Live-streaming for frontline and distant witnessing: A case study exploring mediated human rights experience, immersive witnessing, action, and solidarity in the Mobil-Eyes Us project

by Sam Gregory

As activists have used live-streaming in movements such as the ‘Arab Spring’, Occupy, 15-M, the Vinegar Revolution, and Black Lives Matter, rhetoric around these and other increasingly immersive technologies has frequently focused on their ability to mobilise attention and generate affective and effective connections at a distance. From 2016 to 2019, the human rights organisation WITNESS set out to explore how emerging tools of live-streaming could be used to better enhance connection, collaboration, and co-presence for frontline activists who were organising in the favelas of Rio de Janeiro and documenting state violence. The ‘Mobil-Eyes Us’ (MEU) project looked at how ‘live video, immersive video and other technologies of co-presence can connect people to meaningful frontline experiences of human rights issues and to meaningful actions that support activists in real time’.[1] This paper explores the project in the broader context of live-streaming, solidarity, immersive witnessing, and mediated suffering. MEU focused on both ‘frontline witnesses’, activists and community members streaming, as well as ‘distant witnesses’ in networks of viewing, action, or solidarity.

Live-streaming is often metaphorised as a portal between frontline witnesses and distant witnesses. In MEU, WITNESS explored methods to bring distant witnesses – both within Brazil and internationally – closer to being actors in solidarity without perpetuating the ‘improper distance’ that immersive witnessing can create. Explicitly the project centred the agency and control of actions around the livestreaming with the frontline witnesses, not the distant witnesses.

In this paper I highlight the diversity of human rights livestreams over the past decade, and the range of approaches to understanding these streams as witnessing and their relationships to other new and old media technologies and practices as well as the political moment in which the project took place. I explore how theoretical frameworks informed the MEU project from the start, and how they were addressed in the project.[2] Responding to the landscape around human rights, witnessing, and solidarity these underpinnings were centred on questions of mediating distant suffering, of self-expressive spectatorship,[3] of the role of co-presence as explored in discussions of virtual reality and immersive experience,[4] and the evolution of the framework of collective, connective, and distant/frontline witnessing.[5] I connect this discussion with the concept of denial of human rights violations.[6] I explore how a livestreaming practice of frontline/distant witnessing steps outside paradigms of mediated ‘suffering’, and following Ong and others[7] de-centres a relationship of suffering, peak moments rather than daily challenges, and the viewer looking from the West to the majority world.

Human rights live-streaming practices of witnessing, evidence-gathering, and mobilisation over the past decade

Human rights live-streaming builds on the affordances and the production methods of previous technologies used both in mainstream media and in activists’ and people’s media. It relates to existing practices of human rights NGO and activist witnessing, activism, and journalism where frontline witnesses communicate information to a distant audience. Live-streamers also inhabit the overall witnessing shift Mortensen and others describe from ‘performing in the media, to witnesses producing and distributing their own pictures’[8] as well as challenge the dominant mass media model-based assumptions about what counts as worth documenting and sharing. The expansion of more widespread live-streaming in the decade of the 2010s is also embedded in a broader socio-political setting of powerful and fragile digital protest[9] and increased digital human rights documentation as well as increasing authoritarian government surveillance of digital activity and repression of online and offline civic space.[10]

Live-streams are used in a range of ways. The experience of Occupy, the Vinegar Revolution, Black Lives Matter,[11] Standing Rock, #NoMuslimBan, and the 2020 movement for racial justice in the US provide insights into protest live-streaming as connective witnessing per Mortensen.[12] This witnessing creates, maintains, generates, and engages publics, as also articulated in Richardson’s ‘Black witnessing’ discussion of the Movement for Black Lives.[13] In the Occupy movement, Kavada and Treré explore how live-streaming reflects movement politics – its ‘immediacy, rawness, liveness and embedded/embodied presence connect with the movements’ understanding of how democracy should be practiced’.[14]

During the ‘Arab Spring’ and subsequent civil wars, Syria Pioneer in Homs and Mo Nabbous of Libya Ahurra TV illustrate how individuals or small organisations initiated uses of livestreams to engage global audiences.[15] During the period of increased refugee flight to Europe journalists similarly tapped into the perceived immediacy and proximity of live-streaming to document visceral experiences such as refugees crossing from Turkey to Greece in a boat by night,[16] while other individual activists such as Elin Ersson, as related in Khattab, used live-stream in rapid micro-mobilisation around specific incidents such as forced deportations of an asylum-seeker.[17] Faced with barriers to physical presence, activists at Standing Rock engaged millions with live-streams from the ground and the air. This frontline witnessing created what Martini describes as ‘a strong sense of proximity with what happens on the ground, the actions of the filming-users and the media presence of other watching users’.[18] In certain scenarios, the novelty or the widespread prevalence of livestreaming has enabled a proliferation of simultaneous broadcasts of singular events – for example, the attempted coup d’état in Turkey in 2016 and the coup in Myanmar in 2021.

As Wendy Chun observes in talking about new media studies, we should focus on the ‘importance of shifting from “what is new” to analyzing what work the new does’.[19] New media formats exist in relationship to older practices, in ongoing feedback loops. In one example, live-streaming acts in intersection with practices of evidence-gathering and accountability that focus on the integrity of specific video records.[20] The live-stream acts as a record a government cannot seize as easily as an activist’s camera. Other usages have explicitly explored particular practices of live-streaming as sousveillance. The DeGuarda work of Nossas Cidades in Rio involved collective monitoring via livestream of a school chosen for demolition by the city government. Copwatchers use live-streams in day-to-day practices of monitoring police encounters (see Fan).[21]

It is important to avoid what Treré terms ‘technological presentism’ as well as related issues of ‘one-medium fallacy’.[22] Live-streaming must be contextualised alongside other forms of media production and media technologies and within existing activist practices and social movement media cultures. This mitigates a ‘technological presentist’ focus on the newest and shiniest technology and the ‘one-medium fallacy’ of focusing on the ‘use of single technologies without disentangling the whole media spectrum with which activists interact’.

Contextualising human rights live-streaming as immersive witnessing, spectatorship, mediated suffering, and as a response to denial

Human rights live-streaming within a witnessing paradigm, per Pantti, requires construction of authority, authenticity, and moral responsibility.[23] It does this amidst a competition for attention where witnessing has shifted from a rare, existential struggle to speak to a quotidian status; and where different acts of digital witnessing themselves, as Chouliaraki notes, ‘vie for visibility’.[24]

In the idealised frame presented by the technological promoters of livestream innovation, livestreams effectively mediate suffering and enable action via empathy. As one of the early VC investors in the Twitter live video tool Periscope described when meeting the app’s founders: ‘I was struck by their desire to spread truth and empathy by enabling anyone to experience the world through someone else’s eyes.’[25] The reality of live-streams is more complicated. As multiple participants in the MEU project noted, people would join their live-streams to look for the violence, and leave if there was none. Or they noted that ‘haters’ were often the first to arrive, not the allies.

The search for violence[26] and the nature of the audience in live-streams reflect critical underlying questions of the MEU project and its relationship to both so-called ‘mediated suffering’, as well as ethical frontline and distant witnessing. First, how to re-centre the agency of the frontline witnesses, and their presentation of lived experience, the gaps and the needs. How does one challenge the constraints of ‘who is/is not considered human and under what circumstances’[27] when frontline witnesses are being observed and when they speak?[28] Second, what narrative approaches, technology affordances, and action opportunities help distant witnesses move from a spectatorship perspective to a solidarity perspective[29] and from self-expressive activism to allyship?

Explicitly the MEU project took an empirical approach to how people actually respond to human rights livestreams, rather than how they should normatively respond. Research into mediated distant suffering has been a growing field since the conception of the MEU project in 2013.[30] While it provides valuable insights, I also push back on the framing of this field of research as focused on mediated suffering. Frontline witnesses present a far more complex and nuanced representation of community resilience, agency, and loss, in which visible violence and suffering is contextualised in other realities and narratives.

Theories of mediated suffering and witnessing informed both the public framing and the implementation of the MEU project. Immersive witnessing technologies such as virtual reality and live-streaming provide an enhanced sense of presence and agency in a human rights context per Gregory and Nash.[31] An analysis of live-streaming practices that look to increase the co-presence of frontline and distant witnesses must attend to Chouliaraki’s questions of ‘improper distance’ and the elevation of ‘truths about ourselves’ over the experience of others. Chouliaraki’s work on the depiction of extremity and suffering focuses on differentiation between ‘representations of suffering that may simply bring a tear to spectator’s eye and those that may actually make a difference in the sufferers’ lives’ or deliver ‘truths about ourselves’, informing our thinking on the relationship between content, producers, and ‘audience’.[32] Chouliaraki’s ‘improper distance’, which identifies how audiences ‘subordinate the voice of distant others to our own voice and so marginalize their cause in favour of our narcissistic self‐communications’, highlights a key risk in a performative and emotive medium like live-streaming.[33] Live-streaming is commonly metaphorised as a portal between people geo-located and bodily integrated into struggles, and the broader networks of activists locally and globally. In this case, how do those distant witnesses move closer to being an actor in solidarity without breaching the ‘improper distance’ which immersive witnessing techniques may amplify?

Work on denial of atrocity informed this project, drawing on Cohen’s work as further developed by Seu.[34] This research considers how denial is sustained in the face of human rights appeals. Denial can operate at multiple levels of cognition and morality. Bystander action is compromised without active experience of the reality of the rights violation. Seu further explores the rhetorical and psychological strategies through which people shirk responsibility for responding to rights appeals while managing to protect their self-identity as a righteous activist, such as by attacking the messenger rather than the content.

The project responded to particular emerging areas in research on mediating distant suffering as highlighted in the survey work by Huiberts, Seu & Orgad, and Ong.[35] The first generation of scholarship – while asking rigorous questions – primarily looks at the phenomenon through mass media and humanitarian communications. Often the perspective is that of a distant other looking from outside, and deploying a gaze from the West to Global South in moments of peak violence. The MEU participants were not external journalists, but ‘small media’ actors such as local journalists and activists. Their audiences were also not always distant in the terms on which the first generation of scholarship focused. Ong and Richardson emphasise reception of mediated suffering by related communities.[36] In a project like this the audiences are not necessarily a West and ‘other’ but include audiences within Brazil, and in differing relationships of power and proximity to the situations in the livestreams.

Frontline witnesses also have agency. Scholarship on mediated suffering frequently presumes limited agency of those in the situation depicted. However, in the livestreaming encounters described in this paper the frontline witnesses and their communities have significant agency to engage with immediate and structural harms. Following Ong, a key research question was also how attention to the ordinary details of activism and resistance, and particularly, how the experience of joy and community resilience formed part of witnessing.

Methods and sources

This paper is based on contemporaneous records, interviews, notes, videos, livestreams, and convenings as well as the author’s own participation in the project. I analyse MEU in relation to themes of solidarity, mediating distant suffering, and confronting denial that were central to its conceptualisation and were utilised in the presentations and discussion of the project. In the accounting of lessons learned I interweave a chronological progression with a focus on key elements of learning in relation to interaction, connection, and action.

Although the author had access to information that is not visible in the public presentation of the project, I do not share information that is confidential or carries security risks, and avoid naming specific participants.

Context to favela-based media activism and the Mobil-Eyes Us project

Mobil-Eyes Us (MEU) was first conceptualised during a WITNESS staff member’s fellowship focused on considering how evolutions in technology, activism, and human rights provided signals of potential futures. The ‘Together Now’[37] research was workshopped with local and international participants. It explored conceptions of ‘co-presence for good’ centred on ‘using the sense of being together with other people in a remote environment to drive concrete, productive actions, engagement and understanding across barriers of geography, exclusion and time zones’.

A critical backdrop to the initial project development was the popular protest movement – the so-called ‘Vinegar Revolution’ – that emerged in June 2013 in Brazil, sparked initially by calls for revocation of a price increase for bus travel. Thousands of people joined protests in the streets in Rio and Sao Paulo, and subsequently in other cities.

Live-streaming played an important role in the Vinegar Revolution ‘not only of engaging people to reclaim the streets politically and creating a process of co-presence and support between the streamer and their audience, but also of producing evidence in legal courts, by sentencing or proving the innocence of activists. Many of them went to jail or were targeted by surveillance schemes or court lawsuits.’[38] One prominent group was the media collective Mídia Ninja which saw a rapid growth in its viewership over the summer of 2013. Another group was Rio Na Rua, a collective of social communications professionals and students.

In the favelas of Rio De Janeiro, community-based media collectives were also among early adopters of live-streaming as a tool for exposing human rights violations. A number of these collectives were in dialogue with NGO activists and others who participated in the 2013 protests. One of these networks, the ‘Straight Talk Collective’ (Coletivo Papo Reto, CPR), is based in the Rio favela of Complexo do Alemão. It has functioned as a hub for documentation of violence by residents, a cultural facilitator, and as a mobiliser for responses to state violence. Its members have played hybrid roles as photojournalists and reporters for mainstream media as well as autonomous community reporters.

Police, military police, and paramilitaries kill young, predominantly Afro-Brazilian residents of favelas at very high rates. These extrajudicial killings typically go unpunished. The WITNESS presence in Brazil works closely with CPR members and other human rights and social justice networks and community organisations both on more experimental projects such as those discussed in this paper and on evidence gathering and building advocacy cases.

In 2015 WITNESS began to advance an exploration of the emerging technological trends – live-streaming and other immersive technologies of co-presence, task-routing to assign work within a network, and new forms of solidarity activism identified in the ‘Together Now’ approach. An earlier paper by the author explores these issues in greater conceptual depth, laying out the underlying technological trends.[39] As a practical project WITNESS framed its approach in the following terms (here from a public presentation outlining the work).[40]

WITNESS explicitly positioned the work in terms of both witnessing and solidarity in our framing of the approach, noting that it envisioned two user groups. These were identified as:

‘Frontline witnesses’. Storytellers and activists on the frontlines of change – particularly in locations without strong local networks. They do the livestreaming and decide on/make the asks of distant witnesses. They are in control.

‘Distant witnesses’. Millions of people in distributed activist networks countrywide and worldwide who care and want to act. They contribute skills, presence, voice and actions based on the requests of frontline witnesses.

Explicitly the project did not imply that participation in live-streams implied participation in decision-making. Distant witnesses could participate directly only in the experience of being there and in the actions, but not in the power to make decisions about what actions to prioritise.

While research began in 2013, and preparatory work took place in 2015-16, the activities covered in this paper took place between summer 2016 when WITNESS led the initial Alternative Rio Live! project during the Rio Olympics, through mid-2019 when WITNESS ended active work on the project. MEU was small-scale in scope, focused on working alongside existing efforts and technologies, and guard-railed by the security and resource constraints of partners. MEU should also be placed in the context of a range of larger favela-based and urban initiatives to counter and document police violence, such as Nós por Nós and Defezap, and Fogo Cruzado, a reporting tool for shootings and police violence. Equally important contexts were a broader global movement expanding in these years to respond to systemic racism and state violence and a stark contrast in mainstream media coverage of the ‘other’, both within societies such as Brazil as well as in European responses to the forced migration crisis from the Middle East and North Africa.

The project included in-platform use of Facebook Live and Periscope as well as a dedicated app that embedded Facebook streams in a third party tool. Within the app, a small test group of distant witnesses could view aggregated livestreams focused around an issue or a campaign. They could provide translation or context. A producer role could push notifications to select groups of participants asking them to amplify content, or make it more visible at particular moments.

Engaging distant witnesses in solidarity: Learning from the Mobil-Eyes Us project

On a wintry Rio morning in mid-2016, a WITNESS consultant began a livestream as part of the Alternative Rio Live! Project. The project was taking place during the Rio Olympics 2016. Behind the haphazard surface glitz of the Olympic venues, activists had documented the harms of forced evictions leading up to the Games. They had focused on the inequities in transportation planning and investment, the enrichment of corporate interests at the expense of the poor, and observed with horror an increased militarisation of the favelas and a spike in military police killings of poor residents and Afro-Brazilians during the Olympics lead-up and event itself.

The WITNESS consultant and a local activist with one of the media collectives in Rio did a series of public live-streams from human rights contexts in Rio using both Facebook Live and Periscope. These were shared organically on social media platforms; people were notified who followed WITNESS social media accounts, and a direct SMS was sent to people who had opted in to a witnessing group that allowed for different levels of potential interest and participation (e.g. only around a major event, frequently, and with what opportunities to act). Because this pilot was conducting during the Rio Olympics, both locations and hashtags also resulted in substantial accidental audiences – for example when the activists streamed from the Vila Autódromo site, where forced evictions had taken place to make way for the Media Centre at the Olympic site. A Facebook Live Map functionality available at that time on Facebook also made it easy for casual viewers to zero in on live-streams taking place in the Olympic area.

Curation as a practice to address problems of presenting fragmentary experience and episodic violence and to resist casual spectatorship was embedded in the project.[41] Project participants began by developing ethical standards they would apply to sharing of both livestreams and recorded videos, using a framework for ethical video sharing that centred privacy and consent, intended use, and informed choices about dissemination.[42]

Curation informed how subsequent stages of the project were designed, both to manage information overload that can characterise livestreams and obscure accountability,[43] and to foster a relationship between distant and frontline witnesses. During the initial Olympics work, the project tried-out an experimental livestreaming curation tool focused on single events.  Deepstream.tv could embed commentary and prompt boxes for maps, videos, texts, and links beside a livestream and enabling switching in one interface between different livestreams.[44] Later in the project post-Olympics, in addition to curation of information around single livestreams, the team looked at how to create arcs of experience across multiple live-streams. This would provide cohesive engagement for consistent groups of distant witnesses rather than focus on single spectacles and one-event audiences. It would allow opportunities to shift focus from distant witnessing of ‘peak tragedies to exploring how political solidarity manifests in the quotidian’.[45] This focus on arcs also attempted to de-personalise live-streaming as focused on a streamer identity and instead focus on shared realities and broader themes.

Working with activists, one of the initial steps was to group types of live-streams that would reflect opportunities for ‘experience and solidarity’, for ‘action’, and for ‘shared joy’. How could one engage an audience over time by combining a walk-along in a community impacted by forced evictions with an expert perspective on the situation, participation in an iconic space or site of power like an official hearing, and a planned non-violent direct action? Story arcs intersected with actions that were identified with frontline witnesses. Building on adhoc action approaches in previous movements,[46] MEU initially considered how to leverage the presence and power of distant witnesses in specific situations and on advocacy targets, utilise rapid media response and amplification, and secure distributed analysis and expert guidance. In one early experiment, the team used what they termed ‘actiontags’ in an experiment explicitly focused on how to address bystander inertia (following Cohen[47]) while avoiding spontaneous action detached from the frontline communities’ needs. These actiontags from the streamers, also shared on social media, encouraged people to take specific actions that would be helpful for frontline witnesses in that moment. Actiontags included ^MAKEVISIBLE to ask people to share the footage, ^BEPRESENT to request a crowd, and ^CONTEXT and ^TRANSLATE to highlight requests to contextualise and translate. These approaches also recognised the one-to-many cognitive overload of the streamer in providing non-verbal information such as links,[48] and the challenges of building solidarity across language barriers.

As Cohen identifies, denial is easiest in de-contextualised settings where observers do not experience the lived rich, personal, social reality of a situation nor feel they have any options to act.[49] Fragmentary witnessing, such as when audiences stumble upon de-contextualised experiences of suffering or violence via the Facebook Live map or via a Rio Olympics hashtag, compromises solidarity. The project looked to focus explicitly on less high-profile situations and day-to-day oppression rather than typical high-profile protest livestream scenarios. In these lower-profile cases they expected even less contextual understanding by a spontaneous audience.

Our approach built on a range of live-streaming formats that were seen in use in existing activism. An analysis by Medeiros, a consultant on the MEU project, of over 100 livestreams and the activities of 18 Rio-based activists over the period 2016-18 identified five characteristic formats: ‘call-outs’, solidarity march coverage, pre-planned direct action, rights violations registered on video streams to be used as evidence, and long-duration streams.[50]

Call-out live-streams asked local residents to be present physically at an event, and also provided advance warning on ‘walk-through’ opportunities for individuals who would not be able to participate in an action.

This video above was one among twelve live broadcasts produced by CPR in a campaign to denounce violations in Alemão.[51] In these videos, favela-based activists walked through the favela noting the impact of military police activities, such as the occupation of civilian homes as firing bases around the Praça do Samba. Here ‘accompaniment’, a traditional human rights protection strategy, includes both the individuals online as well as representatives from the Public Defender’s Office, media, and other human rights groups participating in-person. At times, as in an August 2018 walk-through livestream from CPR during a large-scale action by military police, walk-through streams also help confirm whether it is safe for residents to come out into the streets, and provide a comment feedback loop for residents to determine safety.

Immersive witnessing, often via virtual reality, as a tool for ‘walking in someone’s shoes’ or more naively, an ‘empathy machine’, was a prevalent idea at the launch of the project.[52] In the WITNESS work on immersive witnessing they had refocused attention on the growing wave of live media that contained similar elements of immersion, presence, and agency to the more technically advanced VR and 360 experiences.[53] In walk-throughs, fundamentally one walks alongside, and often literally one step behind those you are in solidarity with (echoing the filmic physicality of activist video as discussed by Snowdon in his analysis of streams and videos from the ‘Arab Spring’).[54] The project rejected the idea of ‘walking in someone’s shoes’ with the implication of its ability to fundamentally experience someone’s differing reality.

Witnessing does not easily fit into particular genres and often bridges roles and intentions.[55] Often these walk-throughs of favela situations bridged into another approach to livestreaming noted by Medeiros. The use of live-streams to document and preserve evidence of violations is often associated with protest documentation where it serves to provide an instantaneous archive that the state cannot seize by taking your camera or phone. It also characterises situations of prolonged conflict such as the favela. In live-streams streamers document bullet holes in walls, show evidence of occupied homes, or explain the presence of spent ammunition on the ground.

Within MEU the project attempted to integrate these call-outs and walk-throughs into arcs of additional streams that provided contextual information as well as entry into the formal mechanisms that try to address the issues, such as a public hearing. One example was documentation and live-streaming around issues facing the residents in a sub-district of Alemão known as Favela da Skol. A series of evictions had taken place in Skol and adequate replacement housing had not been provided. From this MEU constructed an arc of livestream stories that included a series of interviews and a walk-through of the rubble at the site of the demolished homes.[56] Distant witness collaborators added further information in the comments documenting the background of the issue as well as links to action.

This approach to story arcs was also a response to the typical narrative tropes of outside journalism on favelas, and the desires of frontline activists to assert another reality. Outside media coverage of favelas is episodic, often implicated in parachute journalism and fixated on particular tropes. Talking of the phenomena of favela-based media production more broadly, Scott notes how favela-based activists (in contrast to the work of professional journalists), rather than ‘reaffirming the need for state intervention, as previous iterations of violent imagery often achieved… introduce an alternative view of digitally reproduced images as a source of counter-hegemonic authority that is capable of deconstructing the favela’s oppression, suffering and chaos’.[57] Videos of individual acts of violence in favelas must be grounded in the broader matrix of structural neglect that legitimises state violence and fails to invest in community-based initiatives to improve health, jobs, and education issues.

Enabling this counter-voice is particularly challenging with livestreams. As one participant noted in a debrief conversation: ‘people join fast to broadcasts of (military) operations, but then leave quickly; for events, people join via notification but then when they see it’s not violence they leave’. Media activists faced an ongoing dilemma about how to bring the right audience to the streams and the right framework to that audience’s engagement. As one participant put it: how do we move from talking about police violence to talking about ‘activities interrupted by police violence’ and how do we ‘create spaces and energy, not just show another shooting?’ Live-streams of violence overshadow everything else. As Poell and Kavada & Poell discuss in highlighting the simultaneous temporal dimensions of ‘contentious publics’ in live-streaming:

This emphasis on the now focuses attention on the violence and spectacle of events occurring in the streets. Real-time reporting often concentrates on sensational forms of protest and confrontations with police and security forces, rather than on structural problems and inequalities.[58]

Simultaneity risks, building on Chouliaraki, an ‘insulation between the public and private dimensions of the vulnerable other which renders his/her suffering public but keeps his/her history and aspiration out of view’.[59]

US Congressional representatives streamed themselves during the #NoBillNoBreak actions in 2015, and environmental movements use live-streams during direct action occupations. These practices of bringing distant witnesses into occupied sites of power inspired MEU. In the case of Favela Da Skol and the house occupations by the military police in Alemão, public hearings[60] provided opportunities for community leaders to address authorities. Within the MEU project they focused on these more day-to-day sites of institutional power, incorporating live-streams from public hearings and inquiries into arcs of narratives for distant witnesses. Remote translators provided translation and context either directly in Facebook as a resource in the moment and for later viewers, or in the dedicated MEU interface designed for remote solidarity watchers.

Research on live-streaming across genres including gaming, commercial live-streaming, and day-to-day spontaneous interaction highlights the pressures on a single streamer to keep pace with an incoming flow of comments. In the experimentation in MEU as well as actiontags, the team looked at roles a distant witness could play as a producer pushing tasks to different people and also as a contextualiser providing additional insight into a livestream based on their distinct expertise. In a walk-through of the Vila Autódromo site an anthropologist who had worked with the community provided additional context. In Alemão during the early piloting the project developed a curated contextualisation approach, where we brought in community activists and human rights defenders to layer in context, links, and translation as the broadcast progressed, rather than rely on the frontline witness or a spontaneous audience.

Contextualisation is neither apolitical nor self-effacing. Over the course of the project participants provided feedback contemplating the obligation to ‘perform’ when they did translation and provided context. They worried how these roles could be politicised and remove agency from the grassroots frontline witnesses, displacing solidarity. Distant witnesses were acutely aware of the possibility of what Chouliaraki describes as ‘narcissistic self-communication’[61] substituting for the reality and voices of the frontline witnesses. One participant pondered if ‘context providers/researchers eventually just use the opportunity to promote what they are most connected or aligned with’.[62]

In ‘Rules for Radicals’, Saul Alinsky identifies joy as a key component of radical activist tactics. Cultural events and community celebrations such as music or DJ events are a key element in collectives’ social media. MEU live-streams included a children’s Easter celebrations in Alemão in 2017 and an Easter event for children and their parents in 2019.[63] In the latter example, distant witnesses from Defend Democracy in Brazil (a US-based group of Brazilian solidarity supporters in the US) and Datalabe (a Rio-based non-favela group) added context and translation to enable structured interaction in solidarity by international audiences. A facilitator within the app asked distant witnesses if the discussion on housing and policing issues ‘resonates with experience closer to home’, and prompted people to provide background, ask questions, and identify points of solidarity.

The choice to centre celebrations reflected considerations both from WITNESS as well as the frontline activists. Joy and the power of savouring ‘small victories’ is an intrinsic part of most activism and we wanted to see how this fitted alongside the call-outs, the protests, and the direct evidentiary documentation of livestreams. Witnessing should be an embodied insight into joy as well as trauma. Distant witnessing also carries the same intrinsic needs for renewal as frontline witnessing and the same concerns about vicarious trauma as we see in other online communities exposed to the suffering of others. This has been a focus recently in the related community of so-called OSINT investigators who search online videos and sources for human rights information.[64]

Outside of the MEU mechanics the focus on community events also reflected two other elements. First, these were the same live-streams that the collectives used to engage local audiences. This reminds of the importance of ongoing intra-community witnessing of structural and direct violence. An analysis of solidarity witnessing must not make assumptions about distant audiences or dismiss the practical benefits of real-time witnessing for proximate audiences. Ong notes in his work how the archetypal witness is construed as a ‘Western middle-class witness’, distanced on multiple social and geographical scales from the mediated experiences of suffering. However, in his research on how different socio-economic groups in the Philippines responded to media witnessing of disaster, ‘low-income Filipinos…’ establish ‘an intense emotional connection to other people’s stories of suffering’.[65] Appropriate or improper distance to violence, along with how such viewing leads to vicarious trauma, will also depend on a person’s racialised, gendered, and intersectional existing exposure to it, as Richardson notes in relation to Black Americans seeing images of police violence in the US.[66]

The use of these community event live-streams also reflected security concerns about the more targeted human rights formats of live-streams. Frontline witnessing always carries risks. In the US, Richardson highlights the ‘weighty baggage’ of witnessing whereby in the ‘risks with bearing witness while black are manifold’ for the primarily Black and Latino frontline witnesses who document police violence’.[67] Witnesses and media activists in Rio de Janeiro face even greater physical threats to their safety in the moment of filming as well as subsequently. Platforms are also used against them – for example, the creation of fake profiles that appear to implicate them in criminal activity.

Live-streaming practices can focus on radical transparency and volume, as practiced in the Occupy and M-15 movements. However, neither existing usages nor the affordances of the MEU project focused on a ‘radical kind of transparency that entailed a near-total visibility of every movement activity, including internal meetings’[68] identified by Kavada and Treré, but instead on a selective presentation. In the security context of the favelas, it was not a viable option to enable further ‘self-surveillance’. The perceived transparency within livestreaming that invites solidarity occurs within the ongoing management of parameters of visibility and invisibility. This boundary management is a core security practice of activists highlighted by Ganesh & Deutch & Schulte[69] as well as key to non-activist Instagram and short-form video influencers who manage a public face and create ‘perceived interconnectedness’ per Abidin.[70] It also reflected a WITNESS-lead advocacy perspective that livestreaming could function within structured engagement, and that information overload inhibited action and could in fact be used as an active excuse for disengagement (following Seu on psychology of denial).[71]

Events in 2018 increased activists’ risk profile, including the military police take-over of urban policing in early 2018, the assassination of city councilwoman Marielle Franco in mid-2018, and the election of President Bolsonaro in late 2018.  Favela-based grassroots groups were less frequently live-streaming in public because of the associated security risks, and also considering what activity to do in private live-streaming, Stories, or closed messaging apps. The final MEU activities reflected both this insecurity as well as the clear consensus from activists to try and de-centre state violence as the narrative about favelas and re-emphasise community integrity and joy within their own content production.

In 2019, WITNESS Brasil  organised a convening that included participants from the MEU project and others using live-streaming. Two key themes emerged; one was this broader context of risk. One collective noted that it no longer accompanied protests with live-streams because their media was stolen by the mainstream and it was the ‘haters’ and the bad actors who joined first. Another participant chimed-in that in public videos, ‘haters and the military police were the first to comment’ and it was tiring to be ‘giving a class’ to them. Even within a framework of ‘surveillance realism’ where live-streamers expect adversaries to join their streams,[72] this went beyond even ‘self-surveillance’ into self-harm. Participants wanted to use live-streaming to contextualise broader societal issues such as drug policy and health and housing inequalities rather than show violence or perpetuate stereotypes. At the very least they wanted to contextualise that violence. Participants were focused on more controllable formats – the short studio live broadcast as used by Voz da Comunidades or as later used by CPR during the pandemic, and the Gabinete de Crise ‘Um minuto’ broadcast. They mixed live studio work with managed street live-streaming, and other social media approaches to ambient co-presence such as the Stories format of Instagram.

The pessimism about haters also reflected the evolution of expectations around new media tools and the understanding that the ‘aesthetics and strategies of amateur video are being used regularly nowadays not only by human rights or progressive groups, but also by status quo agents, army forces and conservative networks’,[73] starting at the top in Brazil with what Simon describes as the ‘politics of digital intimacy’[74] practiced in the informal Instagram and Facebook Live streams of the election campaign of President Bolsonaro. Live-streaming usage was increasingly even more non-cosmopolitan in aspiration and non-human rights driven in practice.

The chicken or egg of live-streamed violence, and looking ahead

It was in one of the final debriefs on the MEU project that one participant posited the experience of live-stream engagement as a ‘chicken and egg’ problem. Is it that people come to the live-streams because we have set up an expectation that they will show violence? Or is it that people come because they are looking for violence?

Human rights live-streaming has been a central part of the shift to a more diversified witnessing practice over the past decade, serving movements, media-makers, and individual activists for diverse purposes including mobilisation, evidence-gathering, and witnessing for distant publics. It sits aside other technologies, novel and older, and practices emergent, established, and evolving. Live-streaming, like all witnessing, struggles for attention in the overall media ecosystem and vies for visibility and trust with other human rights content. The experience of the haters and the violence-seekers coming first is common to many streamers.

Within MEU the project looked to empirically grapple with how to move participants from spectatorship to solidarity, and from self-expressive communication to interaction (following Chouliaraki). It did this with a consciousness of the power of denial (Cohen and Seu), and the weaknesses of a turn towards immersive witnessing in potentially creating ‘improper distance’. It also wanted to move beyond a focus on distant ‘suffering’ to reassert the agency of frontline witnesses. Their narratives of ordinary life, of joy as well as pain embody a counterpoint to media narratives of favelas, and potentially generate stronger solidarity and action. Following Ong and Richardson we also reasserted the impact of intra-community frontline witnessing.

Across a range of research into live-streams that identified characteristic types, as well as a series of campaigns and pilots that integrated existing formats and approaches, we learned that curation was useful to engage distant witnesses, to confront denial and pre-existing tropes of favela violence, and to address both the safety and engagement rationales for managing the representation of community life. Live-stream simultaneity combined with violence perpetuated problematic representations. Live-stream transparency was not appropriate for security or action. Arcs of ‘discovery’ in walk-throughs, access to power spaces, and ‘shared joy’ provided entry points for distant witnesses, but also revealed fissures when those distant witnesses were given responsibilities in solidarity, challenging them to ask about their political power and whether they were acting as an ethical distant witness. Increasingly over time, security concerns came to the fore, raising more questions about when ‘self-surveillance’ leads to self-harm.

Interaction, connection, and risk were constant discussions. There are multiple areas of potential research into live-streaming as an approach to mediating human rights relationships and generating solidarity and action. One area that would be valuable is empirical research into the responses of both distant and frontline witnesses to the experiences of narrative arcs and constructed campaigns as opposed to spontaneous live-streams. Further exploration of how ‘ordinary life’ is presented as a counterpoint to spectacle in live-streaming could build on other research areas around live-streaming in other domains. Much research to-date around mediating ‘distant suffering’ is grounded in broadcast or non-live media,[75] so approaches looking at not only live-streaming but also how social media tools of ongoing ambient co-presence, such as Stories formats, potentially create different relationships to people on the front lines of human rights contexts is critical.

Author

Sam Gregory is Program Director of WITNESS, which helps people use video and technology for human rights. He works on emerging challenges and opportunities in authenticity and evidence, live-streaming, narrative and media manipulation, deepfakes and AI. From 2010-18 he taught at Harvard’s Kennedy School, where he earned his Masters in Public Policy. He has  published in Journal of Human Rights Practice, Information, Communication and Society, Fiber Culture, and American Anthropologist, and was lead editor of the handbook ‘Video for Change’. He co-chairs the Partnership on AI’s Expert Group on AI and Media, and is on the Technology Advisory Board of the International Criminal Court.

Acknowledgements

I am grateful for the opportunity to collaborate on the MEU project with key favela-based activists and media makers, particularly at Coletivo Papo Reto. I thank the WITNESS team and consultants who worked on this project including Clara Medeiros, Adriano Belisario, Eric French, Shawn Van Every, and Victor Ribeiro.

References

Abidin, C. ‘Communicative Intimacies: Influencers and Perceived Interconnectedness’, Ada, 8, 2015: 1-16.

Belisário, A. ‘Rio Goes Live: Video Activism and Human Rights in Brazil’, WITNESS Media Lab, 2019: https://lab.witness.org/portfolio_page/rio-goes-live-video-activism-and-human-rights-in-brazil/ (accessed on 1 February 2021).

Chouliaraki, L. The spectatorship of suffering. United Kingdom: SAGE Publications, 2006.

_____. ‘Improper Distance: A Critique of Solidarity as Irony’, Mediation, Ethics and Otherness special issue in the memory of Prof. Roger Silverstone, edited by L. Chouliaraki and S. Orgad, International Journal of Cultural Studies, 14, 4, 2011:363‐382.

_____. The ironic spectator: Solidarity in the age of post-humanitarianism. New Jersey: Wiley, 2013.

_____. ‘Digital witnessing in conflict zones: the politics of remediation’, Information, Communication & Society, 18:11, 2015:1362-1377.

Chouliaraki, L., Orwicz, M., and Greeley, R. ‘Special Issue: The Visual Politics of the Human’, Visual Communication, 18, no. 3, August 2019: 301-309.

Chun, W. and Keenan, T. New media, old media: A history and theory reader. United States: Routledge, 2006.

Cohen, S. States of denial: Knowing about atrocities and suffering. Oxford: Blackwell Publishers Ltd, 2001.

Dubberley, S., Satterthwaite, M., Knuckey, S., and Brown, A. ‘Open Source Investigations: Vicarious Trauma, PTSD, and Tactics for Resilience’ in Digital witness: Using open source information for human rights investigation, documentation, and accountability, edited by S. Dubberley, A. Koenig, and D. Murray. Oxford: Oxford University Press, 2020: 271-291.

Fan, M. Camera power: Proof, policing, privacy, and audiovisual big data. United Kingdom: Cambridge University Press, 2019.

Fricker, M. Epistemic injustice: Power and the ethics of knowing. United Kingdom: Clarendon Press, 2007.

Ganesh, M., Deutch, J., and Schulte, J. Privacy, visibility, anonymity: Dilemmas in tech use by marginalised communities. Report, Tactical Technology Collective, 2016.

Gregory, S. ‘Co-presence: A New Way to Bring People Together for Human Rights Activism’, WITNESS Blog, 2013: https://blog.witness.org/2013/09/co-presence-for-human-rights/ (accessed on 1 February 2021).

_____. ‘Ubiquitous Witnesses: Who Creates the Evidence and the Live(d) Experience of Human Rights Violations?’, Information, Communication & Society, 18, 11, 2015a: 1378-1392.

_____. ‘See, Hear…Act?: Co-present storytelling and action’, WITNESS Media Lab, 2015b: https://lab.witness.org/portfolio_page/see-hear-act-co-present-storytelling-and-action/ (accessed on 1 February 2021).

_____. ‘Immersive witnessing: From empathy and outrage to action’, WITNESS Blog, 2016a: https://blog.witness.org/2016/08/immersive-witnessing-from-empathy-and-outrage-to-action/ (accessed on 1 February 2021).

_____. ‘Live video, curation, and action at #Rio2016 Olympics’, WITNESS Blog, 2016b: https://blog.witness.org/2016/08/live-video-curation-action-rio2016-update-witness-media-labmobil-eyes-us-projects/ (accessed on 1 February 2021).

_____. ‘Cameras Everywhere Revisited: How Digital Technologies and Social Media Aid and Inhibit Human Rights Documentation and Advocacy’, Journal of Human Rights Practice, 11, Issue 2, July 2019: 373-392.

Huiberts, E. The distant audience? A multi-method study about people’s reactions towards mediated distant suffering. PhD thesis, Ghent University, 2018.

_____. ‘Watching Disaster News Online and Offline: Audiences Experiencing News about Far-Away Disasters in a Postbroadcast Society’, Television & New Media, 21(1), 2020: 41-59.

Kavada, A. and Treré, E. ‘Live Democracy and Its Tensions: Making Sense of Livestreaming in the 15M and Occupy’, Information, Communication & Society, 23, no. 12, July 2019: 1787-1804.

Kavada, A. and Poell, T. ‘From Counterpublics to Contentious Publicness: Tracing the Temporal, Spatial, and Material Articulations of Popular Protest Through Social Media’, Communication Theory, 2021: 1-19.

Khattab, M. ‘Micromobilization through live streaming: the Elin Ersson case’, The journal of media innovations, 6(1), 2020: 50-62.

Kumanyika, C. ‘Livestreaming in the Black Lives Matter Network’ in DIY utopia: Cultural imagination and the remaking of the possible, edited by J. Hunsinger, J. Foley, L. Daily, L. Doyle, C. D’Ignazio, S. Duncombe, G. Aiello, L. Giannini, and C. Bergère. United States: Lexington Books, 2016: 169-188.

Licoppe, C. and Morel, J. ‘Visuality, Text and Talk, and the Systematic Organization of Interaction in Periscope Live Video Streams’, Discourse Studies, 20, no. 5, April 2018: 637-665.

Martini, M. ‘Online Distant Witnessing and Live-Streaming Activism: Emerging Differences in the Activation of Networked Publics’, New Media & Society, 20, no. 11, April 2018: 4035-4055.

Medeiros, C. ‘Introduction to Mobil-Eyes Us’, WITNESS Media Lab, 2019a: https://lab.witness.org/portfolio_page/introduction-to-mobil-eyes-us/ (accessed on 1 February 2021).

_____. ‘Calls-outs in livestreamings: learnings and practices’, WITNESS Media Lab, 2019b: https://lab.witness.org/portfolio_page/call-outs-in-live-streamings-learnings-and-practices/  (accessed on 1 February 2021).

Medeiros, C. and Belisário, A. ‘4 Tips to Have Co-Present and Active Audiences in Livestreams’, WITNESS Media Lab, 2019: https://lab.witness.org/projects/livestream-co-present-storytelling-action/  (accessed on 1 February 2021).

Medeiros, C., Belisário, A., and French, E. ‘Storytelling in livestreamings: The Favela Da Skol case in Brazil’, WITNESS Media Lab, 2019: https://lab.witness.org/portfolio_page/storytelling-and-narrative-arc-in-livestreamings-the-favela-da-skol-case-in-brazil/ (accessed on 1 February 2021).

Mortensen, M. ‘Connective Witnessing: Reconfiguring the Relationship between the Individual and the Collective’, Information, Communication & Society, 18, no. 11, July 2015: 1393-1406.

Nash, K. ‘Virtual Reality Witness: Exploring the Ethics of Mediated Presence’, Studies in Documentary Film, 12, no. 2, July 2017: 119-131.

Ong, J. ‘”Witnessing” or “Mediating” Distant Suffering? Ethical Questions Across Moments of Text, Production and Reception’, Television & New Media, 2012.

_____. ‘Witnessing distant and proximal suffering within a zone of danger: Lay moralities of media audiences in the Philippines’, International Communication Gazette, 2015: 1-16.

_____. ‘Toward an Ordinary Ethics of Mediated Humanitarianism: An Agenda for Ethnography’, International Journal of Cultural Studies, 22, no. 4, February 2019: 481-498.

Orgad, S and Seu, I. ‘The mediation of humanitarianism: Toward a research framework’, Communication, Culture & Critique, 7, 2014: 6-36.

Pantti, M. ‘Journalism and Witnessing’ in Handbook of journalism studies, edited by K. Wahl-Jorgensen and T. Hanitzch. United Kingdom: Taylor & Francis, 2019: 151-164.

Poell, T. ‘Social media, temporality, and the legitimacy of protest’, Social Movement Studies, 2019: 1-16.

Richardson, A. Bearing witness while black: African Americans, smartphones, and the new protest #journalism. United States: Oxford University Press, 2020.

Ronzheimer, P. Periscope Channel, ongoing: https://www.periscope.tv/ronzheimer/1DXGyAyDyELGM (accessed 15 May 2021)

Scott, J. ‘Death on Repeat’, Journal of Legal Anthropology, 3, no. 1, June 2019: 21-40.

Seu, I. ‘”Shoot the Messenger”: Dynamics of Positioning and Denial in Response to Human Rights Appeals’, Journal of Human Rights Practice, 3, no. 2, June 2011: 139-161.

Simon, F. ‘The Politics of Digital Intimacy’, 10 August 2019: https://medium.com/viewpoints/the-politics-of-digital-intimacy-2da2f0622088 (accessed on February 1 2021).

Snowdon, P. The people are not an image: Vernacular video after the Arab Spring. United States: Verso Books, 2020.

Treré, E. Hybrid media activism: Ecologies, imaginaries, algorithms. United Kingdom: Taylor & Francis, 2018.

Tufekci, Z. The power and the fragility of networked protest. United States: Yale University Press, 2017.

WITNESS Brasil. ‘Brasil: Redes de streaming compartilham experiências dos protestos de 2013’, 2013: https://portugues.witness.org/brasil-redes-de-streaming-compartilham-experiencias-dos-protestos-de-2013/, 2014 (accessed on 1 February 2021).

WITNESS Media Lab. ‘Announcing WITNESS’s Ethical Guidelines for Using Eyewitness Footage in Human Rights’, 2015: https://lab.witness.org/announcing-witness-ethical-guidelines-for-using-eyewitness-footage-in-human-rights/  (accessed on 1 February 2021).

WITNESS Media Lab, ‘Live Easter in Alemão, Rio de Janeiro’, 2019: https://lab.witness.org/portfolio_page/live-easter-in-alemao-rio-de-janeiro/ (accessed on 1 February 2021).

WITNESS Media Lab ‘Live-streaming and co-presence storytelling for action’, n/d: https://lab.witness.org/projects/livestream-co-present-storytelling-action/ (accessed on 1 Feb 2021).

---

[1] WITNESS Media Lab n/d.

[2] Gregory 2015a; Gregory 2016.

[3] Chouliaraki 2006 and subsequent.

[4] See Gregory 2016a and Nash 2017.

[5] Following Mortensen 2015, Gregory 2015, Richardson 2020.

[6] Following Cohen 2001, Seu 2011.

[7] Ong 2015, 2019.

[8] Mortensen 2015, p. 12.

[9] Tufekci 2017.

[10] Gregory 2019.

[11] Kumanyika 2017; Richardson 2020.

[12] Mortensen 2015, p. 5.

[13] Richardson 2020.

[14] Kavada & Treré 2019, p. 1.

[15] Snowdon 2020, pp. 215-225.

[16] Ronzheimer 2015, noting that the livestreams shot with refugees are no longer visible on the account.

[17] Khattab 2020.

[18] Martini 2018, p. 45.

[19] Chun & Keenan 2006.

[20] Gregory 2015.

[21] Fan 2020.

[22] Treré 2018, pp. 9-11.

[23] Pantti 2019, pp. 151-164.

[24] Chouliaraki 2015, p. 11.

[25] Shontell 2015: https://www.businessinsider.com/what-is-periscope-and-why-twitter-bought-it-2015-3

[26] And perhaps also activist ‘riot porn’ per Rasza 2013.

[27] Chouliaraki & Orwicz & Greeley 2019, p. 8.

[28] Although not explored in this paper Fricker 2009 and concepts of epistemic injustice are highly applicable.

[29] Following Chouliaraki 2013.

[30] For a field survey see Huiberts 2018.

[31] Gregory 2015a, 2016; Nash 2017.

[32] Chouliaraki 2006, p. 7.

[33] Chouliaraki 2011, p. 372.

[34] Seu 2011.

[35] Huiberts 2018; Seu & Orgad 2014; Ong 2019.

[36] Ong 2015, 2019; Richardson 2020.

[37] Gregory 2013.

[38] Belisario 2019; WITNESS Brasil 2014.

[39] Gregory 2015.

[40] See also similar versions, Gregory 2015b.

[41] Live.witness.org 2016.

[42] WITNESS Media Lab 2015.

[43] Kavada & Treré 2019.

[44] Gregory 2016b.

[45] Ong 2019.

[46] See WITNESS Brazil 2014.

[47] Cohen 2001.

[48] Licoppe & Morel 2018, among other HCI studies on live-streaming.

[49] Cohen 2001, p. 18; 68.

[50] Medeiros 2019b.

[51] Although some of these videos are no longer viewable, other examples are here, here, here and from Defezap here.

[52] Gregory 2016a.

[53] Ibid.

[54] Snowdon 2020.

[55] Gregory 2015a.

[56] Medeiros & Belisário & French 2019.

[57] Scott 2019, p. 6.

[58] Poell 2019; Kavada & Poell 2019, p. 14.

[59] Chouliaraki 2013, p. 187.

[60] Medeiros & Belisário & French 2019; Medeiros & Belisário 2019.

[61] Chouliaraki 2011, p. 368.

[62] WITNESS Media Lab 2019.

[63]  Ibid.

[64] Dubberley & Satterthwaite & Knuckey & Brown 2020.

[65] Ong 2014, 2015.

[66] Richardson 2020, pp. 153-178.

[67] Ibid., pp. 83-84

[68] Kavada & Treré 2019, p. 8.

[69] Ganesh & Deutch & Schulte 2016.

[70] Abidin 2015.

[71] Seu 2010.

[72] Kavada & Treré 2019.

[73] Belisário 2019.

[74] Simon 2019.

[75] Huiberts 2020.