagery to news organizations (Nilsson and Wadbring 2015). It is also possible that some eyewitness content was deemed too explicit, a question that will be addressed in the following pages. Nevertheless, the analysis found a variety of uses of eyewitness visuals with featured as well as enhancing placements, in contrast to previous research that found a marginal presence of eyewitness imagery in news coverage (Nilsson and Wadbring 2015; Pantti and Bakker 2009). Furthermore, one of the newspapers, Dagens Nyheter, relied solely on eyewitness visuals for its early breaking-news coverage from the scene of the attack. While the newspaper’s staff photographers covered other aspects of the story, its early edition did not include the newspaper’s staff imagery of casualties and those injured.

Visualizing the Human Impact

All four newspapers covered the event and its aftermath extensively. The impact of the attack, the search and the capture of the suspect, and public and official responses were part of the breaking-news coverage. During the following days, coverage focused on terrorism and the motivations of the suspect, the victims, public reactions expressed at the site, and the response of civil society. Continued coverage addressed causes and political responsibility, the ongoing investigation and terrorism. The biographies of the victims, published after families had been notified, was another focus, as were public displays of emotion and interviews with witnesses and first responders relating their experiences. The four newspapers relied on staff photographers, freelancers and agency imagery as well as citizen contributions for their visual coverage. This analysis focuses primarily on imagery made at the scene of the attack.

Images of casualties were published in digital and print editions of the four newspapers. Drawing on Hanusch’s categories of explicitness, almost all images of victims showed implied death. Bodies of those killed were shown either covered or in body bags, shielded from view or with most exposed body parts or identifying personal objects pixelated. In some cases, explicit video was only published in part or as selected freeze frames, such as a series of screen grabs from an eyewitness video published in Dagens Nyheter digitally and in print along with an article headlined: “Mobile film from the attack shows scenes of horror from the street” (8 April 2017). Each frame was edited and digitally pixelated, rendering bodies un-sharp in a kind of partially unveiled death. The sharpness of the original video could be discerned from a thumbnail-size freeze frame from the video published on another page the same day. Placed inside a map of the route of the attack, the thumb-nail was sharp, though reduced in size so as to not expose the casualties depicted in it.

All four newspapers pixelated faces of injured people in pictures published in digital and print editions. There were some exceptions to this implied and masked treatment found in digital editions, including in a svd.se image of an injured woman seated on the pavement peering at her blood-covered foot, her face exposed and unveiled to the viewer, with the remains of a dog killed in the attack laying in the foreground. In later digital editions and in the Svenska Dagbladet print edition the following day, the dog had been cropped out and the woman’s face pixelated. This adjustment shows how the visualization became more discrete as the coverage progresses and images were re-published, adjusted, cropped and edited, a common picture editing technique. The two tabloids showed the impact more explicitly, such as on the Aftonbladet 8 April cover where the aforementioned image of the injured woman included the blood-covered pavement in the foreground, though without the dog. The Aftonbladet print cover was full-frame, wrapping front and back and including other witnesses looking on in apparent horror.

A notable exception to the predominantly implied death appeared in an Aftonbladet publication of an image of a dead suicide-bomber who had attempted to plant a bomb on the same street, Drottninggatan, in 2010. That attack had failed, causing the suicide bomber to die from injuries sustained in the blast but resulting in no other casualties. This archival image was published on 8 April 2017 alongside an article about the history of terrorism in Sweden noting that the 2010 attack was the first suicide bombing though not the first terrorist attack in the country. The re-publication of this birds-eye view of the dead bomber—his blood-covered body partially covered and his face exposed—stood out in the examined coverage as the only image of contorted death. ¹³ A difference in treatment of casualties in distant settings appeared in another publication of archival imagery, in a Dagens Nyheter article on the history and psychological effects of terrorism. The article was illustrated with a reproduction of the paper’s coverage of past terror attacks, including the 2016 attack on the Brussels airport. The image dominating that day’s cover depicted two injured women, one of them with apparently serious injuries, her face and injuries uncovered and bloodied and her clothes torn, leaving her body partially exposed (dn.se, 8 April 2017). The contrast to the implied injuries in the paper’s coverage of the Stockholm attack reveals the more restrictive boundaries with greater geographical proximity (Hanusch 2012).

Relationships within the photographic frame were a significant part of the newspapers’ visualization. Casualties and injured people were consistently shown accompanied by others. First responders were shown either administering life-saving treatment to individuals shielded from view, or as having abandoned those efforts, indicated by a body bag covering the shape of a body or by the first responder walking away. Injured individuals were shown being carried to safety or tended by other members of the public, carried into an ambulance by medical staffers or assisted by the police. The faces of people helping those injured were frequently in focus in these photos, showing a range of expressions in reactions to the event. Onlookers not participating in the efforts also were shown reacting to the impact. Taken together, these physical and spatial relationships visualized the impact as including the public’s response. Furthermore, the response of members of the public and first responder came into greater focus since the faces of the victims were shielded through photo editing, pixellation or cropping. The complexity of these visualizations underscores Hanusch’s argument that it is important to consider how casualties and the impact of a crisis are represented, rather than merely focusing of frequency or explicitness.

The most explicit images showing the immediate impact and casualties or injured people, were less prominent in the coverage in the following days as imagery from 7 April was republished in a continued coverage. The most repeated images were of first responders and members of the public aiding those injured. Other prominent visualizations were of the shrines and floral offerings appearing and growing once the cordoned off area had been opened to the public the day after the attack. The boarded-up windows of the Åhléns department store where the truck had struck quickly became a focal point for people filling the wall with post-it notes and flowers. Furthermore, the print-edition covers from the examined dates visually foregrounded this shifting visual theme, from an initial focus on human impact and trauma to resilience and unity. In an analysis of the U.S. coverage of 11 September, Zelizer found a relatively narrow, thematic visualization in the journalistic coverage. This “play to a narrowed visual aesthetic” (Zelizer 2002, 116), Zelizer argues, reflects a society’s desire not to see the gory details, but rather to focus on symbols and places, most prominently the Twin Towers as a way to process trauma. A similar process might have been at play in Stockholm where the sites of commemoration became a theme in the coverage. In an analysis of media frames in Swedish news coverage of 7 April, although not focusing on visual frames, Nord, Färm, and Jendel (2018) found a unity frame to be more prominent than a chaos frame, which they interpreted as linked to a political discourse of unity. The visual themes found in this study resonate with those findings.

An Ethics of (not) Showing

The newspapers’ selection process for visual content was addressed by two articles, one in Expressen and the other in Dagens Nyheter, both focusing on decisions to withhold visual materials. Decisions to actually publish explicit imagery were not addressed and explained in writing in the examined materials. Expressen’s editor-in-chief on 8 April addressed the image-selection process in commentary running alongside a staff photograph dominating a two-page spread of the print edition; the image showed the scene of the attack with first responders attending to victims while crowds of eyewitnesses looked on. Noting that the staff photographer who made the picture had not been on assignment but rather happened to find herself on the scene when the truck struck, the editor reflected: “Thankfully she is alright.” Commenting on her unpublished photographs, he continued: “It is not possible to publish some of her pictures. They have great news value and the public interest is obvious, but–No.” While not fully explaining what tipped the scale in favor of privacy, the phrasing of the explanation given suggests a difficult deliberation process.¹⁴

A Dagens Nyheter article from 8 April accompanying the publication of the eyewitness video discussed previously addressed the explicitness of the materials, which according to the article was offered to the newspaper by a Turkish news agency. “The images are so strong that DN has chosen not to publish the film,” the journalist explained, while at the same time describing its content, noting details of the damaged bodies of several of the victims. This article was also published in the digital edition dn.se on the day of the attack, along with one unsharpened freeze frame. Thus, Dagens Nyheter made it clear, in writing as well as in the out-of-focus visuals, that eyewitness footage was available though not deemed appropriate to share in its entirety.

The trauma of witnessing and the ethics of image circulation on social media were addressed in a total of three articles, prompted in part by the virality of images of casualties over the weekend following the attack. A Svenska Dagbladet article cautioned against sharing imagery of those injured since it may lead to their identification before families had been notified (8 April 2017). Dagens Nyheter advised parents about how to speak to children who may have witnessed the attack or seen or shared viral imagery on social media (10 April 2017). In another article, Dagens Nyheter (also on 10 April) addressed fact checking and asserted the integrity of its own coverage by showing—in writing and by providing visual examples—how news photographs were being distorted and spread as propaganda in forums on social media.

The responsible posture of the news organization and their coverage of the event were further asserted by two editors-in-chief, who in their respective articles criticized social media providers for allowing viral content of victims to circulate on forums accessible to children. A Svenska Dagbladet editor writing as a parent referred to his 14-year-old son and the fear of his exposure to harmful content on social media (10 April 2017). The Expressen editor asserted high public confidence in news coverage based on a recent survey (10 April 2017). While not specifically referring to visual coverage of the 7 April attack, this statement nevertheless contrasts the ethical boundaries set by news organizations with an absence of boundaries on social media.

The examined articles, seven in total, expressed an ambivalence about photography and the impact of images found in previous research (Zelizer 2005). However, in contrast to journalistic motivations for publishing controversial imagery (Mortensen, Allan, and Peters 2017), editors discussed withholding certain visuals from 7 April, not publishing them. This might explain why few articles addressed the ethics of publishing explicit visual content. Another possible explanation is a cultural consensus about the appropriateness of withholding content (Hanusch 2012). However, journalists addressed boundaries and ethics for showing in articles about the virality of imagery on social media and the manipulation of photographs for propaganda purposes. While not specifically acknowledging a weakened journalistic gatekeeping power (Schwalbe, Silcock, and Candello 2015), the examined articles in effect posit that the public are today’s gatekeepers and that this brings a responsibility connected not just to viewing but also to sharing images.

Conclusions and Discussion

Research has shown that proximity to an event has an impact on image selection in crisis reporting (Allan 2014; Hanusch 2012; Zelizer 2002). The analysis of the events of 7 April 2017 in Stockholm confirmed those results. However, the degrees of explicitness found in the analyses revealed a tension between showing and shielding. For instance, one image may be both shielding (a victim) and exposing (the visible face of an eyewitness aiding a victim), creating a visual tension where readers are guided to connect with the eyewitness reacting to a scene partially hidden from our view. However, the photoshop functions visibly used to shield readers from the sight—which the eyewitness appearing in the image obviously has seen, as has the journalist editing the picture—exposed the journalistic boundaries set for showing and seeing. The decision to actually publish such images rather than withholding them may be seen as a visualization of the power asserted by journalists to determine the appropriateness of showing “what terror looks like” (Allan 2014). In this respect, the findings suggest that journalists asserted gatekeeping powers rather than ceding them to practices and visualizations outside the newsrooms. In short, they can be said to advocate an ethics of not showing.

At the same time, published commentary about the selection process and the ethics of photographing and showing indicate an awareness that the journalistic gatekeeping power is being weakened. Anyone who wishes to see what terror looks like can google it on the internet or connect to certain social media forums. Furthermore, in a crisis covered as local news, numerous readers have personally experienced and perhaps visually recorded the event. Thus, they have seen and perhaps even shown others what terror and its impact look like to them. This may explain why the newspapers commented on the decision to withhold visuals from publication.

The findings raise questions more broadly about the responsibility of journalism and its claims of credibility. One strategy used by journalists on 7 April was to cast themselves as citizens affected by the trauma rather than as journalists. This was done by foregrounding affect and journalists’ reaction to being on the scene and to viewing explicit images during the selection process. Thus, it can be argued that journalists sought to claim credibility by assuming a civic perspective. Yet another strategy, which may be understood as pedagogical, was expressed by discussing the morality of sharing visual content, highlighting the responsibility of parents, and creating awareness of how news coverage is used in propaganda. Thus, journalists were making the argument that the public should adhere to an ethics of seeing and showing (Linfield 2010) that is outside the control of journalism. Furthermore, through a restrictive selection and presentation of visual content in published context, newspapers created a contrast between their stance as ethical in contrast to that of actors on social media.

Explicit visual content circulating outside the examined newspapers did not appear to have a significant impact on how the newspapers visualized the event, in contrast to findings in the literature on citizen witnessing. This may be due to a cultural consensus found in previous research regarding the appropriate of explicitness, which in turn may lead eyewitnesses to choose not to offer images to news organizations (Nilsson and Wadbring 2015). An argument raised in the literature on citizen witnessing is that it offers alterative perspectives and forces a kind of transparency by exposing assumptions in the journalistic production process and framing of events (Allan 2014). This could be discerned in the examined articles addressing decisions to withhold images and in articles where journalists positioned themselves as members of the public first. However, the pressure for transparency seemed to emerge from a wider visual culture—such as image sharing on social media—rather than from materials submitted to the newspapers. Further research into citizen witnessing as a practice would be welcome to re-evaluate discourses in the literature about the motivations and practices of citizen witnesses.

A newsroom policy protecting the privacy of victims and cultural perceptions of appropriateness appear to have guided the presentation and treatment of visuals in the examined materials. A few explicit images of victims of international crises stood out in contrast, an indication that this protection is weakened with increased distance. More broadly, 7 April exemplifies an embedded contradiction in how journalism visualizes crises depending on context. A visualization of resilience and a functioning society emerged already on the first day of coverage alongside visuals of the trauma in the local coverage examined here. Further research might look into whether a similar sense of resilience is visually represented in reporting of distant crises or whether the focus, determined by high news value, emphasizes chaos over unity.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. The digital editions were viewed through links to the newspapers’ coverage of the attack published on the respective sites: (dn.se, svd.se, aftonbladet.se, expressen.se). The print editions were viewed in the digitized newspaper archive of the Royal Library in Stockholm, and through supplementary searches on the Media Arkivet digital archive that contains an archive of Swedish news media content.

2. Photographs, video and digital TV-news entries published in the digital editions of the newspapers were analyzed. Each video clip was counted as one entry, and each TV-news segment was also considered one entry.

3. Nearly all print-edition images were published the previous day in the digital edition. Furthermore, while both print and digital editions republished various images, and images were published across platforms and across titles, the total number of unique published visuals for the selected dates is estimated to be below the noted print-edition number.

4. Some eyewitness content was likely unaccounted for since the contributions were not always identified as such or bylined in the examined newspapers. In cases of uncertainty, the imagery was excluded and not counted as eyewitness content. Yet, in some case it was possible to identify the provenance of imagery not bylined, for example through interviews with eyewitnesses published along-side an eyewitness video clip. In those cases, the imagery was counted as eyewitness content.

5. Each eyewitness entry was counted as one even though a TV-news segment most likely included several visuals. However, when eyewitness content appeared in a TV-news segment, each image or clip was counted separately. The rationale for this choice was the study’s aim to identify the function of eyewitness visuals rather than the total number of such visuals. The consequence of this strategy is that the total number of visual entries in the corpus does not accurately reflect the actual number of visuals.

6. It was not possible to examine the layout of the digital editions since the available materials consisted of links to individual articles.

7. Headlines, captions and, where relevant, articles were included in the analysis for contextual information and in order to trace the function of the visual content in its published contexts.

8. For instance, one image might be used in several forms or titles, either tightly cropped or with a detail enlarged or foregrounded.

9. Relevant articles were identified by examining all print editions for the selected dates, and through a complementary search in newspaper archives designed to find any overlooked materials. The search terms used for the digital search focused on social media, ethics, photograph, video, and image or photo sharing.

10. Svd.se and dn.se, the respective digital editions of Svenska Dagbladet and Dagens Nyheter, did not run web-TV content during the examined time period—an indication of differences in the allocation of resources.

11. The image was published on the digital news site on the day of the attack, appearing next to a staff image showing first responders and police on the by-now closed-off street. On 8 April, as previously noted, the same eyewitness image, blown up full-page size, covered the front page of the Dagens Nyheter early print edition as the featured visualization of the event.

12. While svd.se also published slideshows from the attack, they were exclusively made up of staff imagery.

13. A more explicit photograph of the dead suicide bomber won an Aftonbladet photographer the Swedish Picture of the Year award in 2011. While the choice of the tabloid-content dailies to publish images of the deceased bomber with his face uncovered caused some controversy, the community of photojournalists, with this award, supported its publication and recognized it as the most important news image of the year.

14. The same article defends the newspaper’s decision to on 7 April publish the CCTV imagery of the suspect—a decision also made by the other three newspapers—noting that normally, this would be a controversial decision. The comment may be understood in reference to the extraordinary situation and, in the editor’s view, the news media’s responsibility to depart from the usual policy of Swedish news organizations to protect the privacy of suspects as well as victims of crime. The privacy of the suspect was protected in examined coverage throughout the two days following the attack, as private imagery published in the coverage was digitally pixelated and the developing story referred to him as “the 39-year-old suspect.” On 10 April, three days after the attack, the tabloids Aftonbladet and Expressen published his name and photographs with identifying details uncovered, while Dagens Nyheter and Svenska Dagbladet protected his privacy.