MEDIATIONS OF OUTRAGE: HOW VIOLENCE AGAINST PROTESTORS IS REMEMBERED

Abstract:This paper examines how the memory of violence against demonstrators is culturally produced. In line with cultural theories of trauma, it assumes that the collective meaning of violence is socially mediated. However, it critiques the primacy of trauma as an analytic frame because its gravitation towards victimhood and negativity occludes the dynamic of action and reaction that defines contentious politics. Drawing on cultural memory studies to identify complex processes of mediation and remediation that render some events more memorable then others, the paper shows how the memory of outrages committed against demonstrators is made meaningful and imbued with a mobilizing capacity.

and 2008; Tilly and Tarrow 2015;Tarrow 2011).What is the impact of collective memory on claimmaking in the present?More specifically, how does the recollection of earlier events inform, shape, or inhibit collective action to change the world?
In an initial mapping of this "memory-activism nexus" (Rigney 2018), I distinguished between memory activism (how actors struggle to produce and change memory; exemplified in Gutman 2017;Wustenberg 2017;Altınay et al. 2019), the memory of activism (how earlier struggles for a better world are culturally recollected; exemplified in Reading and Katriel 2015; Eyerman 2016; Rigney 2016), and memory in activism (how the cultural memory of earlier struggles informs new movements in the present; exemplified in Harris 2006;Zamponi 2013 and2018;Della Porta et al. 2018;Chidgey 2018).There is a feedback loop between these different practices, meaning that an integrated approach to their interplay over longer periods of time is needed.This basic mapping of the field has since been echoed by Priska Daphi and Lorenzo Zamponi (2019) in a recent special issue of Mobilization where they invoke a comparable tripartite distinction between movements about memory, memories of movements, and memory in movements.Although their terminology is different, the result is a convergent research agenda that calls for new collaboration between scholars working in memory studies and those working in the field of social movements.
The present article is written as a contribution to this larger agenda, which it approaches from the perspective of cultural memory studies; that is, the subdiscipline within memory studies that brings expertise in the analysis of cultural artifacts and of storytelling practices to the understanding of how memory is produced and how it affects social action.This special issue on "cultural trauma" provides a welcome occasion to address an important aspect of the memory-activism nexus: how the memory of violence against demonstrators is collectively remembered, and how that memory feeds into later movements.
As Anna Reading and Tamar Katriel rightly point out in Cultural Memories of Non-Violent Struggles (2015), both public commemoration and memory studies have long focused on victimhood and war.In a new version of Jane Austen's famous complaint in Northanger Abbey (1817) that history books have been dominated by "the quarrels of popes and kings" (1923,108), they argue that this gravitation towards military states of exception has effectively submerged "the cultural memories of struggle and agency in non-violent contexts" (2015, 1).The critique offered by Reading and Katriel can itself be seen as a form of academic memory activism, designed to feed back into practices of public commemoration by providing new frames of reference and a new empirical focus.Their collection reflects this agenda in alternating between, on the one hand, essays that analyze how activists (especially in the women's movement) have been remembered by others and, on the other hand, studies that salvage more or less forgotten episodes of "non-violent struggle" and, in doing so, bring these hitherto occluded histories out of the archive and into society's "working memory" (Assmann 1999, 130-45).

Cultural Memories of Non-Violent Struggles has undoubtedly helped open up research into
alternative sites of memory production and reframe the agenda of memory studies by putting active citizens rather than victims at center stage.Recovering the history of nonviolence as an instrument of social struggle-what Martin Luther King in his Nobel prize acceptance speech famously called a "weapon" that "cuts without wounding" (1964)-provides a welcome reminder that not all of history has been violent, and that memory is produced in the civic realm as well as in theatres of war.
Nevertheless, to study the memory-activism nexus exclusively through the lens of nonviolence is shortsighted.To begin with, because nonviolent resistance continuously shades into more militant tactics, the border between the two is a structural matter of debate within movements themselves and those who remember them.More importantly for the present argument, nonviolence is regularly met by state violence, with the power of the people pitted against the force of the army or police (Arendt 1969).Since nonviolence has a history that is inseparable from the violence deployed against it, it follows that the memory of nonviolence is also entangled with the use of force.
As I began the first version of this essay (January 13, 2020), reports were emerging of the firing of live ammunition at demonstrators in Tehran, who had gathered by the thousands in the city to protest against their government's role in the shooting down of a civilian aircraft.A day earlier, the US president had tweeted: "DO NOT KILL PROTESTERS: The World is watching" (Sullivan and Cole 2020).While presidential tweets have to be taken with a huge dose of skepticism, this one reflects a reality; namely, that hundreds of demonstrators have been killed across the world in the last 12 months, and that the possibility of a civilian bloodbath is a scenario that shapes and overshadows both the actions of demonstrators and the mediated reactions of the international community.The killing of protesters is as old as contentious politics.But is it always remembered as traumatic?<A>FROM TRAUMA TO OUTRAGE Lucy Bond and Stef Craps (2019) have shown in a recent survey how the concept of "trauma" came to prominence as an academic tool of analysis in the post-World War II era and how its meaning has been inflected in different disciplines.In Empire of Trauma (2009), Didier Fassin and Richard Rechtman show how a victim-centered and therapy-oriented trauma discourse has also come to permeate public discussions and policymaking regarding post-conflict societies.As these works indicate, "trauma" has been used ubiquitously both as a category of practice and a category of analysis (Bourdieu 1972), and has ended up meaning different things to different people.This is illustrated by the divergent pathways that "trauma" has taken in the adjacent fields of cultural studies and cultural sociology, as well as in the evolution of the more specific term, "cultural trauma" (not to be confused with cultural trauma theory; Bond and Craps 2019), within the field of cultural sociology itself.
There have been both continuities and modulations in the concept of "cultural trauma" over the years (Alexander et al. 2004;Eyerman 2004;Alexander 2012;Eyerman et al. 2011;Eyerman 2019).At the core of all definitions, however, is the idea that the collective meaning of shocking events that "threaten society's existence" (Smelser in Alexander et al. 2004, 44) has to be socially mediated using cultural forms: "To transform individual suffering into collective trauma is cultural work" (Alexander and Breese in Eyerman et al. 2011, xii).Accordingly, the concept of cultural trauma theory entails the "analysis of the symbolic processes through which suffering-real or perceived-is inscribed with a compelling meaning" (Sciortini in Eyerman 2019, 11).The cultural production of a compelling narrative allows people to overcome the past and to come to terms with the irrevocable challenge to identity that suffering has occasioned (Alexander and Breese in Eyerman et al. 2011, xii-xiii).
At first sight, this cultural trauma theory as developed in sociology is very much at odds with trauma theory in cultural studies.Where the former approach has been constructivist (emphasizing the power of language and discourse to make social facts speak, if only after some effort), the latter has emphasized instead the limits of sense-making-"ineloquence" (Berlant 2011)-in the face of certain experiences.From the perspective of cultural studies, the traumatic is a function of events rather than of their representation, specifically of devastating events that radically disturb our capacity to represent the world and hence understand our place in it.Events experienced as traumatic are all the more profoundly disturbing precisely because they exceed our capacity to articulate them within the available models for meaning making.As formulated in the influential work of Cathy Caruth, a "trauma is the confrontation with an event that, in its unexpectedness or horror, cannot be placed within the schemes of prior knowledge" (1995,153); for this reason, it continuously returns in memory as an "unclaimed experience" (1996) that cannot be fully captured in language and hence integrated into a therapeutic process.Where theories of cultural trauma in sociology have focused on the outcome of mediation, trauma theory in cultural studies has emphasized instead the cognitive, aesthetic, and moral difficulties entailed in making sense of events that profoundly undermine identity and show up the "limits of representation" (Friedländer 1992).In line with this point of departure, cultural scholars have turned their attention to the ways in which writers and other artists have sought to creatively work around these limits.As the study of literature in particular has shown, the unprecedented cruelty of the Holocaust has provoked ever-renewed efforts to testify to its horror.It ushered in the "era of the witness" (Wieviorka 2006), in which testimonies at least bear witness to atrocity even if they can never fully represent it.
Cultural sociology and cultural studies thus take opposing approaches to the subject of trauma (one emphasizing how the meaning of events is collectively produced, the other emphasizing the difficulties in representing certain events).The two approaches converge, however, in positing a tension-filled gap between experience and representation when it comes to events that are an affront to shared understandings of the world.Even more fundamentally, the two schools of thought converge around the assumption, encapsulated in their shared use of the term "trauma," that suffering is the primary frame for understanding how societies and those who analyze them make sense of their histories.
But is trauma the only analytic frame available, or the most productive one?If we stick to the principle, central to all culturally inflected theories of trauma, be these in sociology or cultural studies, that the meaning of events is culturally produced, then it follows that collectively significant events might also be rendered meaningful within alternative frames and have protagonists other than "suffering" collectivities.After several decades of framing history in terms of trauma, the time seems ripe to reflect on the underlying cost of the analytic frame itself: that it structures events in terms of their negativity, highlights the figure of the victim, and occludes the visibility of alternative histories, including the history of nonviolent struggle and other collective expressions of hope in the possibility of achieving justice.
An emphasis on victimhood sits uneasily with the figure of the active citizen that is at the heart of contentious politics.For this reason, "outrage" may be more fruitful than trauma when it comes to framing violence perpetrated against demonstrators.The word "outrage" evokes the egregious violation of a norm or law combined with a moral judgement on the part of an aggrieved party.While outrages entail suffering, they also trigger a rebound in the form of righteous indignation or "apt anger" (Srinivasan 2018).Those who witness them are not just victims stunned into inaction, but also rights-claiming subjects who are "outraged" by the violations committed.Using 'outrage' as an analytic frame makes visible how the violence perpetrated against protesters is remembered within a dynamic of action and reaction or, to recall Manuel Castells, within a dynamic of anger and hope (2012, 1-2).This paper will accordingly carry over the idea, central to cultural trauma theories, that the collective meaning of events is produced through "symbolic renderings" (Alexander 2012, 4), but will apply this principle to mediations of outrage.

<A>THE DYNAMICS OF MEDIATION
Before specific cases are discussed, more needs to be said about what "symbolic renderings" entail; or, to translate this into the terms of cultural memory studies, about how meaning and affect are produced through cultural forms carried by media technologies (Erll and Rigney 2009;Erll 2011;Rigney 2016b).A factor of importance that deserves attention in its own right, mediation in this broad sense enables memory to find expression in publicly accessible "cultural carriers" (a term formulated here by analogy with "carrier groups"; Alexander 2012).It involves multiple operations whereby information is structured, circulated in society, and made available in material forms with which people both individually and collectively engage.It will suffice to highlight some key features of this dynamic here.
To begin with, mediation is a multimedia phenomenon in the sense that many different media technologies, and not just language, are brought into play in the representation of events and the elaboration of their meaning.Each medium has its own mnemonic affordances, as studies of photography, film, literature, performance, and the Internet have shown (Zelizer 1998;Erll 2008;Taylor 2003;Van Dijck 2007).Crucially, however, collective memory does not emerge from any single medium, but takes root in society by virtue of being reproduced across multiple media platforms.It forms a "plurimedial" constellation (Erll 2011, 138) that extends from verbal accounts to commemorative performances, to the materiality of monuments, and back again."Remediation" is thus a structural feature of mediation, a memory retaining its valence only as long as attention to it is renewed through new versions (Erll and Rigney 2009).The fact that an event is iteratively recalled in new cultural carriers can be taken as an observable measure of its social meaning (Rigney 2012).
Secondly, mediation is multilayered in the sense that structuring information entails more than the application of a single "narrative template" (Wertsch 2002).Multiple cultural forms help to make sense of specifics, "form" being understood here as the "ordering, patterning, or shaping" of information by calling on recognizable models of meaning making (Levine 2015, 3).Inherited "models" (Rigney 2005) or "schemata" (Erll 2011) make experience legible to a new public and hence "claim" it for memory.A systematic mapping of the cultural repertoire of mnemonic forms, which would bring together insights from narratology and cultural history, is still a desideratum.
Pending such an overview, it will suffice here to flag the fact that information modelling can be identified at multiple levels of any given representation.At the most generic and abstract level, all narratives revolve around a clash between what a (collective or individual) subject wants and their power to achieve it-or, to use Hayden White's definition of "narrativity"-between desire and the law (1987,(23)(24)(25).If an underlying conflict constitutes the "deepest" level of structuration, other models affect the organization of the plot (for example, as romance, tragedy, comedy, and so on) or the roles assigned to actors (as villains, heroes, victims, and so on).Moving closer to specifics, one can also identify the use of particular genres (models for combining certain themes with certain forms) that are continuously evolving by virtue of imitation and adaptation.While new genres emerge thanks to new media technologies (YouTube memorials, video testimony, for example), others fall into disuse while remaining in the cultural archive for possible future reactivation (the elegy and the epic, for example).Finally, particular stylistic models are activated at the "surface" level of representation in, for example, sentence structure or visual montage.In this multileveled way, unique actors and actions in particular representations are rendered legible by reference both to deep-seated traditions of storytelling and to more ephemeral aesthetic styles and genres.The applicability of old models roots new events in a deep memory of storytelling (Lotman 2019, 137), endowing them with a "resonance" (Armstrong and Crage 2006) that enhances their memorability and hence the likelihood that they will inspire new versions of themselves.
Thirdly, the mediation of memory is multi-scalar (De Cesari and Rigney 2014).Broadly held views about society's past take root only thanks to repeated instantiations of a collective narrative across different media, as well as their uptake by members of the public.This means that discourse analysis at an aggregate level to identify widely held beliefs needs to be complemented by the close study of the production of singular cultural carriers and how they mobilize individuals.Particular cultural carriers have the potential not only to inform others but also to mobilize affect in the intimacy of watching, listening, reading (Hirsch in Altınay et al. 2019, 14-5;Rigney 2016b).Meaning making and affect work together, especially in the case of "experiential narratives" (Erll 2008) that give information in an attention-holding form and offer a vicarious experience of the pain and joys of others.Aby Warburg's idea of Pathosformeln (Warburg [1924-29]  Finally, mediation is structurally selective.In the very process of producing meaning and affect in the ways described above, it works as a filter on what can be remembered.It tends to crystallize in particular "sites of memory" (Nora [1984(Nora [ -92] 1997) that become the focus of intense reproduction and semantic investment at the inevitable cost of sidelining other events (this is the collective equivalent of "retrieval-induced forgetting"; Hirst et al. 2018, 443).Such selectivity can be largely explained by the fact that convergence on a limited number of topics is a condition for their becoming communal points of reference (Rigney 2005).But what makes one topic more memorable than another?Memorability is constituted, on the one hand, by what people consider relevant (Irwin-Zarecka 1994) or politic to remember; on the other hand, it is dependent on the resonance between events and the inherited models we have for giving expression to experience.The principle of "differential memorability" (Rigney 2016a) means that events or situations that are easier to capture using the available cultural repertoire will "speak more" to the public, and hence be more easily picked up and remediated, where others are overlooked.
Using the idea of differential memorability, I argue in what follows that violence against protesters has played a prominent role in the history of nonviolence because it is more memorable than peaceful action.

<A>DIFFERENTIAL MEMORABILITY: (NON)VIOLENCE
When big crowds, especially unprecedentedly large crowds, demonstrate to claim their rights, they make the headlines.Cases in point are the enormous February 2003 global protests against the Iraq War, which gave rise to at least one documentary (We Are Many 2015); and the Million Man March on October 10, 2015, which gave rise to multiple variations on the "million" theme and at least one film (Get On the Bus 1996).More work is needed to identify the conditions under which peaceful demonstrations become memorable and how the positive celebration of that memory inspires later movements.The immediate focus here, however, is on peaceful protests that end in violence.The hypothesis is that the suppression of protest is not just newsworthy but also highly memorable in the long term, arguably more memorable than demonstrations that ended peacefully, because the combination of violence and nonviolence brings together political relevance and cultural resonance in a powerful way.For this reason, the memory of nonviolent movements has often ironically crystallized around violent reactions to them.The fact that the names "Peterloo" (Thompson [1963] 2013, 734-68), "Sharpeville" (Lodge 2011, 280-349) and "Tiananmen" (Lim 2014; Cheng and Yuen 2019) are internationally recognizable, not just as actual locations but also, and above all, as sites where demonstrators have been killed, is testimony to the particular memorability of outbursts of concentrated violence against citizens.How to explain this?
As the case of "Bloody Sunday" shows (Rigney 2016a), civic massacres are highly memorable, evidenced in the number of remediations they have occasioned and in the multiple crossreferences between them.The recurrent labelling of events as "Bloody Sunday"-starting with the killing of demonstrators in Trafalgar Square in London in 1887 and continuing down through Selma (1965) to Bloody Sunday in Derry, Northern Ireland, in 1972-reflects an awareness of the resonance between them on the part of the stakeholders in their memory.Each Bloody Sunday has generated multiple representations in different media over a longer period of time, while each new Bloody Sunday-like event has resonated both locally and internationally with earlier moments in which the democratic rights of protesters were violently denied.The name "Bloody Sunday" itself helps capture the idea of a dramatic reversal of fortunes as people exercised their right to demonstrate on their free day.The moment when a scene of nonviolence is transformed into bloodshed provides a highly dramatic "figure of memory" (Assmann [1992(Assmann [ ] 1997, 37-38), 37-38).It encapsulates in a nutshell the essence of "narrativity" as a clash between desire (a collective will) and the law (the power of the state).
The "slow violence" (Nixon 2011) of structural poverty or of an oppressive regime is difficult to capture in terms of a story with a beginning, middle, and end.In contrast, as I have argued in detail elsewhere, the transformation of a peaceful demonstration into a bloodbath is not only highly dramatic, but also melodramatic (Rigney 2016a).Described by Peter Brooks (1995) as the aesthetic form par excellence of post-revolutionary modernity, melodrama mobilizes emotions in presenting a dramatic opposition between good and evil in a sensational and visually powerful way.The figure of a civic massacre resonates with the aesthetics of melodrama by capturing in a condensed form the opposition between right and might, between the rights of citizens to make their voices heard and the repression of that right on the part of the state.It is repeatedly recalled in new contexts because it makes visible deep-seated anxieties about the limits of democracy and the power of the people.The memory of such sensational reversals of fortune has been carried by dramatic images of confrontation and dispersal, both photographs and drawings, which work as Pathosformeln to evoke collective movement.Citizens confronting the state: this is meaning at its most compelling.
More case studies are needed to show how the collective memory of civic massacres is culturally produced.But their importance as sites of memory can already be adduced from their recurrent use in popular culture for capturing structural inequalities in a vivid and concentrated way.
The massacre of civilians at Sétif (1945), for example, forms the starting point for Rachid Bouchareb's film Hors-la-loi (2010; Outside the Law) about the Algerian struggle for independence: the placing of the massacre at the outset of the movie provides a narrative motivation for the liberation struggle that follows.The indiscriminate killing of student protesters also features prominently in Alfonso Cuarón's film Roma (2018): that a key episode in the life of the characters should be played out to the backdrop of the Corpus Christi massacre on June 10, 1971, when 120 students were killed by "paramilitaries" in Mexico City, testifies to the endurance of the memory of that event (along with that of the Tlatelolco massacre of October 2, 1968;Sorenson 2002).Within the overall narrative of Roma, the violent suppression of the student demonstration mirrors the mistreatment of the female Mixtec protagonist by her boyfriend, who also happens to be one of the paramilitaries.Finally, the massacre of French-Algerian demonstrators in Paris on October 17, 1961, has likewise generated repeated mediations in literature, film, historiography, memorial plaques, and commemorative events (Rothberg 2009, 227-308).The many remediations of this massacre since the mid-1980s were in the first instance an activist reaction to the concerted attempts made by the police to suppress its memory (just as the multiple remediations of the Bloody Sunday in Derry in 1972 were a response to the state's denial of responsibility; Conway 2010).The recurrence of the bloodbath across different media reflects the slow and difficult overcoming of France's "colonial aphasia" (Stoler 2011); the fact that a bloodbath rather than other forms of injustice should become a central focus of contention, however, derives from its capacity as a highly dramatic event to capture structural inequalities in a single figure.
The potential memorability of the event is realized through its fit with the model of melodrama.

<A>MOBILIZING MARTYRDOM
Not all events lend themselves to capture in the form of a dramatic reversal on the lines of a Bloody Sunday.When police repression has been spread out over a longer period of contention, and the violence against demonstrators is episodic rather than concentrated, other models are brought into play.These include most notably the model of civic martyrdom, in which memory is captured through the figure of the individual demonstrator killed in the line of action.This again yields a melodramatic contrast between right and might that, in many cases, is further enhanced by the heartrending contrast between the youth of the victims and the brutal foreclosing of their lives.The salience of dead demonstrators in the memory of activism is evidenced, for example, in the case of Luigi Trastulli, a 21-year-old Italian steelworker killed in March 1949 by police on the occasion of a rally against the signing of the North Atlantic Treaty by the Italian government (Portelli 1990); Francesco Lorusso, a 25-year-old student member of a militant left-wing organization, who was killed by police during a demonstration in Bologna in 1977 (Hajek 2013); Carlo Giuliani, a 23-year-old shot dead during clashes with the police at the G8 summit in Genoa in 2001 (Caffarena and Stiaccini 2011;Memou 2013, 48-63); Benno Ohnesorg, a student killed by police during a demonstration in West Berlin in 1967 (Klimke 2010); Neda Agha-Soltan, a student shot dead on the occasion of a demonstration in Tehran in 2009 (Assmann and Assmann 2010); Pouya Bakhtiari, a 27-year-old electrical engineer who was shot in the head while protesting in Karaj, Iran, in December 2019 (Siamdoust 2020); and the list continues to grow.The memorability of these deaths is observable in the intensity and frequency with which their lives have been recalled in songs and poetry (see Siamdoust 2020) and, most importantly, in the reproduction of their images in the form of photographs or drawings derived from them.Individual victims give a face-often literally so in the form of portraits-to the memory of protest, making it concrete, condensed, portable, and as a result, easy to recall at other moments and other locations.
Remediated over and over again in the form of stencils, graffiti, and posters, the visual and material presence of the portraits, usually based on snapshots taken at happy moments, turns the memory of the dead into a living force.The youthfulness of their faces is an ever-renewed reminder of the unnaturalness of their deaths and the loss of their potential.The photograph of Neda Agha-Soltan's fatally wounded body, followed by photographs of her when she was alive and happy, went viral, and offers an exemplary case of the importance of images as carriers and catalysts of memory (Assmann and Assmann 2010).Displaying the faces of the victims of state violence has become a key strategy on the part of those agitating for the recovery of historical truth --exemplified by the weekly gatherings of the Mothers of the Plaza de Mayo in Buenos Aires (Taylor 2003, 161-88), as well as by the annual commemorations of Bloody Sunday (Conway 2010, 68-94).Carried onto the streets in the arms of demonstrators, the display of portraits expresses an embodied commitment to remembering the dead person on the part of the stakeholders in their memory.In the case of victims who were themselves demonstrators or who were killed in the context of a demonstration, the display of their portraits also serves to reinvigorate the struggle which occasioned their deaths.Thus, images of the young men known as the "Gezi Park martyrs" were regularly displayed as part of ongoing protests in Istanbul in 2013.

MP:
INSERT FIGURE 1 HERE Often accompanied by the collective cry "They are among us," the images of the dead of Gezi were displayed in order to assert the "transcendence of organic decay within the growing matrix of oppositional presence" (Gruber 2017, n.p.)."They are among us," echoed the slogan."Francesco is alive and fights along with us," which had been used in the commemoration of Lorusso (Hajek 2013, 90), suggesting a pattern in the evocation of the mobilizing presence of the dead.In the case of the Gezi martyrs, the regular remediation of photographs into stenciled drawings underscored the sense of a personal attachment to the dead while it amplified their status as collective icons by rendering the images more abstract than a photograph.Moreover, as figure 2 shows, deaths at different locations were connected visually to form a transnational pantheon, with the Greek student Alexis Grigoropoulos, killed in Athens in 2008, linked to Gezi "martyr" Berkin Elvan (as he had earlier been linked to Lorusso; Hajek 2013, 125): Although these figures did not die intentionally for a cause, the legibility and performative force of their images was arguably enhanced by their resonance with the narrative model of martyrdom.This model is deeply rooted in Christianity and Islam (meaning also that its use can blur the distinction between religion and politics in sometimes problematic ways; Bargu 2016).But the idea of civic martyrdom does not just have religious roots; it has also become part of a secular tradition of commemoration in left-wing activism going back at least to Peterloo in 1819 (Thompson [1963(Thompson [ ] 2013, 660-780) , 660-780) and the Chicago Martyrs of the 1880s (Streeby 2013).In this secular context, a martyr is a person whose memory is venerated because in the very act of becoming victim to state violence they bore witness to the ongoing validity of a particular cause or set of beliefs.The etymology of "martyr" in English and in other languages such as Turkish (şehit) and Arabic (shahid) brings together the double idea of sacrifice and witnessing.Remembering martyrs, often crystallized in a portrait, is tantamount to a renewed act of witnessing to the legitimacy of a particular cause and, by way of rebound, a way of mobilizing others to continue promoting it.
The cultural model of martyrdom can also be said to resonate in the remembrance of protesters who died not as a direct result of state violence during a mass demonstration but because of self-immolation or hunger-striking.The shocking death of Czech student Jan Palach, who set himself on fire in the center of Prague in January 1969, offers a case in point.Remembering his violent demise is a way of bearing witness to his refusal to give up the struggle against the Soviet domination of his country and hence reasserts over and over again his resistance to tyranny.The memorability of Palach's self-immolation was from the outset enhanced by its resonance with the national memory of Jan Huss, who had been burnt at the stake, also in Prague, for his heretical religious beliefs in 1415 (Lederer 1982;Sabatos 2009).More recently, the memory of the self-immolation of the 26-year-old street vendor Mohamed Bouazizi, widely interpreted as a protest against the willful destruction by the police of his fruit cart and hence his livelihood, became an important catalyst in the Tunisian revolution (Alexander 2011).
Finally, the memorability of violent deaths that are not directly related to protest but nevertheless bear witness to an injustice that needs to be challenged is also enhanced by the model of martyrdom.Cases include Khaled Said, the young man who was beaten to death by police in Alexandria, Egypt, in 2011 and whose memory is believed to have played an important role as a "visual injustice symbol" (Olesen 2013) in the run-up to the Egyptian revolution.They also include Emmett Till, the African-American teenager murdered in 1955, whose brutalized body fed into the civil rights movement through the many reproductions of his image (Tell 2019), and most recently, George Floyd, whose brutal killing by police in May 2020, became the central reference point in an international wave of Black Lives Matter protests in which his portrait figured prominently.
Mediated by images displayed across different media, as well as carried by demonstrators, the memory of state violence becomes itself a mobilizing force in the further conduct of nonviolent protest.Such deaths are best framed, to recall my earlier discussion, in terms of outrage rather than trauma: they are not merely violent occurrences in the past but reasons to act in the present.In her study of the contested memory of the 1970s in Italy, Andrea Hajek has shown how Italian activists, like those discussed by Gruber with reference to Gezi Park, deliberately mobilized the deaths of protesters as a rallying call for new action, as a reminder that people have good cause to resist the state (2013,(65)(66)(67).Commemorating the deaths of demonstrators provides a new occasion for mass demonstrations that, while protesting in the first instance against those particular deaths, also renew calls for structural reform.Mobilizing the memory of civic martyrs feeds into the idea that their deaths can and should be retrospectively "redeemed" as a milestone on the path to transformation.The idea of "redemption" is invoked here following Walter Benjamin's discussion of how shocking moments of defeat can be revisited in memory as the prequel to a delayed victory (1970,.The situational logic suggests that remembrance involves not merely mourning the irrevocability of their deaths but also giving them a prospective meaning by continuing the struggle for rights.In this way, the memory of the outrage does not necessarily inhibit action.Instead, it can rebound into renewed action based on anger but also a resilient hope in the possibility of change.In such cases, remembrance itself works to counter defeat and not just express it.

<A>CONCLUSION
In the preceding sections, I have shown how memory of nonviolent protest has tended to crystallize around the memory of its violent suppression and explained this by the fact that violent confrontation encapsulates structural inequalities in a compelling form that resonates with deep-seated cultural models.Using "outrage" as an analytical frame, I have shown how the memory of civic violence becomes caught up in a dynamic of action and reaction whereby memory works as a mobilizing force and becomes itself a renewed act of resistance.
More research is needed into the durability of outrage and its ability to sustain protest in the long term when hope starts to fade in the possibility of fundamental change.More work is also needed to explain the cost of this gravitation towards violent deaths in the "mnemohistory" (Assmann [1992] 1997) of nonviolent struggle, and to identify the "retrieval induced forgetting" it occasions on the part of activists' groups and the public at large.Finally, future work should investigate why certain forms of violence against civilians have little uptake in memory despite their potential memorability.Is it because of censorship or strategic forgetting on the part of stakeholders?Or because the sheer scale or unrelenting character of the violence has left the public indeed traumatized into ineloquence (to recall the terms of trauma theory in cultural studies)?Proposing outrage as an alternative frame thus inevitably feeds back into a consideration of instances where righteous indignation fails to find expression using the models outlined above.
In the last year, many hundreds of demonstrators were killed across the world, in Iraq, Chile, Iran, Sudan, to name just a few of the countries affected.The melodramatic and martyrologic models described above seem less conducive to capturing a violence that is neither short-lived nor limited to a small number of victims, begging the question if and how the recent killing of protesters on such a scale will ever be constituted as memorable.Where some instances of violence against protesters seem to be immediately captured, allowing their remediations to feed back, like the Gezi martyrs, into ongoing protests, finding an adequate form to express violence that is "slower" or unprecedented in scale seems to require more time and more creativity.It was only some 30 years after the killing of hundreds of students in Gwangju in 1980, for example, that South Korean writer Han Kang produced the harrowing and heartrending novel Human Acts ([2014] 2016) as part of a gradual coming to terms with this violence that included archival projects, as well as public apology.This time delay suggests that future research on mediations of outrage could be combined in fruitful ways with approaches based on trauma theory that are sensitive to the psychological and cultural limits of representation while also accounting for the power of people to rebound.

]Figure 1 :
Figure 1: Portraits of Berkin Elvan held up by protesters in Ankara (March 13, 2014).The 15-yearold died as a result of being hit by a tear-gas canister on his way to buy bread during antigovernment protests in Istanbul in 2013.Photograph Umit Bektas/Reuters.Reproduced under license ADB096026011NL.