Agonistic peace agreements? Analytical tools and dilemmas

Abstract Later years have seen the growth of a vibrant theoretical discussion on agonistic peace and the importance of creating space for contestation, plurality, and dissensus post-accord. However, there has been very few attempts at embedding agonistic theory in empirical analyses of peace agreements. This study attends to that lacuna by investigating how agonistic principles can be integrated and investigated in peace agreements. We suggest a threefold set of indicators for assessing the degree to which peace agreements are invested with agonistic dynamics: (1) what types of spaces for interaction are offered post-accord; (2) what forms of inclusion are stipulated; and (3) how is the peace agreement framed in terms of conflict termination and consensus/dissensus? We illustrate how the various indicators could be put into motion in concrete analyses applying them to examples from the Oslo Accords, the Belfast Agreement, and the Colombian Peace Agreement. Finally, we discuss four dilemmas and problematiques of integrating agonistic ideas in peace agreements; the issue of power, the mixing of agonistic and liberal ideals in peace agreements, the principle of ‘nothing is agreed until everything is agreed’ and the related double-edged potential of constructive ambiguity and finally the challenges of implementing peace agreements.


Introduction
The literature on conflict resolution has traditionally been characterised by a liberal logic of finding rational, win-win solutions to conflict that meet the needs of all parties and transcend incompatibilities. 1 While it is often stressed that conflicts are neither good nor bad, but potentially both constructive and destructive depending on how they are handled, efforts to deal with conflict in constructive ways rarely leave space for passions, conflict, and contestation. 2 Oftentimes the remedy to conflicts has been sought in the creation of peace agreements and postconflict strategies striving for consensus and univocal agreements on understandings of the past. Such an approach underlines the necessity to go beyond parties' official positions and make all involved actors elaborate on underlying interests, which can be negotiated into a commonly agreed framework with the possibility to bring advantages to all.
To illustrate how agonistic principles are to be understood and applied in practice, we bring in examples from three peace agreements: the Oslo Agreements, 12,13 the Belfast Agreement 14 and the Colombian Peace Agreement, 15 which function as testing grounds for the indicators. These three cases have been chosen because they display quite different degrees of agonism and therefore serve as useful illustrative vignettes when developing analytical tools for the analysis of agonism in peace agreements. Others have shown that Colombia is somewhat of a prime example in terms of an agonistic peace agreement. 16 The Oslo Agreements, in contrast, display quite low levels of agonism, 17 whereas the Belfast Agreement exhibits a more mixed picture in terms of agonistic ideals. 18 The advantage of using these three agreements also lies in their temporal and procedural differences: The Oslo Agreements were signed in 1993 (Oslo I) and 1995 (Oslo II) and outlined an open-ended gradualist-approach to ending the Israeli-Palestinian conflict. The Belfast Agreement was signed in 1998 and used a comprehensive approach to put an end to 'the Troubles' in Northern Ireland. The newest agreement, signed to end violence in Colombia in 2016, is the most ambitious of the three, in also outlining detailed measures for transitional justice. While none of the three peace accords were designed with formally conceptualised agonistic elements in mind, they all include elements of agonism and importantly give insight to the limits and potential of designing agonistic peace agreements.
In the following section, we unwrap main objectives and principles of agonistic peace. Thereafter, we develop the study's conceptual framework, where the discussion of each indicator is followed by practical examples from the Oslo Agreements, the Belfast Agreement, and/or the Colombian Peace Agreement. Finally, we discuss dilemmas and problematiques of integrating agonistic principles in peace agreements and outline the study's conclusions.

Agonistic peace
In an agonistic peace, former enemies ideally come to see themselves as 'sharing a common symbolic space' where conflict is played out and regulated by procedures and platforms accepted by the conflicting parties such as, in this case, a peace agreement. 19 These spaces are characterised by competitions over hegemonic power within a specific agonistic framework. New relationships are to be fostered in which groups no longer see each other as targets for destruction, but instead view one another as contenders whose views they disagree strongly with, yet with respect for the adversaries' need for contestation. 20 In that sense, the shift from antagonistic towards agonistic relations constitutes a 'radical break with the social order that underpinned the violence of the 12 'Interim Agreement on the West Bank and the Gaza Strip (Oslo II)' (Israel-Palestine, signed 28 September 1995); 'Declaration of Principles on Interim Self-Government Arrangements (Oslo I)' (Israel-Palestine, signed 13 September 1993). 13 The Oslo Agreements refer to the Oslo I Accord, officially called the Declaration of Principles (DOP), signed in Washington, DC, in 1993 and the Oslo II Accord, signed in Taba, Egypt, in 1995. From here onwards, we refer to both documents whenever the Oslo Agreements/Accords are mentioned in the article. 14 'The Belfast Agreement' (United Kingdom-Ireland, signed 10 April 1998 past'. 21 Such processes do not only concern the transformation of enemy 'others' into adversary 'others' but also enable the transformation of societal relations and/or ideologies, ideas, and concepts. In that sense, agonistic peace also stresses the importance of altering relations of power between the former antagonists through the agonistic struggle. 22 The creation of agonistic peace is a very fine balancing act between the need for political unity and popular demands for plurality. 23 All participants in the unfolding peace process must respect a plurality of views regarding the shaping of institutions and rules, meaning that 'all is contestable', but the very framework put in place by peacemakers for negotiating and developing relations in war-to-peace transitions cannot and will not be overhauled. 24 In this way, the agonistic approach to peace is both more and less ambitious than traditional approaches. Less ambitious because it accepts and even encourages conflict to continue non-violently but also highly ambitious in its requiring of a certain level of openness to difference. The latter process is highly unsettling and a 'riskful engagement' as it tampers with perceived boundaries of identitylaying bare our inherent vulnerabilities and insecurities. 25 This points to not only the psychologically challenging character but also to the intrinsically politically charged nature of these endeavours. 26 While the present analysis focuses on how to inscribe agonism into peace agreements, it is nevertheless relevant to consider that 'agonism itself unfolds within a larger political dynamic that informs the very terms of its encounter.' 27 Thus, questions arise of who has the power to structure the peace negotiations and who gets to determine if and how agonistic principles are written into peace agreements. Therefore, agonistic respect among former enemies should not only be the outcome of a peace agreement, but an integral part of the overall peace process. Subsequently, agonism should be understood as an ontologically shared commitment between former antagonists on societal as well as political level. While power relations are deemed to be omnipresent, agonism describes a relationship of simultaneous recognition and struggle. 28 A key term here is subjectivity indicating that there can only be power relations if 'the other' is recognised as a subject who is able to act and resist against the hold placed on him/her and to open up new possibilities within that struggle. 29 Thus a 'game of power [played] with as little domination as possible'. 30 While a variety of aspects of agonistic peace have been theorised, 31 little research has investigated the translation of agonistic ideas into practical propositions. With this article we take a first 21 Schaap, 'Agonism in divided societies ', p. 272. 22 Michel Foucault, 'The subject and power ', Critical Inquiry, 8:4 (1982); Shinko, 'Agonistic peace: A postmodern reading'. 23 Paulina Tambakaki, 'The tasks of agonism and agonism to the task: Introducing Chantal Mouffe: Agonism and the politics of passion ', Parallax, 20:2 (2014)   in that regard by conceptualising how peace agreements can reflect agonistic ideals and thus lay the ground for an agonistic peace. This conceptual development has double benefits: on the one hand, it can help analyse the degree of agonism in a peace accord, which is a crucial part when assessing implications of the agreement long term; on the other hand, it can serve as vital input for policymakers and practitioners when thinking about alternative routes when drafting peace accords.
Agonistic peace can be investigated on different levels of analysis, from elite to grassroots' endeavours. 32 In this study, we take formal peace agreements as a starting point, which, although negotiated at the elite level, have decisive consequences for the whole society post-accord if/when implemented, including for potential grassroots activities. In that regard, we suggest a primary focus on the 'who' and the 'how' in creation of agonistic spaces in peace agreements. That implies a discussion focused on the groups (for example, minorities, paramilitaries, or armed groups) to be included in peace agreements and the principles according to which they are to interact. Through this theory development we contribute to filling the 'institutional deficit' that for long has been an inherent weakness within agonistic theory. 33 Furthermore, our development of tenets of agonistic peace corroborates knowledge on specific dynamics of that kind of peacea peace that comes into being through contestational dialogue, pluralism, and incorporation of dissent. By doing so, we substantiate the peace concept with tangible elements, in contrast to the sometimes bluntly put positive/negative peace divide, where peace is mainly defined in negative terms, as either lacking direct violence (negative peace) or being freed of structural and cultural violence (positive peace). 34

Agonistic dimensions of peace agreements
In what follows we unwrap three central aspects of peace agreements: agonistic spaces, agonistic inclusion, and agonistic framing. Spaces are material in terms of actual output of the peace agreement, whereas framing is more ideational and may have important consequences for how conflict is understood post-accord societies. Inclusion is both ideational and material in concerning who is included and how those included are framed. The three elements are interconnected but also distinct, and a peace agreement may well 'score high' in one category and not the others. We will demonstrate the utility of each dimension by using examples from Colombia's, Northern Ireland's, and/or Israel-Palestine's peace agreements.

Agonistic spaces
One of the most central elements of an agonistic approach is the proposal of platforms and institutions where conflict can be continued post-accord by means of dialogue and passionate contestation. 35 Institutional development is an underdeveloped area in agonistic research, 36 which too often has been solely preoccupied with the broader ethos and spirit of societal relations. 37 avenues for agonistic peace', Journal of International Relations and Development, 23:4 (2020), available at: {https://doi.org/ https://doi.org/10.1057/s41268-019-00176-6}. 32 Strömbom, 'Exploring analytical avenues for agonistic peace'.  Manon Westphal draws up an intriguing way of thinking about agonism as conflict regulation and how mechanisms of regulation can be understood in relation to institutions. Inspired by that approach, we argue that peace agreements can be drafted and analysed with an eye to how and in what way they provide institutional spaces for agonistic interactions and relations. The goal is not to meticulously spell out how institutions should be designed according to principles of agonism, since there will always be a substantial element of context-related circumstances that need to be considered. Instead, one can distil ideas conductive for the creation of institutions that will increase the agonistic character of the polity from agonistic theory's specific visions for political responses to conflicts. 38 For hegemony to be continuously subjected to dissent and contestation, spaces for interaction must be provided and safeguarded. Formal institutions to be rearranged in the aftermath of violent conflict should include representative bodies on national and municipal levels and arrangements of meetings and negotiation-forums on mid-and grassroots levels, which can bring input to the peace process from social movements, non-governmental organisations, and religious movements. For many years, conflict resolution has diverted its elite focus to encompass multitrack peace processes, where track 1 (elite), track 2 (civil society actors) and track 3 (grassroots) are seen as valuable forums which need to be equally addressed and linked in various ways to create a lasting peace. 39 For a peace agreement to incorporate agonistic features thoroughly and not just at surface level, all these three levels must be addressed and invested with agonistic dynamics.
For an agreement to promote agonistic peace, it is not only important that it enables former enemies to interact but also how this interaction is envisioned, shifting focus to the dialogical characteristics of agonistic spaces. Recently, scholarship has developed the idea of agonistic dialogue. In order for dialogues to be invested with agonistic features, they should be intense, that is, occur frequently, sustained over time, meaning that they should have a long-term and continuous character, and relational, 40 implying that all parties ought to engage in exchanges even though they might not at all times agree with the content or direction of the dialogue. This conceptualisation invites each participant to 'incorporate concerns of oneself and of the other in their own perspective, even when they continue to disagree' 41 and entails the principle of 'always listen to the other side', 42 with respect, no matter how conflictual that claim may be. Agonistic, dissensual dialogue, or 'multilogue' 43 thus harbours a plurality of views and voices and bears the characteristic of being open-ended and unfinished. Hence, people engaging in agonistic dialogues are invited to have a say about the different norms and inclusion principles under which they are to be subjected.
Finally, peace agreements should propose not only formal, political platforms but also ensure space for resistance, such as allowing peaceful protest and activism and providing a space for dissident voices to interact with the political elite level, that is, promoting bottom-up, agonistic public spaces 44 where the conflict and societal relations can be contested and discussed. This conception invites us to think of space not as singular but as a multiplicity of discursive areas where contention can be carried out. Among these spaces, there is neither predetermined centre nor an underlying principle of unity. In this way, a variety of demands can be turned into claims that challenge existing relations. 45 Actors in those agonistic spaces can create a 'chain of equivalence', consisting of a (temporary) connection between actors that have experienced disadvantages, marginalisation, and exclusion by existing power relations due to their individual subject positions within society. Thus, agonistic public spaces can enable the entrance of new counterhegemonic imaginaries into formal institutions.

Spaces in the Colombian Peace Agreement and the Oslo Accords
A prime example of how agonistic spaces can be written into a peace agreement is visible in Colombia. In many ways, that agreement reflects ideals of agonistic peace, especially when it comes to public and institutionalised agonistic spaces. The suggested platforms for continuing conflict post-accord in Colombian society are: (1) the political system, the Congress in particular; (2) the media; (3) dialogue forums where the government is obliged to go into dialogue with social movements. All three agonistic spaces are envisioned as inclusive and participatory, which is promoted through concrete institutional and procedural changes. As regards the Congress, institutional changes have been implemented so that the former resistance movement, FARC, is offered seats in congress and moreover 'political movements and parties can acquire and retain legal status' and, importantly to promote 'the transition of social movements and organisations with a political role towards their establishment as political movements or parties'. Regarding media as a platform for conflict, measures are taken to promote social movements' ability to disseminate information and reach the broader public with their views and perspectives. The agreement stipulates that 'the Government undertakes to provide a closed institutional TV channel designed for political movements and parties with legal status for the purposes of the dissemination of their political platforms within the context of respect for ideas and difference.' Finally, the agreement promises 'guarantees concerning dialogue as a state response to demonstration and protest, through the establishment of dialogue mechanisms and participative spaces …, with mobilisation and protest being addressed democratically'. This means that social movements are promised the right to discuss their concerns with decision-makers. Here it seems that the dialogical principles of intensity and relationality are fulfilled. When it comes to the durability aspect however, the accord offers little information on how long these processes are supposed to last.
In the Oslo Agreements, the linkages to agonism are much more limited, even though they include all three formal tracks (elite, civil society actors, and grassroots) that allow for negotiations on different societal levels. Both the DOP and Oslo II are predominantly engaged with institution building and are seemingly the results of official diplomacy and negotiations between Israeli and Palestinian representatives but also a result of 'backchannel diplomacy', informal and unofficial exchange and contacts between non-state actors. DOP and Oslo II define the task and structure of the Palestinian council, direct the establishment of a strong Palestinian police force, and elaborate on the transfer of authorities from the Israeli military government to 'the council' in specific areas. A second strong focus lies on the economic growth and economic cooperation between Israel and the Palestine Liberation Organization (PLO) as well as Israel, PLO, and the wider region. Several clauses call for creating mechanisms for Israeli-Palestinian cooperation on a wide variety of issues such as water, energy, and labour. While all these initiatives provide institutionalised spaces and platforms for interactions, the phrasing and language of the accords do not indicate a construction of any agonistic public space. Furthermore, the accords do not mention the creation of any platforms designed to pick up on contention or plurality from the public. However, when it comes to the People-to-People programme, the accords provide a slightly different framing. Annex 6 of Oslo II seems to include a relational perspective as it discusses dialogue and cooperation on a civil society level that should 'facilitate the efforts leading to full reconciliation based on the 45 Tambakaki, 'The tasks of agonism and agonism to the task'. agreed political process'. Here, the document mentions cultural and educational cooperation and cooperation in enhancing dialogue and relations between the two peoples. While the DOP only mentions a Human Resource Development and Cooperation Plan, 'providing for joint Israeli-Palestinian workshops and seminars, and for the establishment of joint vocational training centres, research institutes and data banks' and refers to the establishment of a Joint Israeli-Palestinian Liaison Committee (Article X) to deal with issues requiring coordination, other issues of common interest and disputes, Oslo II explicitly refers to the significance of dialogue and interaction to increase peaceful coexistence and encourage 'exchanges in the field of education and by providing appropriate conditions for direct contacts between schools and educational institutions of both sides'. However, there is no explicit reference on how these dialogues and interactions are to be facilitated and whether they are designed to be intense, sustained over time and relational. Furthermore, Article XXII (chapter 4) explicitly states that 'Israel and the Council shall seek to foster mutual understanding and tolerance and shall accordingly abstain from incitement, including hostile propaganda' and that Israel and the council 'will ensure that their respective educational systems contribute to the peace between the Israeli and Palestinian people'. Here, the documents mention mutual understanding and tolerance for the first and only time and only Annex 6 elaborates on its implications. This however does not mirror profoundly agonistic principles. Rather, it displays a thin pluralism in its focus on mutual understanding rather than contestation. Thus, even though there is a focus on dialogue in the people-to-people programmes, the spaces for interactions as expressed in the Oslo Accords bear very little resemblance of agonistic ideals.

Agonistic inclusion
Inclusion has become somewhat of a standard formula when trying to create political orders in the transition from war to peace, 46 and by studying principles for inclusion in peace accords, we can discern whether these are likely to increase the agonistic character of relationships. However, inclusion offers no inherent references with regards to whom to include or in which ways inclusion should be carried out. 47  inclusion. An open understanding of inclusion implies ambiguous, universalistic references to the groups to be included in the peace, such as for example 'conflict parties', 'stakeholders', or 'local actors'. There are no self-evident members of such groups and the collective identification referred to as context dependent. Often these sorts of references are used to build legitimacy and as a mean to achieve that end. In one sense, open framings can be seen as favouring agonism as no specific group or collective is left outside. On the other hand, establishing platforms of interaction and safeguarding that all interest groups have had the possibility to partake in these means that one has to face the task of specifying the groups that should be engaged in institutions post-accord. 49 One deficit with open references is thus that they may be too vague in pointing out the groups to be included to safeguard ideational pluralism. The notions of pluralism and inclusion in agonistic thought go beyond a thin liberal approach of simply allowing for multiplicity, entail challenges to prevailing recognition forms, 50 and are more generous as they open up for a view in which adversaries are respected and maintained in a mode of 'agonistic mutuality' 51 . This implies a far more dynamic and ambiguous conception of social interdependence than any passive, tolerant 'letting the other be' 52 as in more liberal understandings of the concept. 53 Sometimes references are instead made to closed reference groups, identified according to such categories as ethnicity, gender, or religion. This type of referencing is often related to the safeguarding of certain group rights for the sake of empowerment and protection. 54 Defenders of such formulations of identity in peace agreements often claim 'strategic essentialism' 55 in order to strengthen, for example, a marginalised group or emancipatory project. However, such formulations also risk losing sight of differences and varying interests within groups and obscure the fact that the closely identified group might share certain and important traits and/or identities with other collectives excluded from that specific identification. 56 This type of identity framing essentialises identities in its disregarding of the notion that identities are malleable and in constant flux. Agonistic theorists would underline the need for a plurality of voices and interests in order for agonism to be possible. However, if groups and their interests are essentialised, then agonism's transformative potential risks being lost, and the possibility to disrupt hegemonic identity constructions vanishes. In the agonistic conception of relationships, the idea of identity as 'wholeness' must be overcome, 57 since it may imply essentialised identity constructions that might stand in the way of agonism, with a risk of impeding the imagination of a potential plurality beyond the existing social order. 58 Lastly, relational principles for inclusion are the most conducive when it comes to agonism. This type of identity reference considers power relations among actors and formulates strategies of inclusion based on that understanding. According to this logic, collectives should be framed in terms of their relationships to other groups. Such formulations often derive from the sociopolitical context and can be expressed in terms of, for example, 'marginalised actors' or 'underprivileged groups'. This formulation of identity in peace accords emphasises that conflict resolution must not only take direct violence seriously, but that it is as important to also address structural violence that is often seen as underpinning conflict. 59 This relational view of peace rejects essentialised understandings of identities and contest that specific political interests would be linked to certain collectives. Instead, the fluctuating nature of identities is stressed together with the fact that interests cut across boundaries of for example gender, class, and ethnicity. To capture the complexity of identity formation, peacemakers must make sense of deep-ranging sociopolitical aspects of conflicts, challenging asymmetric power-relations. 60 Since it illuminates power imbalances and opens up for their possible transformations, the relational view is thus seen as conducive to societal transformation, from which linkages to antagonisms channeled into agonistic relations can become visible.
The relational aspect of inclusion is very closely related to an intersectional approach to peace. 61 It investigates mutual interaction between groups but also how power is differentiated within identity groups, emphasising power dimensions inherent in peace agreements. Studying intersectional identity frames within peace agreements, 62 implies searching for the mentioning of, for example, aspects of gender, class, and ethnicity within the remits of one peace agreement, where these categories are non-essentialised and potentially overlapping. If that is the case, it should contribute to increasing the agonistic character of the imagined polity. Moreover, an intersectional approach to peace accounts for inequalities along multiple identity lines embedded in the interplay between different power systems. Therefore, peace agreements that neglect how people are affected by power structures based on their multiple identities cannot be considered conducive if the aim is to envision a move towards agonistic relations. Importantly, the concept of intersectionality does not only refer to gender but includes many more facets of identity and difference, such as class, nationality, disability, and/or sexuality. 63 Moreover, it is equally crucial to acknowledge the multiple and intersecting boundaries along which various and fluctuating identities coalesce and are contested. 64 According to that logic, peace processes and documents should make inclusion of previously unconsidered identity constructions possible. Below, we exemplify these dynamics in the Oslo Agreements and Belfast Agreement.

Inclusion in the Oslo Agreements and Belfast Agreement
In the Oslo Agreements closed identity framings seem to prevail. Group membership is identified by nationality or ethnicity. Israelis and Palestinians are the only group categories that are mentioned in the DOP and Oslo II. Except for the Annexes (3 and 4 in the DOP and 6 in Oslo II), the mentioned identities are mainly related to the political/elite level, excluding civil society or any other potential differences and varying interest within the respective groups.
The Belfast Agreement is quite ambiguous in terms of the identities that should be included in negotiations and postconflict work. On the one hand, the agreement is generous with open identity framings, on the other hand, at the times when closed identity framings are mentioned, they are quite dualistic with a focus on Unionist and Nationalist identities. When reading through the document it seems that identity-framings are predominantly expressed in open terms with some substantial exceptions. The agreement continuously emphasises the necessity of cross-community support for all new institutions and decisions taken by those. However, it does not specify who is to be included in the notion of cross-community indicating an open identity framing.
The need for integrated education and mixed housing is stressed several times and indicates an ambition to invest in institutions such as the school-system and municipal planning-projects. Again, we are not told which groups or types of education are to be integrated or how or for whom housing is to be mixed. It is moreover stressed that the police service should be representative of the community as a whole, without pointing out any specific identities or groups that need such representation. The open identity-framings also prevail in linguistic termsthe agreement specifies that linguistic diversity and various ethnic communities should be viewed upon with tolerance, respect, and understanding, but no specific languages or ethnicities are pointed out. However, the identity aspect of nationality is frequently mentioned as an important one. The text refers to the birthright of 'every person born in Ireland to be part of the Irish nation'. Again, it is not specified if those belonging to the Irish nation should claim belonging based on ethnic or civic traits, yet nationality seems to be a strong marker of identity and within this broad identification lays the possibility to harbour plural and possibly intersecting identities. Since the Irish nation is described in open terms, it is hard to distinguish who does and does not belong to that identity category, which goes well in line with vagueness in pointing out the actual collectives to be included in the post-accord society. The agreement has a whole section devoted to the creation of a victims' commissions and the importance of acknowledging the suffering and meeting the needs of the conflict's victims of violence. Again, there is no pointing out who the victims are or any effort to evaluate who have suffered the most which could aggravate some, but it also leaves the question up for discussion, and could definitely be ground for heated agonistic exchanges within the framework of the victims' commission.
Thus, the Belfast Agreement mainly uses open identity markers, however, throughout the document several overlapping identity frames are also expressed. For example, when it comes to more institutional arrangements it seems to favour the conflict's main parties in terms of Unionists and Nationalists, which is a closed identity framing. However, under the heading of human rights, the text stresses that public authorities in Northern Ireland should 'carry out all functions with due regards to the need to promote equality of opportunity in relation to religion and political opinion; gender; race; disability; age; marital status; dependants; and sexual orientation. Public bodies would be required to draw up statutory schemes showing how they would implement this obligation.' Here, it is clear that the parties who have signed the treaty are aware of many different identities and interests besides the ones which might obviously be channelled from 'unionist' or 'nationalist' political aspirations. Moreover, intersectionality is underlined in several passages, even though it is unclear how exactly intersecting identities could be expressed in political interest groups, for example when the parties affirm in particular 'the right to equal opportunity in all social and economic activity, regardless of class, creed, disability, gender or ethnicity', and 'the right of women to full and equal political participation'.

Agonistic framing
Since a peace agreement essentially consists of words, it is highly critical which words are chosen and how the conflict issues at stake are described. While peace agreements are commonsensically understood to put an end to conflict, an agreement enabling and laying the ground for agonistic peace should preferably be formulated in more open terms where the possibility of continuing conflict nonviolently is present. Quite often, peace agreements use wordings like 'ending conflict' and the like, suggesting a formal starting-point of 'peace', when conflict is supposed to be terminated. While merely being a rhetorical step, when assessing agonistic features in a peace agreement, it may be pivotal to investigate if and how the conflict is described in ways that acknowledge the continuation rather than the termination of the conflict post-accord. This may be especially important if the agreement is communicated to the larger population as this will shape people's expectations about the peace to come. Framing conflict continuation or transformation rather than ending may even potentially make the agreement more digestible for the larger population and potentially bring about less (ontological) anxiety about the future, as the parties can continue to fight for their cause, however by other means.
Likewise, agonistic framings in peace agreements should be analysed with an eye to whether it stipulates consensus, as is often the case within traditional reconciliation scholarship, or if dissensus serves as an ideal, where pluralist and conflicting views might remain over time, feeding and driving further conflict. Along the same lines, peace agreements often offer general frames of the conflict and draw up overarching understandings of future relations. Indicators to look for in this regard are how the agreement stipulates goals like unity, shared, common or mutual understandingswhere the common denominator is an underlying effort in support of striving towards the creation of unified views or narratives of the conflict or political preferences. On the opposite end of the spectrum, agonistic principles entail safeguarding rights to voice dissent or disagreement leading to an acceptance of a multiplicity of potentially conflicting narratives about the conflict.
The latter search for dissensus-oriented goals goes well in line with the plea for pluralism inherent in agonistic theory. 65 Peace agreements can potentially be framed in ways that allow plural perspectives of a given phenomenon, event or decision to be acknowledged. An agonistic, pluralistic approach would enhance possibilities for voicing different views of history as well as different views on root causes of conflict. In that sense, peace agreements can create meeting places where different narratives can be discussed and ventilated, such as victims' commissions or commemoration ceremonies or agencies that are seen as investing the process with agonistic elements. The way the peace accords frame the substance of the conflict in unitary (one view of the conflict) or pluralist (plural understandings of the conflict) terms thus helps discern between drawing up a future where plural understandings of the conflict, politics, and history can be ventilated or a future that holds efforts to create unitary narratives where difference is eradicated. In substantial terms, we can look at how the sources of conflict are described: does the text indicate a common view of the main reasons for the conflict, or are different groups' views taken into consideration? How is the history of conflict described? Is room left for dialogue and contention regarding the conflict's past, or should one common story of its history be established? Here, arrangements for discussion on core issues of the conflict are important to investigateis there space to discuss such issues or are they left aside for later due to their difficult and conflicted nature? Also, what arrangements in terms of dealing with the past are put in place? Are transitional justice procedures mentioned in terms of, for example, compensation of victims, and in that case, what victims? Furthermore, is the creation of memorial forums and the like mentioned and if so, according to which principles?

Framing in the Colombian Peace Agreement and the Belfast Agreement
While the Colombian peace agreement promotes difference in opinion and cherishes participation and critical engagement of social movements, the framing of the agreement in many ways seems to imply that the conflict between the FARC and the government is meant to end with the signing of the agreement. The chapter regarding reintegration and transformation of FARC into a political party and social movement is titled 'End of the Conflict' and numerous times throughout the agreement, the initiatives proposed are said to take place 'at the end of the conflict'. The overall impression of the agreement is that the conflict between the FARC and the government is now solved and the agreed solution is a pluralist and inclusive societyrather than the pluralism and participation being framed as measures to continue the fight and dispute the common future by political means. Perhaps this is related to the practice of crafting agreements where the 'script' of signing a peace agreement somehow presupposes the end of conflict and might be received with wonder and raised eyebrows if the document frankly states that it marks the heated yet political continuation of conflict.
Interestingly, the peace envisioned in the Colombian Peace Agreement is not only conflictual but also 'harmonious'. This tension between, on the one hand, respecting difference and giving space to diversity, and on the other hand, striving towards harmonious relationships is captured well in the phrase 'united in diversity', which is mentioned in the beginning of the agreement as part of the 'new vision of a Colombia at peace'. Similarly, it is stressed elsewhere that the parties are committed to 'the promotion of relations of social harmony and coexistence, on the basis of tolerance and respect for differences, especially differences of thinking, including critical thinking, in order to establish the bases for reconciliation, nonrecurrence, and peacebuilding'. Thus, while critique, opposition, and difference are emphasised, the ideals of harmony and unity also prevail.
The Belfast Agreement is likewise ambivalent when it comes to consensual or dissensual framing. When the principles for decision-making within the newly established British-Irish Council (BIC) are drawn up, it is stipulated that interactions within it should be consensus-oriented. Also, in a passage in the agreement the signing parties pledge to 'endeavour to strive in every practical way towards reconciliation and rapprochement within the framework of democratic and agreed arrangements'. Here, the term rapprochement also points towards an ideal of unification of views, which we associate with a consensus-oriented view on reconciliation. It is furthermore stressed that 'relations should be characterized with mutual respect rather than division'. This indicates an inclination towards co-existence and not eradication of difference, alongside potential conflict, which according to our previous theoretical elaborations can be understood as an agonistic objective. However, the agreement also includes several formulations that possibly clash with the consensus-orientation mentioned above. For example, the importance of acknowledging 'substantial differences between our continuing and equally legitimate political aspirations' are emphasised. This could be interpreted as an aspiration to keep differences intact, which would be quite the opposite of consensus-oriented reconciliatory goals. This also goes for the continued emphasis on respect of differing identities and ethos as an overarching value in all interactions. When reconciliation is mentioned, the accord specifically states that the 'promotion of culture of tolerance at every level of society, is an essential aspect, again underlining the inevitable aspects of co-existence of differences'. This points to the aim of cherishing difference rather than forging unified narratives and identities. It is furthermore stated that the achievement of a peaceful and just society is the true memorial to the conflict's victims of violence. Thus, the agreement has little traces of essentialised or unified memory aspirations.

Dilemmas of agonism in peace agreements
As evident throughout this article, all three peace agreements allow for different degrees of agonistic principles. In what follows, we further discuss the three peace agreements and reflect on more general dilemmas of integrating agonistic principles in peace agreements following from our elaborations above.
First, one such problematique regards power relations in terms of who gets to form the postaccord society. In that sense, the inclusivity of a peace agreement is indeed subject to agonistic power struggle in and of itself. Here, questions arise such as which parties should be allowed at the peace table and which groups should be mentioned in the final text? Analysing agonistic dimensions of any peace process should therefore imply scrutinising power relations and their potential transformation. Likewise, the post-accord situation comes with new power struggles, in which subordination and domination are to be negotiated and rearranged so as to create legitimacy for a new societal order. Agonistic principles such as contestation, dialogue, pluralism, and dissent should therefore be incorporated into the negotiations of the peace agreement in order to alter the overarching structures of power and relations among parties to conflict. The recognition of the subjectivity of the former antagonist alongside an understanding of peace as an agonistic political activity might then be reflected in the nature and structure of the peace agreement. However, whether an agonistic peace agreement is sufficiently able to challenge larger overarching structures of power within which the conflict emerged remains to be seen.
A second, potentially problematic element regards the combination of agonistic and liberal principles in peace agreements. The three accords, perhaps most notably the Colombian one, mix principles pertaining to dissensus and consensus. From a more practical perspective this is highly interesting, as it raises questions of how to understand the character of a peace agreement if it harbours ideals of both dissensus and consensus. 66 In one sense, the agreement then is more ambiguous in terms of agonistic content. On the other hand, it is not necessarily the 66 See also Walsh and Murphy, 'Agonistic transitional justice'. fact that consensus principles undermine the agonistic ideals in the agreement. Peace accords and the practice of peacemaking have a long tradition of consensus-ideals and thus it is not surprising that the language of consensus is quite sticky and therefore prevalent in more agonistically inclined agreements such as the Colombian one. For a more fully-fledged agonistic peace to be realised, forums where key actors can discuss the future of peace, be it on elite or grassroots level are imperative. If those forums are initiated, it might be unproblematic if also some parallel initiatives are set up according to consensus principles. In that sense it is important that dissensus can permeate a peace process, but exceptions from that rule should not be expected to overthrow the whole process.
Third, an interesting element of agonism in peace agreements regards the aspect of leaving issues open for further discussion post-accord. Whereas the Colombian and Northern Ireland peace talks followed the logic of 'nothing is agreed until everything is agreed', the Oslo Accords reversely left core elements of the conflict to be discussed at later stages, and now almost thirty years post-agreement, these issues remain unresolved. In one way, the latter logic corresponds to agonistic principles to not agree on everything but agree on the basic institutions and leave space for further disagreement and discussion of central issues. However, if there are no permanent institutions set up for continued discussions, the act of leaving areas open for further discussion can have very problematic consequences, as in the case of Israel-Palestine, where it has proved impossible to move beyond status quo. Hence, a gradual approach to peacemaking can be understood as closing down agonistic potential by putting a lid on difficult and symbolic conflict issues and, as in the case of Israel-Palestine, destructive dynamics among parties spurred by the peace process may result in difficult issues never allowed to resurface. In that regard, the avoidance of conflicted issues can cause not only the end of agonistic potential but also the end of the peace process.
While the peace talks in Northern Ireland followed the logic of 'nothing is agreed until everything is agreed' during the negotiations, it is famous for its use of constructive ambiguity, that is, the deliberate act of formulating something vaguely in an agreement, first coined by Kissinger. 67 At the time of signing the agreement, the constructive ambiguity was celebrated as a genius way to get everyone on board the agreement but since then, it has engulfed heavy criticism. For example, James Dingley 68 harshly criticises the application of constructive ambiguity in the Belfast Agreement, arguing that 'if an agreement is ambiguous it is not an agreement'. On the one hand, constructive ambiguity holds agonistic potential, since it keeps aspects of the agreement open for interpretation and hence for further discussion and conflictual engagement postaccord. On the other hand, however, if the processes and platforms for continuing discussions on the aspects left open with vague formulations, constructive ambiguity can be very problematic and pose great problems for the implementation of the agreement, since it remains unclear what exactly is to be implemented. 69 Fourth and last, we want to emphasise the conundrum concerning the implementation of peace accords. It is of course not enough that agonistic principles are reflected in a peace agreement for them to be realised post-accord. It takes continuous commitment to dialogue and respect from all included parties as well as a broad acceptance from the people so as to not leave room for peace spoilers to act on possible mistrust that may ruin prospects of further respectful engagements. The implementation of the three peace agreements investigated here all have its problems. Most problematically, in the case of the Oslo Accords, the political efforts at Palestinian statebuilding, have been dominated by very technocratic ideas of institution building that were sketched out in the agreement. While the Oslo Agreements did not open much space for agonistic encounters on different levels, the past developments indicate that possibilities for any agonistic space are ever shrinking. 70 While some instances of agonistic engagement exist, they are unrelated to the elite peace process, if such a process can even be said to persist. 71 On the contrary, the Belfast Agreement has been implemented relatively successfully. Yet, the agreement has resulted in a dualistic rather than pluralistic agonism, 72 and the society remains heavily segregated. On the political level, interactions between the former conflict parties do not reflect agonistic principles in terms of pluralistic, vibrant, and intense confrontations. Rather, contestations about past atrocities and the future status of Northern Ireland are oft set aside in the name of seeking consensual 'common ground' on everyday political issue. 73 While the Colombian Agreement to a large extent reflects agonistic principles, 74 its implementation is not that straightforward. We have seen a relatively successful transition of the FARC, the former guerilla, into a political party as well as emerging agonistic public spaces. For instance, primarily peaceful protests in 2019 and 2021 against the government have been unprecedented in terms of the scope, scale and transformative potential. 75 However, several problems with implementing the Colombian Peace Agreement remain, most notably in terms of exacerbated violence with increasing murders of social leaders, insufficient re-integration of ex-combatants as well as a top-down approach to peacebuilding overlooking local participation. 76 Hence, there are many challenges related to implementing even a quite comprehensively agonistic agreement like the Colombian one. At the end of the day, turning words into practice and transforming an entrenched (culture of) war remains one of the most difficult tasks. Hence, the Colombian case illustrates the limits of agonism and how difficult it is to confront power inequalities even when equipped with radical, agonistic principles. In fact, radical elements in the peace agreement may even be considered one of the problems, as it created a lot of backlashes in the conservative parts of Colombia, not least in terms of resistance to women's inclusion and LGBT rights leading up to the referendum on the peace agreement in 2016. This is considered one of the reasons for the agreement not getting the majority of the votes, which again created problems for the implementation of the agreement, even as a revised version was approved in congress. 77 In that way, radical ideas in peace agreements need to be weighted carefully and preferably framed in manners that resonate with the larger society.

Conclusion
In this article, we have developed indicators for analysing the extent to which peace agreements draw up potential for agonistic peace: spaces for interaction, inclusion, and framing of contested issuesall which can be described in more or less agonistic ways. By so doing, we have attended to the institutional deficit in agonistic theoryespecially by elaborating on, how agonistic spaces 70 can be understood in terms of formalised and public platforms and by bringing in discussions on how processes of inclusion can be invested with agonistic features. Furthermore, our discussions pointed out a weakness in many peace accords: that they do not specifically discuss what principles shall guide interactions once forums like victims' commissions or dialogue groups are set up. Lastly, we suggested a focus on how the conflict and its contested issues are framed. Our illustrative vignettes stemming from the three peace agreements show that our indicators can serve to tease out agonistic content in peace accords, pinpointing their agonistic potential as well as ideals pointing towards more liberal, consensus-oriented ideas, which might create hurdles when trying to implement various parts of the accord after the formal elite negotiations leading up to the accord have ended. Moreover, we have discussed various dilemmas and problematiques of crafting, formulating, and implementing agonism in peace agreements. Important relational foundations for agonism are laid during negotiations and in the accords themselves, but they need to be continuously reenacted for agonism to be sustained in the volatile post-accord phase. We invite further studies on how agonistic principles are implemented in postaccord settings, as well as discussions on the respective weight of certain indicators, pointing to how to measure and assess different degrees of agonism in various peace agreements. Moreover, future research could move beyond elite peace accords and venture into the myriad of peace initiatives on mid-and grassroots levels (sometimes initiated from the elite level) and study how they align themselves in accordance with agonistic principles.