Divergent Disaster Events? The Politics of Post-disaster Memory on the Urban Margin 46

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

by addressing the processes of remembrance in relation to disaster risk reduction (DRR) and disaster risk creation (DRC) (Lewis & Kelman, 2012;Tierney, 2014;Wisner & Lavell, 2017). My objective is to analyse critically the idea of 'disasters as events' based on the politics of disaster memory, applying the 'landscape as arena' perspective (Alderman & Inwood, 2013). This approach attends to open-ended and potential confrontations within the production of a landscape and considers calls for social justice. My focus therefore lies on how disaster memory processes reflect divergent discourses and practices that might question the (re)production of vulnerabilities.
Empirically, I focus on the post-disaster landscape of memory on the margins of Santiago, Chile. This landscape emerges from the action of different actors, including the state and the survivors of the rain-induced landslide that occurred on 3 May 1993. Following ethnographic methods, I recount the emergence of symbols, memorials and practices of commemoration, and analyse what these initiatives mean for different groups and their contrasting interpretations of disaster risks. The research question guiding the chapter is: How do post-disaster landscapes of memory reveal divergent interpretations of disaster 'events' and how can these contest disaster risk creation?
The chapter is organised into five further sections. I begin by reviewing the literature on disaster culture and landscapes of memory before describing the methodology. Section four contains my findings, organised into two dimensions of memory: emergence of the post-disaster landscape and practices of commemoration. In section five, I discuss the findings in relation to the literature. Section six concludes the paper.

culture, memory and landscape 6.2.1 the politics of interpreting disasters
For decades, understanding disasters has centred on human experience and social causation, along long-standing academic interventions (Gaillard et al., 2014;O'Keefe et al., 1976;Smith, 2006). Interpreting the causes of disasters can be categorised according to a threefold typology: acts of God (i.e., supernatural forces punishing humanity), acts of nature (i.e., external and non-human entities) and acts of men and women (Quarantelli, 2000). In the latter view, an earthquake is "a physical happening that does not have any social consequences unless there are human beings who by their decisions and actions create built environments that can be impacted" (Quarantelli, 2000, p. 4). This typology provides a general progression towards social causation but has been criticised for its Eurocentric worldview that sees historical development as linear and some societies as more advanced than others (Greg Bankoff, 2018). A further problem is that certain forms of the naturalistic paradigm persist in academic approaches to disasters (Hewitt, 1983b;Sliwinski, 2017). This paradigm, developed particularly within the physical sciences and engineering, remains dominant in disaster studies (Gaillard, 2019;Hewitt, 1983b).
In this frame, disasters are something abnormal, the result of some 'un-ness': they are unmanaged, unexpected, unprecedented, always deriving "from natural processes or events that are highly uncertain" (Hewitt, 1983b, p. 10 emphasis in original). The notion of an 'event' reinforces the idea that disasters are bounded in time and space, implying that hazards "are not viewed as integral parts of the spectrum of man-environment relations or as directly dependent upon these" (Hewitt 1983. p.10). More recently, Sliwinski (2017) recognises that despite a formal assertion of the importance of social inequalities in addressing and managing disasters, a 'physicalist' interpretation of the latter persists. This is critical, as these interpretations are ingrained in DRR initiatives, many of which still view disasters as "exogenous objective things instead of social processes" (Sliwinski 2017, p.274, emphasis in original). These naturalistic views ignore wider social processes and fixate upon technological solutions. Evidence shows how politicians and the media continue to blame nature when disasters occur (Gould et al., 2016;Marks, 2015;Tierney et al., 2006). This view depoliticises the latter by understanding nature as an essential entity that is not affected by power relations (Castree, 2001;Pelling, 2001;Robbins, 2012) and has practical consequences in terms of relief, recovery, and vulnerability reproduction (Collins, 2008;Marks, 2015;Mustafa, 2005).
Other approaches develop a more critical frame that positions disasters within complex human-environmental relations, denaturalising their explanation (O'Keefe et al., 1976). This view contests the notion of disasters as natural by linking them to social development and highlighting different levels of causation (Hewitt, 1983b;Lavell & Maskrey, 2014;Oliver-Smith et al., 2017;Wisner et al., 2004). Acknowledgement of the root causes of disaster risks is particularly important, as it enhances a critical understanding of more distant and structural factors. The Pressure and Release (PAR) model developed by Wisner et al. (2004) and the forensic framework advanced by Oliver-Smith et al. (2007) are examples. Here, it is critical to focus on disasters as "systemic processes that unfold over time" and whose "life history (…) begins prior to the appearance of a specific event" (Oliver-Smith et al., 2017, p. 471). Efforts must be centred on identifying factors in long-term, complex processes, including histories of exploitation and colonialism, in order to understand the impacts of a disaster. This is assessed by Oliver-Smith in his study of Peru's 'Five-Hundred-Year earthquake' (1999a). Temporally, disasters are neither odd nor unlikely catastrophes, but phenomena that unfold depending on a particular society's vulnerabilities; as Oliver-Smith and Hoffmann (2002, p. 3) state, disasters are "processual phenomena rather than events that are isolated and temporally demarcated in exact time frames". In delineating this tension, Oliver-Smith and Hoffmann (2002) define a disaster explicitly as 'process/ event', accounting for the combination of conditions of vulnerability with technological or natural hazards as destructive agents that are embedded in both human and natural systems and which unfold over time.
According to anthropological work on disasters (e.g., Hoffman, 2016;Oliver-Smith & Hoffman, 2002;Sørensen & Albris, 2016), culture reflects the ambivalence of 'eventness' and processual views in particular ways. The extent to which vulnerable groups, disaster victims and people at risk in general establish different interpretations of their experience of disasters is relevant to their behaviours. Symbols and practices describe the "perpetual presence" of disasters through "political practices, social organization, landscapes, bodies, language, rituals, material culture, memory, etc." (Sørensen & Albris, 2016, p. 75). Disasters bring cultural continuity and change, although it is hard to assess them in short time spans (Hoffman, 2016). In addition, assuming that communities are heterogeneous and do not necessarily share a culture, their cultural framings are traversed by power relations that generate potential clashes between sub-cultures according to assets and resources (Engel et al., 2014;Warner & Engel, 2014). Finally, acknowledging the importance of culture for DRR, Lavell and Maskrey (2014, p. 273) address the need to replace "an imaginary of extreme, exogenous events with an imaginary of managing risks in day-to-day development processes". They acknowledge the need for cultural changes in values and aspirations toward inclusive risk management for vulnerable groups. This requires attention to disasters' 'perpetual presence', including, as I describe next, the embodiment of cultural framings by memory symbols in space.

memory and space in disaster processes
Memory as a collective phenomenon relates to the work of Maurice Halbwachs (1992, p. 38), for whom "individual thought places itself [in] social frameworks for memory". Through memory, social groups emerge as individuals united by commonalities such as class, religion, gender and location, invoking shared identities. Studies on memory expand how national sentiments are constructed through public memorials and monuments that help to recollect a common past (Huyssen, 2003;Johnson, 2004). Similarly, disaster studies researchers have addressed the importance of memory in understanding how groups share experiences of trauma (Hoffman, 1999;Ullberg, 2013). Studies have shown the relevance of memory to recovery, disaster education, risk reduction and resilience (Boret & Shibayama, 2018;de Vries, 2011;Garde-Hansen et al., 2017;Harms, 2012;McKinnon, 2019;Wilson, 2015). Some show that memorials go beyond mere 'mnemonic devices', relating their presence to greater resilience (Boyet & Shibayama, 2018), or show how social memory helps communities to cope with ongoing socio-environmental crises through an 'ethics of endurance' based on their experiences (Harms, 2012).
As a cultural product, disaster memorialisation (i.e., processes of constructing memorials and/or performing commemorative rituals in relation to past disasters) transcends myriad political issues. As anthropologist Susana Hoffman asserts, culture "can be utilized and manipulated by different factions involved in a disaster and thus become political" (Hoffman, 2002, p. 115). Nationalist, gender, religious or ethnic tensions can emerge between dominant and dominated groups through such memory processes (David, 2008;Simpson & Corbridget, 2006;Simpson & de Alwis, 2008). Critically for my study, the role of disaster victims -including their collective struggles and memory strategies -is central to the outcomes of memorialisation. For example, Bos et al. (2005) analyse the role of communities of victims in setting and maintaining disaster issues on the political agenda, which depends on their resources and capacity to institutionalise particular claims. Zenobi (2014), in his study of how families of victims and survivors of a fire present themselves publicly and mobilise to fight for justice, shows the extent to which they associate with other organisations (e.g., human rights bodies or political parties) and embrace different levels of politicisation or depoliticisation.
Memorialisation also entails understanding the role of memorials and commemoration. Simpson and de Alwis (2008, p. 12) compare the processes of constructing memorials in Gujarat and Sri Lanka and their associated disputes, concluding that memorials act as "nodal points of contestation" for and against the state. Sandrine Revet (2011) shows the coexistence and clashes of two parallel commemorative processes related to a series of mudslide disasters in Venezuela. In her account, "official" events supported by institutions such as the Catholic Church were contested by alternative performances and even served as a stage for protesters to address wider national political tensions. Hence, commemorations themselves are political: as Katharyne Mitchell concludes regarding a 9/11 commemorative event staged by George W. Bush, this was "not just a memorial speech concerning a political, terrorist act, but (…) an intensely political act in itself; an act building on the collective memory of the recent past, but also producing that memory's future through a highly particular form of aestheticized, spectacularized politics" (2003, p. 443, emphasis added). Commemorations raise questions relating to justice and political mobilisation, to identities and belonging, and to collective definition of strategies and the terms of the struggle.
Many authors emphasise the spatial character of memory through concepts such as sites -or lieux, in Nora's formulation (1989) -or landscapes with long-standing usage in historical and geographical scholarship (Lowenthal, 1975;K. Mitchell, 2003;Till, 1999). A landscape is "an ordered assemblage of objects" acting as "a signifying system through which a social system is communicated, reproduced, experienced, and explored" (Duncan, 1990, p. 17). In this sense, a landscape is a text in which a set of meanings are inscribed. The 'text' is not fixed, however, nor is it the material backdrop of a particular culture. Rather, "the spaces themselves constitute the meaning by becoming both a physical location and a sightline of interpretation" (Johnson, 2004, p. 316). Therefore, analysis of a landscape reveals both presences and absences: it is simultaneously about remembering and forgetting -a permanent struggle to include and exclude certain accounts (Huyssen, 2003). Moreover, in order to attend to the power relations embedded in the landscape, it is critical to understand it as a social product that encompasses particular agendas developed by certain actors (e.g., the state in the case of war memorials, or disaster survivors in the construction of a memorial to victims), resulting from a pragmatic set of actions (D. Mitchell, 1996).
My analysis of post-disaster landscapes of memory follows this pragmatic and spatial character, while considering their contentiousness. The notion of 'memoryscape' used by Susan Ullberg (2010,2013,2017) in her analyses of the 2003 floods in Santa Fe, Argentina, centres on these dynamics. For her, this concept highlights the 'spatiotemporal' nature of memories, including their interactions and contestations. In her research on social memory concerning the floods, there is an asymmetry in that not all memories are equivalent and, moreover, oblivion is also relevant to the memoryscape. The role of flood victims (or inundados) according to their multiple identities (e.g., as victims and vulnerable, but also as politically active) is especially relevant to memory production (Ullberg, 2013). She suggests the concept of 'embedded remembrance' (2017) to assess how memory of the flooding is inscribed in the everyday practices of the most vulnerable residents of hazardous areas. What is critical there is that remembrance emerges not only through public commemoration of past events, but also as part of the everyday livelihoods of inhabitants of this part of suburban Santa Fe. For example, their engagement in informal economies and practices of barter, reminiscent of the 2003 aid and relief, renders the process of remembrance less visible but still present.
As with disaster vulnerability and exposure, post-disaster memorialisation exists on top of historical processes of injustice and marginalisation of certain social groups. In an effort to explicitly account for social justice issues in memory spaces, Alderman and Inwood (2013) propose the concept of 'landscape as arena'. The idea of the 'arena' addresses the multiplicity of actors, processes and discourses at play, which are constantly in contrast or opposition. This is a particular way of describing "the contentious nature of remembering the past", not by addressing landscapes as finished products, but as "open-ended, conditionally malleable systems" that are always "under constant reconstruction and reinterpretation" (Alderman & Inwood, 2013, p. 193). From this perspective, a landscape of memory is a place in which social groups with different levels of power and resources "actively interpret and debate the meaning of the past as part of larger struggles over recognition and the legacies of discrimination and dispossession" (Alderman & Inwood, 2013, p. 193). They suggest that a dualist notion between the dominant and dominated "blinds us to the multiple agendas and identities that are often at work in memorialisation processes" (2013, p. 194). This includes overlooking the function of memory spaces as 'sites of counter-memory' developed by ignored groups that challenge dominant narratives, as well as other competing discourses between and within communities.
In this paper, I use the notion of 'landscape as arena' to examine memorialisation in the foothills of Santiago. By focusing on the politics of interpreting disasters, this approach seeks to understand how memory questions the processes that generate disaster vulnerability, or whether this process assists in their reproduction and why. As such, the landscape I describe involves both open confrontations concerning disaster risk creation and other subtle claims regarding the past and the future. I therefore use the notion of 'landscape as arena' to address contrasting views within disaster memory processes and thus describe how symbols in space and practices embody divergent interpretations of the same historical disaster. As shown below, interpretations of the causation and temporalities of disasters as events and/or processes take centre stage.
On 3 May 1993, a combination of high temperatures and heavy rain elevated snowfall level, causing a landslide that dragged a considerable amount of mud and debris down the Macul Ravine. Within a couple of hours, roughly 10,000 square metres of debris had been deposited in the populated area (Corvalán et al., 1997;Naranjo & Varela, 1996). The highly exposed low-income settlements along the sides of the Macul Ravine were the hardest hit. According to the National Emergency Office (Oficina Nacional de Emergencias, ONEMI), five poblaciones (neighbourhoods) were particularly affected: El Progreso, Fernando Dominguez, El Esfuerzo, La Higuera and Ampliación La Higuera (ONEMI, 1995), all located in the comuna of La Florida (one of the 52 districts in the Santiago Metropolitan Region). The aluvión (or debris and mudflow) destroyed 307 housing units and damaged 5,610, and left 3,800 victims including 26 people dead and eight missing (ONEMI, 1995). While the mud and debris affected many eastern comunas, the localised impacts make this a spatially concentrated disaster whose associated practices remain local and intra-district.
Two historical characteristics of the area are contextually important. First, the communities affected in 1993 share a history of social cohesion. Some were the result of long-term struggles to access urban space, involving squatting, self-construction and community organisation -methods that were firmly embedded in left-wing political parties (Garcés, 2002) -during the 1960s and 1970s (Muñoz, 1990. Many inhabitants participated in grassroots resistance during the dictatorship (Hardy, 1986) and the area even became a focus for left-wing people returning from exile during the 1980s (Biskupovic & Stamm, 2016). The second detail is the area's considerable transformation since 1993. In the aftermath of the disaster, victims were temporarily relocated to an emergency encampment and later permanently to a new housing project, the Santa Teresa Village (SEREMI-MINVU, 2013). A year later, the Metropolitan Regulatory Plan for Santiago (PRMS), the main land use planning instrument that regulates all 52 of the region's comunas, extended the urban limit further east to encompass the foothills (MINVU, 1994). This triggered intense capitalist urban growth as real estate companies densified this formerly rural landscape with high-income residential projects and gated communities, similar to other areas of eastern Santiago (Sabatini & Salcedo, 2007;Salcedo & Torres, 2004). As a result, the semi-rural population of 5,600 people has grown to 25,000 since the late 1980s (INE, 2017;Muñoz, 1990).
This context is critical for three reasons: (a) certain local symbols, places, practices and interpretations are rooted in these communities' shared past, including their grassroots experiences during the dictatorship; (b) densification of the foothills led to the settlement of a large number of middle-and high-income residents who had not lived through the disaster and were unaware of local hazards; and (c) the real estate-led urban development process triggered major local mobilisation that criticised this outcome and pushed for more democratic urban planning (Biskupovic, 2015b;Ferrando, 2008). The memory landscape described below references these features.
In order to develop contextualised knowledge from the perspective of the disaster victims and inhabitants of the foothills, my research is based primarily on ethnographic methods. During eight months of fieldwork in 2018-19, I conducted 48 in-depth interviews with current residents of the foothills and informants such as politicians and people who experienced the hazard first-hand. In many cases, the interviews sought to construct inhabitants' life stories, with emphasis on their attachment to the area, their experience of the disaster and of cultural symbols, and their engagement with commemorative events. I followed a snowballing approach for the selection and invitation of participants. As shown below, emergent conflicts exist between communities, and since I wish to avoid intensifying this friction, participant's names -and in some cases their neighbourhoods of origin -are not revealed. I also conducted participant observation, attending community meetings and public events. Finally, I assembled a number of secondary documents to gain more information about the disaster and associated memory processes. Newspaper and web articles, policy documentation and historical archives were useful for understanding wider urban development and the post-disaster memory process. Based on an inductive analysis of the data and drawing on the notion of landscape as arena, in the next section I begin by accounting for the production of a landscape of memory and the contestations that emerge in the process, before describing commemorative events and divergent discourses from different social groups.

findings: cultures of disaster memory on the foothills 6.4.1 perpetuating 3 may in space: emerging landscapes and identities
The inhabitants of the foothills remember vividly the odd weather on the morning of 3 May 1993, when high temperatures accompanied heavy rainfall. Informants also recall the time of the landslide or aluvión, as the majority of residents were out at work on that Monday morning. As the flow descended, the houses affected were occupied mainly by elderly people and children. This situation is highlighted by informants and scientific research alike in reference to the potential harm had the hazard occurred during the weekend or, worse, at night (Naranjo & Varela, 1996). Returning from work, many inhabitants remember the ruined houses and neighbourhoods surrounding the old Catholic church of Santa Teresa, which remained standing and has become an icon of the disaster (see Figure 6.1). From the start, inhabitants assigned a hybrid responsibility for the disaster. As an 87-year-old informant states: "This was part human, part nature". For him, the disaster relates to other historical environmental processes: "every 30 years, the [Macul] Ravine 'takes' water" from the mountains, "activating itself". This processual understanding of the disaster is well established among foothill inhabitants, who refer constantly to the settlement of families in the area and account for the environmental and human histories of the foothills (Biskupovic, 2019). Following that day, the area has undergone multiple transformations. The state's approach to risk management focused on the physical dimension of the hazard, leading to initiatives involving (1) a number of land use exclusion zones through the local master plan (the Plan Regulador Comunal, PRC), and (2) mitigation infrastructure, namely seven ponds that should retain future flows (MINVU, 1994;MOP, 2006). While these physicalist interventions have so far protected the foothills, they make part of a dominant mode of governance rendering risks a technical issue that neglects the vulnerabilities that arise from ongoing real estate development (see next chapter). In contrast, long-time inhabitants of the foothills share a common discourse relating to their historical rural upbringing, yearning for a past in which they were closer to nature and to their neighbours -a way of life that was dramatically disrupted by the disaster. In that sense, in addition to physicalist interventions, a particular post-disaster landscape of memory emerged in which many symbols -names, objects, memorials and practices -perpetuate the 1993 disaster. The landscape helps residents to maintain certain traditions based on their long-standing attachment to the area. However, beyond the aluvión itself, the real estate-led transformation of the foothills also works as a source of detachment from their past. The process has not only shifted their relationship with the environment and with disaster risks, but has changed their social environment and transformed their traditional bonds. In this sense, alongside a common identity as long-time, pre-disaster inhabitants of the foothills, there is resentment of new residents for the loss of certain traditions.

Map 6.1 The current landscape of the research area in the foothills of La Florida
Source: Author based on IDE.cl and Corvalán et al. (1997) This landscape of memory is diverse but concentrated in the area impacted by the aluvión. Map 6.1 highlights in dark blue the current course of the Macul Ravine and in light blue the area affected by debris in 1993. The neighbourhoods hit are shown in yellow, including Fernando Dominguez and El Progreso to the south of the Macul Ravine, which no longer exist. The space surrounding the Macul Ravine is currently excluded from development by the PRC, creating a set of localised injustices for inhabitants (as described in Chapter 5). Also shown are places relevant to the memorialisation process, such as Santa Teresa Village (the housing project where disaster victims were relocated) and the CIQMA community centre, where a commemoration event is held annually (see next sub-section). The relief process influenced how victims and inhabitants now remember the aluvión.
The state made efforts to clean up the debris and rehabilitate the area (ONEMI 1995); however, some of the scattered rocks were overlooked and local residents now view them as testaments to the disaster. Such items can be found around the foothills, many 'decorating' public areas (Figure 6.2 (a)). While long-time inhabitants recognise the presence of the rocks as a consequence of the 1993 disaster (and a failure of subsequent relief efforts), new residents do not make this connection. The rocks do not have plaques explaining their origin, so they remain unknown and 'silent' to newcomers. Although there are no open contestations concerning these artefacts, they reveal contrasting interpretations of the materiality of the landscape.
The state also intended more explicit forms of remembrance during the recovery process. After almost two years in an emergency camp, the majority of victims were relocated to a new neighbourhood developed by the housing authorities. This project, located roughly 200 metres from their original site, was named '3 de Mayo' (3 May) by the Housing and Planning Service (SERVIU) in reference to the date of the aluvión (MINVU, 1995). Built in a number of stages, it consisted of more than a thousand housing units to accommodate around seven hundred families affected by the disaster (SEREMI-MINVU, 2013). The name was questioned by residents from the outset and informants claim never to have liked it, referring to the neighbourhood as 'Santa Teresa Village' instead. As a current resident remembers: "We're just recovering from a huge problem that started on 3 May, we all know and commemorate 3 May, and then they go and name the población '3 de Mayo'! So the residents decided, without any official change, to rename it 'Santa Teresa'. We do not live in '3 de mayo' -we live in Santa Teresa Village. (…) We always wanted to change the name officially but nobody calls it '3 de mayo' anyway. It's very strange for someone to call it that -only electricity and water bills use it, no one else (…). We changed the name immediately (…); nobody liked it…it was too obvious".
The inappropriate choice of name, resulting from the lack of participation permitted by the neoliberal housing policy instruments applied, became a contested matter for disaster victims and was exacerbated by a sense of forced resettlement. Many people resent the recovery process to this day. While they apparently want to leave the disaster behind, their built environment is a daily reminder of the landslide as a pivotal moment in their long-standing inhabitation of the foothills. After losing their homes in 1993, they did not need help from the state to remember the disaster by pointing to them when it happened. This example shows how state efforts intended to memorialise a disaster can be resisted by those who suffered its consequences.

Figure 6.2 Artefacts and spaces of memory in the foothills
Sources: a, c, d photo by author and b (URBATORIVM, 2018) These state initiatives contrast with other efforts by local organisations, which have constructed memorials to explicitly honour and remember the victims. Just like the overlooked rocks, these memorials are located in public parks and accessible spaces, allowing people to mourn. Figure 6.2 (b) shows a memorial constructed by the Neighbourhood Associations (Union Comunal) of La Florida -a gift from an organisation from outside the impacted area that represents all of the neighbourhood associations from the district of La Florida. The memorial is located in Solidarity Park, which now occupies the location of the destroyed El Progreso población. Built four years after the disaster, it reads "In memory of our brothers who fell during the aluvión of 3-5-93 due to the adversity of Mother Nature 47 ". This explicit condemnation of nature illustrates a naturalistic interpretation of the disaster that is less commonly held by disaster victims and long-time foothill inhabitants who, as mentioned, attribute responsibility more evenly to environmental and human factors (Biskupovic, 2019). Figure 6.2 (c) shows another memorial, in this case constructed directly by the committees of the neighbourhoods affected by the disaster. The inscription reads "We will always remember you" and is accompanied by the word aluvión, the date of the disaster, and a list of the 23 local people who died. This is regarded as the 'official' memorial and is located within the CIQMA (Centro Intercultural Quebrada de Macul, "Macul Ravine Intercultural Centre") compound, an important place for memory practices in this area of Santiago ( Figure 6.2 (d)). The CIQMA is housed inside the Fernando Dominguez neighbourhood's Catholic church which, as mentioned previously, was left standing by the mudflow and immediately became a place of great significance for local inhabitants.
After 1993, the area surrounding the church was abandoned and used as a refuse dump, but in the year 2000, residents of the El Esfuerzo and Las Perdices neighbourhoods sought to turn the space into a community centre. Squatting it at first, they later regularised the tenancy with the state and founded the CIQMA. Given its trajectory, the Centre is deeply ingrained in the history of the La Florida foothills and the 1993 disaster. Although it is also used for recreational, cultural and educational activities, it hosts myriad symbols from the disaster. Besides the 'official' memorial, paintings on the building's façade remember the disaster, and 23 trees were planted by the community in memory of the people who died that day, although the latter initiative has been criticised as a stunt by the local right-wing mayor (GerminaLaFlorida, 2014). Ultimately, the CIQMA is the focus of the memorialisation of the 1993 disaster, and remembrance practices and discourses emerge on each anniversary, as covered in the next section.

performing memory: practices and discourses for remembering the disaster
Every 3 May, the CIQMA invites people from the foothills to come together in commemoration of the aluvión. Given the area's religious roots, people gather to celebrate a Catholic mass involving various speakers. The ritual is organised by the CIQMA's managers, who are members of families affected by the disaster, and assisted by other neighbours and municipal officials. Many residents of this part of La Florida feel obligated to attend the mass, and each year the venue is packed (Figure 6.3 (a)). The service has gradually become an 'official' and 'institutionalised' ritual of remembrance. Members of NGOs, public sector workers and politicians accompany residents from all of the neighbourhoods in the foothills. With a few exceptions, particularly figures who worked alongside residents throughout the emergency, relief and recovery phases, there is generalised criticism of attendance by politicians. As one organiser recalls: "When there're elections, a

lot of people turn up [to the mass] who want to address the public…and we've never allowed them [to speak] (…), out of respect to the bereaved. (…)
This is not something political. We feel it, we live it…". The latter is particularly important, as the mass is presented as an event organised 'by' and 'for' the people, involving an emotional response to the tragedy and a number of rituals (e.g., Figure 6.3 (b)).

Figure 6.3 Practices of Remembering the 3 May on the Foothills
Source: author CHAPTER 6 (b) (a) Some research participants question the unifying character of this commemoration. One informant recalls that, traditionally, a long procession led from Santa Teresa to the CIQMA, and that people used to bring candles and sing and pray their way to mass: "It was very beautiful. It felt special; it was emotive". However, with the passing of time "we are losing all of that and [these feelings] are gradually being lost". For her, this is because the "official mass" has become a "mechanical" celebration where politicians go to "seize the opportunity" and "appear in the photo". In response, parallel gatherings are organised with the aim of recovering a more traditional sense of community and where local people can share their emotions and openly mourn the aluvión:

"We light a fire in the park… (…) we meet around a rock, we light candles and torches and ask [the park's managers] that they don't turn on the lights. (…) We pray; we have the priest there and a guitar as well. And there the people cry… there's feeling".
These parallel practices continue today but are less popular than before and remain secondary to the 'official' mass. However, during the 2016 service, a group of people staged a protest outside the venue against the unsustainability of development in the foothills, triggering a division of opinion. The protestors were particularly concerned with defending El Panul, a forested area of La Florida that a network of citizens and organisations seeks to protect from real estate development in order to turn it into a public park (Biskupovic, 2015b). They strongly criticised La Florida's right-wing mayor for his ambivalence regarding conservation of the forest, although the organisers of the mass took a critical stance against the protest, which, according to them, blurred the unity of the commemoration.
The organisers subsequently became stricter in keeping the ceremony peaceful. Despite attendance by important decision-makers, their role is now to appear united as a disaster community, and they have adopted a conservative position, more closely controlling what is said and done during the event. For them, "it is a moment of prayer, reflection and unity in pain", and while they agree that there are problems in the area, they "believe that any [public] manifestation will only blur the solemnity of the act and will not bring solutions". As such, before the commemoration ceremony that I attended in 2018, the organisers asked community members to refrain from any "political, social or personal manifestations" during the service. For them, the ceremony should be a cultural event in which they collectively contemplate their tragic past. They should be united as a community, bound by this event that occurred almost three decades ago.
The schism led to further disagreements between groups regarding disaster memory. More conservative stances regularly entail a pessimistic view of overcoming common tragedy, especially considering the newcomers and wider changes in the foothills. One research participant states that remembering helps them to "never forget that this disaster took some of our neighbours and some of our houses… and that's something you never really recover from". This somewhat defeatist discourse arises from decades of exclusionary policies that trap them in a situation of social disadvantage by comparison with new residents of the foothills. While long-time inhabitants do not explicitly exclude newcomers, they highlight a difference in terms of how they experienced the disaster and are joined by that, unlike new residents. This type of discourse is also shared by the managers of the CIQMA. For them: "This is not something that happened one 3 May and a year later it was over. People become impoverished and lose hope; families dissolve and marriages separate. There're people dying from sadness one, two, three years later. So that's why we remember [the disaster] every year. And when the next 3 May comes and the church is full, you'll see that we're united by that tragedy".
In these interpretations, there is a temporal understanding of the disaster that goes beyond 3 May 1993. There is a sense of unity that arises from a common negative experience that endures to this day. As organisers of the mass, however, their current practices of controlling speeches and keeping the event peaceful ensure that the dominant discourse of commemoration is anchored in the past. Here, a particular power tension arises: in their effort to 'not politicise' their practices of memory, they depoliticise the foremost developmental trend that keeps them at risk. By using the platform as a means of looking solely back to 3 May 1993, they disregard disasters that are currently incubating.
In contrast, another group of long-time foothill inhabitants have pushed to politicise the commemoration, making explicit claims regarding their situation. Ahead of the 2018 event, they approached the organisers, requesting permission to read a speech during the service, claiming that "this is the time to catch the attention of the authorities" present. In the speech, they would criticise the state for its role in the current condition of the foothills, emphasising, for example, a number of ongoing problems and shortcomings in terms of road infrastructure, environmental degradation and conservation, issues relating to water and flood management, and disaster risk management in general. In the words of a group informant: "We need to stand up. After 25 years, this territory remains abandoned. (…) Imagine if there were a new aluvión -how are they going to evacuate all these people? In 1993 there were four hundred houses and a thousand people; now there're four thousand houses and fifteen thousand people. Where? How? They'll all be trapped (…) So the problem of the disaster continues, you know?" In their view, the annual commemoration should serve as a platform for warnings about the future of the foothills and to educate residents in the risks they face; however, "there's nothing like that [because] the mass is not organised by all of us". These differences were made clear in the days prior to the 2018 event, and the ceremony itself then became an arena for contrasting understandings of disaster risk. Although for each group the 1993 aluvión is 'ongoing', this does not mean the same thing for both, and the latter group seeks to use the commemoration to criticise the perpetuation of risks and vulnerability by wider governance structures.
Regardless of these criticisms, the commemorative mass of 2018 followed the organisers' plan. The speeches highlighted a consensus arising from a common tragedy, with frequent references to 3 May 1993 and the shared misfortune that unites them. A handful of people spoke, including politicians and representatives of the organisation. The speakers yearned for a lost past, called for remembrance of the dead and care for their family members, and made explicit demands for respect for the mass as a solemn act. A few appealed to the strength of the community and their efforts in standing up against adversity, particularly in the context of the severe weather on the day of the tragedy. The current urbanisation trend, the lack of disaster preparedness measures, or any other contentious or divisive issues are absent from the event.
A year later, the more critical group of residents prepared and circulated a pamphlet during the 2019 mass that reads: "Today we remember our 23 neighbours whose lives were so abruptly cut short by the aluvión of 3 May 1993. We remember them with love and respect. But we must also remember other things. A people without memory is a people destined to suffer the same calamities again and again. We want to remember this and to remind those who make the decisions of their devastating past, present and future effects on our lives".
The pamphlet goes on to describe a number of urban and environmental problems that emerged after the disaster and ends with a call for change to the PRC and for the group's participation in the process. The document sought to launch a more critical discourse and to politicise post-disaster development in the context of over two decades of perceived exclusion. Moreover, it explicitly assigned blame to real estate firms and to the local government and its planning instruments, identifying present conditions as a result of specific political decisions. As such, its purpose was to generate a discourse of justice involving allocation of responsibilities, participation in urban governance processes, and reparations for what has been endured.
Ultimately, while the two groups view the events of 1993 as extended over time and both refer to an 'ongoing disaster', they differ in their approach to the political and practical elements of their shared temporal discourse. Whereas the organisers of the commemoration service struggle to maintain unity within the community based on the tragedy they are still living, the other group mobilises a similar processual understanding of the disaster, but one that is more critical. For them, the 'ongoing' disaster is a call to transform their current situation and resist the creation of further disaster risks. Although the commemoration takes place every year, the demands and voice of the foothills have yet to be truly heard.

discussion: convergence and divergence of discourses
The landslide that occurred on 3 May 1993 profoundly changed the foothills of La Florida. The findings of this paper describe this from the perspective of memory processes, focusing in particular on symbols and practices of remembrance. While some of the literature on post-disaster memory asserts that the presence of such symbols and practices is a critical dimension of community resilience (Boret & Shibayama, 2018), I have shown that this is not always the case. Symbols and commemoration events can lead to contestation in which different ideas are at play. As assessed by Ullberg (2013), remembering is a heterogeneous and unequal process. In her analysis of a 'memoryscape' in Argentina, she shows that while some practices might enhance resilience, others contribute to social vulnerability (Ullberg, 2010). This is consistent with my findings: In the foothills of Santiago, despite similar interpretations denoting an 'ongoing' disaster, divergent meanings entail either a continuation of the status quo or critical reflection on the root causes of risks. Discourses and practices of memory are thus ingrained in various urban political claims regarding the trajectories and futures of the foothills. These findings have relevance for the literature on landscape, memory and DRR/DRC.
The landscape of memory in the foothills is spatially concentrated in this area of south-eastern Santiago. This contrasts, for example, with Revet's (2011) account of La Tragedia in Venezuela, when mudslides destroyed multiple settlements in the north of the country, influencing the emergence of symbols and rituals in a number of places. On the margins of La Florida, meaning is associated with certain rocks and built memorials, but also with absences. Examples include memories of settlement in the foothills and shared struggles to access housing, of sharing a rural livelihood that contrasts with the dense urban project of Santa Teresa, and of the day of the aluvión and the ensuing recovery. All of these are carried by inhabitants throughout their daily lives and alternative commemorative practices. The ability of long-time residents to 'decipher' the origins of scattered rocks illustrates how memory is embedded in this historical trajectory of inhabitation. It is in this sense that the alternative commemoration is perceived as 'theirs', contrasting with the 'official' one at the CIQMA. This heterogeneity of rituals is also assessed by Revet (2011) in her description of an institutionalised commemoration led by a Catholic priest and one led by communities whose collective narrative is based on building an 'us'. Thus, post-disaster landscapes are not only multiple and diverse, but also open and remade through these anniversary practices.
The landscape of memory in this marginal area of Santiago also includes narratives that are pressured by power, an issue upon which I have expanded regarding the unintentionally political annual commemoration service. The event brings together a diverse group of residents and representatives of public and private organisations in remembrance of a tragedy that scarred the area. However, its failure to address enduring hazard conditions means that this conservative ceremony does little more than perpetuate the status quo, a problem that lies at the centre of emerging disputes between local groups. As addressed by Zenobi (2014), depoliticisation is a strategy of some memory groups depending on their interests and networks. As such, attendance by the local right-wing mayor might be a motivator of such positions, as the CIQMA organisers work closely with the local government. Hence, the contested organisation of the ceremony -as well as the wider landscape of memory -also reflects different attitudes towards the ideologies that currently govern the foothills (from the right-wing Major of La Florida to the wider capitalist urban development and technical management of risks).
Herein lies the importance of the idea of landscape as arena. It is a frame that explicitly aims to examine the contested claims emerging at the level of 'dominated' groups. Disputes over the official commemoration have increasingly made the event a space of friction between divergent discourses. In particular, growing efforts by the dissident group to disrupt the mass illustrate a desire to challenge a dominant narrative. The post-disaster memory landscape is thus contested in relation to critical concerns regarding the disaster itself and the ongoing production of vulnerabilities. Disputes over disaster interpretations and calls for justice are at the centre of the landscape of memory. While the idea of still living the 1993 disaster is shared by the groups in dispute, they differ in their views of the commemoration's potential to improve their situation. The 1993 aluvión is an 'ongoing event' that enables some to lament a common tragedy but that mobilises others to contest the production of risks. Although there are no struggles for (re)defining the past tragedy, contestations exist in regard to governance responsibilities for recreating disaster risks.
In this sense, the disaster is interpreted as both 'event' and process in line with Oliver-Smith and Hoffman's (2002) definition. While some naturalistic discourses exist in the foothills, such as on the stone memorial in Solidarity Park, the bulk of the communities that experienced the disaster interpret it in processual terms, alluding to both long-standing social and natural features. The differing approaches of the commemoration organisers and the protest group illustrate how interpretation of disasters as socially-embedded processes does not necessarily constitute a more just way of dealing with disaster risk. Despite apparently overcoming the naturalistic paradigm in disaster memory, some interpretations still address disasters as exogenous things, as Sliwinski (2017) reminds us. It is hence possible for interpreting social drivers of risks while externalising responsibility for the creation of those conditions. Such a depoliticised view, as shown here, maintains the status quo and existing vulnerabilities.
Denaturalisation of disasters through memory processes is not sufficient per se to deliver more just risk management, and this has implications for DRR and DRC. The question is: How can memory processes contribute to disaster preparedness, enhance emergency warnings, or educate the local population on their risks to climate hazards? If DRR initiatives should raise concerns about the complexity of risks -as the United Nations Office for DRR states, risk is 'everyone's business' (UNDRR, 2017a, 2019) -then cultural practices are an integral part. As shown in my findings, particular discursive formations that explicitly confront naturalising narratives are more likely to raise awareness. The issue is therefore how disaster memory reflects the past or the future of disaster risks -how the production of a landscape of memory brings together a post-disaster community to deal with a shared tragedy but also to question their current risk conditions. Such a process must recognise the slow and complex cultural change needed to enhance DRR while resisting DRC. Consequently, the competing interpretations that emerge within disaster memorialisation can constitute a first step towards transforming the structural conditions that create vulnerability.

conclusion
In this paper, I have presented a particular perspective on the politics of disaster memory. My aim was to understand how a landscape of memory reflects divergent interpretations of disaster temporalities. I focused on the foothills of Santiago and its symbols, practices and interpretations of memorialisation, showing how disaster memory differs between groups. Drawing on the idea of landscape as arena, I examined the diverse agendas and contentious views involved in the construction of disaster memory. Here, I described the presence of ongoing disagreements between communities over disaster memory, considering how these disputes address long-term development processes. Moreover, I argued that thematisation of such processes is a relevant concern for memorialisation in terms of how questions of social justice are put forward by post-disaster communities.
Post-disaster landscapes of memory are not stable but an arena where multiple views contrast. By following an ethnographic approach, we can see that memorialisation processes are open and that post-disaster communities are diverse rather than homogeneous. Groups of inhabitants who suffered the same disaster can participate in memory practices while exhibiting varied interpretations of their current risk experience. Hence, by showing these contested cultures on the ground, memory researchers are able to reveal that despite concurring observations regarding disaster experiences (e.g., 'we are still living the disaster'), these can have manifold implications. As such, interpretation of disasters as 'unnatural' or as 'processes' does not necessarily enhance resilience. Therefore, the divergent interpretations shown here do not concern the temporality nor the causation of risks, but the extent to which groups of inhabitants engage in the commemoration to protest their current conditions. To make such claims publicly and politicise disaster risk creation is an important dimension to potentially advance towards a more just and safe urban area.
In this sense, my work is socially relevant to understanding how post-disaster communities deal with their ongoing vulnerabilities. Views of disaster risk are heterogeneous across landscapes of memory. Symbols of past disasters are present in different places and help to make sense of this past. This, however, does not translate into an active discussion of a current hazardous state. Different groups involved actively in a post-disaster landscape of memory might describe the persistence of disaster experiences at the local level but will not necessarily discuss and seek to reduce the risk to which they are exposed. I argue that to do so would constitute a more socially just form of DRR, as claims concerning recognition, redistribution and reparations are made in relation to current disaster risk production. Ultimately, a strict focus on memorialisation is limited, as symbols and practices change only slowly, while other community-based organisations might enhance DRR but fail to engage in memory. It therefore remains to be seen how these organisations relate to disaster risks in general and to the landscape of memory in particular, either maintaining the status quo or seeking to disrupt the root causes of risk.