The Politics of Spatial Testimony: The Role of Space in Witnessing Martyrdom and Shame During and After a Widely Televised and Collectively Perpetrated Arson Attack in Turkey

Introduction

On July 9, 1993, a group of activists in Turkey launched a campaign for the authorities to preserve the Madımak Hotel’s charred remains as “a witness” to the arson attack it had seen the week before and to declare it a “museum of shame” (Ekinci, 1993). Located in the central-eastern part of the city of Sivas, the hotel had been set ablaze by tens of assailants before an inactive law enforcement, thousands of onlookers, and live TV cameras. The arson had killed 2 hotel workers and 33 individuals who had been invited to Sivas as part of a culture festival organized by an association representing followers of Alevism, one of Turkey’s demographically minor faiths. A mass funeral held a few days after the attack had seen mourners commemorate each of the 33 victims as şehit, regardless of their confessional affiliation. Şehit is Turkish for martyr and is cognate of şehadet, which denotes the act of testifying as well as the status of martyrdom.

What are the political and analytical possibilities and limitations of mobilizing testimony and its cognates as spatial mechanisms for condemning mass violence? This question preoccupies a growing body of the literature today across the fields of architecture, anthropology, archaeology, and media studies, as sites of violence and/or their images feature prominently in the documentation, communication, and litigation of such human rights violations as the Sivas arson attack. Some have welcomed this prominence as an antidote to the post-Holocaust tendency to premise testimony’s authority on an ethics of compassion rather than epistemic verifiability (Schuppli, 2013, 2014; Weizman, 2010, 2014). Others have criticized it for running the risk of perpetuating deep-seated assumptions permeating the politics of violence’s representation, such as the synonymy of looking, seeing, and intervening (Herscher, 2014); the political legitimacy of intervention (Ewalt, 2011); the immediacy of mediation (Feldman, 2015); and knowledge production’s immunity from power structures (Hauser et al., 2018). This article not only draws on but also contributes to these criticisms by problematizing another such assumption: that the ways testimony and subjectivity shape each other are assimilable into discrete moments with unambiguous subject positions such as those of victims and perpetrators.

Problematizing this assimilability, I argue, is important for at least two reasons. First, it acknowledges that testimony entails not only witnessing but also bearing witness, where subjectivity is a much less temporally and spatially delineable phenomenon than that of subject position (Oliver, 2000, 2001, 2003; Rothberg, 2013, 2019). Second, it helps to better understand and unlock the sociopolitical potential of spatial testimony at a time when images of violence continue to proliferate, and the urgency of engaging with various violent histories intersectionally rather than in isolation from one another is ever more evident. I unpack this argument further through an ethnography of memory activism around the Sivas arson attack. A prevalent approach to remembering the arson attack has been to mobilize the site where it took place and televisually mediated images thereof. This mobilization has relied on the subject-positional extremes of guilt and innocence and has sought to rigidify and morally hierarchize these through the affective trope of shame, as evident in the on-site “museum of shame” being campaigned for. In-depth observations among, and conversations with, memory activists indicate the limits of this reliance as well as its possibilities. They show that, among the activists, televisual mediation has led not only to strict identification with victimhood but also to self-reflection about how it might have been prevented, and that sites associated with the victims’ martyrdom have prompted reflection on not just the arson attack but also other violent episodes in the history of Sivas and its environs, ranging from sixteenth-century uprisings against the Ottoman administration to the Armenian genocide of 1915–16.

Testimony, Space, Subjectivity

Space’s emergence as a focus in the scholarship on testimony in the context of political violence can be dated to the early 2000s when the post-Holocaust “era of testimony” (Felman, 1991) or “era of the witness” (Wieviorka, 2006) became subject to criticism for relying on individual witnesses in the production of historical truth. The criticism, more specifically, was that this reliance conflated the personal and psychological with the collective and sociopolitical (Douglass & Vogler, 2003); overshadowed numerous other resources through which to piece together historical truth (Sarlo, 2005, pp. 9–26); cast the witness as a sweeping category that overlooks significant subject-positional differences between various witnesses (Fassin, 2008); and treated survivors as self-evident embodiments of veracity, and thus obscured the politically charged processes of exclusion, selection, validation, and analysis that in fact govern truth production (Givoni, 2011). Architectural theorist Weizman (2010, p. 14) has suggested that the post-Holocaust paradigm deprived testimony of its epistemic function by employing it “primarily as an ‘ethical’ resource” whose “function was in being delivered in the first place” rather than in “revealing knowledge or authenticating claims of historical injustice.” Against the grain of this paradigm, Weizman (2010, p. 14) has heralded a decisively object-centric one termed “forensic architecture” where, thanks to recent technological and legal developments, artefacts serve as “object-witnesses” on a near par with their human counterparts. According to Weizman (2014, p. 29), whereas “the era of the witness” prioritized ethics over epistemics and thus engendered an “anti-universalist” notion of truth as an “inherently relative, contingent, multiple, or non-existent” category of knowledge, “forensic architecture” promises to reclaim “truth as a common project” by mobilizing the epistemic potential of artefacts towards registering and conveying violent histories in collectively palpable ways.

However, such mobilization has not been free of contention, either. Consider architectural historian Herscher’s (2011, pp. 130–41) discussion of how human rights advocacy’s increasing reliance on surveillance transforms space “from a target of violence to a witness to violence.” He criticizes this reliance for assuming truth to be readily locatable in violence’s spatial effects rather than in its markedly human causes, driving forces, aims, and intentions. In fact, the human continues to significantly shape the summoning of artefacts as witnesses, as evident in the work of experts both before and during court hearings. Before hearings, experts manipulate the materiality of objects so that they can serve as witnesses (Schuppli, 2014), and in court, it is their illocutionary acts that help artefacts “speak of” violence (Schuppli, 2013, p. 165). The expertise at work here is also worthy of scrutiny for its entanglement in knowledge-based power structures. Contributors to a recent forum on “archaeology as bearing witness” have problematized archaeologists’ claiming center stage in the condemnation of violence for exacerbating the disempowerment of victims, survivors, and their descendants (Hauser et al., 2018, pp. 537–540 and 542–544). Herscher (2014, p. 496) hints at a similar problem; he argues that “looking,” or visually mediated encounter with spatial testimony, ought to be approached neither as a neutral means of verification nor as an inherently emancipatory act, but with an awareness of the “distribution of power and knowledge” in which it is entangled.

The problem here concerns not only looking as such but also the registers through which violence is looked at. Rather than just problematize the assumption that seeing violence means acting upon it, Ewalt’s (2011) work on contemporary humanitarianism’s reliance on real-time satellite mapping questions why an intervention from outside is necessary or legitimate at all. To associate activism with such intervention, for him, is to reproduce the colonial gaze as the only possible register through which to approach contemporary conflict. Similarly, Feldman (2015, p. 82) criticizes recent forensic approaches to political violence for reducing victims’ and survivors’ “bodies, topography, and demography” to “raw material” and thus furthering the violent dehistoricization to which they were subjected in the first place. Invoking filmmaker Godard’s aphorism that, gradually throughout the second half of the twentieth century, “[t]he Jews became the stuff of fiction; the Palestinians of documentary,” Feldman (2015, p. 137) problematizes “the shuffling and assignment” of peoples into separate epistemic “regimes” of “sensibility and insensibility.” At stake in mobilizing space and its images as part of processes through which to testify to violent histories is therefore not just the content of testimony but also the extent to which it might help challenge the dehistoricization resulting from the violence being testified to.

How, then, might the spatiality of spatial testimony be employed, if at all, to contest this dehistoricization, and what sorts of historicization might emerge in its stead? These questions guide the empirical analysis offered in the rest of this article. The empirical material derives from my ethnography of memory activism around the Sivas arson attack over the two years that followed the government-sponsored and partly commemorative architectural transformation of the Madımak Hotel in early 2011. To approach commemoration as an ethnographic object helps problematize the scholarly tendency to consider remembrance as indexing the aftermath of the event being remembered and therefore to overlook its participation in the processes through which events continue to unfold. This tendency is especially problematic when dealing with violent events; focusing narrowly on how violence is represented in its supposed aftermath assumes a shared understanding of its being acknowledged as violent or as having come to an end (Çaylı, 2018, 2019). Avoiding this assumption helps examine the extent to which commemoration challenges or reproduces “present-day” power relations structured by the “past” it commemorates (Davidson, 2016; Lavrence, 2005) and helps warn against its totalizing potential (Bishop, 2016; Curtis, 2004).

Televisually Mediated Spatial Testimony during the Sivas Arson Attack

As the event itself is at stake in what purports to be its post-event representations, my empirical analysis of spatial discourses and practices of witnessing and martyrdom requires that I revisit the context of the arson attack. The culture festival whose participants were killed in the blaze was named after Pir Sultan Abdal, a sixteenth-century minstrel venerated in Alevism as a martyr who was hanged by the governor of Sivas for refusing to submit to his authority. Pir Sultan is a prominent figure in Alevi “martyrology,” a sacred lineage based not on birth, but on tragic death during nonviolent resistance against tyranny (Hess, 2007). This lineage originates in a series of mid-seventh-century events known to have engendered the schism between Sunnism and the rest of Islam. These events include the assassination of Ali—prophet Muhammad’s son-in-law and the fourth caliph (651–61), whose name many Alevis consider the origin of the term Alevi, that is, “pertaining to Ali” (Dressler, 2013, p. 4). They also include the Battle of Karbala (680) where Ali’s son and his six dozen supporters were killed by the much larger forces of the second Umayyad caliph, whose dynasty’s claim to the caliphate they repudiated for being based on tyranny rather than on rightful heirdom (Zırh, 2016).

Pir Sultan’s name became embroiled in a state-funded “Minstrel’s Monument” placed just outside the main festival venue to mark the occasion. The local press flagged certain features of the monument to speculate that it honored Pir Sultan rather than all “minstrels” (Bu Anıt ‘Ozanlar Anıtı’, 1993), labelling the minstrel “a foremost rebel in Anatolian history” (Bozgeyik, 1993). Portrayals of the festival as rebellious embroiled the Minstrels’ Monument in the arson attack in physical ways, too. The monument was defaced by assailants as they surrounded the main festival venue around noon on July 2. Having faced resistance from those inside the venue, the assailants then went for various other buildings hosting the festival or its sponsors (e.g., the Governorship), shuttling back and forth between these sites and reaching their thousands in the process. They ultimately surrounded the hotel where the festival’s guests had sought refuge. When the authorities demanded that the assailants peacefully disperse, the latter reportedly presented three prerequisites, one of which was the removal of the Minstrels’ Monument (Tüleylioğlu, 2010, pp. 488–492). The authorities complied and brought the toppled monument to the crowd outside the hotel. This proved to only encourage the assailants; they seized the statue and burnt it on the spot, moments before setting fire to the hotel itself (Tüleylioğlu, 2010, pp. 186–87).

The media’s involvement in this chain of events was not just limited to local newspapers. It also included a conservative-leaning nationwide network launched just six months ago, thanks to the privatization of TV broadcasting. The network’s continuous reporting from Sivas throughout the day of the arson attack ostensibly amounted to a live broadcast, albeit a selective and monitored one. Shortly before the attack, “the unrest in Sivas” was reported on as having been “brought under control,” to the utter dismay of those trapped inside the hotel who watched some of this reporting on their TVs while also witnessing the assailants outside grow both in number and aggression (Özbakır, 2010). Once the building was set ablaze, near-live footage of it surrounded by thousands of onlookers arrived on screens. The resulting sociopolitical impact was such that many 1980s children born to Alevi families discovered their connection to the faith while watching this footage live; hearing their parents cry “they’re burning us,” these children would ask who “us” is, to then be told, “Alevis” (Özer, 2015). The collective televisual witnessing soon informed the 33 victims’ incorporation into Alevi martyrology, regardless of their confessional affiliations (Yildiz & Verkuyten, 2011). In speaking of the 33 as martyrs, many Alevi activists and intellectuals have emphasized each martyr’s having testified to values central to the faith “before the eyes of the whole world” (Demir, 2015; Öz, 2013).

Scholars have tended to approach the Sivas victims’ being remembered as martyrs in light of the opposition between Alevi and Orthodox Islamic—in Turkey’s case, Sunni—notions of martyrdom, associating it squarely with the former (Hess, 2007; Yildiz & Verkuyten, 2011; Zırh, 2016). Some (Yildiz & Verkuyten, 2011) have been readier than others (Hess, 2007; Zırh, 2016) to acknowledge that the arson attack did not just dovetail with a pre-existing notion of Alevi martyrdom but also shaped it significantly due mainly to its inclusion of non-Alevis. Still, their explanation for why memory activism around the arson attack and Alevi martyrology have so decisively influenced each other has been virtually identical. The explanation is ontologically oriented; it suggests that certain qualities characterized both the historical Alevi martyrs’ and the Sivas martyrs’ ways of inhabiting the world at the time of death: “passive” subjection instead of “active” aggression (Hess, 2007), unawareness of the likelihood of death as opposed to clear awareness of it (Yildiz & Verkuyten, 2011), and sociopolitically motivated resistance against majoritarian tyranny rather than ethnicist combat against the religious other (Zırh, 2016).

The genealogical validity of this explanation notwithstanding, another one that considers the Sivas arson attack as generative rather than just derivative and does so for spatially charged reasons surfaced during my ethnography among experienced memory activists. What justifies an amalgamation of the thoughts of these otherwise distinct individuals is not limited to their sustained engagement in commemorating the Sivas victims and includes two other biographical similarities. First, all have long identified personally with Alevi Sivas martyrs due not only to shared values (religious or otherwise) but also to generational affiliation as most of these victims, like the activists in question, were young adults at the time of the arson attack. Second, they all hail originally from Sivas province but lived in Ankara since the late 1970s, to which they or their families migrated, due partly if not entirely to the period’s wave of violence that targeted central and eastern Turkey’s Alevis, including a pogrom in Sivas’ reputed “Alevi neighborhood” Alibaba (Jongerden, 2003, p. 83). It is worth quoting at length an activist in his late forties (personal communication, July 6, 2011) as emblematic of how this generation of Alevi activists recall finding out about the arson attack:

Although we were not in Sivas at the time, I can say we experienced the massacre minute by minute. Word was out already in the afternoon that reactionaries had begun to harass our dear ones. Soon, news emerged about things starting to get physical. But, then, the situation seemed to be calming down or, at least, was not worsening according to the news. Towards the evening, though, we were confronted with images of the hotel being besieged, stoned and set on fire. Hundreds, maybe thousands, holding a handful of people—the country’s best and brightest—captive and then setting them on fire. . .this is not something you can easily get over. Of course, first, you place blame for what the martyrs went through where blame is due: the state, the law enforcement, the arsonists. But then you begin to introspect. Could we have done something? Why did we just sit and watch? In fact, some of us were in touch with Alibaba, which is just a 20-minute walk from the hotel. While the news kept fluctuating, Alibaba’s residents were on tenterhooks—naturally so, given the neighbourhood’s recent history. Some even tried to convene and march to the city centre. But they were stopped in their tracks by the gendarmerie who went up to the neighbourhood to quell the reactions. Meanwhile, the assailants were given a free hand. Still, you ask yourself, could more have been done to prevent what happened?

Reiterating that the material and the visual co-constitute the experience of urban space (Wells, 2007, pp. 142–43), these remarks render televisual mediation a factor that needs consideration in spatially focused analyses of witnessing and martyrdom in the Sivas case. Consider the spatially charged way in which footage of the attack was broadcast. As the activist indicates, reasons that make spatiality central to this footage are twofold. The first involves extensive use of long shots, through which the vastness of the crowds surrounding the hotel was foregrounded. The second reason concerns how differentially the footage was experienced across the viewership, where the differentiation derived from spatial knowledge of central Sivas and its history.

What, then, might be the relationship between this twofold spatiality and the need for introspection regarding the events? Here it is useful to think with Ellis, the prominent theorist of media witnessing. Writing in the immediate aftermath of the rise to prominence of live-from-the-scene news, he suggested thus:

The feeling of witness that comes with the audio-visual media is one of separation and powerlessness: the events unfold, like it or not.. . . At once distanced and involving, it implies a necessary relationship with what is seen. The relationship is one of complicity; [that] you know about an event. . .implies a degree of consent to it. (Ellis, 2000, p. 11)

Ellis’s repeated use of the verb “imply” to convey a certain sense of complicity recalls Rothberg’s term “the implicated subject.” The term refers to “various modes of relation” involved in many contemporary contexts of violence, which are unassimilable into the categories of victim, perpetrator, and bystander, and which therefore “move us away from overt questions of guilt and innocence and leave us in a more complex and uncertain moral and ethical terrain” (Rothberg, 2013, p. 40). Indeed, the truth testimony produces always necessarily involves not just an epistemic but also an ethical component. The first establishes guilt or innocence based on “witnessing” an event at a specific time and place and communicating this experience to others, whereas the second concerns “bearing witness. . .to that which cannot be seen” (Oliver, 2000, p. 31). While witnessing orientates one’s “subject position” in space and time, bearing witness constitutes “subjectivity”— or one’s sense of agency and “response-ability”—by continually encouraging one’s “encounter with otherness” (Oliver, 2003, p. 137)—or one’s quest to address others and receive their response (Oliver, 2001, pp. 85–106). This quest intensifies for implicated subjects who are left continually oscillating between the two extremes of guilt and innocence.

The arson attack’s televisual mediation has produced just such implicated subjects through both form and content. Formally, the extensive use of long shots that foregrounded the crowd outside the hotel sought to portray the arson attack as based on popular dissent—a portrayal bolstered periodically throughout the footage by supportive commentary from anonymous individuals near the camera. The reporting’s near-live feel—still a novelty for Turkey at the time insofar as breaking news was concerned—engendered the popular assumption that millions knew what was happening and yet failed to intervene. In fact, its content was selective. As recalled by the earlier-quoted activist, some, like those in Alibaba, did indeed attempt to intervene but were suppressed both physically—by the very law enforcement that was missing from the hotel’s forecourt—and epistemically—by way of exclusion from televisual reporting. Therefore, spatial knowledge of both the suppression of counter-assailants and Sivas’ previous violent histories such as the 1978 episode served to further exacerbate the helplessness felt by those identifying with the 33 victims. This is how televisual mediation amplified violence’s sociopolitical impact: by turning the very individuals who remember the victims as martyrs into “implicated subjects” haunted by continual introspection regarding what more might have been done.

There is cause to argue that the affective focus of the memory activist campaign for an on-site “museum of shame” is informed by the “implicated” subjectivity in question rather than by the dichotomous subject-positionality of innocence versus guilt on which the activism has appeared to rely in its public discourse. According to Keenan (2004), global and live transmission of human rights violations was central to shame’s rise to prominence in late-twentieth-century humanitarianism whose approach to putting sociopolitical pressure on violators was marked by a shift from reason to affect. But in mobilizing shame as such a televisually driven mechanism, the humanitarianism of the period overlooked that the very human rights violators they sought to bring to account were often also relying on the same televisibility, as they committed violations consciously for the cameras (Keenan, 2004, p. 446). If Keenan casts doubt on the sort of outward mobilization of shame that memory activism around the arson attack seems to pursue in its public discourse, Agamben helps conceive it as a catalyst for self-reflection. “Shame,” Agamben argues in his analysis of Levi’s eponymous essay, is not easily frameable within the guilt/innocence binary, therefore resonating with Rothberg’s “implicated subject;” the shame of having witnessed but survived violence results in “desubjectification” (Agamben, 1999, pp. 105–06). Passivity and sensibility co-exist in shame; it derives from being aware of and indeed being moved by one’s own passivity (Agamben, 1999, pp. 109–11).

This section has suggested that martyrdom’s centrality to memory activism around the arson attack derives from the spatially charged way it was experienced as a media event rather than just from its ready assimilability into Alevi martyrology. Adopting Rothberg’s term, I have argued that this experience has turned certain activists identifying personally with the martyrs into “implicated subjects” who are left in much more ambiguous territory than that permitted by the guilt/innocence binary. What might “implication” as such mean for in-person encounters with sites that are linked to the arson attack and are spoken of through the concepts of martyrdom, witnessing, and shame?

In-person Spatial Testimony in 2010s’ Sivas

My fieldwork in Sivas commenced in 2011, when memory activism of the sort that the Sivas case has pioneered in Turkey had just elicited an official response. In 2009, the government had launched an initiative dubbed Demokratik Açılım (Democratic Opening), convening with nongovernmental actors and representatives of historically underrepresented groups such as Kurds and Alevis. Senior ministers had called this an opportunity for “the state to revise its memory” (Ministry of State for Religious Affairs, 2010, p. 6 and 17). Madımak Hotel’s state-sponsored transformation into a “museum of shame” had been among the five demands raised in the meetings with Alevis. This had been followed in summer 2010 by the first-ever ministerial-level commemorative visit to the site. Come the year’s end, the authorities had expropriated the hotel, then refurbishing it in the spring, albeit with no publicity. A press preview in June 2011 had revealed the building’s new program: a commemorative-cum-educational institution open to the public on weekdays. My fieldwork began that summer and took place at intervals over the following two years. Part of it concerned the anniversary commemorations held in central Sivas, which I attended twice with Alevi activists, travelling overnight by bus from the capital Ankara.

The hotel’s state-sponsored transformation turned half of its ground floor into a “Memory Corner,” whose centerpiece is a 10-by-15-feet stainless-steel structure that displays a name list. The list comprises 37 names, therefore including not only the 33 culture festival participants and the two hotel workers who died during the arson, but also the two individuals killed in the melee outside the hotel as the crowd was belatedly dispersed. One of these individuals tops the list, as his name starts with A and the list is in first-name-alphabetical order. A statement displayed in the Memory Corner speaks of the names listed as “37 of our people” killed “in the painful incident that took place on 2 July 1993,” while a second one invokes “national unity and unison” as remedy. Those remembering the 33 festival participants as martyrs have repudiated the Memory Corner’s sweeping name list and have performed this repudiation by refusing to enter the building (Öztürk, 2011). Meanwhile, state authorities have defended the list as a “human-centric” refusal “to discriminate between the dead” (Yalçınkaya & Ceylan, 2011).

The Memory Corner’s “indiscriminate” victim list is symptomatic of the mainstream ways the arson attack was historicized in the early 2010s. Examples include a ruling party MP’s statement in parliament that “the majority of those taken to court” for the arson attack “are as innocent as those set ablaze” (130’uncu Birleşim, 2012); one of Turkey’s most popular political TV shows where a survivor and a former suspect, whose acquittal many of those upholding the victims’ legacy have continued to contest, were presented together as “two witnesses”; and another such show where a commentator opined that “being set ablaze” and “setting ablaze” each deserves recognition as a legitimate form of “victimhood” (Çaylı, 2012). The broader historical narrative promoted here is that the arson attack was but one episode in a string of conspiracies whose known perpetrators were in effect naïve pawns used by clandestine political-criminal networks seeking, as also implied in the Memory Corner, to damage Turkey’s national unity and unison (Kenanoğlu, 2013). The late 2000s and early 2010s were, therefore, a context in which state-endorsed historical revisionism abused the notion of witness and, even more controversially than the latter, that of victim, obfuscating subject-positional differences involved in violent histories.

The theme of martyrdom, as explained in the previous section, had long enabled those upholding the 33 festival guests’ legacy to underscore the very subject-positional differences in question. Faced with attempts to obfuscate these differences, the theme began to mark ever more visibly the anniversary commemorations held in Sivas. This annual event, which since the mid-2000s has seen increasingly larger crowds flock into the city from across Turkey and beyond, drew thousands in the early 2010s. Throughout this period, the event typically proceeded along a two-kilometer route stretching from the aforementioned “Alevi neighborhood” of Alibaba down to the site of the arson attack where flowers were laid, thereby retracing the trajectory that law enforcement had denied Alibaba’s residents in the lead-up to the blaze. It was invariably led by “martyrs’ families”—relatives of the 33 victims—alongside a bus equipped with powerful amplifiers broadcasting songs and anthems associated with the “martyrs”—the historical Alevi ones as well as those killed in Sivas. A speaker on the bus periodically interrupted the music to make brief remarks, many of which concerned the Memory Corner: “today we march against those who listed murderers alongside our martyrs!” The speaker also cued the crowd on what slogans to shout, most of which featured the arson attack as directly linked to numerous other historical atrocities that are now the subject of Alevi martyrology.

The martyrdom theme was highlighted in 2011 by a group of around 30 young marchers dressed in uniform-like attire comprising red vests and balaclavas, who held up cardboard cut-outs of zülfikar, an icon known as “the sword of Ali,” the inaugural Alevi martyr. The same year, commemoration participants coming from Istanbul had prepared a sign that read “Museum of Shame.” Their plan was to hang it atop the entrance of the site of the blaze to materialize their disapproval of the outcome of its recent transformation and especially the commemorative name list. The police turned out to have gathered intelligence about this intervention and used it as pretext to barricade the building’s forecourt. Only the martyrs’ families were allowed past the barricade for the flower-laying ceremony. When the “Museum of Shame” sign was passed forward from hand to hand to ultimately verge on going past the barricade, the police reacted by teargassing the crowd. The ensuing melee led Alevi leading figures to dedicate their closing speeches to emphasizing that what had just taken place was “a continuation of previous massacres and the centuries-long tyranny of hegemonic powers that slew our martyrs in history.” Pir Sultan Abdal—the sixteenth-century minstrel after whom the 1993 festival targeted by the arsonists was named—featured prominently in these speeches as one such martyr. One activist in her twenties (personal communication, July 2, 2011) drew the link between Pir Sultan, the arson attack victims, and the commemoration participants not just ideologically but also spatially. She suggested that “he was murdered just around the corner from where they teargassed us today and where our martyrs were slain yesterday,” pointing to what is now known in Sivas as Dikilitaş Square.

The materially charged impact these martyrological connections have made on memory activism around the arson attack is evidenced by a set of statues commemorating certain Alevi victims in their ancestral villages in Sivas province’s Emlek region. Known to locals as “Martyrs’ Monuments,” these statues are modelled on a 1978 monument to Pir Sultan Abdal located in Banaz, renowned as the minstrel’s native village in rural Sivas. This is where the Pir Sultan Abdal Festival was held thrice before moving to central Sivas in 1993; the monument honoring the minstrel was built as part of preparations for the festival’s inaugural episode (Demir, 2008). The late 1970s was a period when rural-to-urban migration was completing its first generational cycle in Turkey (Zürcher, 2005, pp. 226–72). This wave of migration not only included Alevis as a substantial demographic but also profoundly shaped their identity through an unprecedented encounter with and/or categorization as the socio-religious other (Shankland, 2003). Alevis encountered an urban Turkey that had little or no room—both socially and physically—for their rituals and practices. Such profound encounters led newly urbanized Alevis to re-establish links with their ancestral villages through initiatives like summer festivals, which offered an opportunity to not only reunite with fellow believers but also conduct religious service denied in cities (Langer et al., 2011, p. 112). In the mid-1990s, such initiatives re-emerged in response to troubles urban Alevis faced, which followed from, but were much more violent than, those in the third quarter of the century. One tangible result of this re-emergence comprised the Martyrs’ Monuments, alongside the Pir Sultan Abdal monument.

Part of my fieldwork took place in Banaz and villages across the Emlek region, which I visited together with activists of the generation discussed in the previous section. I observed that similarities between the Martyrs’ Monuments and the Banaz monument to Pir Sultan Abdal were not just formal. They also included the various meanings the monuments were ascribed by villagers. Some of these meanings derived from collective labor and knowledge in the village, which enabled the monuments’ construction in conditions of climatic and economic hardship and which the villagers considered reminiscent of the martyrs’ perseverance in the face of persecution. Other meanings concerned surrounding landscapes. Many villagers in Banaz spoke of the Pir Sultan monument as spatially referencing land forms nearby that they considered witnesses to crucial events in the minstrel’s life or in Alevi cosmology. Such was the way in which the villagers across the Emlek region made sense of the location of each Martyrs’ Monument.

Some of these conversations around the monuments and the landscapes nearby transitioned into histories other than those of Alevism. An activist in his fifties hailing from the Emlek region referred (personal communication, June 30, 2012) to a tributary bordering one village as sevkiyat (deportation) brook and to another one nearby variably as kanlı (bloody) or karanlı (corrupted from karanlık; dark) brook. Sevkiyat brook, he explained, is where the region’s Armenians were gathered in 1915 and “deported” eastward by orders of the then Ottoman government. Around this time, kanlı brook witnessed the massacre of Armenians from another part of the region; the stream ran blood red for months. Having to elaborate on these vernacular names prompted the activist to comment on his ancestors’ role in them: “My grandpa was tasked with taking them further east; that’s all he did—he didn’t kill anyone.” “The killers always came from Sunni villages,” said another male activist of the same generation (personal communication, June 30, 2012) who accompanied us. Still another admitted, “Alevis did what they could—many Armenians sought refuge in our villages—but, had we been able to do more back then, what we subsequently experienced might also have been different.” Such conversations would then return to the Sivas arson attack, implying its being among the subsequent experiences that could have had a different course had 1915 unfolded differently.

Further instances when sites associated with the Sivas victims prompted segues from the arson attack to various other not necessarily Alevism-related violent episodes like the Armenian genocide took place during the 2012 episode of the commemoration in Sivas. As an activist of the above-mentioned generation and I chatted on the way to Sivas about the city’s historically significant buildings, he lamented that, while there were many in the past, most had by now become extinct. “Of course, one can always blame insatiable contractors,” explained the activist, “but, if you ask me, the real damage was done long time ago, in 1915, when the Armenians perished; the artisans and master masons were all Armenians, and in their wake, no one was left to build anything of quality” (personal communication, July 1, 2012). I asked him where his knowledge of and interest in the topic came from. The activist responded by referring to his ancestral village in the southern region of Sivas province and the stories he had heard from its elders. “Thousands of Armenians used to live in the region; they were exiled, massacred in 1915,” he explained, referring to the Armenian genocide:

Our elders say Alevis were not involved; all killers were Sunnis. Some even protected the Armenians by hiding them. Be that as it may, could more have been done to prevent what happened? After all, this was not just a catastrophe experienced by a particular people, but a curse upon an entire geography. Imagine, if central Sivas’ population were a third Armenian, as it had been before 1915, could a massacre like Madımak take place?

The activist then went on to relay further village-borne stories relating to the Armenian genocide. “The region around our village is famous for its fruit trees—especially apricots—but it is said that, after 1915, even they became useless,” he recalled. “Of course; who would eat the fruits of a geography awash with blood even if the trees yielded them,” he concluded.

These remarks continued to reverberate the following day, when the commemoration in Sivas saw police barricades move further up from where they had been the year before by about half-a-kilometer. An hour-long sit-in ensued throughout which tension levels gradually rose, at times bordering on physical confrontation between the police and the activists. I noticed the senior activist shuttle jumpily between the barricades and the crowd performing the sit-in a few dozen meters in front. Ultimately, the police conceded and moved the barricades back to where they had been in 2011. The procession resumed. On the way back, the activist recounted to me his experience during the sit-in:

This is so unfair, I thought to myself. Had there been barricades in 1993 right here in these streets, the massacre would have been prevented; yet, there were none. With such thoughts going through my mind, I was about to snap and lunge at the police. Then suddenly a kid appeared from between buildings lining the street and walked toward me. He said, “uncle, calm down. My family has a wonderful apricot grove. It’s very calming; let me take you there.” I remember strolling with him for a while. Then someone else called out to me; I looked their way. When I turned back, the kid was gone. (Personal communication, July 2, 2012)

It was not possible or even analytically relevant to ascertain whether the activist’s experience was a daydream or an actual encounter. What rendered it empirically significant, however, was the conversation he and I had had on the way to the commemoration, in which apricot trees had prompted him to reflect on the magnitude of the Armenian genocide’s social and spatial impact. Now, in the activist’s experience during the impasse in front of the barricades, the same feature of the landscape resurfaced as indexing a space in which to find wisdom vis-à-vis violent histories’ likelihood to recur precisely where they have previously occurred.

The activist’s remarks therefore recall my observations across the Emlek region and do so for at least two reasons. First, whether in the case of the site of the arson attack and the barricades mounted to render it inaccessible or in that of the Martyrs’ Monuments and the landscapes nearby, the extent to which sites associated with the arson attack serve to remember the 33 victims concerns much more than just their formal features. Rather, it derives from the social processes characterizing the construction and/or use of these environments—put differently, from the extent to which these processes are seen to resemble or contradict the values known to have marked the events surrounding the victims’ martyrdom. Second, both cases demonstrate that, once these sites are rendered a medium through which to evince such cross-historical resemblances and contradictions, the histories across which they operate become capable of engulfing numerous others than just the arson attack and/or events significant to Alevi cosmology and/or martyrology—other histories such as the Armenian genocide.

Conclusion

This article’s aims have been twofold: first, to analyze memory activism around the Sivas arson attack, especially in terms of its references to space as a “witness” to violence, its remembrance of victims as “martyrs,” and its campaign for an on-site “museum of shame”; and, second, to contribute to the growing literature on the politics of testifying to violence spatially. The analysis has shown that the spatial import of the notions of “witness,” “martyrdom” and “shame” in question has been informed by televisual mediation, which amplified the presence of assailants outside the hotel and erased that of counter-assailants nearby in the “Alevi neighborhood” of Alibaba. But awareness of this amplification and erasure required further spatially charged knowledge—knowledge, that is, about the hotel’s central location in Sivas and its being within walking distance of Alibaba, about the role of violent histories in making Alibaba the neighborhood that it today is, and about how its residents were treated in the lead-up to the arson attack. Certain televisual witnesses who possessed this sort of knowledge were left oscillating between the two subject-positional extremes of guilt and innocence. It was especially knowledge of how Alibaba’s residents were treated in the hours leading to the arson attack that heightened these viewers’ sense of enforced passivity—the “shame” through which they witnessed the victims’ martyrdom.

The experience of shame invoked so emphatically by the campaign for an on-site “museum of shame” is therefore not only the stuff of a subject position grounded in a purportedly universally relatable pro-victim stance that a discrete physical or visually mediated encounter with sites linked to the arson attack can alone secure. It has also to do with a self-reflexive subjectivity informed by a geographically and socially specific body of spatial knowledge deriving from multiple encounters. Indeed, working from this specificity promises to benefit the very universality the activist campaign seeks to engage. For it taps into an array of historical experiences much wider and more universally relatable than can be afforded either by an unambiguous notion of guilt for which one must repent or by that of a victim whose identity is solely premised on suffering. In the case of the senior memory activists featured in this article, these historical experiences pertained not only to the Sivas arson attack but also to the Armenian genocide and indeed entangled the one in the other. While such entanglements have so far featured only in personal exchanges, the social pattern they signify indicates their suitability to the sort of objectives that memory activism around the arson attack has pursued.

Following from this analysis, the article proposes to reconsider a tendency characterizing the spatial turn in scholarly explorations of testimony and its cognates. The tendency is to limit testimony’s function to the production of knowledge about violent events whose eventhood is considered discrete and spatiotemporally predetermined. The article’s second section where this tendency was surveyed had ended with a question regarding its potentials and limitations in accounting for the dehistoricization that violence inflicts on victims, survivors, and their sociopolitical heirs. The question meant to acknowledge testimony as a practice that is oriented towards not just epistemics but also ethics—that it is an attempt to reclaim the very agency undermined through violence and its dehistoricizing impact. The empirical analysis offered earlier has shown this question to be about much more than just the various registers employed in historicizing a violent event, including the documentarian urge to preserve the site of violence “as is,” the memorialization of victims as “martyrs,” the performative attempt to resuscitate through commemoration the very spatial knowledge obscured by violence and its televisual mediation, and daydreamy encounters with alterity triggered by limits imposed on such commemorative attempts. Nor is the question explorable only by engaging discretely with each of the diverse types of individuals employing these registers, such as survivors, members of professional organizations, and those who sociopolitically consider themselves victims’ heirs. The extent to which spatial testimony challenges violence’s dehistoricizing impact also involves asking precisely what sorts of historical moments constitute the violence being commemorated, where and when its eventhood begins and ends, and, most importantly, how the answer to each of these perennial questions affects the extent to which previous answers have challenged or confirmed power relations structured by violent histories.