Agonistic Memory and the UNREST Project

This paper reflects on some of the findings from a Horizon 2020 research project, Unsettling Remembering and Social Cohesion in Transnational Europe (UNREST, Horizon 2020, funded 2016–2019, http://www.unrest.eu/), which aimed to test and apply an agonistic mode of remembering in different settings. The analysis focuses on the potential advantages of promoting agonistic representations of past conflicts in museums through the adoption of ‘radical multiperspectivism’, as opposed to the ‘consensual multiperspectivism’ informing most contemporary exhibitions and displays. The paper argues that such an approach, which foregrounds socio-political passions by drawing on both artistic interventions and contrasting narratives, can deepen visitors’ understanding of violent conflicts and help counter the growing shift towards antagonistic memory, by turning enemies into adversaries.

enemy-other, with the purpose of reaching a conflict-free society -typically imagined in the image of a fictionalized past of ethnic purity. It applies the moral categories of 'good' and ' evil' to the agents involved in the narrative, and as identities are morally essentialized, it cannot reflect upon its own constitutive role in the construction of identity. The cosmopolitan mode of remembering builds on an understanding of the world as one big and potentially harmonious entity, united by a common culture based on the recognition of human rights. The moral categories of 'good' and ' evil' are applied to abstract systems such as democracy and dictatorship, and the cosmopolitan mode is highly self-reflexive, in so far as it considers all identities to be constructed and is able to reflect on the perspectives of both the Self and the Other as victims. Finally, an agonistic mode of remembering recognizes conflict as an ontological and fundamental characteristic of human society, but it tries to deconstruct the moral pitting of the other as an enemy on moral grounds through social and political contextualization of the historical conflict. An agonistic mode of remembering is also highly conscious of its own responsibility as a social discourse in the construction not only of the identity of the 'we' position, but also of that of the ' adversary'.
Our conceptualization of agonistic memory has obviously been informed by the wider concept of agonism. Although this concept has been subjected to many different interpretations by, among others, Hannah Arendt, William Connolly and Bonnie Honig (Connolly; Honig; Lederman), we decided to take our point of departure in Chantal Mouffe's interpretation, because, unlike the other thinkers, she fully acknowledges the existence of uneven power relations and hegemonic ideologies, recognizing the ontological condition of antagonistic relations (Mouffe, On the Political; Agonistics). According to Mouffe, society is composed of asymmetric power relations, and collective identities are constituted through the political relations between an US and a constitutive outside in the form of an ' other' or THEM. Collective identities are therefore always already inscribed in potentially antagonistic relations (Mouffe, On the Political 10-16). For Mouffe, agonism is mainly a quality related to the political public sphere as a way to recognize but also to mitigate antagonism. This has two consequences. Firstly, that an agonistic approach to the uses of the past will have to oppose or unsettle hegemonic ways of understanding as well as reveal the socio-political struggles characterizing the public sphere both in the past and in the present. This is an entirely relational definition and any kind of discourse that disrupts or unsettles hegemonic discourse could be said to have an agonistic function in this sense. Secondly, an agonistic approach is supposed to mitigate or sidestep the antagonistic propensities for violent conflict that are inherently related to the asymmetric power relations. As stated by Nico Carpentier (148), discourses can be defined as either antagonistic or agonistic according to the nodal points that characterize the internal, textual relation between the 'self' and the ' other'. This definition is not entirely relational; rather, it is a definition of what we could call the minimum defining traits of agonistic discourse. In her latest book, Mouffe herself (For a Left Populism 18, 49) mentions the defence of equality and radical democratic processes, and she favours the expression that agonism is characterized by a relation of ' conflictual consensus' between parties -a relation where the opponent is recognized not as an enemy to be destroyed but as an adversary to be engaged with.
In line with Mouffe's conceptualization, we argued in our 2016 paper that the cosmopolitan mode of remembering, by focusing on the perspectives of the ' other' as a victim, 'is unable to incorporate the perspective of the opposed "Other", the perpetrator as a subject in his own right' (397). By contrast, we subscribed to Olick's argument (148) that learning from the past 'means listening to both victims and perpetrators, and not to judge absolute truth or even to sympathize with either but to learn from their experiences and perspectives'. Agonistic memory, therefore, incorporates the perspectives of the perpetrators. Whereas Mouffe's democratic theory envisions the ' other' as an adversary and not an enemy, this obviously does not make sense when we are dealing with representations of perpetrators of mass atrocities. In such cases we do not advocate 'taming' the representation in the image of an adversary, but to facilitate an understanding of the contexts and narratives that made such cruelty socially and politically possible. Does that mean that we will have to understand the Nazi perpetrators responsible for the Holocaust? Yes and no. Yes, because we need to understand what kind of social and political conditions it takes to make normal people turn into war criminals, believing they are doing the right thing. If we do not, we will be unable to see the same conditions emerging in contemporary society. And no, because we cannot allow this understanding to become an excuse or legitimation of the crimes committed.
How do we ensure that understanding does not turn into legitimation? There are probably many different ways to avoid this, but one particular way we put forward in the article, and have since further elaborated, is through a specific kind of multiperspectivism. Multiperspectivism, as defined by Erll (Memory in Culture 151), refers to a narrative being simultaneously focalized through different subject positions. If we combine this concept with the basic distinction of the ethico-political relation between 'us' and 'them', we are able to distinguish between ' consensual' and 'radical' multiperspectivism. In consensual multiperspectivism, often applied by cosmopolitan memory discourses, voices and perspectives belonging to characters who basically agree, or at least believe in the possibility of rational consensus, coincide. In radical multiperspectivism, voices and perspectives belonging to antagonistically opposed enemies, typically victims and perpetrators, meet, alongside those of bystanders, traitors, collaborators and so on (Hansen 16). The former could be compared to dialogue in Habermas's sense, while the latter, which offers us the perspective of the perpetrator in dialogic interaction with the perspective of the victim and other historical agents, could be compared to a Bakhtinian form of dialogue (Bakhtin; Gardiner; Roberts). We would argue that radical multiperspectivism is an efficient strategy to provide understanding without legitimation.
In the rest of this paper, we reflect upon the development of the concept of ' agonistic memory' in light of the findings from the Horizon 2020 research project, Unsettling Remembering and Social Cohesion in Transnational Europe (UNREST, Horizon 2020, funded 2016-2019, http://www.unrest.eu/). The project was theoretically underpinned by our 2016 article, and it was launched at a moment when it became more and more obvious that the cosmopolitan memory mode, hegemonic in the EU's consensual approach to traumatic memory and conflict, was being seriously challenged by the antagonistic mode openly adopted by the rising extreme nationalist movements. The explicit aim of the project was to pursue a third, agonistic memory way, which would embrace political mobilization and conflict as an opportunity for democratic development. The project was thus concept driven and it aimed at theory building, as well as theory testing.

Learning from the UNREST project
Logical systems and abstract categorizations tend to fall short when confronted with reality. One of our premises in the UNREST project was therefore the need to treat our categories as ideal types and to fully accept that we should conceive them as fairly fluid and dynamic in social and cultural contexts, as opposed to enclosed memory regimes. To test this and other hypotheses, we opted for two case studies. One focused on social agents and modes of remembering relating to mass grave exhumations in Spain, Poland and Bosnia (Ferrándiz and Hristova). The other case study analysed methodological approaches and modes of representation in five European war museums (Berger et al.; Cento Bull et al.). While both case studies helped us revise and refine our conceptual framework, in this section we will discuss the main findings from the analysis of the five museums (Historial de la Grande Guerre in France; the Kobarid Museum in Slovenia; the German-Russian Museum and the Military History Museum in Germany; Oskar Schindler's Factory in Poland). Our choice is due partly to the relevance of the theme of cultural representations in this volume, and partly to illuminating exchanges and debates we have had with museum practitioners over the course of the project -not least since we organized, in collaboration with the Ruhr Museum, a new war exhibition inspired by agonistic memory theory which was publicly launched in Essen, Germany, on 11 November 2018. Fieldwork by UNREST researchers at these museums relied on qualitative methods, including archival research, interviews with museum curators, as well as in-depth group analysis of permanent exhibitions. Interviews with visitors and analysis of the visitors' books at each of the five museums were also part of the fieldwork and informed the analysis.
Our analysis of the five war museums confirmed one of our hypotheses, namely that a cosmopolitan approach to war and conflict in museum settings has become widespread in the European context, albeit often mingled with traditional antagonistic representations. The cosmopolitan approach was especially evident in the Historial and the Kobarid museums, both of which dealt with the First World War, as they opted to represent both civilians and soldiers as victims, independently of the army they fought for. While the emphasis on victims was also prominent at Schindler's factory, here the exhibition, which focused on the Second World War, clearly positioned the Poles (both Catholic and Jewish) as the victims of the Germans, who were portrayed antagonistically as the aggressors and perpetrators. Less often we found examples of agonistic representations of past conflicts. The latter were especially visible at the Military History Museum in Dresden, consisting of what we would term agonistic interventions, such as provocative art installations, side-by-side representations of contrasting war myths or indeed thought-provoking displays of beautifully crafted yet deadly war objects. These kinds of interventions can effectively unsettle visitors and disrupt hegemonic discourses. For this reason, to a large extent we replicated this type of approach in the Ruhr Museum exhibition, provocatively entitled 'Krieg. Macht. Sinn.' (War. Power/Makes. Sense/Meaning). Covering both twentieth-century and contemporary wars, such as the conflicts in Afghanistan and Syria, the exhibition used objects, images, texts and videogames to present visitors with contrasting representations and understandings of various aspects of war-making, including aerial bombing and genocide. For instance, on display is ' a group of images and exhibits that juxtapose the subjects of honouring and remembering the heroes of the First World War with the remembrance of fallen German army soldiers and IS suicide bombers' (Berger et al. 28).
Our analysis of war museums and our own engagement with a war exhibition helped us distinguish between agonistic interventions and agonistic memory discourse. While single agonistic interventions can be both unsettling and provocative, they cannot in themselves provide radical multiperspectivism in the sense of helping visitors understand the historical conditions, social grievances and political passions that made ordinary people turn into perpetrators, bystanders or collaborators. As Stiem has contended in relation to the statues of former perpetrators, it should not be a question of choosing whether to tear them down or modify them, but of adopting a multilayered, thoroughgoing and complex approach. In a museal setting, this would mean complementing single provocative agonistic interventions that rely on art installations or authentic objects with narrative-based radical perspectives on past conflicts, for instance through oral history retelling. A successful example of agonistic multiperspectivism through oral history is provided by a recent exhibition, entitled 'Voices of '68', which opened at the Ulster Museum in September 2018. The exhibition made use of oral narratives in providing contrasting perspectives on 1968 and the 'Troubles' in Northern Ireland (Reynolds and Blair). Yet oral history in museums is very rarely employed in this way; more commonly, it is used to offer plural narratives in terms of gender, age and social background within a consensual overarching perspective. Mouffe's agonism, therefore, may be placing too much emphasis on artistic interventions and too much responsibility on the artistic community to disrupt hegemonic discourses.
As museums do not operate in a vacuum, but are heavily influenced by political and even diplomatic constraints, we found that this helps to explain the scarcity of agonistic approaches in such institutions and the propensity to opt for consensual, as opposed to radical, multiperspectivism, despite scholarly arguments in favour of radical new approaches to museal representations. Museums, in fact, have been theorized as ' agonistic spaces' where hegemonic narratives can be disrupted and conflicting voices can/should be heard (Pozzi). Back in the 1990s, Doering and Pekarik argued that visitors, although they access museums with pre-established ' entrance narratives', should find their views challenged and unsettled, as this would lead to deeper questioning. In this context, our exchanges with museum curators and practitioners brought to light the importance of conceiving an exhibition not as an end in itself, but rather as a lever to develop new ways of engaging with the public and promoting debate. We also agree with Mouffe that, despite being mainstream and established institutions whose role is still to represent hegemonic memory cultures, museums are also affected by social movements and protests at local, national and transnational levels, and should therefore continue to be targeted, together with public monuments and statues, in order to promote agonistic representations.
Finally, it is important to understand visitors' and audiences' emotional and cognitive reactions not only to agonistic but also to cosmopolitan cultural memory representations. Our reception analysis is still ongoing, but preliminary findings indicate that a cosmopolitan representation of war -even of a distant one like the First World War -can successfully inspire among visitors strong revulsion to bloody conflicts and compassion for the other. This is highly relevant, since visitors need to care and empathize about what happened in the past if they are to engage with deeper understanding of the dynamics of collective struggles and conflicts at different historical periods. Thus, as our visitors surveys at the Kobarid Museum of the First World War clearly indicate, its cosmopolitan message resonates strongly among visitors of all nationalities, eliciting compassionate feelings for those who fought in the trenches, regardless of the army they belonged to, as well as a clear revulsion against all wars. Nevertheless, our analysis also reveals that these feelings tend to be generic and superficial, unable as such to promote understanding of the other's perspectives or indeed of the struggles and passions that lead to conflict. In fact, an in-depth analysis of Italian and Slovene visitors to the Kobarid exhibition indicated that many also harboured a strong sense of national belonging, often accompanied by indifference, and at times even antagonistic feelings, towards one's own former enemy, rather than greater understanding and empathy. Nor did these visitors appear to have gained an understanding of the socio-political factors and agents that engendered the war.
Hence, revulsion and compassion are welcomed reactions for both cosmopolitans and agonists, but from an agonistic point of view they can be effective only insofar as they facilitate visitors' engagement with radical multiperspectivism and socio-political passions. We might say that the cosmopolitan memory mode provides a good starting point in terms of its effects on stakeholders, but falls short of promoting complex understandings of the conflictual constructions of the collective 'We' and 'They' identities underpinning violent antagonism, as well as of what is required to turn enemies into adversaries.
To conclude, we found much support, from both empirical evidence and engagement with heritage practitioners, for the proposal that heritage institutions can effectively promote agonistic representations and understandings of contested pasts. Such an approach draws on radical multiperspectivism, promotes an agonistic discourse through both artistic interventions and contrasting narratives, and engages with emotions, representing and eliciting stronger affects than just compassion and revulsion.