Anti-refugee Mobilization in Social Media: The Case of Soldiers of Odin

Mediation and Protest

Social movement researchers have developed the concept of political opportunity structure to understand the various mobilization strategies and opportunities of social movements in relation to established political structures and institutions (Tarrow, 1994). Political opportunity structure refers to the “dimensions of political environment that provide incentives for people to undertake collective action by affecting their expectations for success or failure” (Tarrow, 1994, p. 85). It primarily stresses the importance of factors external to activists within social movements (such as economic, social, political, symbolic, or situational factors) that influence and affect collective action and the development of social movements (Cammaerts, 2012). For example, mainstream media coverage has historically had a profound impact on the development of social movement practices and on the political outcomes of protest mobilization (Gitlin, 1980; Rucht, 2004). Therefore, to further develop the understanding of (digital) media and communication in relation to the political opportunity structure of social and protest movements, Cammaerts (2012) advances a model that explicates the dynamics of media and communication, both external and internal, to movements and mobilization. The model distinguishes between three different forms of mediation opportunity structures: (1) the media, (2) the discourse, and (3) the networked opportunity structure. These are interrelated in practice but could be analytically separated for the purpose of explaining different rationales and strategies of social and protest movements in various contexts. The three forms of mediation must also be understood in relation to three contemporary logics of protests ¹ existing in the repertoires of networked contentious action (Cammaerts, 2012): the logic of numbers (mass demonstrations and the spectacle of numbers), the logic of damage (property destruction or large-scale disruptions which attract media attention and demonstrate seriousness), and the logic of bearing witness (public performance and civic disobedience which include staging media spectacles and invading established media structures). These three logics of protest cut across the three forms of mediation opportunity structure, that is, they can be strategically deployed in relation to various forms of mediation and networking tactics. Cammaerts’ model foregrounds the agency of social movements (particularly in relation to the public), but it also stresses the constraints of protest logics in relation to mediation opportunity structure. In the following section, the relation between mediation and protests is discussed in detail.

Mediation Opportunity Structure

Media opportunity structure refers to the mainstream media representations of protests and social mobilization, and its importance for mass mobilization (i.e., attracting numbers) and political legitimization. However, because media representation tends to “neutralize dissent” (Cammaerts, 2012, p. 122), thereby defusing the radical dimension of political struggle and making it yet another “everyday life practice” (Berglez, 2006, p. 180), social movements face a problem from a mass mediation perspective. One recurrent strategy, to avoid neutralization, has been the use of violent action (predominantly property damage and clashes with police) to attract more sustainable mainstream media attention (Ekman, 2011). Although this opportunity structure tends to make a social movement visible to the greater public over a longer period of time, a constraint is that it entails a loss of public support and, ultimately, political legitimacy. The notion of mediation opportunity also conveys the protest logic of “bearing witness” (Cammaerts, 2012, p. 124), that is, of activists staging media spectacles or other forms of performances. Here (digital) mass self-mediation for the purpose of providing content to mainstream media (visuals, video, etc.) could also be included (Uldam, 2013). Historically (i.e., pre-digitalization), news media coverage has been seen as pivotal in gathering public support and extending the mobilizing capabilities of social movements—or to put it in the words of political scientist Raschke, “A movement that does not make it into the media is non-existent” (as cited in Rucht, 2004, p. 29). Since the arrival of mass self-communication (Castells, 2013), multimodal self-mediation on various platforms has become increasingly important for social movements, but at the same time these online self-mediations “often serve the purpose of accessing traditional mass media” (Uldam, 2013, p. 63). This brings us to the second form of mediation opportunity structure.

The discursive opportunity structure refers to the various “counter-narratives to mainstream representations,” produced by protest movements “contributing to a global archive of protest artefacts and constructing collective identities through connecting the personal to the political” (Cammaerts, 2012, p. 130). Self-mediation is the locus of communication strategies deployed by far-right movements to achieve public visibility and attract new activists (Ekman, 2014a). The focus on self-mediation also originates from the hostile attitude toward mainstream media prevalent among far-right activists (Ekman, 2014b). Self-mediation includes producing and disseminating counter-discourses on various political and social issues and showcasing the ability to use force (cause damage). It also refers to the “inward looking” (Cammaerts, 2012, p. 130) mediated practices of protest movements. Self-mediation becomes a device for producing in-group solidarity and thereby contributes to shaping the identity of movements of dissent (e.g., Caiani, della Porta, & Wagemann, 2012). “Witnessing” (Cammaerts, 2012) through self-mediation is pivotal to constructing collective identities within movements. Moreover, self-mediation connects the various forms of mediated “performances” (e.g., representations of street action, spectacles, and group cohesion) to opportunities for participation. This implies that self-mediation is a sort of mobilizing tool in which visible evidence of concrete action is manifested. Self-mediation also functions as an archive for movements, preserving counter-discourses and various representations of contentious politics in time and space.

Finally, networked opportunity structure refers to the mobilizing and organizing capabilities of (digital networked) communication technologies. In particular, various uses of the Internet are emphasized in relation to protest actions, but “older” media practices, such as texting and mobile phone use, have been important in mobilization processes (Cammaerts, 2012). In recent years, commercial social media and content-sharing platforms, such as Facebook, Twitter, and YouTube, have become increasingly significant in organizing, mobilizing, and representing protest activities (Bennett & Segerberg, 2013; Kavada, 2010). Social media platforms are used, among other things, to attract attention to protest actions, mobilize activists, circulate counter-narratives, generate networked counter-publics, and create visual archives (cf., Kavada, 2012; Penney & Dadas, 2014). The use of communication technology during protests also enables various forms of “sousveillance,” that is, monitoring from below. This is particularly evident when activists film police conduct during protests and demonstrations—primarily for the purpose of exposing police violence and civil rights violations (Cammaerts, 2012).

Taking the three forms of mediation into account, Cammaerts (2012) shows how protest logics (numbers, damage, and witnessing) intersect with mediation practices in various ways. Moreover, the model outlines how protest logics can be found at all levels of the mediation opportunity structure (Uldam, 2013)—that is, in relation to strategically selected media and communication practices in and around protest events. However, since Cammaerts’ model adheres to the opportunity structures of social movements operating within the realm of civil society (i.e., having democratic and inclusive goals—albeit not always using democratic means), when repositioning the concept of mediation opportunity structure we need to take into account the specific form of politics prevailing in extremist and far-right movements. Hence, to understand the inverted logic of a mobilization with an implicit uncivil rationale—propagating violence against individuals, fomenting politics of fear and hatred, and reflecting reactionary patriarchal values—the article will now address the concept of “politics of emotion” (Ahmed, 2004b) inherent in the far-right anti-refugee discourse and practice.

Politics of Emotion and Masculine Ideals

The street politics advocated by vigilante gangs such as SoO is shaped and augmented in communicative interaction among anti-refugee groups and activists in social media. In line with Papacharissi’s (2015) argument that all social media networks are affective publics, it is important to assess the particular characteristics of both individual and structured feelings in networked communication of un(civil) character. So, to understand the underpinnings of anti-refugee discourses and sentiments, Ahmed’s (2004b) discussion of “politics of emotion” will be related to the role of masculine ideals and the aestheticization of political messages in (historical) fascism (cf., Back, 2002; Benjamin, 1968, 1930/1979). Far-right politics is shaped to fit collective emotions of loss and insecurity—creating an affective combination of (ethnic) belonging and a fear of the “other” (Ekman, 2017). Within the contemporary discursive construction of a community with closure (cf., Atton, 2006), the threatening other is embodied by the external immigrant/refugee or the internal ethnocultural/racial other. Following Ahmed’s (2004a) argument, feelings of love and hate are central to aligning the bodies of certain individuals within a community while simultaneously excluding others (p. 117). So, by feeding off of emotive responses to real events and societal conditions, affective public sentiments (such as the reaction to the attacks in Köln on New Year’s Eve 2015 or the terrorist attacks in Paris earlier the same year) are appropriated by far-right actors to legitimize street politics and violence aimed at refugees. Mirroring an increase in nationalist sentiments across Europe, a general politics of fear is advocated by various populist and far-right actors. Common to these are their emphasis on cultural differences that are conceived as essentially “insurmountable” (Balibar, 1992, p. 21). Moreover, contemporary far-right discourses and, indeed, also populist ones stress the “demographic threat” posed by immigrant groups and refugees across Europe. Far-right discourses attach essential values, alien to “ours” (patriarchal, sexist, etc.), to immigrants and stress that immigrants’ behavior is inherently violent (e.g., Said, 1981/1997). For example, violence against women is framed as an “immigrant problem” (Miles & Brown, 2003, p. 52) and immigrants and refugees thereby embody an immanent threat, “which is approaching” (Ahmed, 2004a, p. 125). Thus, by constructing objects of fear—refugees—the far-right politics of emotion can produce an endless number of bodies, onto which networked structures of hateful emotions are projected. This generates a societal climate where violent attacks against refugees can be tolerated or even be considered justified.

The political practices of the extreme right have always been nourished in milieus dominated by men, and its political ideals are related to specific forms of masculine culture. The relationship between fascism and male violence has been thoroughly discussed in relation to historical fascism (cf., Bauman, 1989; Drake, 1989/2004). Moore (1966/2004) argues that fascism’s “stress [on violence] goes far beyond any cold, rational appreciation of the factual importance of violence in politics, to a worship of ‘hardness’ for its own sake” (p. 266). In regenerated fantasies of disciplined and obedient men, far-right activists consider themselves the vanguard in defending “European” or “Nordic” values. For example, by using attacks against women as a recurrent topic in propaganda, the violence deployed (or advocated) by activists appears as an honorable, rational reaction (Ekman, 2014b). So, by claiming to protect Swedish women from alleged sexual attacks by immigrants and refugees, the far-right discourse locates the rationale of violence in narratives of self-defense and the protection of innocent (White) women. This can be traced back to the masculine ideals of “heroic realism” prevailing in the Third Reich, which idealized a hyper-masculine (historicized) naturalism (cf., Karlsson & Ruth, 1984, p.165). Kimmel’s (2007) research on former neo-Nazi activists who renounced the movement found that camaraderie, male bonding, and group cohesion are more important than actual political beliefs in the recruitment of new activists (Kimmel, 2007). So, the relationship between masculinity, violence, and far-right street politics is evident, and far-right actors feed off of a purported “crisis of masculinity” (Connell, 2002). This means that the far-right milieu also encompasses a strong aversion to feminism. Within far-right online communication, users express a perceived loss of male identity, stressing the betrayal of “weak” (White) feminized men in mainstream society (Ekman, 2014a).

Finally, the online strategies of far-right actors can be understood as a form of aestheticization of politics (Benjamin, 1930/1979). By combining an affect-laden discourse (violence against women) and performances of masculinity (the public practices of vigilante gangs), the online communication of far-right actors resonates with what Benjamin (1968) defined as the aesthetic lure of fascism. By “rendering politics through aesthetics” (Back, 2002, p. 628), far-right actors deploy new technology to create visual images of (violent) expressions—for example, images of intimidating activists dressed in uniforms and ready to patrol the streets. Back (2002, p. 633) argues that new technology “allow[s] new horizons for the expression of whiteness” and race politics, but it also enables contemporary reenactments of historical street politics embodied by the proto-fascist Freikorps in Germany, Mosley’s fascist Blackshirts in England, and the equally fascist Swedish Munckska kåren, during the 1920s–1940s. Consequently, it is important to underline the historical lineage within far-right street politics and its self-appointed task of public “policing.”

Method and Material

To assess the three levels of mediation opportunity structure, the case study will draw on material from three different forms of mediation: (1) the social media communication of SoO, (2) “alternative news” platforms pertaining to the populist and extreme right, and (3) newspaper coverage of SoO. The overall methodological approach is rooted in critical discourse analysis (Fairclough, 2010), emphasizing the relation between online communication and political practice. The three types of mediation are analyzed using different methods to assess the various genres, modalities, and sizes of the material.

First, social media content is analyzed through a purposive sample of material relating to the communication on social media platforms where SoO are active. It also includes and discusses some of the material published by monitoring organizations and investigative journalists who gained access to the “closed” domains of the network’s online communication. One of the methodological problems when assessing actors such as SoO is that their political practices and street tactics are mainly communicated in closed online groups. However, by including second-hand data, gathered by monitoring organizations and investigative journalists, the article will provide some insights into the networked strategies and the relation between the front stage and the backstage of online communication, among activists and supporters. Thus, the social media data of SoO belong to the areas of both internal networking and self-mediation.

Moreover, the analysis includes material from online platforms belonging to the populist and far-right milieu. It includes all articles (n = 29) published in four “alternative news” platforms from January to September 2016. These platforms are “alternative” in the sense that they challenge the mainstream media (e.g., Bailey, Cammaerts, & Carpentier, 2008, p. 72). However, as Downing (2000) notes, they are better understood as “repressive radical media” since they reflect discursive strategies of “authoritarian populism,” portraying themselves as “underdogs” speaking on behalf of a silenced majority (pp. 88–95). The analysis reveals how the network uses alternative media platforms to propagate an alternative discourse on vigilante politics and the use of violence targeting refugees in various cities in Sweden. The analysis focuses on the contextualization of key discursive themes (Fairclough, 1995) in relation to the SoO. The assessment of the use of alternative news platforms as a strategy of self-managed ² public visibility—filtered through the pseudo-journalism of the online platforms (e.g., Ekman, 2014a)—makes visible the discursive opportunity structure of SoO.

Finally, to assess the media opportunity structure of SoO, the last section includes a quantitative content analysis (Neuendorf, 2002) of all news articles (n = 90) covering SoO, published in the seven most widely circulated newspapers in Sweden from January to September 2016. Since mainstream coverage of protests and social mobilization is crucial for achieving broad public visibility and also for gaining (potential) political legitimacy, the representation of SoO in mainstream news becomes an important element in the process of gathering public support. The data were drawn from the newspapers’ print versions. ³ The reliability of the final coding process has been ensured in two steps of inter- and intra-reliability procedures. First, a small sample (15 units) was coded by an additional researcher for inter-reliability measurement, yielding an inter-coder reliability level of .94. In the next step—after some minor adjustments of the coding frame—the coder randomly selected 15 units and recoded (test–retest) them for intra-coder reliability. The final test showed no intra-observer inconsistencies.

Overall, the analysis reveals how mediation opportunity structure can be understood in relation to a concrete example of (un)civic mobilization and (street) politics in a period of perceived insecurity among the general public. Moreover, the transformation of immigration policies and changing state politics targeting refugees—in connection with the “refugee crisis”—provides an important context, because discursive changes or shifts (Krzyżanowski, 2013) are more tangible in times of crisis or public insecurity. The analysis is structured as follows. First, the background of SoO is explained and the social media networking of SoO is analyzed, with particular emphasis being given to the role of Facebook in the international spread of the network and how private social media communication facilitates recruitment, mobilization, and group cohesion. In the subsequent section, the self-mediation practices, both on social media and “alternative news” platforms, are analyzed. In the final section, the (mainstream) news media representation of SoO is assessed. The three sections follow Cammaerts’ (2012) model, and the mediation opportunity structures are discussed in relation to the above-mentioned logics of protest.

SoO and Networked Opportunity Structure

The SoO first emerged in the small town of Kemi, Finland, in October 2015, having been founded by a neo-Nazi activist with prior convictions. ⁴ Dressed in uniforms with SoO logos on the back of their jackets (in the same fashion as One Percenter Biker gangs), they began patrolling the streets of Kemi, claiming to protect Finnish citizens from “Islamic invaders” (Finnsiö, 2016). The group quickly attracted attention, and SoO groups began to emerge in other Finnish cities in the latter half of 2015. They became especially noticeable in cities where groups of asylum seekers were located as a result of the ongoing refugee crisis. The group was supported by some officials from the right-wing populist party The True Finns, and they were actually greeted positively by the Minister for Justice (a member of The True Finns). The minister not only expressed approval but also stressed that the real problem was not SoO, but that the “Finnish people felt unsafe.” The organization was condemned, however, by his governmental colleague, the Interior minister (Finnsiö, 2016). By profiling themselves on social media platforms such as Facebook, SoO quickly spread to other Nordic cities, and in January 2016, the first Swedish group appeared in the nation’s capital, Stockholm. In a short time, the network set up Facebook groups for different cities in the country as an attempt to engage activists in mobilizing and coordinating street patrolling. The first street patrolling activity took place in Stockholm and attracted a lot of public attention. However, the network was only loosely coordinated by a Swedish leader and had problems sustaining activities in more cities. The network conducted street patrolling in various cities during the first half of 2016. Even more local groups were visible on Facebook, reaching about 20 local chapters. The number of “supporters” on Facebook exceeded 7,000 and the national group remained active during most of 2016. However, in the latter half of 2016, the activities of the network declined, and at the time of this writing, there only seem to be minor coordinated public actions. This also points to the volatility and temporality of sociopolitical activities centered on networked communication—in contrast to more traditional forms of grassroots mobilization and organization building.

Social media communication (particularly on Facebook) has played an important role in engaging activists and facilitating contact between those willing to participate in street patrolling and those organizing the activities of the network, thus creating opportunities for recruitment and mobilization (Cammaerts, 2012). Facebook groups have also functioned as important environments for in-group discussions. However, these groups have been private, with access being granted by the (informal) network leadership. Within these closed Facebook groups, racist sentiments have been widely disseminated among members. Two journalists from the largest Swedish tabloid Aftonbladet infiltrated SoO and revealed various conversations from closed SoO groups on Facebook (Röstlund & Wiman, 2016). The conversations between members and sympathizers contained utterances familiar to observers of neo-Nazi extreme right discourse. Examples are statements that “traitorous politicians deserve to be hanged” (a recurrent element within extreme right discourses) and discussions about how to “deal with” Muslim immigrants. In one closed chat group, accessible only to the national leadership and some trusted local members, the conversation was both exceptionally misogynist and unambiguously racist; jokes about Black people being beaten to death and thrown in ditches were cheered on by activists in the chat, and equally popular were jokes about dead boat refugees from Somalia (Röstlund & Wiberg, 2016). Racist sentiments and animosity aimed at refugees are augmented in interaction between the top leadership and local activists. These racist conversations play an important role in fostering communal sociopolitical identities among the participants. Affective talk between participants mediates between individual experiences (e.g., one woman activist’s personal experience of a rape) and group cohesion; thus, the online interactions create a sense of communal experience (cf., Ahmed, 2004a). The conversations reflect what Papacharissi (2015, p. 101) defines as a “deeply personal ideology of a networked self,” a combination of the private and collective realms of networked politics. Discursive interaction between users on platforms harboring racist sentiments—with no alternative or opposing voices—can spiral out of control. These conversations often foment discursive violence—for example, graphic (verbal) descriptions of violence targeting refugees and immigrants. User interaction among activists is partly structured by the sociotechnical affordances of Facebook, where personalized flows of information and closed conversation groups contribute to a personalization of reality; that is, reality is constructed through flows of information where any contesting explanations have been blocked out, creating a malevolent echo chamber (e.g., Hirvonen, 2013). This also points to the constraints of an inward-looking community of activists (e.g., Cammaerts, 2012). On the other hand, the anonymity in closed group conversations with an explicitly uncivil character has been one of the key factors in sustaining an active extreme right-wing online community. This form of communication is rare in open online spaces, where personal identities can be revealed to the greater public. However, you find similar conversations on platforms where anonymity or pseudonymity is possible.

Self-mediation

The use of social media platforms enables various forms of self-mediation practices for protest movements. SoO uses their open Facebook group “Soldiers of Odin Sweden Official” to promote the network and to display various activities undertaken by activists. Unlike the closed groups deployed by activists for networking, coordination, movement building, and group cohesion, the open platform is far more sanitized. The open page mainly consists of two kinds of visual messages—images and short narratives of street activities (such as helping out homeless people or patrolling city streets) and images of (almost exclusively male) activists standing together with their backs turned to the camera. The latter imagery is typical of the visuals usually portraying One Percenter Biker gangs. The various group images of (anonymous) activists in different locations, dressed in the SoO-uniform jackets, create an intimidating appearance. Moreover, the self-mediation of activists is an attempt to create images of popular mobilization (cf., Uldam, 2013). However, there is still a palpable gap between “connective” and “collective” action (Papacharissi, 2015, p. 128), that is, it is still much easier to “like” an image on Facebook or publish a cheering comment than to engage in real activities in public spaces (e.g., Cammaerts, 2012).

The discourse on the open Facebook page is toned down, and when textual messages appear, they are merely descriptive reports of activities. The commentary is exclusively supportive and seems to function as an “inward looking” (Cammaerts, 2012, p. 130) communicative practice, mainly constituting in-group solidarity. There are two exceptions to this. In their feed, SoO have published two statements (one in English and the other in Swedish) rejecting all allegations of their being a racist organization. In the statement, SoO claim that they “believe in protecting the streets with observe-and-report styled patrols, and if necessary defend/protect anyone who may need us and/or report to the local authorities.” Moreover, they state that “it is time to take back our streets, cities, states and countries,” thereby producing an immanent body of the threatening “other” (Ahmed, 2004b). There is no explicit mention of refugees in the statement, but it carries a resilient presupposed meaning—the “streets, cities and countries” are going to be “recaptured” from immigrants and refugees. The statement functions as a counter-narrative to the framing of SoO by mainstream news outlets (see next section), thereby echoing the recurrent idea that the “politically correct” mainstream media conceal the truth from ordinary citizens. This is also evident in the commentary published by activists and supporters on Facebook. Consequently, SoO’s antagonism toward mainstream media is a constraint for the group, and since their self-mediation on Facebook is limited and mostly defensive, SoO lack the potential to achieve resonance in public space (Cammaerts, 2012).

The second form of self-mediation takes place in the flourishing online milieu of “alternative news sites” associated with the populist and extremist far-right. As an alternative to traditional news organizations, a number of platforms emulating traditional online news have successfully been created by actors on the far-right (Ekman, 2014a; Hirvonen, 2013). Since the network distrusts mainstream media, it has been utilizing and, to some extent, working together with far-right news platforms in order to reach out to larger audiences. The analysis includes all articles on SoO published by four platforms belonging to the far-right in Swedish politics: Fria Tider (FT), Avpixlat (Av), Nya Tider (NT), and Nordfront (NF). In total, the online news sites published 29 articles covering SoO during the first 9 months of 2016. The first discursive theme is the rise and spread of SoO, with a focus on the “success” of the network. FT published two articles with the headlines “Soldiers of Odin are now established in eighteen countries” and “Our growth is totally exploding.” NT published a longer feature story entitled “The citizen’s guard from Finland is now established across Europe,” and Nordfront pointed out that SoO “is now established in Sweden.” These articles focus on the “positive” development of “securing” the safety of Swedish citizens in public spaces and include various endorsements from the public and expressions of tolerance from police officials. The articles characterize SoO as a solution to the ongoing “refugee problem.” One article (NT) refers to crimes committed by asylum seekers, describing, for example, a “rape on a 14 year-old girl in Finland” and that an “Afghan [was] sentenced to life imprisonment for raping a 17 year-old girl and burning her to death.” These “stories” adhere to the larger idea that immigration is the key root of sexual violence, that refugees are inherently violent, and that street patrolling and the use of force are mere acts of self-defense.

Several articles cover the street patrolling activities of SoO and describe the activities as protecting and serving the public, which forms the second discursive theme. Some of the articles focus on citizen’s arrests undertaken by SoO activists, whereas some articles highlight the public support for SoO, such as the article headlined “Girls think that Soldiers of Odin are great” (FT). In a couple of articles, activists from the network explain why they are needed in public spaces. Articles also decry the “unfair” framing of SoO in mainstream media and include images and videos of (anonymous) activists patrolling in uniforms. The only activist identifiable in the imagery is the national “spokesperson.” Overall, the published images and video clips depict groups of physically intimidating activists dressed in black uniforms and making themselves a conspicuous presence in the streets. This imagery embodies the hyper-masculine, an exaggerated and violent male identity prevalent in the extreme right’s sociocultural practices (Ekman, 2014b; Kimmel, 2007), and expresses an aestheticization of (immanent) violence.

The third discursive theme is defined as public antagonists and counts for half (15) of the published articles. Three categories of articles depict three different kinds of antagonists in the material. The first category comprises articles about SoO being “harassed” and “frisked” by police when patrolling the streets or the police expressing concern about the presence of activists in public space. The second group of articles describes “left-wing extremists” attacking activists. In these articles, SoO appear as victims, evoking sympathy. The articles describe the attacks as unprovoked and the anti-racists as violent assailants. Finally, there are articles describing how the “criminal gang Bandidos” is reacting violently to the presence of SoO in the streets. Taken together, these articles frame SoO as a victim of unfair treatment from different actors in society—constructing an image of the network as being targeted while merely trying to “keep the streets safe” for the general public. In all, they create an image of victimization (e.g., Atton, 2006), revealing the constraints of self-mediation within an isolated movement.

Mainstream Media Representation

During the first 9 months of 2016, the seven largest newspapers published 90 articles mentioning SoO. This publicity (in addition to television coverage and wider Internet publicity) made SoO known to the wider public. Since the network was geographically distributed, with local participation in various places, even more articles were published in the local and regional press. The following analysis uses descriptive content analysis to characterize the representation in the seven newspapers with the largest circulations. Of all the articles, 62% appeared in April and May, when SoO was most active and new local chapters were beginning to surface around the country. A total of 8% covered the Finnish organization, but most of the articles (62%) were published in the domestic news sections of the newspapers. There were no articles portraying SoO in a positive light, and 60% had a neutral attitude toward the organization. Of the 40% of the articles that framed SoO negatively, more than half were published in opinion and editorial sections or as columns in the culture sections. SoO were associated with violence in 41% of the articles. Representatives of the group were present in 24% of the articles, and they were quoted directly in 11% of the articles. SoO were characterized in quite different ways in the newspapers. In total, five different prefixes (or descriptive terms in the sentences describing the network) were used. The most common designation was “right-wing extremist” (32%), followed by descriptive term (28%), “vigilante group/citizen’s guard” (23%), “Nazi” (12%), “xenophobic” (3%), and “street patrol” (1%). There were notable differences between the newspapers, with the two national tabloids mainly using the terms “right-wing extremist” and “Nazi” to describe SoO and the large broadsheets Göteborgs-Posten and Sydsvenskan using their own self-descriptive term “citizen’s guard” or no particular descriptive term. In relation to the variables measuring whether the newspapers mirrored public concerns about refugees in public spaces and whether the articles presented any legitimate arguments for vigilante patrolling, the results were decisive. In only three articles (3%) was public concern represented, and in two cases the concerns were articulated by ordinary members of the public. No articles expressed any legitimate arguments in favor of vigilante gangs (such as the failure of the police or other official agencies to fulfill their responsibilities). So, the public “spectacle” of street patrolling and its association with violence attracted plentiful media attention (Cammaerts, 2012)—but it also resulted in a lack of public support. The absence of both acceptance and understanding in the mainstream media was clearly an obstacle to gaining public legitimacy. Some of the articles covered instances where politicians were expelled from the right-wing populist party The Sweden Democrats because of their association with SoO. So, even if SoO gained plenty of media visibility (particularly during the spring of 2016), the growth of the organization was primarily facilitated by recruitment among football hooligans and activists with a background in the far-right milieu—rather than being a result of visible public support in the media. Another problem for the network is its harsh attitude toward journalists, as previously mentioned. Only the group’s self-declared national leader was allowed to speak with journalists, which partly explains the low frequency of actual participation in the news material. On the other hand, at least half of the existing quotes from activists in the material originate from the articles written by two journalists who went undercover to report on SoO (Röstlund & Wiman, 2016). In these articles, the quoted activists made racist, misogynist, and violent statements, mirroring the discourse flourishing in the anonymous and uncivil part of the Internet. These statements clearly had delegitimizing consequences. The opportunities and constraints of the various mediations of/by SoO are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Summary of the Mediation Opportunity Structure of SoO (Modified From Cammaerts, 2012).

	Numbers	Damage	Witnessing
Networking	Opportunities: Recruitment Mobilization Group cohesion Constraints: Transferring online engagement to offline engagement Problems controlling activists’ communication on SM platforms	Opportunities: Educating members on legislation and possibilities for citizens to act as “police” Constraints: Discursive violence Risk of being exposed	Opportunities: Experience- and strategy sharing Combining private and collective experiences Constraints: Inward-looking Gap between collective and connective action
Self-mediation	Opportunities: Producing counter-narratives and counter-frames Archiving street politics Controlling narrative Constraints: Speaking to the already convinced Creating padlocked siloes of communication (Papacharissi, 2015)	Opportunities: Deeds that speak Decisiveness in a time of insecurity Showcasing force Aestheticization of violence Constraints: Risk of repression and violent escalation Risk of confrontations with/in the public, victimization	Opportunities: Constructing collective identities Instilling confidence in supporters and potential activists Constraints: Inward-looking No engagement with formal politics Visual style connoting criminal gangs and political extremists
Media representation	Opportunities: Visibility Potential geographic spread Constraints: Aversion toward mainstream media and strict policies prohibiting contact with it	Opportunities: Street patrolling and violence attract media attention Utilizing general sentiments of fear Constraints: Public condemnation of violence and vigilante behavior Minor public support	Opportunities: Visibility The tangibility of street actions Constraints: Little focus on actual politics Focus on individual members of SoO and their (violent) background

SM: social media; SoO: Soldiers of Odin.

Concluding Discussion

The analysis of the mediated opportunities and constraints of SoO points to the significance of uncivil society actors in online mediated spaces—and also to the tangible threat that refugees and ethnic minorities in European countries are facing from groups such as SoO. Since uncivil actors use the same commercial social media platforms as everyone else and deploy similar affective technologies of mass-self communication, there is a need for greater scholarly understanding of the relations between anti-refugee mobilizations, online racism, and digital communication practices. Although the concept of mediation opportunity structure was developed in relation to social and protest movements with democratic goals (Cammaerts, 2012), the model explicates important strategies for any social or political actor who seeks to impact the public mindset (and eventually policy making) outside institutional politics. The analysis has shown that the rapid (geographic) growth of the SoO was facilitated by coordinated networking communication and that public visibility was made possible by the combination of various forms of self-mediation and by a rather large mainstream media coverage. So, the analysis of SoO partly validates the importance of mediation opportunity structure in building contentious political networks, and particularly the significance of attracting mainstream media coverage (Cammaerts, 2012; Uldam, 2013). It is important to stress, however, that the mediated constraints, visible in the analysis of SoO, indicate that uncivil engagement in networked communication has obvious limitations. The SoO faced “structural constraint” (Cammaerts, 2012, p. 130) because of the focus on violence and the total absence of public acceptance in the mainstream press. Moreover, the many examples of victimization in the “alternative” news platforms belonging to the far-right reveal the problem for SoO to construct a meaningful counter-discourse (see Atton, 2006; Papacharissi, 2015, p. 122). And since the “alternative” online platforms of the far-right (still) have a limited trust among the general public, it was difficult to reach beyond the already convinced. Finally, and most important, the internal violent racist discourse and the various public confrontations (with police, anti-racists, etc.) add to the constraints of mediated opportunity in the case of SoO. So, despite active communication in social media and the possibilities of sustaining a broader network outside already convinced activists and supporters, SoO failed in attracting any significant public support. Instead, by reproducing well-known self-mediation strategies of right-wing extremist actors (Ekman, 2014a) and by creating communities with closure (Atton, 2006), SoO did not manage to impact public mindset in any significant way. In the end, the intimidating imagery of anonymous activists used in the organization’s self-mediation almost certainly contributed to the public’s renunciation of the network. Instead of counteracting the negative framing in mainstream media, the self-mediation strategies enhanced the prevailing perception of the network as far-right extremists—rather than creating new expressions (e.g., Back, 2002).

Nonetheless, the case of SoO reveals the (perilous) engagement of violent actors in uncivil online public networks and its political opportunities emerging in periods of political instability and public insecurity. Uncivil actors, with explicit racist and anti-democratic goals, benefit from the emergence of commercial social media platforms. Since they adopt networking tactics and self-mediation strategies on platforms with millions of users (i.e., Facebook), they can possibly—if the political and sociocultural contexts change in a more favorable direction for uncivil actors (e.g., Ruzza, 2009)—expand their political agenda to new publics and arenas. Consequently, this calls for more research into the (critical) relationship between uncivil engagement and social media networks—particularly in other political and sociocultural contexts—to explicate not only the positive and progressive potential of social media networks and contentious politics but also the growing threat posed by uncivil actors in contemporary Europe and elsewhere.