No Digital “ Castles in the Air ” : Online Non-Participation and the Radical Left

This article presents results from a study of online presence in activist milieus associated with the radical left in Sweden discussed from a perspective of non-participation. With the aim to further the understanding of digital nonparticipation as communicative strategy in activism, it builds upon empirical findings and argues that the online practices and use of social media, as could be observed in milieus associated with the radical left, indicates active nonparticipation and that this, in turn, is related to the ambition to claim autonomy. The article draws from existing scholarship on critical perspectives on protest movements and social media as well as empirical examples of online content published by radical leftist groups. Furthermore, it analyses how these activities could be understood in terms of active and passive non-participation, abstention or adaptation to social media affordances, as well as implosion of the social in digital media. The findings suggest that much of the activities in the material could be described as active nonparticipation and that this media practice relates to ideological positioning and values in the milieu.


Introduction
Ten years ago, "cyber-protest" usually meant active engagement with digital online media by protest-or activist groups (e.g.van de Donk, 2004).However, with the present development and proliferation of online digital media, the question of how to understand political agency in relation to online media practices has become less obvious.Following heady narratives about "Facebook revolutions" and "revolutions 2.0", there has emerged a literature where more critical perspectives on protest movements and social media are presented (e.g.Cammaerts, Mattoni, & McCurdy, 2013;Dencik & Leistert, 2015).This approach has turned its attention to a number of tensions and contradictions that come with the use of social media for social change, including issues such as surveillance and cen-sorship (Hintz, 2015;Redden, 2015), affordances of individualism (Fenton & Barassi, 2011), commodification of users and systemic constraints (Fuchs, 2011;Leistert, 2015), and, affordances of connectivity versus the need for anonymity in activist groups (Treré, 2015).
This article is based on the findings from a previous study of online presence of the radical left in Sweden (Swedish Media Council, 2014).The ambition is to address some of the conflicts above and connect them to a recent discussion about online non-participation as a strategy (Casemajor, Couture, Delfin, Goerzen, & Delfanti, 2015), with the purpose to contribute to our understanding of non-participation.My argument is that the online practices and use of social media as could be observed in milieus associated with the radical left indicates active non-participation and that this, in turn, is related to a political ambition to claim autonomy.
The argument is structured as follows: After a brief background about the nature of the study I will discuss the theoretical concepts participation, empowerment and autonomy and present the analytical framework, followed by notes on method.The analytical part is structured from three themes: active and passive nonparticipation, abstention and adaptation, and implosion of the social.The paper ends with a discussion about the possibilities for the concept online non-participation and its implications for further, empirical studies.

Background
In contrast to much of what has been published about social movements and online media, the study from which the material in this article was taken could be described as a propaganda study (Jowett & O'Donell, 1999) designed to chart "anti-democratic and proviolent messages on the internet" (Swedish Media Council, 2014).It was conducted by the Swedish Media Council and initiated by the Swedish government as part of an "Action Plan for Protecting Democracy against Pro-Violence Extremism".The task was to describe where and what type of messages could be found online that publicly propagated for ideologically motivated violent action, and what could be said about recruitment strategies in this material.The question was not what these groups do or why, but how they present their activity and cause publicly.Much of the aim, design, and scope of the study were defined by the commissioner, which meant that the findings are limited to a quite narrow area of investigation.It is not a study of social movements in general, but of a small autonomous or anarchist fraction of the radical left who publicly display an apologetic sentiment toward violence.Nor is it a study exclusively about corporate social media, but more broadly defined online media that includes organization web sites.
The findings showed that the radical left, at least in the sense as was defined by the assignment, were sparse users of online digital media for dissemination of propaganda; in contrast to the other groups included in the study, especially the extreme right.Instead, they seemed to be quite reluctant in adapting to online communication.While the original study did not include measures for examining non-participation, the results were of a kind that activates questions about participation/non-participation dynamics that motivates a review of these findings in a new theoretical context.

Online Non-Participation
Lately, the scholarly interest in the issue of online nonparticipation and disconnection has increased (Cammaerts, 2008;Kaun & Schwarzenegger, 2014;Portwood-Stacer, 2013;Selwyn, 2003).While much of the focus of these studies has been on individual motivations for abstention, attention has also been drawn to nonparticipation as a form of activism (Casemajor et al., 2015), as well as an interest in the decline of certain types of online radical politics (Wolfson, 2014).Still, the greater part of the literature that has surfaced since the Arabic spring, the Occupy-, and Los Indignados-movement tends to focus on connectivity, online presence and participation (Bennett & Segerberg, 2013;Gerbaudo, 2012;Hands, 2011).In this section I will look into some of the theories and models of participation and discuss theoretical foundations for understanding digital non-participation in terms of strategy or protest.Online media here refers to those channels and services where messages and content is made public on the web; either in the case with social networking sites, or in other forms of user-generated content, what used to be called web 2.0.Hence, forms of private or non-public online communication such as e-mail or direct messages are not included in this definition.The reason is that the original study focused on public messages that were easily accessible and published with the intent to reach an audience.

Social Implosion and Interpassivity
The idea of non-participation as empowering in relation to media has been part of what could be described as a postmodern critique of social theory, especially as articulated by Jean Baudrillard (1988).Even if Baudrillard's work does not present a scholarship that offers verifiable theories, some observations and concepts have proven useful figures of thought and Baudrillard was early to assess the affordances of an abundant information society.One such concept that he addressed both politically and philosophically was the issue of media-related non-participation.In the essay "The Masses: The implosion of the social in the media" (Baudrillard, 1988), Baudrillard introduced the idea that non-participation should be understood as a rational and effective response to a power structure that fostered a hypocritical form of non-communication: "I would no longer interpret in the same way the forced silence of the masses in the mass media.I would no longer see in it a sign of passivity and of alienation, but to the contrary an original strategy, an original response in the form of a challenge."(Baudrillard, 1988, p. 208) Non-participation in this context is not understood as hampered citizenship, but as rational and empowering, a position that finds resonance in more recent accounts of non-participation (Casemajor et al., 2015).The challenge is directed toward a system that requires of its population a symbolic engagement: to have opinions, to be well-informed, to make conscious choices.What Baudrillard opposes is the equation of participation with empowerment-a theory of empowerment that only acknowledges active participation is seen as hypocritical and quite useless; a more efficient form of opposition would be to refuse to participate.While Baudrillard's level of analysis makes grand claims with little-to-none empirical grounding, his change in perspective should also be added to the recent debate on strategic non-participation also when it comes to digital media.
The backdrop for Baudrillard's argument was a Marxist understanding of mass media as a tool for manipulation of the masses.In the digital era, manipulation and false consciousness may not hold the same position in critical thought, even if the notion of filter bubbles (Pariser, 2011), is an example of updated suspicion about media manipulation ("what the internet is hiding from you").Its legacy is also present in discussions about involuntary participation where notions of social media's affordances of visibility and individualism have been commented on by scholars (Fenton & Barassi, 2011;Fuchs, 2014;Poell & van Dijck, 2015).Furthermore, the challenge in digital media can also be approached through the concept of interpassivity (Dean, 2009;Žižek, 1998), described as the opposite to interactivity.Originally developed as a critical term in order to explain how mediated quasi-interactivity functions as stand-in for real engagement, it has come to use in discussions about online engagement/disengagement and what is sometimes called "slacktivism" (Morozov, 2011), where taking part in online petitions and campaigns become stand-in for actual social engagement and activism.

Participation and Empowerment
Participation is a key feature of digital life (Kelty, 2015).The concept, however, is polysemous and it is quite difficult to define what qualifies as participation.This means that any attempt to build conceptual models for participation/non-participation has to be dynamic (Casemajor et al., 2015).In the field of media studies, at least two theoretical conceptualizations of participation co-exist.On the one hand, there is the tradition that understands participation as a term connected with the political, e.g. in form of democratic deliberation (Dahlgren, 2014).On the other hand, there is the notion theoretically based in cultural studies about active audiences and audience participation, which gained new status with the rise of digital, interactive media.Media studies harbor both these conceptual understandings that sometimes lead to tensions (as could be illustrated by the dialogue in Jenkins & Carpentier, 2013).But what brings them together is that they both address the notion of empowerment.
Definitions of empowerment are manifold and range from individual aspects such as psychological enhancement, and self-awareness, to more social and political where both individuals and groups achieve a status of autonomy (Freire, 2000;Rodriguez, 2001).Sometimes empowerment is described as a scale from self-awareness to political action (Higgins, 1999).In the context of this article, empowerment should be understood in relation to autonomy rather than selfawareness.Much of the literature on mediated participation has concentrated on the question of empowerment (Castells, 2009;Dahlgren, 2009) and participation (Carpentier, 2011;Jenkins, 2008).Some of these scholars have pointed to online social media's role in increasing civic engagement and political participation for citizens (Castells, 2009).One could say that the position taken by these scholars has equated empowerment with participation, and accordingly the idea that more participation equals more autonomy has taken hold.
Notions of empowerment and digital media have been connected with the problem of the unequal distribution of access to the internet, popularly referred to as the digital divide (e.g.Norris, 2001).Lately, with the expansion of internet access on a global scale, interests in various types of digital divides, in terms of social inequalities have gained ground in critical research (Danielsson, 2014;Livingstone & Helsper, 2007).This means that focus has changed from asking if people have the possibility to go online, to asking what they do when online.Here again, the question of empowerment becomes important since there are structural limitations to how digital media can be used.Along traditional lines of active and passive media use, online digital media can be used as a means for political participation (simulating the golden days of public sphere) or as mere passive consumption (the decline of the public sphere).However, there is yet another type of relation to digital media and empowerment that does not fit into these categories, namely the acts of deliberate non-participation.

Models of Non-Participation
An interesting paradox in relation to the discussion about participation and empowerment above is that much of the developments in today's digital media could be described in terms of forced or involuntary participation: online mainstream media invite us to "like" and share content with our social networks, while smartphones gathers data over our mobility and communication (Dahlgren, 2014).The idea of participation has been a central part of all theorization about the internet, in terms of affordances for interactivity or user-generated content (Kelty, 2015).One recurring critique from critical internet scholars has been that power relations have too easily been eradicated from theories on participation and that digital media participation creates new or reinforces existing structures of unequal distribution of power (Beer, 2009;Cammaerts, 2008;Goldberg, 2010).
One reason for taking an interest in nonparticipation is that it could (and should) be understood as a politically significant action (Casemajor et al., 2015, p. 851).This means that in an era of forced participation that serves the interest of existing power structures, one way of opposition is to refuse to participate.This insight is part of the dynamic model of online participation/non-participation proposed by Casemajor et al.: "The framework…is an attempt at going beyond the linear model of participation and showing that non-participation cannot just be considered as hampered citizen action or passivity; it can also be empowering" (2015, p. 863).
The model sketches four possible types of participation: active participation, passive participation, active non-participation, and passive non-participation.And in the case of active non-participation, three ideal types of action are defined.The first one is obfuscation which means to act in ways to diffuse data gathering online, through erratic behavior that renders collected data useless.The second type of active nonparticipation is called sabotage and includes strategies to disrupt digital platforms, for instance, denial of service-attacks on websites.The final ideal type is exodus and simply means withdrawal from digital platforms, sometimes in order to create own.The model proposed by Casemajor et al. (2015) shows some resemblance to a model developed by Rucht (2004) called the "Quadruple 'A'"; a model that presented four ideal types of communicative strategies that social movements have developed in relation to unfavorable mainstream media coverage.Building on data dating back to the 1960s and onwards, Rucht identifies four strategies on an axis between inwardand outward oriented strategies; as well as strategies with low or high resource demands: abstention and attacks on mainstream media are not very resourcedemanding.The former means avoiding publicity and "keeping to oneself" while the other allows interactions with mainstream media but voices mistrust.Alternatives and adaptation, on the other hand, are strategies with quite high resource demands where the former means creating own media outlets and the latter to enact events that might attract the attention of mass media (Rucht, 2004, p. 46).Two of these strategies, abstention and alternative are directed inwards, while attack and adaptation are outward-looking.Rucht (2011) has later commented on the implications of online digital media for his model, acknowledging the possibilities but warning about over-estimating the importance of digital media (p.259).

Activism and Media Practice
The role of digital media in mobilization among radical groups is well-researched (Askanius, 2012;Neumayer, 2013).In a study of online counter publics associated with the extreme right and radical left, Neumayer (2013) argues that in order to understand how these groups make use of digital media, one has to look at three dimensions: technological affordances; strategies, tactics and media practices; as well as political positions and ideology.Consequently, media practices do not exist in a vacuum but should be understood in relation to both technology and ideology.Mattoni (2012Mattoni ( , 2013) ) discusses media practice as part of wider repertoires of communication in social movements.In that context, repertoires include both participation and non-participation in terms of interactions with media technologies, media outlets and media professionals, what Mattoni refers to as "relational media practices" (Mattoni, 2013, p. 49).
When approaching media practices and repertoires of communication from the perspective of ideology, notions of empowerment and autonomy again become relevant.Langlois & Dubois (2005) discuss autonomous media strategies, as attempts to "bypass mainstream media through experimentation with new forms of democratic communication" (Langlois & Dubois, 2005, p. 23).On a similar note, Milan (2013) speaks of appropriation of means of communication by social movements in terms of "liberated technologies" and part of "emancipatory communication practices" (Milan, 2013, p. 2).This, in turn, is related to the practices of the autonomist leftist tradition of creating and claiming autonomous spaces in the social landscape (e.g.Katsiaficas, 2006).Following this, it is possible to understand ideologically motivated non-participation as part of the media practices and repertoires of communication in radical groups.
The perspectives outlined above constitute an analytical framework that turns attention to participation/non-participation dynamics, as well as issues of autonomy and empowerment in an ideological context.

Material and Method
As already mentioned, the original study was designed as a propaganda study, i.e. an investigation into modes of address and means of persuasion as employed by groups advocating violent means for an ideological cause.The method was a qualitative content analysis of digital media and online platforms, with a focus on text analysis (semiotics, discourse analysis).The material consisted of online content produced within three ideological milieus that were pre-defined by the assignment: the extreme right, militant jihadist and the autonomous left. 1 While there are many aspects that 1 The governmental action plan and its use of terms such as "anti-democratic" and "pro-violent" have been contested by social activists and academics, especially because of a tendency to equate extra parliamentary activism and civil disobedience with extremism.For an overview of this debate, see Kaun (2015).separate these types of movements, they had been targeted by officials as environments with a potential to propagate for violent actions for an ideological cause.It should be said that the sample was quite limited since the content up for analysis had to meet the criteria for voicing an extreme stand in favor of political violence.Few groups or online content did this and the material analyzed here was published in milieus in and around militant anti-fascist and militant animal rights groups.Needless to say, this is a marginal phenomenon in context of the radical left activist scene online, but it stood out in terms of communication strategies in comparison to what could be observed in the material published in the extreme right milieu where more aggressive discourses were present.In all, three websites associated with the militant fringe of the radical left were observed: antifa.se,revfront.org,and djurensbefrielsefront.org,and of these only revfront.orgcould be described as active.These are also strictly closed milieus that seldom engage in propaganda or open recruitment."Act without being seen" has been a motto and this might be true also for how participation in digital media is organized.
The analysis consisted of three parts: first, a description of pro-violent communication as could be found in the online milieu associated with the radical left; thereafter followed a charting of the web presence of three groups that had been identified as groups who met the criteria of openly expressing themselves in favor of ideologically motivated violence.Finally, the report included a thematic analysis of online videos published on YouTube (for a full account, see Swedish Media Council, 2014, pp. 132-190).
In the review of these findings, it is important to acknowledge the communicative modalities and affordances provided by online, digital media.What qualifies as online content is not restricted to symbolic expressions in the form of words, images, video and audio, but could also include possibilities for interaction that are part of the communication practice and experience, or a social, or network media logic (Klinger & Svensson, 2015;Van Dijck & Poell, 2013).To like, link, share, and comment are activities that have become part of meaningful communication and by looking into how these activities are activated, one can get the sense of how digital media communication works in the studied environment.The study did not analyze user interaction, but paid attention to affordances for interactivity, and how these were used.

Analysis
The following argument is structured in order to address two sets of questions: what types of nonparticipation can be identified in the online activities of the radical left, and, how can this material be conceptualized in relation to theories on activist media prac-tices?The analysis has been thematically divided into three categories: active and passive forms of nonparticipation; abstention and adaptation; as well as the implosion of the social online.

Active and Passive Non-Participation
The most visible form of non-participation that appears in the material is "exodus", i.e. instances of apparent withdrawal from online communication.The other types of active non-participation that are described in Casemajor's et al. (2015) model, obfuscation and sabotage, are not represented in the material.The withdrawals can come in different guises, either in terms of complete exodus, i.e. closing down web sites, or in semi-withdrawals as in turning off or limit communicative affordances.Both are result of active choice.
When it comes to complete exodus, this can be observed in the case with antifa.se,web site for a militant antifascist group Antifascist action (AFA).It had been on the web since the early 2000s but was apparently disconnected around 2010.Hence, the site was not in operation during the sample period.A similar case of probable exodus was djurensbefrielsefront.org, a web site by DBF, the Swedish branch of militant animal rights activists: Animal Liberation Front.The site was still accessible but gave the impression of having been abandoned, the last updates dated back to 2010.
Online communities and chat forums dedicated to the radical left were filled with speculations on whether AFA still exist at all, since they showed no presence online.The quote below is an example from this type of discussion on socialism.nu,a leftist online community, and is an answer to questions about the existence/non-existence of AFA: "Oh, yes.They're still here.Their activities haven't disappeared with the web site.That line of thinking originates from how the Nazis organize themselves.There, they have tons of Web sites but little or no activity IRL.It's a good thing that AFA does not spend time building digital "castles in the air".Those who need to know about AFA's existence will certainly see it =)" (https://www.socialism.nu/showthread.php?t=6164&page=8)Even if we do not know how well-informed this commentator is regarding the issue, the response bears witness to a notion of active non-participation.Furthermore, the quote highlights another aspect apart from mere presence/absence, as it makes a clear distinction between online activities and activities in the "real" world.By describing online activity as "building digital castles in the air", the commentator gives voice to a view where "real" action is preferred over symbolic action.This could also be seen as an outcome of ideological positioning and an act of distinguishing them-selves from political opponents by means of its view on and understanding of media practices.Haunss (2015) addresses the issue of online/offline-dynamics and notes that much of the recent literature on protest movement's social media practices stresses the role of preexisting social ties in mobilization (p.26).This has been noted in previous studies on autonomous scenes, where personal interaction is preferred above mediated communication (Leach & Haunss, 2009).
The study focused on content that was accessible and could be observed; hence, it is not correct to speak of full non-participation.However, there are examples of what could be described as semi-withdrawals from online communication.One such example is the YouTube channels RevFrontMedia and RevFrontMedia2, both associated with the militant group Revolutionary Front (RF), who, at the time of the study had published in total 19 videos online depicting violent confrontations with neo-Nazis and the vandalizing of homes and properties of people claimed to be associated with the extreme right.While the publication of these videos arguably is an example of online participation, they also display a tension between active non-participation and passive participation.This manifested itself through the use of the comment function, which is pre-set by YouTube to be automatically attached to the published video.This comment function can however be deactivated, and this has been done for all of the videos published by RevFrontMedia.Since the deactivation requires an active choice, it is fair to argue that this also could be interpreted in line with the argument of active non-participation.At least, it does not invite viewers to participate or communicate through the means of the affordances of this type of digital media.

Abstention and Adaptation
While the exodus strategies described above could be filed under what Rucht (2004) named "abstention", i.e. to avoid attention from mass media and keep to oneself, there are examples in the material of strategies that rather meet the criteria of "adaptation".Rucht (2004) describes adaptation as an extrovert strategy with considerable high resource demands where activist groups plan their activities with (mass) media attention in mind.For the adaptation strategy, the distinction between "real" and symbolic action is less articulated which means that actions that are "unsuccessful" in meeting their concrete aims can still be deemed successful if they achieve the right media attention (e.g.DeLuca, 1999).
One example of adaptation to media attention in the material is the Facebook-account and logo that RF launched during the sample period.It could be added that this addition of a Facebook page in late 2012 does not make the group appear as early adopters when it comes to social media.This arguably late awakening does not lessen the enthusiasm; the launch was announced in quite high-flown manner on the website revfront.org:"[Revolutionary Front] is an organization that constantly strives for improvement and to take steps forward in our political work.We understand the importance of being seen in public space, both on the streets and on the Internet.We have therefore chosen to start a Facebook page.A page that, we hope, our readers will appreciate.The page will serve as a tool for spreading our news and to reach out to more people with our political message."(http://revfront.org/?p=5465)With Mattoni's (2012Mattoni's ( , 2013) ) concept "relational media practices" it becomes possible to approach adaptation as not only an adaptation of activist groups to media logics in order to attract the attention of mainstream media professionals, as described by Rucht (2004), but to also understand adaptation in relation to technological affordances.The quote above could be seen as an example of adaptation where the group adapts its action to the requirements or logics of a media technology, in this case online social media.Milan (2015) writes about a "politics of visibility" (p.63) brought about by social media, a politics that allows surveillance and mass scrutiny and makes activist groups vulnerable to political adversaries and the state.There are good reasons to abstain from using these services, but Milan's (2015) research has pointed to a sense of inevitability of corporate social media that leads protest groups and activists to a resignation to the unwanted consequences of being present in corporate social media.The Facebook page in itself could be seen as a form of adaptation.
However, the words by which this Facebook page was announced bear witness to an understanding of Facebook as, not primarily a means for reciprocal communication, but for broadcasting political messages.There are also formulations that declare that this move is partly motivated by the ambition to reach out to outsiders.However, the announcement says nothing about using the Facebook page as a platform for contact and communication, but seems to treat it more as a tool for broadcasting political messages.This tendency becomes even clearer as they describe political messages as "propaganda", a term probably more associated with totalitarian ambitions of manipulation through media than social and reciprocal communication: "After ten years as an organization we are also proud to present the [Revolutionary Front] logo!It will represent the organization and mark our propaganda."(http://revfront.org/?p=5465) The mode of communication employed could be described in terms of a complicated tension between secrecy and visibility.Communication is reduced to one-way channels of information rather than to share and create a sense of community.Comments-functions are turned off, and overall, the communication affordances provided for interaction and reciprocal communication are not put to use in any considerable extent.One way to approach this is to see it as "asocial" media practices.

Implosion of the Social Online
A common definition of social media is that it refers to media services that enable reciprocal communication between humans, and provides some sense of community (cf.Fuchs, 2014, p. 1).When observing the use of social networking sites and digital media in the sample, it appears as if the way that these services have been put to use renounces some of those modalities of communication and community-building that makes them "social".In accordance with the discussion about semi-withdrawal above, it is a complex act of being present without full participation, similar to passive forms of participations such as lurking (Crawford, 2011).
A first example of "asocial" online presence is RF who are active in publishing propaganda on the web and in social media, but seem more cautious when it comes to sociability online.The content that is published is not aimed at recruiting sympathizers; it does little to promote the positive and inspiring sides of being associated with RF or participating in their actions (perhaps it is not be necessary to advertise-potential followers might be attracted or motivated by other incentives than to seek for a social community).It might also be more important for various reasons to keep internal activities secret than it is to put themselves on display (other than through their actions).Anyone wishing to come in contact with the group is directed to e-mail or a pre-paid mobile phone number.There was a @revfront Twitter-account but it was closed for outsiders-only accepted followers could read its tweets and see the number of followers.
The asocial online presence is also true for the other groups; AFA and DBF are even more reclusive, the latter explicitly discouraging people from attempting to make contact with them.Leach and Haunss (2009) have noted how autonomist activist scenes build heavily on personal contacts and direct communication, which might help explain this desire to redirect users from the online platforms to other means of communication that are more direct and personal.
Secondly, many of the services for publishing content online are social per default, and hence set limits for the control over content.As has already been mentioned, RevFrontMedia had switched off the commentingfunctions for their YouTube-videos, but the display of these videos is companied by a list of suggestions for similar videos presented in the right-hand margin on the YouTube web site.On many occasions during the sample period, the viewer was given suggestions for videos produced and published by right-wing extremist groups.Hence, when using YouTube as a tool for broadcasting propaganda, there is a risk that it provides unintentional attention to propaganda from political enemies.
Finally, articles and communiques published in the milieu are written by pseudonyms or unidentifiable collectives.This anonymity feeds into an overarching strategy of secrecy that is complemented by photographs depicting people in ski-masks or a casual custom of blurring out faces from photographs and videos.Who these individuals are remains a secret.One of the things that social networking sites brought to online culture was identity.Where previous forms of online communities had been characterized by a logic where identities were disconnected from physical reality ("On the internet, nobody knows you're a dog") (Turkle, 1995), My Space and Facebook brought back traditional modes of identification ("On the internet, everybody knows you're a dog") (see Gershon, 2011).Fenton and Barassi (2011) have noted how social media is built on a "politics of individuation" that stands in a contradictory position to collective action, solidarity and participation.For groups that traditionally have valued its secrecy, sometimes on good grounds regarding a history of surveillance and infiltration of groups on the marginal left, the priority to keep oneself hidden is in conflict with the visibility paradigm of digital media.Furthermore, Fenton and Barassi (2011) argue that the logic of selfcentered participation promoted by social media can represent a threat for political groups rather than an opportunity.This tension reflects an ideological contradiction within autonomist tradition between individual liberty and collective solidarity.Autonomist politics have sometimes been described as a "politics in the first person" (Leach & Haunss, 2009, p. 262).Related to this, lifestyle politics and slacktivism has long been targeted as a problematic consequence of social media activism.
The process of de-socialization could be understood both as an extension of a tradition of secrecy within the milieu, but also a strategy of how to oppose a technical architecture that reduces the ability to control over content.Put another way, the conflict between control and emancipation as identified by Dencik and Leistert (2015) is relevant here as well.The groups that were included in the study constituted a very secluded milieu.They did not show much interest in using the web to attract new followers or recruit activists.Control over content appeared to be more important than sociability.

Discussion: Non-Participation and Empowerment
Previous research on militant groups associated with the radical left have pointed out that these groups sel-dom take an active part in designing propaganda or make efforts to reach out for new possible sympathizers (e.g.Peterson, 2001).Propaganda is understood as a means associated with the extreme right.It is therefore not surprising to find the type of active non-participation presented in the analysis.Theoretically, we can assume that there are numerous reasons for these kinds of groups to abstain from online communication and corporate social media (Dencik & Leistert, 2015).Still, as was shown in the empirical analysis, abstention is not total, but rather selective and partial.From these findings we can make the following observations: First, empowerment is a complex term (and it is difficult to speak of empowerment without hearing the voice of those assumedly empowered).Still, if we understand empowerment partly as acts to claim autonomy, it is possible to review much of the online media practices described above as active non-participation with claims to autonomy.It is clear that much of the non-participation is a consequence of active choice.The groups are present online, but they have orchestrated their presence in a way that reduces the degree of participation, for instance by disabling possibilities for commenting and keeping strict directives of how to get contacted.
Following this, when these milieus are put in a historical context, it shows that they have a tradition of "acting without being seen" which collides with the affordances of connectivity and visibility of social media.This tradition, in turn, might help explain the tendency to treat online media as channels for the broadcasting of propaganda (rather than open it up for reciprocal communication) as a form of non-participation.

Conclusions
From what we have seen in this study, there are clear instances of active non-participation in digital media in the milieus included in the sample, which opens the field for further inquiries about activism and nonparticipation as media practice.Because this study was limited to analyses of online content, questions that concern intentions are beyond its scope.Hence, further research into the strategy and political implications of disconnection and non-participation would be instrumental to complement these findings, as well as would also efforts to put them in historical and geographical contexts.Such research would require a different methodological approach, preferably including interviews.In addition to this, the question of how a certain ideological position is associated with a media practice that builds on non-participation is worth elaboration, especially considering possible variations between different ideological groups.Given its limitations, the study shows how a non-participationframework can provide a fresh perspective for understanding activism in the digital age.