Introduction
After decades of civil war and failed peace agreements in Mozambique, 2025 marks almost six years since the signing of the Maputo Accord for Peace and National Reconciliation (Maputo Accord) in 2019. This achievement was the result of several years of work by a small mediation team facilitating dialogue between the Government and Resistência Nacional Moçambicana (Portuguese for Mozambique National Resistance), or RENAMO, with the goal of achieving sustainable peace. This process represents an increasingly rare example of internationally supported dialogue and mediation leading to peace. This process has been characterized as ‘doing things differently’.Footnote 1 The process centred the parties themselves and was based on the development of strong webs of relationships that enabled trust and progress in ending a long-running civil conflict in Mozambique. This has been characterized as ‘peace listening’ – an approach that stands in contrast to the more conventional approach of ‘peace talks’.Footnote 2
This article explores the key distinguishing features of the Maputo mediation process that allowed it to achieve success in its agreement and implementation during a period in which peace mediation as a tool of conflict resolution is in decline.Footnote 3 Drawing on empirical research conducted with parties to the process, we outline two framing elements that constitute the practice of ‘peace listening’. The first is human-centred and value-based mediation, which centres the needs of the parties. The second element foregrounds the flexibility of peacemaking actors to create an ‘enabling environment’ for peacemaking, which challenged the structural and hierarchical nature of international peace mediation. The paper presents two novel contributions to the field of peace mediation. The first is to present a qualitative case study of the Maputo Accord process in Mozambique based on the perspectives and testimonies of the participants themselves. Secondly, by centring the participants in the research, we highlight the potential of relationality as an underpinning theory of successful mediation. We ask what made the Maputo Accord process ‘different’ from previous attempts in the long and complex history of peacemaking attempts in the country,Footnote 4 and in so doing, we address an ontological and theoretical gap in the literature on ‘Track One’ processes when it comes to relationality,Footnote 5 which has focused predominantly on mediator strategy rather than the experiences of parties in the process. Here, we develop an interdisciplinary theory of relationality for peace mediation. In the broader context of a world in which power and relationships are shifting,Footnote 6 we argue that relationality offers a way of thinking about the future of peace mediation as a ‘space for dialogic learning’ beyond the contested terrain of Western International Relations (IR).Footnote 7
The article first sets out a theoretical framework of relationality, drawing on the literatures of both IR and peace studies to frame a theory of relationality for peace mediation. The section ‘Mediation in Mozambique: state of the art’ provides the background to the Maputo Accord processes, contextualizing it against a longer history of mediation efforts. The section ‘Methodology’ outlines the methodology and the pragmatic approach taken to data analysis. The section ‘“Peace listening” in the Maputo process’ presents the data and identifies the core framing elements of ‘peace listening’ as an approach to mediation. Finally, the section ‘Peace listening and relationality in mediation’ returns to the core aim of the article, which is to articulate the value of relationality as a framework for Track One mediation processes.
Relationality in mediation: towards theoretical innovation
The approach taken by the mediators in the Maputo Accord process was a conscious departure from dominant models of peace mediation. It eschewed both the model of elite bargaining that had dominated previous efforts,Footnote 8 as well as the highly structured, internationally backed comprehensive process that is now common.Footnote 9 In their own words, the mediators prioritized building trust and relationships with the principal actors as a means of allowing a process to develop.Footnote 10 Strikingly, the interconnections between the different layers of the process in Mozambique demonstrate how neither ‘international’ nor ‘local’ approaches sufficiently capture the dynamics of a Track One mediation process. The commitment of the mediators and donors to genuine national ownership challenged entrenched hierarchies in global peace mediation practice, while the relationship between the mediators, principals, and local constituencies demonstrated the interdependence of actors. Combining insights from the separate literatures of IR and peace studies produces a more holistic approach that avoids the pitfalls of focusing exclusively on either the international dimensions of power and hierarchy on the one hand, or on an essentialized ‘local’ on the other.
Relationality in IR
Relational theories of IR highlight how rationalist approaches to mediation – whereby the focus is on actors operating as rational individuals pursing their own interests within a mediation process – exclude examination of how actors behave in relation to each other. As Qin notes, ‘[m]ainstream western IR has to date focused on the individual actor, and in its repertoire there is no such choice by which “relations among actors and their context” can be a meaningful unit of analysis to start from’.Footnote 11 Relationality rejects the premise that actors can be treated as discrete units operating independently of each other, in favour of a theory of interrelatedness in which actors are connected to each other through a series of social relationships rooted in context. Rather than focusing on individuals, relationality therefore focuses on ‘actors in relations’, making the web of relationships within which actors are embedded the primary unit of analysis.Footnote 12 It is these relationships that both enable and constrain action, and which shape how actors will determine whether certain forms of action are rational or not. This web of relationships should also not be seen as a static state of being but a fluid and ever-changing form of process. For Kavalski, ‘[t]he transformative potential of relationalism derives from its suggestion that “things” do not exist prior to or outside of relations – that is, actors/entities emerge in the process of interactions, and their identities and interests will be different in different spaces and times’.Footnote 13
Jackson and Nexon further propose that the actions of individual actors (such as states) need to be seen not as self-contained, but as part of a ‘web of interdependence’.Footnote 14 Shifting the focus from the individual to the way in which they interact in the global space requires a focus on process, acknowledging that change does not always need to be associated with individuals, but may also occur as a result of interactions. Such interactions occur in the context of shared goals or ‘projects’ that ‘create webs of reciprocal implication’ through shared goals.Footnote 15 These interactions in turn create ‘configurations’, or systems of meaning generated by narrative and dialogue.Footnote 16 A relational approach, therefore, sees actors as being entwined in an ongoing process of negotiation of boundaries that shapes behaviour rather than static entities with fixed interests. The so-called relational turn in IR therefore highlights the tendency to see all problems through a Western lens, leading in turn to the application of concepts such as rationality and power to the practice of peace mediation. Relational processes serve to ‘produce and re-produce identities and define and redefine roles’.Footnote 17 Odysseos frames these interactions as coexistence, highlighting the element of ‘withness’ of relationality, an ‘active attitude for ‘being-other-directed’ (meaning being open to work with difference rather than trying to suppress it),Footnote 18 allowing a move towards more equitable relationships. To connect the IR literature with that of peace studies, we draw from Brigg’s understanding of relationality as an ‘ontological orientation’ that places ‘greater emphasis on relations, processes or interactions than on substances, entities or “things”.Footnote 19
Relationality in peace studies
While interest in relationality as a conceptual frame has grown in peace studies research in the last few years, the shift to a ‘relational sensibility’Footnote 20 in peacebuilding practice has been observed in the minutiae of ‘everyday peacebuilding’ practices.Footnote 21 There has been an observable shift, whereby ‘interveners increasingly “facilitate” or advise rather than [direct] or [do], and the intervened-upon are “empowered” to realise their own inherent capacities […] Relationship moves to the foreground, and is increasingly considered necessary for the realization of program objectives and goals’.Footnote 22
This increased interest in relationality reflects and brings together four different strands in the peace literature. First is the impact of the ‘local turn’, which critiqued the unequal relationship and power dynamics between peacebuilders and ‘local’ communities in peacebuilding and sought to place the local at the centre of research and practice.Footnote 23 Second is recognition of the role of hierarchy and power in the production of knowledge and what ‘counts’ as effective peacemaking.Footnote 24 Third is literature that centres on the culture and politics of difference.Footnote 25 Finally, it reflects emerging literature that recognizes the inherent dynamic, networked, emergent, ‘entangled’Footnote 26 and ‘patchworked’Footnote 27 relationships that make up peacebuilding and peacemaking processes. Conceptualizations of ‘relationality’ in the literature therefore range from focusing on relationships between actors, to relations among entities, to complex and messy entanglements. Here, we draw from Brigg’s overview of ontological approaches to relationality to position our own contribution. For Brigg’s, relationality exists along a spectrum from ‘thin’, ‘thicker’, and ‘thick’, each with its own advantages, disadvantages, and tradeoffs.Footnote 28 Thin relationality refers to relationships among entities, such as states, individuals, and organizations, that do not inquire about how these relationships come into being in the first place. A ‘thinner’ understanding of relationships in peace processes focuses on ‘behavioural interaction, subjective experiences of each other expressed in attitudes, beliefs and opinions’, attitudes such as mutual recognition and trust. Trust in this sense is difficult to measure as it requires insight into how the actors think and feel about the other.Footnote 29 For this reason, it is often measured by proxy – through the existence of objectively observable indicators such as small confidence-building measures that signal good faith to the other side. At the other end of the spectrum, ‘thick’ relationality gives conceptual priority to relations over entities and thus embraces a ‘more fluid and fundamentally dynamic understanding of the social and political world of peacebuilding’.Footnote 30 In his concept of ‘entanglement fetishism’ however, Torrent warns against a ‘projection of a wholly relational world’ where actors in a conflict system relate in a fully ethical and effective way.Footnote 31 This projection does not reflect the fragmented and messy reality of peacebuilding spaces.Footnote 32 In this article we seek a middle ground, positioning ourselves ontologically in the middle of the spectrum, to develop a theory of ‘thicker’ relationality for Track One processes that are most often dominated by ‘thin’ conceptions of relationships, or elite bargains. Our ‘thicker’ approach acknowledges that elite relationships between Track One actors characterize the mediation, but demonstrate that these relationships are mutually formed through subjective attitudes, experiences, and accountability.
Peace mediation research: relationships without relationality
Peace mediation is dominantly described as ‘a process of conflict management, related to but distinct from the parties’ own negotiations, where those in conflict seek the assistance of, or accept an offer of help from, an outsider … to change their perceptions or behavior, and do so without resorting to physical force or invoking the authority of law’.Footnote 33 While over 60 years of mediation research has contributed important theories about who should mediate, when, where, and how,Footnote 34 recent research has demonstrated that the dominant form of knowledge production in the field of peace mediation has been that emerging from the field of IR.Footnote 35 This research has centred on quantitative methods, based on strong underlying preferences for rationalist theorizing whereby conflict parties are treated as rational individual actors within a bargaining process. This, in turn, leads to a focus on how mediators can best influence those actors to reach an agreement to end violent conflict.Footnote 36 This focus on rationality and objectivity has filtered down into practice through the emphasis on expertise and technical process design when it comes to mediation. There is a focus on substance – a pre-defined agenda and key issues to be addressed. Despite being ostensibly about communication, peace mediation has become a structured undertaking characterized by both power and governance.Footnote 37 Schomerus, in her ethnographic study of the UN-led Juba peace talks (2006–2008), describes a ‘regulated process that can be captured in models or conducted along legal guidelines’ but which ultimately entrenches a limited and unsuccessful toolkit.Footnote 38 Seen in these terms, peace talks become a forum for dialogue, the aim of which is to arrive at an agreed form of governance capable of managing conflict.Footnote 39
For Qin this represents ‘a familiar structure of international norms that shape and reshape identities, interests and therefore behaviour’ of rational individual units within the system.Footnote 40 The preoccupation with complying with international guidelines comes at the expense of understanding how parties experience the process.
This system itself is not only structured but hierarchical.Footnote 41 The primacy afforded to rational theories of IR is reflected in the preference within the field of mediation for quantitative studies of the success or otherwise of the mediation process, complete with modelling process-related variables. With some notable exceptions,Footnote 42 what has received far less attention has been the role of ideational or ‘worldview’ differences in how parties engage in mediation. These differences have been subordinated to the external demands of peace mediation policy. As a result, scholars have identified the inbuilt inequalities between the external intervenors (mediators and mediation teams) and the parties who are ‘mediated’. For Jabri, for example, the institutionalized norms of peacebuilding place primacy on Western knowledge systems and norms in a way that displaces the agency of non-Western populations (those most often ‘mediated upon’). Often, for the mediated, the script is already written, and actors perform their roles within the boundaries of that script.Footnote 43 Relationality as an analytical lens enables the identification of hierarchies and as such tools or approaches that can help to flatten them by enabling a relational ‘web’ to function. Relationality challenges the idea of a pre-written ‘script’ of interaction in a peace process, and opens up theoretical space for more meaningful engagement between peace process actors. As such, it grounds a theory of relationality for Track One processes that straddles this middle ground: acknowledging the presence of relationships between the ‘principal’ actors, while treating aspects such as trust and ownership as relational.
Mediation in Mozambique: state of the art
Mozambique has a long history of conflict and peacemaking. After centuries of Portuguese colonial rule (1498–1975), attempts to build a post-colonial order resulted in ideological power-struggles between elites, failed attempts at nation-building and regional political spill-over from Rhodesia (now Zimbabwe) and South Africa, resulting in a civil war lasting between 1977 and 1992. The FRELIMO (Frente de Libertação de Moçambique) government, which had come to power promoting communist ideology, became embroiled in an armed conflict with the opposition group RENAMO (Resistência Nacional Moçambicana). The conflict lasted for 15 years and resulted in the deaths of over one million Mozambicans caught between the warring parties, while devastating the country’s infrastructure and economy.
In 1992, a General Peace Accord (GPA) was signed in Rome, brokered by the Italian Catholic lay organization, Community Sant’Egidio. The GPA sought to end hostilities between FRELIMO and RENAMO and bring a lasting end to the conflict. For 20 years following the agreement, Mozambique was regarded by the international community as a successful example of a negotiated settlement bringing stability and development.Footnote 44 However, in 2012, there was a resurgence of violence related to political power-sharing,Footnote 45 the allocation of state resources, and the disarmament programme that had left many ex-combatants ineligible for pensions.Footnote 46 Violence and insecurity in Mozambique continued to increase in the years following 2012, precipitating a number of national and international initiatives attempting to broker peace.Footnote 47
Between 2013 and 2015, there were over 100 rounds of talks facilitated by 5 mediators from Mozambican religious and academic institutions.Footnote 48 A short-lived peace agreement signed in 2014 broke down following violence after disputed election results. In 2016, a new round of discussions were convened by a set of international actors, including regional actors such as South Africa and multilateral organizations including the EU. These took place in the luxury Avenida Hotel in Maputo. These ‘Avenida’ talks sought to address key issues of decentralization of power, including revision of the constitution, and military issues and bring a definitive peace.Footnote 49 However, the talks were ultimately unsuccessful. Their failure has been attributed to the way in which the process was conducted, with an externally driven agenda, a lack of direct communication between the parties, and a lack of confidence in the mediators.Footnote 50 The Maputo Accord process emerged in December 2016, when the President and long-time leader of RENAMO Afonso Dhlakama announced a one-week truce following talks with Mozambican president Filipe Nyusi. From that point, engagement between the two leaders (hereafter referred to as the Principals) focused on direct and continued engagement to build trust and create a favourable environment for dialogue between the two parties.
In February 2017, Nyusi established a Contact Group to support the nascent initiative, chaired by then-Swiss Ambassador Mirko Manzoni. A few days later, Nyusi launched a new phase of negotiations in a meeting with the Contact Group, international experts, and two newly created commissions: a commission on decentralization with a mandate to propose draft legislation and necessary constitutional amendments, and a second commission on military affairs, including disarmament, demobilization, and social reintegration (DDR), as well as verification and monitoring mechanisms. This nascent process was facilitated by a mediation team of Ms Neha Sanghrajka, then representing the Center for Humanitarian Dialogue, Mr Mirko Manzoni, then-Swiss Ambassador to Mozambique, Jonathan Powell, Inter Mediate, and Eduardo Namburete, a RENAMO MP. The team was provided with technical support by a small Peace Process Secretariat (PPS) based in Maputo. In a complex political environment, with the weight of failed peace initiatives behind them, the direct dialogues and consultations between the Principals continued in an unprecedented manner.Footnote 51 Notable milestones included the creation of a joint monitoring and verification team to oversee the cessation of hostilities in April 2017, followed by an announcement from Dhlakama of an indefinite truce (after consultations with Nyusi) the following month. In August 2017, Nyusi and Dhlakama met for the first time since 2015 in RENAMO’s military headquarters in Gorongosa National Park in a historic meeting to illustrate their commitment to the peace process.
The unexpected passing of Dhlakama in May 2018 created upheaval for the process. A long-time central figure of leadership of RENAMO,Footnote 52 his passing precipitated internal party discord, threats of violence, and subsequent leadership transition.Footnote 53 In January 2019, RENAMO officially elected Ossufo Momade as their leader. In this period of political upheaval, constitutional amendments to deepen decentralization were approved by Parliament, and consensus on joint structures for the implementation of DDR was reached, thereby addressing the two key points of division between the parties. The launching of the DDR process accompanying the Memorandum of Understanding on Military Affairs before an official peace agreement was signed was a key milestone. These milestones were supportive in breaking a long-standing deadlock on the way to a conclusive peace agreement, and later the Maputo Process Accord.
In July 2019, Ambassador Mirko Manzoni was appointed as the Personal Envoy of the United Nations Secretary-General for Mozambique to provide good offices to support the facilitation of dialogue between the Government and RENAMO, and towards the signing and subsequent implementation of a peace agreement. Parliament approved an amnesty bill in 2018, and with the DDR process beginning in the Gorongosa district, the peace process culminated in the Maputo Accord for Peace and National Reconciliation (Maputo Accord). The signing of the Accord was witnessed by the President of Namibia and President of the Southern African Development Community (Hage Geingob), the President of Rwanda (Paul Kagame), former Presidents of Mozambique (Joaquim Chissano) and Tanzania (Jakaya Kikwete), the Personal Envoy of the UN Secretary General and President of the Contact Group (Mirko Manzoni), and the Representative of Sant’Egidio Community (Matteo Zuppi) on 6 August 2019.Footnote 54
Methodology
Methodologically, we conducted a thick analysis of the Maputo Accord process through literature review and semi-structured interviews. Firstly, we conducted a desk-based review of publicly available statements on the process. These included media reporting, reports from Security Council debates, public information documents produced by the PPS, and the reflections of the mediators at the Geneva Security Debate in May 2024.Footnote 55 The aim of the review was to identify key themes in how the process was reported.
Secondly, we conducted a series of semi-structured interviews with those who had been involved in the process. Participants were asked to reflect on their own experience of the process, including why they committed to the process, how the mediators related to them as an individual within the process, and why they felt the process was a success. We interviewed a total of 27 participants across 4 categories: mediators, political participants on both sides, donors, and civil society. Mediators are defined as those who played an intermediary role in the process. The parties included representatives of both FRELIMO and RENAMO who had participated in the process. Donors were representatives of foreign governments who had contributed financially to the process and had participated in the financial administration mechanisms. Finally, civil society included both locally based actors who had been involved in the process and international experts who had supported implementation of agreements or who had expertise on the process in its broader context. Interviews were conducted in person during two research trips to Maputo in September 2023 and February 2024. Further interviews were conducted online in December 2023. Participants were identified in consultation with the PPS. This approach was necessary because the aim of the research was to present the process from the perspective of those who had been involved, and access could only be secured through the PPS. Introductions were provided to participants, but PPS staff were not involved in the research process to maintain academic independence.Footnote 56 The anonymized transcripts and field notes from the interviews were coded using MaxQDA coding software.
In analysing the data, we worked with the principle of ‘pragmatism’, analysing empirical accounts of key parties to the process to understand their behaviours and motivations.Footnote 57 We did not seek to establish a fully objective, ontological or scientific explanation for the success of the Maputo Accord process. Nor do we present a wholly subjective epistemic account. By pragmatism, we mean a middle ground between these two approaches, recognizing that knowledge production is a social and discursive activity,Footnote 58 and that useful knowledge is borne from practical and embodied forms of social engagement.Footnote 59 The initial analysis identified a number of key and connected themes. From this process of inductive reasoning, relationality emerged as a useful conceptual framing because of the emphasis it places on actors’ experiences of interactions.Footnote 60 Placing the relationship between actors at the heart of the analysis helps us understand the process from the participant’s perspective, and assess the extent to which a relational approach contributed to the success of the process. As Jarstad et al note, ‘… it is important to study both lived practices of a relationship and the stories the actors tell about [peace] … Such stories shape expectations (and thus trust) as well as filter how the behaviour of the other should be understood …’.Footnote 61
A pragmatic approach does not seek to establish a universalizable template, but rather to help orient us within the relevant field, allowing for both continuity and change,Footnote 62 allowing us to present the findings of a single case study in a way that is insightful and informative without seeking to be either normative or prescriptive in its contribution.Footnote 63 The aim is to present the learning from Mozambique not simply deductively within the confines of existing knowledge on peace mediation, but rather in a way that can ‘simultaneously inform and transform’ the field without prescription.Footnote 64
‘Peace listening’ in the Maputo process
While peace mediation initiatives are most often described using the term ‘talks’, in the case of Mozambique, the phrase ‘peace listening’ was used by one international ambassador to capture the difference in approach.Footnote 65 This stands in contrast to the attitude of protagonismo of previous initiatives, whereby mediation actors sought centre stage in the process.Footnote 66 We conceptualize peace listening in two main ways. The first is the equal valuing of human-centred concepts like trust, ownership, and humility that underpin the web of relationships. The second is the presence of an enabling environment that deconstructs global scripts of international peace mediation. These two factors enabled the operation of a web of relationships between the parties. We identify three axes around which this web was spun. The first was the relationship between the principals and the mediators. This was significant because of the breakdown in trust that had led to the end of the Avenida process. As Cho notes, ‘[w]ith peace mediation being an inherently relational endeavour, whether a conflict party perceives its relationship with an intervening third party in an equal or hierarchical manner has a significant bearing on their attitudes towards a peace process’. Second was the relationship between the principals themselves and their willingness to enter into direct communication. Third was the relationship between the political process (comprising both the mediators and the principals) and wider constituencies. This category, which includes both the political constituencies of the principals and the international donor community resourcing the process. While the process was not multi-track or ‘inclusive’ in the formal sense, the analysis reveals that even apparently exclusionary talks at the elite level are embedded in broader knowledge systems and networks that interface with externally set policies.Footnote 67 Acknowledging the role of these broader forces is crucial for understanding the aims, motivations, and behaviours of the principals. Finally, by defining ‘participants’ as including mediators, donors, principals, and local constituencies on an equal basis within the relational web, we challenge the binarized approach which treats international and local actors as separate units of analysis. We aim for a truly emic perspective from inside the process as a whole, not just one set of actors.
Beyond protagonismo
Peace listening was characterized by three distinct features of the approach: trust, ownership, and humility. In the following section, we discuss each feature through the prism of relationships between actors involved in the Maputo process, demonstrating how the key elements of trust, ownership and discretion, in the sense of process facilitation rather than direction, underpinned the approach.
Trust
While trust is a central term in peace mediation literature, it is often conceptualized within a game theory and rational actor perspective, in which actors engage with each other based on cost-benefit analysis and imperfect information.Footnote 68 Whereas practitioner literature has long focused on confidence-building measures to build trust between the parties, critical and feminist IR literature has focused on building trust from an ethics of care perspective between the mediator and the parties.Footnote 69 Trust, in this sense, is built on those being mediated upon feeling a sense of dignity, agency, and ownership over their process.Footnote 70 In Mozambique, an often voiced criticism of the Avenida process was that it was conducted in the media. A number of interviewees discussed how the public way in which the talks were being conducted, including the leaking of information to the media, was making it impossible to move beyond political statements or to build any trust in the process.Footnote 71 The Maputo Accord process grew out of a desire for confidentiality, for control of the process, and the need for direct communication between the parties. A primary objective in the early stage of the process was therefore to build relationships of trust and establish confidence in the new mediators. Donors noted how the mediators invested ‘years and years of engagement and trustbuilding with the Principals’, to the point that ‘they were on speed dial’.Footnote 72
The key feature of this stage of the process was the initiation of backchannel communications. These were conducted initially between intermediaries who had access to the principals on each side. Over time, the passing of messages between the parties developed into proper dialogue between the two.Footnote 73 At this stage, parties, mediators, and civil society interlocutors are all agreed that the talks needed to be closed to protect them and to keep the parties on board.Footnote 74 In particular, the question of ‘who’ was invited to participate and when was controlled by the parties themselves.Footnote 75 One mediator noted the significance of the level of control of the process, describing a Track One process in which the parties had decided that the only interlocutors were top-level. No one other than the very top level (meaning the President of Mozambique and his counterpart in RENAMO) was authorized to speak to the mediators. This was interpreted as demonstrating the level of trust and commitment that the parties were placing in the process and their desire to deliver peace through direct dialogue free from interference.Footnote 76 An example of trust building was the strategy of all communicating in a single voice. To avoid the destabilizing effect of parties releasing competing statements into an already polarized political environment, all communication from the process had to be negotiated between the parties and released as the position from the process rather than of one individual side.Footnote 77 This approach ensured that the parties were speaking with one voice and reduced the incentives for competing political messages. One interviewee noted that when a statement was to be released, they had to work together with their counterpart on the other side until they reached a final form of words that was consensual. This was described as ‘our speech’.Footnote 78 In this way, the parties were supported to define their own message in their own voice.
As one donor observed, the emphasis on building relationships with the parties gave Mozambicans ‘the space to define what a peaceful Mozambique look[ed] like for them’.Footnote 79 The sense of time and space was a major theme, which meant creating the conditions necessary for a fledgling process to mature. This approach was contrasted with other initiatives, such as Avenida, where interviewees felt ‘there was a tendency to try and rush things before they were ready’.Footnote 80 Peace listening and the process of building trust through communication using intermediaries allowed the mediators to understand the aims of the parties and how to support them to define their objectives in a non-hierarchical way. One interviewee highlighted the effort that had been made by the mediation team to understand the positions of the parties – based on their different concerns – and how to work with them.Footnote 81 Another respondent, an international donor, noted, ‘the two mediators spent a lot of time listening […] and giving the space for both of the parties to speak their truth, to speak their needs, and to build a process around that’.Footnote 82 This approach allowed for sufficient trust to be built to establish a firmer set of ‘agreed rules of the game’ for talks that provided some certainty for parties going into the process.Footnote 83
The increasing trust between the parties was evident in gradual changes in behaviour and willingness to take steps to maintain confidence in the talks. For example, the decision of President Nyusi to travel to the bush to visit Dhlakama, and Dhlakama’s agreeing to leave the bush to meet him, was cited as an example of a changed relationship between the parties.Footnote 84 Similarly, the fact that at times of tension in the talks the parties were willing to make concessions to build confidence demonstrated the improvement in trust and relationships that enabled these risks to be taken. Examples include the pre-Christmas truce that led to the end of violence in 2016, and the Memorandum Of Understanding (MOU) on DDR that preceded formal agreement.Footnote 85 One interviewee noted how the trust had grown from almost non-existent at the outset to a meaningful process of communication in which parties were able to discuss the issues, smiling and laughing,Footnote 86 demonstrating the agonistic interactions of negotiation in a very tangible way.Footnote 87
Ownership
National ownership of peace processes has become one of the most dominant buzzwords of recent decades. Yet despite the apparent desire to place conflict-affected communities at the centre of peacemaking,Footnote 88 many still do not feel heard, listened to, or taken seriously.Footnote 89 This feeling of not being heard stems from an inherent lack of humility on the part of international mediators,Footnote 90 technocratic approaches to peacemaking that create insufficient channels for listening,Footnote 91 and unintended replications of the hierarchies of racism and power entrenched through centuries of coloniality.Footnote 92 In the case of Mozambique, national ownership refers not to the involvement of grassroots civil society in the process, but to the fact that the agenda for the talks was defined by the national political leaders themselves. The process of building trust through closed communication enabled meaningful ownership of the process. What emerged from the interviews was a sense among the parties of an absence of outside imposition when it came to the substance of the talks.Footnote 93 One interviewee highlighted how the process was ‘exclusively driven domestically’,Footnote 94 with another commenting on how they had valued the fact that the talks took place in Mozambique itself.Footnote 95
The move towards national ownership reflected the personal legacies that the political leaders wished to achieve.Footnote 96 Direct communication between the two on ending violence enabled the emergence of a consolidated agenda, which coalesced around two key issues: decentralization and power sharing, the main concern for RENAMO, and demilitarization, which was the priority of the government. This process allowed the parties to ‘construct a common idea of what [they] want as peace’.Footnote 97 Once this vision had been defined, the task of the talks was to implement it.Footnote 98 As one interviewee noted, ‘we were in control of everything because we knew what we wanted’.Footnote 99
The interviews present a clear picture of how the mediators worked equally with both parties to enable safe communication that focused on the future and what could be achieved through talks rather than bargaining over positions on the past. In this way, the approach differed from previous processes in which it was felt that the mediators and international actors had been seeking to enhance their own profiles,Footnote 100 and helped to move the positions of the parties towards consensus-building. The success of this approach was attributed by one interviewee to the fact that the mediators did not have their own agenda, nor were they ‘chasing their own prestige’ in the process, thus allowing them to advise parties if positions were unreasonable without fear of alienating them and losing their position.Footnote 101 The intentionality of this approach is summed up by one of the mediators themselves, who reflected ‘[w]e tried to really do our best … to give the responsibility about the process to Mozambicans directly …. [B]ecause at the end it is national ownership that makes the process sustainable’.Footnote 102
Discretion
Putting the parties at the heart of the process requires de-centring the mediator, and it was this ability to act with discretion that characterized the facilitative approach. In contrast to previous talks where it had been felt that the mediators brought their own interests to the agenda, the most striking feature of the talks to emerge from the interviews was the facilitative style adopted by the mediation team. This approach was rooted in a belief in the centrality of the Mozambican people themselves to the process of making peace. Interviewees highlighted the way in which the mediators offered their support to negotiations and decision-making but did not seek to direct the outcome,Footnote 103 described by one party as helping to ‘inform thinking’.Footnote 104 Experience was shared as a way of helping with discussions rather than presented as prescriptive,Footnote 105 or in other words, ‘listening rather than answering or devising’.Footnote 106
The overall approach was one of support to the process. The mediators did not push the process according to their own timeline but were willing to wait for the parties to be ready. Interviewees spoke of the time necessary for consultation and persuasion,Footnote 107 to ‘retreat to discuss strategy’,Footnote 108 and identify their core demands.Footnote 109 Many respondents referred to examples of mediators and the PPS requiring flexibility and adaptive capacityFootnote 110 – pivoting plans, process, and operationsFootnote 111; waiting for the Principals to make certain decisions (e.g. waiting for weeks on end for signals from Dhlakama to meet in Gorongosa) and then adapting their processes to fit the principal’s needs.Footnote 112
In addition to creating the time for these processes to happen, the mediators were also willing to lean on civil society interlocutors to help build trust in the process. It was acknowledged by interviewees that at times trust in the process was low, for example, when there were delays in implementation of agreements, and that civil society were able to step in to help support the process.Footnote 113 The use of civil society intermediaries was also seen by interviewees as a way of building and mobilizing local networks to help maintain support for talks at the grassroots level.Footnote 114 As one interviewee noted, in mediation, sometimes it is ‘not about the issues but the dynamic – this dictates how much time will be required’.Footnote 115 The success of the process and the ability to keep it moving forward when faced with challenges such as the COVID-19 pandemic was rooted in the relationships between the mediators and the Principals, a dynamic that could not be replicated by replacing personnel, no matter how eminent.Footnote 116 Several respondents commented on the ‘longevity’ of the mediators, in which staying for seven years was necessary in the Mozambican context.Footnote 117 This demonstrates a ‘very human and respectful way of engaging’Footnote 118 that acknowledged the human dimensions of addressing deep-rooted conflict in which trust between the parties has been eroded through years of political conflict and violence, and the breakdown of previous agreements.
Challenging structure and hierarchy
The notion of peace listening was also characterized through distinct process elements that challenged dominant structures and hierarchies in mediation literature and practice. These were: building relationships beyond the official peace table, navigating questions of participation and inclusion in a nuanced and bespoke manner, and creating an enabling environment of donors, operational staff, and other external actors.
Mediation beyond the peace table
The process of building sufficient trust to enable the talks to succeed meant cultivating relationships beyond the formal space of a peace table. Much of the work happened in informal spaces and with a wide range of people, including civil society.Footnote 119 As a result, as one interviewee noted, there was no timetable because the mediators had no control of what would happen.Footnote 120 The mediators were willing to meet the parties where they were, both physically and metaphorically, including very real gestures of crossing the front lines of conflict and travelling to meet the principals.Footnote 121 As one donor observed,
the fact that they both went out in their wellingtons and stayed up on the mountain in the rain … not taking the helicopter in and out but walking up the mountain and walking down the mountain, and interacting constantly for the time that was needed, with not only the RENAMO leadership, but also with the [normal] men who would escort them up and down …. [T]hese kinds of things are part of the trust-making.Footnote 122
A process centred on relationships also requires the ability to be flexible and to respond to the changing landscape within an ecosystem of peace. Both the mediators and the PPS had to adapt to changing circumstances, pivoting plans, processes, and operations sometimes on an hourly or daily basis.Footnote 123 For example, the mediators spent weeks waiting in a rural area with little connectivity or infrastructure for decisions to be made and for signals to be received that Dhlakama was ready to meet.Footnote 124 They also recognized the need for time for consultation and persuasion that was an essential part of constituency building.Footnote 125 This approach belies the idea that the primary spaces in which peace is made are formal or time-bound. This flexibility in terms of where and when mediation takes place stands in contrast to the increasingly stage-managed formal processes more commonly associated with the UN.Footnote 126 As one donor noted,
putting the [process] into the UN system was probably the biggest risk that they took, … going into a system that’s so central in its ways, and having to manage that system while trying to deliver in what is an unorthodox process.Footnote 127
When the Maputo Accord process was brought under the auspices of the UN, the mediators resisted the creation of a large-scale peace architecture around the talks. The fact that the process began locally meant that when it was eventually transferred to the ownership of the UN, it did so without the same level of conditionality that often characterizes international processes,Footnote 128 one of the primary achievements of the mediators.Footnote 129 At times, this created tension around the flexibility of the process and the desire for defined pathways and outcomes. For example, one donor described the difficulties in adopting an approach to the DDR process in which the process of reintegration would have had to begin before demobilization had been completed because of delays in implementation:
… considering how the political landscapes move here, it wouldn’t have worked. [UN requirements] would have been just to tick a box, in my opinion, because the targets were moving and RENAMO were changing their minds and the government were changing their minds and they came up with new ideas all the time.Footnote 130
Another example given was the pressure to devise a completion strategy for the process in accordance with UN procedure. As the respondent asked, ‘how can you have an exit strategy when you don’t know when it’s going to finish? Of course, you have a planned timeline. But we all know that those timelines are never kept, and it’s not possible’.Footnote 131 Ultimately, however, while the process was officially UN-backed, it did not develop into a large UN operation. The primary relationship remained that between the mediators and the Principals without excessive UN intervention.Footnote 132 Notably, some donors felt that the flexibility of the Maputo Accord process enabled them to shine a light on the limitations of a structured process that is risk-averse and not fully rooted in the context. One ambassador explained:
[When we] went out to community, we would come back and the UN was sitting in their capital saying ‘Oh, it’s so dangerous, what are you doing out there?’, and then we were able to tell them what actually was happening, and which just changed the narrative and the way that we were able to engage with UN agencies and NGOs, because we were able to push them to get out of their comfort zone and get out into community and start working.Footnote 133
Bespoke inclusion
While flexibility was a strength of the process, it also raised potential conflict with established norms of peace mediation.Footnote 134 In particular, resisting dominant practices around inclusion created tension between the objectives of supporting the Principals to reach a deal and including the perspectives of broader society. A secretariat member discussed the ‘delicate dance’ needed to balance confidentiality, transparency and inclusion under external pressure from other stakeholders such as the UN.
[T]iming is everything, knowing when it’s the moment to be discreet, and when it’s the moment to be transparent and forward where then you can involve civil society, the media, or other third parties, maybe who wouldn’t be directly involved. So I think that was kind of like a delicate dance they did all the time, every day.Footnote 135
Here, the respondent is addressing directly the closed nature of the talks between the Principals, and the fact that they were so closely controlled. While the model of dialogue between two (male) leaders is increasingly at odds with the norms of peace mediation policy and practice, particularly under UN auspices, the talks themselves were unofficial in their early stages, and so less constrained by normative requirements.Footnote 136 However, as the process progressed, initiatives were supported in other spaces that allowed for the involvement of society more broadly, reinforcing the idea that mediation does happen only in formal spaces. Actions taken at the local level can be essential for maintaining the conditions necessary for talks to succeed. This is most notably seen in the Maputo Accord process in the decision to begin implementing the DDR process before a formal agreement had been signed. The micro-processes of peacebuilding were supported by the work of the small but operationally very effective PPS, which enabled the direct relationship between the Principals. Implementing the DDR programme before a final agreement had been reached was a key feature of the process. Effective technical implementation, as well as clear communication, was needed to garner public support. When asked about why this process had been effective, PPS staff focused on relational elements, citing a culture that was both human-centred and non-hierarchical, instilling a culture of respect within the working practices of the process.Footnote 137 Implementing the DDR process at the grassroots level helped to consolidate trust in the political process at a time when it was weak.Footnote 138
Inclusion is evident in two aspects of this process. The first is the agreement to pay pensions to former RENAMO combatants. While this was a very difficult political decision that took years to negotiate,Footnote 139 interviewees characterized this as a step rooted both in human dignity and in the recognition that the failure to pay pensions had been a key factor in the resurgence of violence. One PPS staff member who supported the process outlined the symbolism of the pension as part of a transition to becoming a full civilian.Footnote 140 The provision of IDs, bank accounts, and pensions was regarded as a crucial gesture of recognition of people who had lived marginal lives for 20–30 years.Footnote 141 In this way, they transmitted a sense of trust to the local communities who had been supporting the ex-combatants in the Gorongosa mountains, recognizing that while trust at the elite level was important, it was also necessary to build confidence in local communities to create peace at the ‘lowest level of oversight’.Footnote 142
The second element is the way in which the DDR process was designed to be gender inclusive. At the political level, the talks involved primarily men. This reflected the fact that direct talks were happening between the two leaders, with very limited numbers of women involved at this level.Footnote 143 It was therefore necessary to find other ways of ensuring that the approach being taken would meet the needs of both men and women. The mediators have described an approach of ‘bespoke inclusion’ in which outside parties were brought into the process to address an identified need.Footnote 144 One such example was the need to ensure gender sensitivity in the DDR process. Civil society partners were invited to play two roles. The first was to provide thematic advice on gender sensitivity – to ‘ensure that the actual processes of demobilization and disarmament had factored in gender sensitivity’ in more technical and logistic terms.Footnote 145 This included practical arrangements such as making sure that ‘female combatants had their needs taken care of. A lot of thought was put into how the assembly areas could be designed in order to meet both women’s and men’s needs …’.Footnote 146 While there are limitations to the approach, most notable that women made up only a very small percentage of combatants, and those in other roles were not includedFootnote 147 – an effort was nevertheless made to ensure both that women were included and their practical needs were met in the DDR process.
An additional initiative to address the risks of reintegration of combatants was seen in the support provided by the PPS for the Peace Clubs. These clubs were run by churches to support dialogue within communities at the local level about the DDR process. One respondent who facilitated these dialogues highlighted the importance of ensuring that women were actively involved. She noted how she would call ‘anyone from the community who’s a woman to come to the meeting, and we will find that we actually have women’s leadership in a community, but if we don’t look for them, they will never come. And I started to do that, to make sure that the women who will make will be part of the community, will be part of the meeting’.Footnote 148
While the talks between the Principals themselves were closed and closely controlled, the broader context in which they took place over time did engage with civil society and with local communities to involve them in the process that was happening.Footnote 149 This extended beyond technical consultation to include a range of non-traditional approaches, such as art, dance, and even the commissioning by the PPS of a peace song. This was interpreted by one donor as identifiable specifically as Mozambican rather than being taken from a one-size-fits-all master plan.Footnote 150
Supporting relational mediation: creating an enabling environment
The third element of ‘peace listening’ centres on the support provided by international actors for the process. The fact that the process began as a bilateral arrangement with Switzerland through their ambassador meant that the space for dialogue could be protected, both in terms of their teams allowing them to devote their full time and energy into the processFootnote 151 – but equally important, the financial support to keep the process going. As noted by one of the mediators, the source of funding was critical to success,
[T]he big advantage was that we were not an NGO, it was a government. So the funding was covered, insured, secured with no problem … we were not in need of any credit or promotion. So we could protect the space for the dialogue completely, no one had a clue.Footnote 152
This was important as it allowed the work to proceed without fear of failure, or as described by the mediator, ‘with neither honour nor blame’.Footnote 153 In addition to building relationships of trust with the Principals, the mediators and their teams had to establish connections with a range of local service providers, from cell phone companies to accountants and banks.Footnote 154 The financial arrangements were set up outside of the existing reporting structures to allow the process to operate under the law and banking system of Mozambique, reducing the audit pressure from governments.Footnote 155 A series of calculated financial risks supported the consolidation of trust in the process, both from the Principals and from donors.Footnote 156
The protection of the process in its early years meant that once agreement was reached, it could be transferred from Switzerland to the United Nations, thereby giving it credibility at the international level.Footnote 157 When the process was transferred to the United Nations, it was placed under the United Nations Office for Project Services (UNOPS), a body with a technical and operational mandate rather than a political one. The role of UNOPS was to ‘enhance implementation capacity and support national priorities’.Footnote 158 There was a very clear dividing line maintained between the political mandate of the Envoy on one hand, and the operational mandate of UNOPS on the other, which also reflected the national ownership of the process and the lead role of the Principals. As one donor noted, ‘my opinion is that the principals don’t necessarily see this as the UN Mozambique effort. They see it as the Personal Envoy has a straight line to the Secretary General. I think for them, it would be ...making it less important by putting it under the UN here’.Footnote 159
A critical component of the Maputo process was the concept of ‘trust-based funding’. A group of international donors created a ‘basket fund’ that was prepared to offer core funding to the process, creating both flexibility and agility for the mediation team.Footnote 160 As one administrator observed,
… [I]f you need to suddenly fly 10 people to a meeting on Saturday afternoon, and it’s Friday evening, a normal procurement process can never give you a ticket overnight, right? So then do something about it. … [Y]ou just have to take responsibility, take risks, and sign something and make sure it’s happening.Footnote 161
An Ambassador who was part of the donor basket fund reflected on how the strategy of adaptability extended to managing donors: ‘people felt involved but at the same time, not interfering in the way that happens with other processes when the donor group or the international community sets their agenda and their way of doing things’.Footnote 162 This reduced the ‘noise of bilateral demands’ on the process.Footnote 163 The inherent logic of the basket fund hindered bilateral competition by establishing shared objectives.
Trust-based administration not only allowed for flexibility and risk-taking, but was also credited with enhancing the overall financial efficiency of the process. As one donor noted, ‘the Maputo process in terms of cost effectiveness was extremely good value for money’Footnote 164 in comparison to large-scale, multi-track negotiations. Another respondent confirmed that it was an interesting model to explore that went against the norm of heavily funded and expensive processes. They stated that even if the process had failed, donors would not have put ‘millions’ into it, due to the ‘relatively light process from human resources, the whole setup, very few international travels. No per diems paid’.Footnote 165 This reality is linked to the value-based approach of longevity and allowing time and space for the mediation, in comparison to a heavily internationalized setup where facilitators fly in and out of the space.Footnote 166 Throughout the process, care was taken to maintain proper standards of accountability and transparency while also enabling a more flexible approach to what needed to be done. This contributed significantly to the agility and, therefore, the success of the process overall.
Peace listening and relationality in mediation
Seeing mediation through the lens of relationality highlights how concepts such as power and normativity can be reframed in ways that allow for the ‘withness’ of coexistence and different understandings of inclusivity. There are, therefore, three key themes that make relationality useful in the context of peace mediation.
From actors to relations
As the Mozambique process demonstrates, our ‘thicker’ conceptualization of relationality can apply to relationships at different levels – from the interactions between political elites to the relationships within local communities.Footnote 167 Interactions between actors occur in the context of ‘projects’ that ‘create webs of reciprocal implication’ through shared goals.Footnote 168 These interactions in turn create ‘configurations’, or systems of meaning generated by narrative and dialogue.Footnote 169 In Mozambique, it was clear that the talks that led to the Maputo Accord had a shared goal, with the Principals, the mediators and the international donor community all committed to the objective of ending violence that created a web of reciprocal obligation connecting them.
These ‘interactions’ can be observed across the changed attitudes and behaviours of the key actors. A striking feature of the interviews was the way in which the parties described a transformed political mindset. Parties on both sides highlighted the value of dialogue as a means for addressing political differences. A true legacy of the process was regarded by the parties as embedding the capacity for dialogue within the political institutions.Footnote 170 The nonviolent engagement facilitated by mediation created the conditions for acknowledging legitimate disagreement and altering the political discourse to emphasize respect,Footnote 171 acknowledgement, and enabling the parties to work together on shared issues.Footnote 172 One interviewee described the need for dialogue as the necessary form of engagement because of the differences between the parties – highlighting the need for peace not only with friends.Footnote 173 Another described how peace was an ongoing process but that they would ‘fight not with guns but with their brains, words and hearts’.Footnote 174 While conflict parties may not have close relationships, they do accept each other as legitimate actors with whom they can work.Footnote 175 This web of relationships should also be seen not as a static end-state of being, but a fluid and ever-changing form of process. Relational processes serve to ‘produce and re-produce identities and define and redefine roles’.Footnote 176 This focus on coexistence allows a move towards more equitable relationships.
Power in mediation
As a research agenda, relationality speaks to one of the most enduring themes of peace mediation literature – that of power. Rather than seeing power as vested in an individual (such as a conflict party or a mediator), relational power exists in the ability to manipulate the web of relationships within which one is embedded. Increasing the relational circles will increase the potential power of an actor. A relational approach governs the relationship between actors rather than the actors themselves.Footnote 177 This manifests through the ongoing process of building trust around shared understandings of norms and morality that stand at odds with a ‘modernist impulse to fix, stabilise, to order and to manage’.Footnote 178 This relates directly to the understanding of ‘power’ or influence that distinguishes relational approaches from traditional rationalist ones.
Importantly, a relational approach does not mean that parties do not have power, but that this power comes from the elements of interaction. Actors do not need to act without self-interest. Rather, these relationships can be manipulated for instrumental purposes, allowing actors to achieve specific goals through interactions within the web of relationships. The relational web thereby creates the means through which each actor pursues their own instrumentalist objectives. For example, it was well acknowledged in the course of the interviews that external actors such as states have their own strategic interests in being seen to be successful at peacemaking and that this can lead to competition between actors. In this regard, Switzerland also had its own interests. However, the decision to adopt the approach of ‘peace listening’ – a facilitative rather than a directive style that prioritized the voices of the parties – allowed for the successful pursuit of this diplomatic interest. The concept of power shifted from one of vertical power relations to a horizontal one.
Another area of particular interest has been the centrality of norms and institutions as driving forces of the discipline. The emphasis in recent years on set forms of process design and normative guidance in mediation reflects a broader trend to rely on institutions and ethical standards to ensure predictability.Footnote 179 Norms have been treated as discrete objects, as travelling packages of moral information.Footnote 180 However, Qin makes the pertinent observation for the field of peace mediation, that actors ‘make and trust rules instead of trusting one another’.Footnote 181 A relational approach emphasizes how in social life ‘shared understandings are not imposed as rules, rights or obligations, but emerge in, from, and through the very process of interaction’ [emphasis in original].Footnote 182 Relationality rejects the idea of a single or ‘gods eye’ view of the universe in favour of more situated knowledge,Footnote 183 rejecting the idea of the authority of the objective outsider or universal norms or concepts.Footnote 184 Rather than the imposition of external concepts as a form of controlling behaviours, relational approaches emphasize the importance of ongoing commitment to relations and active participation in interactions.Footnote 185 While adherence to norms, particularly around inclusion and the participation of women, was viewed as one of the less strong aspects of the process, particularly by international donors, there is nevertheless an argument that normativity in this case emerged more organically from the needs of the process as it progressed – most notably in relation to DDR.
Normativity may therefore look different when seen relationally, embedded in, rather than distinct from, action, by linking ends and means recursively (meaning that the means can influence the ends and not only the other way around); crystallizing through the stabilization of practice rather than independently of it.Footnote 186 When seen through this lens, norms emerge through a processual view of social relationships and ‘normative configurations’ are established, enforced and interpreted through ongoing processes of social interaction.Footnote 187 This fits with the outcome-focused and ‘bespoke’ approach to inclusion adopted in Mozambique.
Centring coexistence
Finally, a relational approach allows for greater sensitivity to historically constructed knowledge and understandings.Footnote 188 Peace listening is a process of mutual complementation rather than competition when it comes to approaches to international interactions. Seeing mediation through the lens of relationality highlights how concepts such as power and normativity can be reframed in ways that allow for the ‘withness’ of coexistence and different understandings of inclusivity. Listening requires ‘freeing dialogue from external authority’.Footnote 189 The goal is not to arrive at a destination of rationally pre-determined rules and institutions, not simply to use mediation to secure acquiescence in the dominant form of political arrangement, but rather to be open to different ways of doing things more akin to a ‘global conversation’.Footnote 190 The rejection of the internationalized approach to mediation in Mozambique should have the effect of opening up a dialogue about how these processes can be better managed in a post-liberal world order. Peace listening as a concept promises emancipation and sovereignty from peace initiatives that rely on global racialized hierarchies and structures, and provides a lens for also deeper examination of what a decolonial approach to relationships between mediators and conflict parties would look like – requiring a higher degree of recognition and respect from the supply side of mediation processes.
Conclusion
The experience of the process that led to the Maputo Accord yields significant insight for the field of peace mediation. By shifting the emphasis from a structured process to one that centred the ability to listen and respond to what was needed, a strong web of relationships was established between parties, mediators, donors, and civil society that enabled progress towards agreement and implementation. Key factors in the success of this approach were the protection of the process through strict control on participation and communication in the early years, as well as the willingness of key domestic and international actors to trust the mediators and to take risks to support them.
There are, of course, caveats to this success story. First is the closed nature of the talks themselves, which reflects in many ways the existing tradition of elite bargaining that has characterized previous efforts in Mozambique. Yet in this context, the culture of negotiation between leaders was cited by one mediator as being key in enabling the talks to happen at all.Footnote 191 Working within the cultural norms made it easier to establish the relationships necessary to challenge and guide the thinking of key parties. However, there was a high degree of agreement across all categories of interviewees that a challenge for the success of the process would be the extent to which the process could be opened up to be more inclusive of Mozambican society more broadly. The absence of peace institutions in which to embed the peace was a particular concern that has proved prescient in light of the contested election result of 2025. It was evident that there is lingering mistrust among political parties in Mozambique; they could not yet be characterized as ‘friends’. Yet all were agreed that the process of engagement that led to the Maputo Accord – including participation in the technical commissions – had left a legacy of capacity for and commitment to dialogue within democratic institutions as the best way to address common challenges such as economic development and peace for all of society. This capacity will be needed, and indeed tested, in the wake of the general election.
We do not make a claim to have discovered a replicable template for mediation, nor to have set out an objective criterion for success. Rather, what the Maputo Accord process demonstrates is the importance of listening and being open to other ways of approaching mediation. It is in the ability to ‘listen’ rather than just ‘talk’ that new learning happened and progress was made.
Video Abstract
To view the online video abstract, please visit: https://doi.org/10.1017/S0260210526101909.
Acknowledgements
The research would not have been possible without the support of the Geneva Centre for Security Policy (GCSP), who hosted the initial debate on this theme and facilitated the research. Our particular thanks to Dr Mark Knight (Associate Fellow, GCSP) and to Ms Fleur Heyworth (GCSP) for their support. The authors also thank the staff of the Peace Process Secretariat in Maputo who hosted research visits and assisted with logistics. A draft of the paper was presented at the Conflict Research Society conference in Edinburgh in September 2024 and the authors thank the participants of that panel for their insightful feedback. The views presented in this article are those of the authors and do not represent those of the PPS or the GCSP. All errors remain the authors own.
Open access
