UN Stabilisation Operations and the Problem of Non-Linear Change: A Relational Approach to Intervening in Governance Ecosystems

In recent years, the United Nations (UN) has increasingly turned towards stabilisation logics in its peace operations, focusing on the extension of state authority in fragile, conflict-prone areas. However, this concept of stabilisation relies upon a series of binaries — formal/informal actors, licit/illicit activities, governed/ungov-erned space — which often distort the far more complex power relations in conflict settings. As a result, UN peace operations tend to direct resources towards state institutions and ignore a wide range of non-state entities, many of which are crucial sources of governance and exist at the local and national level. In response, this article places the UN’s stabilisation approach within a recent trend in peace research focused on the hybrid nature of socio-political order in conflict-affected regions, where non-state forms of governance often have significant and legitimate roles. Rather than replicate misleading state/non-state binaries, the article proposes a relational approach and develops a novel analytical framework for analysing a wide range of governance actors in terms of different forms of symbiotic relationships. It then applies this approach to specific examples in Mali and the Democratic Republic of the Congo (DRC), demonstrating the highly networked power arrangements present in conflict settings. The article posits that a relational approach would avoid many of the false assumptions at the heart of today’s stabilisation interventions and would instead allow the UN to design more effective, realistic strategies for pursuing sustainable peace in modern conflict settings. It concludes that relationality could be used more generally, including to explain the waning potency of the so-called ‘third wave’ of democratisation.


Introduction
The UN has turned towards stabilisation logics in its peace operations, increasingly rationalising its operations in terms of stabilising fragile settings and enabling state institutions to flourish. 1 This is illustrated in the cases of the DRC, Mali, Haiti, the Central African Republic (CAR) and (for a time) South Sudan. Yet nowhere in UN policy or doctrine is the term 'stabilisation' formally defined, despite repeated calls for its clarification Gorur 2016). In practice, the UN's approach to stabilisation has favoured institution-building, extending state authority into areas where the state is seen as either absent or lacking sufficient governance stability capacities, and directing resources into national reform and capacity-building programmes. This concept of stabilisation posits that a strong, effective state is the solution to the major risks facing countries and societies where UN peace operations are deployed. As such, it relies upon a series of binariesstate/non-state, licit/illicit, governed/ungoverned space -which replicate the Weberian notion of the state as a monopoly on the legitimate use of force and exclusive control of governance. While this is especially true of stabilisation, state-centrism is common across peacekeeping.
The unintended consequences of statecentric approaches are most evident in countries with a visibly predatory and/or corrupt state, such as the DRC or South Sudan. UN support to state institutions in such settings runs a high -and relatively well-recognised -risk of elite capture, misuse, and outright theft, potentially fuelling the kind of kleptocratic regimes that tend to perpetuate violent conflicts. While the UN has identified a range of mitigation measures against these risks (e.g., UN 2013), it has not generated a more fundamental questioning of the statist nature of stabilisation. If anything, the UN's approach has been to ' double down' on the state, focusing even greater resources on those countries with the most serious institutional weaknesses and few prospects of transformation in the short term. Today, after more than ten years of experimentation with stabilisation, major UN peace operations now appear mired in intractable conflicts, with little likelihood of achieving their benchmarks for success. 2 What can explain this consistent failure of stabilisation to achieve its stated objectives, and is there a more empirically based approach to such conflict settings?
This article places the UN's tendency towards state-centric stabilisation within the recent upsurge of critical peace/conflict scholarship that calls for greater attention to the hybrid nature of socio-political order in conflict-affected areas (Wallis et al. 2018;Mac Ginty 2013;Staniland 2012). It also argues for less state-centrism and more understanding of heterogeneous socio-political orders, where non-state forms of governance often have significant (and legitimate, in the eyes of the population) roles to play when it comes to addressing the trajectory of a conflict. While this thinking has gained some traction in other areas of the UN's work, there has been little attempt to apply it to the UN's stabilisation-focused operations. Rather than reifying binaries between the state and 'non-state' actors, we posit a theory of relationality that recognises contestation, accommodation, and cooperation between a range of state and non-state actors in the provision of public goods and the production of socio-political order. While disruption of binaries has been a long-standing topic for Foucauldian scholars (e.g., Gruffydd Jones 2013), the introduction of relationality in the stabilisation sphere is an innovation.
To demonstrate relationality as a form of ordering in conflict, the article develops an analytical framework and applies it to two UN peace operations contexts -Mali and the DRC -which were chosen as representative examples of UN-led stabilisation missions. While other cases, such as Haiti and the CAR, could have served to demonstrate many of the key characteristics of UN stabilisation, Mali and the DRC offer important opportunities for cross-case comparison: both are large countries with extremely poor governance in the peripheries, and both are considered instances of large-scale stabilisation by the UN with expansive and innovative authorisation relating to the use of force. The case studies focus on the interdependent, symbiotic relationships between parties to conflict, those who exploit the disorder, and indeed those who are working to build the peace, together illustrating how governance is formed and reformed in fragile settings. The studies are based on a range of primary and secondary data, author interviews at UN headquarters in New York with peacekeeping practitioners and country experts, as well as our own in-country field research interviews between 2016 and 2018.
We argue that UN stabilisation missions struggle to deal meaningfully with this empirical reality, which has important consequences for the way stabilisation is pursued and its likelihood of success. Our approach argues for a better understanding of the networked governance arrangements in which the UN intervenes, including recognition of the ways in which support to the state may work against fundamental objectives of lasting peace. Rather than divide conflict settings into artificial compartments of state/non-state and local/national, 3 a relational approach seeks to account for how power is distributed through and within relationships. Analysing the roles and relevance of alternative sources of authority will allow the UN to go beyond the state-centric orthodoxy that currently prevails and generate more realistic and achievable strategies to address conflict settings. 4 This article is not meant to offer an alternative conceptualisation of stabilisation, nor is it possible to fully apply a relational approach to the complex settings of Mali and the DRC here. Our primary purpose is instead to demonstrate the utility of relationality as a lens through which to understand these conflict settings, in particular by dislocating the state from the overriding centrality of the UN's thinking. With relationality as a starting point, the UN can craft a more reflective, realistic, and strategic approach in settings characterised by chronically weak governance and high levels of violence. 5 More generally, and beyond the scope of this article, relationality offers a potential lens to comprehend the waning power of the so-called 'third wave of democratization' (Huntington 1991), where states may be less and less central to governance in conflict-prone settings.

Western visions of the failed state
Modern concepts of stabilisation cannot be separated from the Western project to establish liberal national institutions in a range of so-called 'failed' or 'failing' states (Ghani and Lockhardt 2009;Rotberg 2004). From the Stabilisation Force in Bosnia-Herzegovina in the early 1990s to interventions in Afghanistan and Iraq in the early 2000s, the overriding objective of stabilisation has been to build and bolster state institutions in areas where the central authority has collapsed or did not really exist previously (Bellamy and Hunt 2015). National doctrines of course vary widely, 6 and there is a palpable difference between the 'hot stabilisation' of military-led counter-insurgency in Afghanistan and the far cooler forms of reconstruction and development touted in other settings. 7 But they all share a common Manichean narrative: state institutions are the legitimate source of authority and order, and must be shored up with force against the non-state forces of disorder and chaos, regardless of how abusive or illegitimate the state may be (Curran and Holtom 2015).
The UN's practice of stabilisation -largely driven by the same Western states that led the above interventions -has followed a similar trajectory. On the one hand, stabilisation has been cast as a positive transition from active peacekeeping towards state reconstruction and development, or 'projects to reduce state fragility' (Gorur 2016: 7). This logic is reflected in the transformation to a stabilisation mission in the DRC in 2010, a move that took place at a time of relative calm where the Security Council was signalling a shift towards institutional support to the state (UNSC 2010). In the CAR, the mission mandate similarly positions stabilisation as a pairing, 'stabilisation and development' or 'stabilisation and reconstruction' (UNSC 2017a: 6), placing it squarely into the realm of state-building, while the Stabilisation and Recovery Section of the UN's Mali mission has described its work as focused on 'state authority restoration.' 8 On the other hand, experts have also frequently referred to stabilisation missions as part of the robust turn in peacekeeping, characterised by the use of force in stabilisation (Mac Ginty 2012). The mandate for the United Nations Multidimensional Integrated Stabilization Mission in Mali (MINUSMA) partially follows this logic, 9 suggesting stabilisation as a counter to asymmetric threats (UNSC 2016;Boutellis 2015), while the United Nations Organization Stabilization Mission in the Democratic Republic of the Congo (MONUSCO)'s offensive use of force in the DRC (the most robust mandate in UN peacekeeping) is seen as a crucial addition to its original stabilisation mandate (de Coning 2018a). This force-driven conception of stabilisation appears somewhat at odds with the more development-focused articulation above and also fails to account for the UN missions that are not called stabilisation missions but which nonetheless use robust force (e.g., South Sudan, Côte d'Ivoire) (Karlsrud 2015). The cacophony grows louder on the ground, where peacekeepers interviewed as to their understanding of the term 'stabilisation' have variously found it to be synonymous with restoration of state authority, protection of civilians, addressing the root causes of conflict, and/or a coordination role across a range of other activities (Gorur 2016).
Despite the widely diverging notions of stabilisation within UN peacekeeping, a familiar thread appears to run throughout: stabilisation's purpose is to support and extend the authority of the state against the many threats posed by illegitimate non-state actors (ibid.). 10 Even sceptics of stabilisation within the Security Council have tended to reinforce this statist aspect of the term. The Russian Federation, following a long practice of defending sovereignty, has often emphasised that national governments should control stabilisation within their own countries, rather than permit external players to meddle in their sovereign space.

Towards a polycentric understanding of governance
There is, however, a large and growing body of evidence running directly counter to the liberal state-building model (Woodward 2017). At a national policy level, resistance to state building has been driven by pragmatism and the desire to extricate from quagmires like Afghanistan and Iraq, where the relatively new term 'good enough governance' implies a speedier exit for international interveners (Grindle 2005;Muggah 2014). 'Good enough' recognises that basic governance transformations can take 40 years or more, and that overly technocratic, top-down interventions in fragile settings have failed to produce the kinds of liberal, democratic institutions envisioned by Western intervention (International Dialogue on Peacebuilding and Statebuilding 2011). Within this, there is a recognition also that 'local governance' (United Nations Development Programme [UNDP] 2011: 12; UN and World Bank 2018) may be a more suitable target for international support, and that participatory processes may generate more legitimate systems and governance modalities than nationally driven ones (World Bank 2011).
Beyond the pragmatism of Western interveners (e.g., Karlsrud 2019), however, a vibrant body of evidence now supports the conclusion that governance -the provision of security and basic services to a population -is not the sole prerogative of states (Risse 2013). Particularly in so-called 'fragile' states -characterised by relatively weak state institutions -a panoply of other actors shapes the governance dynamic. Local chiefs operate as respected judiciaries; criminal organisations regulate the flow of goods and services to communities (de Waal 2015; Cockayne 2016); private companies secure the livelihoods and safety of individuals (Hall and Bierstecker 2002); and rebel groups orchestrate subnational statelets with all the trappings of a government (Mampilly 2011;Podder 2014). These are not discrete systems of governance, but exist in simultaneous modes of overlap, cooperation, complementarity, and competition.
Vincent Ostrom (1999) terms this phenomenon 'polycentric governance' -a system of intersecting orders where formal institutions are complemented and/or substituted for by a range of other entities deeply involved in providing (but also often threatening) the everyday needs of the population. Far from the notion of 'ungoverned spaces' (Clunan and Trinkunas 2010), this work understands that power abhors a vacuum and that conventional notions of state sovereignty are poor epistemes for the hybrid forms of socio-political authority that have developed in areas where the central government does not rule alone (Boege, Brown and Clements 2009).

Disrupting the dichotomies in conflict settings
While hybrid forms of governance exist even in stable settings, armed conflict tends to provoke the most extreme forms of contestation over the right to govern. Indeed, the raison d'être of many of today's conflict parties is to reconfigure constellations of power as a means to exercise authority over different and often expanded territories and populations. Violent conflict triggers large-scale shocks to the socio-political order, mass displacement, rapid distribution of resources, and uncertainties as to the legitimate locus of authority. The fluidity of conflict opens opportunities for bargains to be struck, alliances to be made and unmade, and actors to pool resources and work cooperatively towards shared goals. Even more than in 'post-conflict' peacebuilding and development settings, the landscape of ongoing violent conflict demands a nuanced understanding of relationships between key powerbrokers and governance providers.
Ironically, however, violent conflict also tends to provoke the most stubborn and ubiquitous use of Manichean dichotomies of the state. Territories under the control of non-state actors are called 'black spots,' and groups arrayed against the state are characterised in Hobbesian terms as the forces of chaos and anarchy (Mampilly 2011: 40). In eastern DRC, the concept of 'islands of stability' evokes a sense that state authority is under constant threat by rebel-infested waters (Barrera 2015). The resurgence of the term 'warlords' in Afghanistan and Iraq hints at a recognition of the state-like function of such armed groups (lords). However, these are still posited as anomalies in the eyes of major Western powers, as 'hotbeds of terrorism' rather than emergent -and often highly legitimate at the local level -forms of governance (Malejacq 2016;Victor 2010).
This state-centric lexicon matters for UN stabilisation operations. In Mali, for example, the telling nomenclature used by the military component of UN stabilisation missions of ' compliant armed group' indicates participation in a formal peace process with the government, whereas a 'terrorist armed group' is, by definition, excluded, regardless of its status with the local population. In fact, across a wide range of conflict settings, armed groups tend to be defined by their 'non-state' status, while a range of other crucial powerbrokers (e.g., entrepreneurs, criminal networks, traditional authorities, etc.) are largely ignored (Hofmann and Schneckener 2011). We view this reductive binary not only as inaccurate in terms of its descriptive power, but as a damaging starting point for planning and executing stabilisation interventions in conflict settings. In response, we offer an alternative analytical framework.

Analytical framework: symbiotic war economies
To reveal the modes of polycentric governance in conflict settings, we argue for a relational approach based on the concept of symbiosis. In complex socio-political systems, as in the natural world, relationships 'shape, empower and at times constrain the extent to which social systems are resilient or otherwise' (Hunt 2017a: 217). The nature of the relationships between key actors -rather than their identity, status or resources -matters most in determining how power is distributed within a given society. A community may have a police station, a judge, and an electricity grid, but its residents may receive their sense of security from a rebel group, justice from a chief, and basic services from a criminal cartel.
The relational approach is not new in other disciplines; in fact, it has recently been proposed as a lens through which to understand peacebuilding in post-conflict settings (Brigg 2016). Moreover, the burgeoning field of complexity theory offers enormous potential in examining how socio-political order can change via shifting relationships (Brusset, de Coning and Hughes 2016;de Coning 2018b;Kleinfeld 2015). But while the emphasis on relationships may seem commonsensical to social scientists, the design and implementation of the UN mission's stabilisation strategies have not, to date, considered this approach. In fact, current stabilisation dogma appears to foreclose such thinking from the outset.
To apply this approach to the cases of Mali and the DRC, we draw on an analytical framework composed of four forms of symbiotic relations (see Table 1). 11 The first, mutualism, describes a cooperative relationship whereby each party benefits from the activity of the other. Relations can be obligate, where actors require each other to exist, or facultative, where cooperation is beneficial but not necessary for survival. Both bees and flowers survive from the act of pollination; digestive bacteria and the host animal benefit from each other; community-based armed groups may offer protection to citizens in exchange for resources.
The second category, commensalism, captures situations where one actor benefits while the other is unaffected. Typically, this situation arises where a larger, more powerful entity provides a form of inadvertent protection to a smaller one without gaining anything for themselves. Remora fish ride alongside sharks for protection, without offering anything to the shark; small birds perch on the backs of large herbivores to feed off insects without fear of attack; local farmers use roads built by international oil companies.
The third category, amensalism, reflects relationships where one symbiont is harmed while the other is unaffected by the act of causing harm. These relations appear conflictual or confrontational, often characterised by asymmetrical interactions, but coexistence can be achieved through iterative cycles of learning and adaptation of behaviour. A large tree casts a shadow, limiting the growth of smaller trees nearby; penicillium kills bacteria on bread but is itself unaffected; chemical runoff from illicit mineral extraction pollutes local groundwater.
The fourth category, parasitism, involves one symbiont living off the other, causing it some harm in the process. Ticks, tapeworms, vampire bats, and various forms of fungi are all examples of parasitic relations, where both symbionts have reached different forms of accommodation that allow for the survival of both parties. Likewise, a criminal network may demand rents from community members, or local politicians may graft from community resources.
Finally, predation offers the most familiar form of symbiosis, where one actor consumes and destroys the other. Rather than treat this as a distinct category in our analysis, however, we consider the moments when other forms of symbiosis -amensalism, in particular -tend to become predatory in conflict settings. In situations where armed groups both live in communities and fight with state actors, a form of predatory amensalism may arise.
In this article we adapt these forms of symbiosis into a heuristic to capture how conflict actors relate and interoperate to produce emergent (dis)order, examining the case studies of Mali and the DRC. In both, we offer the beginning of what such a framework would offer -a starting point for a new way to understand how a stabilisation mandate could be adapted to such conflict settings.

Mutualism Commensalism Amensalism Parasitism Predation
Each party benefits from the activity of the other.
One actor benefits while the other is unaffected.
One symbiont is harmed while the other is unaffected by the act of causing harm.
One party lives off the other, causing it some harm in the process.
One actor consumes and destroys the other.

Mali
The conflict system in Mali is complex due to the involvement of so-called terrorist groups, local militias, inter-ethnic competition, and long-standing disputes between regional actors. Both greed and grievance motivate fluid alliances between parties to conflict as well as those who suffer its consequences. For a long time, this web of conflict relations was understood in simple dichotomies, framed by two dominant narratives. The first depicts a monolithic Tuareg independence movement as a threat to the territorial integrity of the Malian state. The second identifies 'terrorist armed groups' as a threat to its monopoly on legitimate violence. While both narratives capture elements of the conflict, they inaccurately present the dynamic in state/non-state terms (Høyer 2013). The periodic conflict analysis contained in UN reports has increasingly recognised a degree of overlap and fluidity across these entities (e.g., UN 2018c and 2019a). Nevertheless, in order to stabilise the situation and fill what are perceived to be governance vacuums, the logic of international intervention has been to restore (or for the first time establish) central government authority over the far reaches of the Malian territory. However, contrary to the idea of 'ungoverned space,' socio-political order in Maliparticularly beyond the urban centres of the south -is the product of linkages between an array of state institutions and ethno-cultural leaders and authorities (Lecocq et al. 2013). Over time, and particularly since the armed insurrection in 2012, this has become intertwined with organised criminal networks and jihadist organisations, as well as powerbrokers beyond Mali's boundaries (Bøås 2015). Rather than fall into strict state/non-state patterns, violent conflict in Mali tends to be manufactured between and across groupings and communities, creating settings where a political leader can be a chief, a warlord, a criminal, and a state actor simultaneously. The following section briefly explores the trends propagating conflict in order to then reveal the symbiotic relationships between key stakeholders.

Separatism/Balkanisation
After multiple failed peace agreements, 12 a coalition of Tuareg groups fighting for an independent Azawad 13 mounted a rebellion against the Malian state in 2012. The Algiers Peace Agreement between the Coordination des Mouvements de l'Azawad, the Malian government, and pro-government armed groups was signed in 2015. However, implementation has faltered, generating growing frustration and resumption of hostilities by some groups that sit outside the framework agreement. Significant figures from within the Azawad independence movement have since aligned or entirely amalgamated with so-called terrorist groups. 14

Violent extremism
The Tuareg rebellion was effectively hijacked by a loose coalition of jihadist groups that controlled large swathes of Malian territory by the end of 2012. While French intervention pushed back their advance, these groups continue to control territory in northern Mali, launching attacks and influencing conflict dynamics in the central regions. 15 While their specific aims and foci vary, they share an interest in the creation of an Islamic state and exploiting instability in Mali to consolidate and grow their operations. This includes involvement in trafficking, kidnapping for ransom, and informal taxation (Hussein 2013).

Organised criminality
Much of the instability in central and northern Mali in the past decade can be traced back to trafficking through the Sahel, including long-standing drug routes from Latin American cartels (see e.g. International Crisis Group [ICG] 2016; Lacher 2012; UNODC 2012). Organised crime, which includes the trafficking of people, weapons, vehicles, cigarettes, and drugs, is a defining economic motivation for armed groups, bandits, and government officials. Indeed, these activities give rise to unlikely alliances and cross-over between these actors (Alda and Sala 2014). Land issues and instrumentalisation of local conflict Disputes between nomadic pastoralists and sedentary farmers over access to land are a long-standing feature of Malian society (McGregor 2017). With population growth and migration, as well as climactic shifts, this competition over scarce resources has become increasingly pronounced in recent years (ICG 2016). Furthermore, local disputes have been instrumentalised by parties to the conflict at the national level, creating interdependent spheres of conflict among communities, armed groups, and political entities (Fulton and Nickels 2017).

Politicisation of violence
Though a feature of Malian patronage politics since independence, presidential elections have contributed to the politicisation of violence. Community 'self-defence' -militia broadly supportive of the government in Bamako -has been used to shore up support and in some cases target opposition strongholds, exacerbating mistrust in government and entrenching political polarisation. Interpenetration between armed groups and the formal political domain increases the potential for political violence but also draws armed elements closer into the governance landscape and calculations.
These trends intersect and continue to evolve, creating new interdependencies and fluid relationships across the state/nonstate, local/national/regional divides. In fact, all of the trends above have transnational effects, where Mali's neighbours influence dynamics within and beyond the country. 16 Taken together, they form the backdrop to the following illustration of several symbiotic relationships in the Malian Sahel.

Symbiotic relationships in the Malian Sahel: The case of Jama'at Nusrat Al-Islam Wal Muslimeen
The above section has followed the traditional approach of identifying key root causes of the conflicts in Mali. However, a relational approach looks also at how these different dynamics intersect, how the identities of different actors are fluid and context specific, and how (dis)order is an emergent rather than a static process. The following offers a brief application of a relational lens to one of the major non-state armed groups (NSAGs) in Mali, Jama'at Nusrat al-Islam wal Muslimeen (JNIM). 17 JNIM is an umbrella organisation formed in March 2017 through a merger of four major players in Mali's NSAG landscape -Ansar Dine, Al-Murabitoun, Macina Liberation Front, and Al-Qaeda in the Islamic Maghreb. Otherwise known as Al Qaeda in Mali, JNIM espouses a jihadist ideology with the political aim of instantiating a caliphate across the Sahel. The group therefore represents a threat to the territorial integrity of the Malian state but is also founded on an underlying ethno-political grievance.
Since the formation of JNIM, which has denounced the 2015 peace agreement, the erstwhile attacks of its constituent members on the Malian Defence and Security Forces (MDSF) as well as on international forces have increased (Joscelyn 2017). JNIM therefore now constitutes the main target for the anti-terrorist operations conducted by MDSF, as well as Barkhane and G5 Sahel, with the support of MINUSMA. 18 While JNIM does not strategically target civilians, civilians are certainly at risk of being caught up in collateral damage in its attacks on security forces (ibid.).
Importantly, the group gains leverage and asserts authority through a range of relationships with other actors. JNIM (or rather its constituent parts and members) are embedded in the communities from which they come. Many members have been recruited locally while 'foreign fighters' have assimilated through intermarriage, making JNIM part of the fabric of these communities with familial as well as economic and political linkages. As a result, they are not typically viewed as 'foreign,' nor are they necessarily seen as a threat (Alda and Sala 2014;Sandor 2017). JNIM may no longer control major cities but it does provide public goods -in particular, offering protection to certain populations from the excesses of the Mali army (Sandor 2017).
Parts of the JNIM grouping have intervened in inter-as well as intra-communal conflicts, instrumentalising local conflicts for their own purposes. 19 This is particularly visible in inter-generational community and clan rivalries between increasingly militant nomadic herdsmen and sedentary farmers in central Mali, over access to land and freedom of movement. The spread of violence to central regions of Mali fuels anti-government sentiment, increases recruitment into JNIM, and stretches the MDSF's limited capacity more thinly. JNIM's involvement in the fighting also provides leverage over emplaced authorities such as the chiefs (McGregor 2017).
JNIM also plays indirect roles in influencing formal politics. Through cooperation and contestation with elements of the independence insurgency, JNIM contributes indirectly to Tuareg secessionism deemed to be an existential threat to the state. 20 At the same time, attacks on the MDSF actively undermine the state in the eyes of locals, highlighting the inability of national authorities to protect civilians (Human Rights Watch 2017; ICG 2016). In this context, JNIM bolsters those parties that reject the state in favour of communal and/or religious self-determination. While on the surface JNIM rejects the formal political sphere as illegitimate, it nevertheless influences politics through its relationships with self-determination groups.
These examples of the linkages between JNIM and other stakeholders in Mali's conflict environment demonstrate the types of symbiotic relations in our analytical framework. First, the relationship between JNIM and the Malian state may appear to be purely predatory: in persistently attacking the MDSF, JNIM is attempting to degrade and disrupt the government, while highlighting the inability of the state to provide basic protections and services. However, a reciprocal benefit for the state has emerged, as the Malian government has used JNIM to justify national security narratives to mobilise political and financial support for its anti-terrorism agenda. 21 Rather than a strictly zero-sum predatory game, the dynamics between the JNIM and the state sometimes appear almost mutually beneficial. In these instances, both sides can achieve occasional benefit through the antagonism. We thus refer to this relation between the state and JNIM as one characterised by predatory amensalism, where both actors can achieve some of their objectives by iteratively degrading each other. This demonstrates that relationships are not static, but can oscillate between different forms over time.
Second, JNIM can be seen as having parasitic, amensal, and even commensal relationships with local communities. On the one hand, it at times instrumentalises certain inter-communal conflicts, supporting some and ostracising other populations, while exploiting these rivalries for recruitment and influence. As de facto governors, JNIM constituents were often abusive to local populations through the strict application of Sharia. On the other hand, reliant on their support and legitimation, JNIM has acted as a parastatal actor, providing public goods and protecting some parts of the population against the excesses of the MDSF (Sandor 2017). In doing so they at least partially work to preserve the overall well-being of the community while at the same time benefiting from its political support and resources.
Third, JNIM has mutualistic relationships with other powerful actors in Mali via organised crime (Kfir 2016). Shared interests in the spoils of smuggling put the terrorists, the separatists, the government-friendly armed groups, and parts of the political elite in an unholy alliance: those with influence over state institutions allow for a certain degree of illicit activity; 22 armed groups acquiesce safe passage through territory they control; and transnational criminal groups provide access to established smuggling routes and distribution networks. This mutualism is obligative in nature, meaning that continued collective enrichment is dependent on the comparative advantages of all involved. The extent to which JNIM may be tacitly colluding with other actors -including the stateillustrates the value of a relational approach to analysing these linkages and their impact on conflict dynamics.
As this last discussion illustrates, JNIM is just one example of symbiosis among governance actors that influences conflict dynamics in Mali, where armed confrontation is but one of many ways relationships are formed. Others include the outsourcing of public security to pro-government militias (Peacekeeping and Stability Operations Institute [PKSOI] 2018), and the relationships between so-called selfdefence militias and local populations. 23 Across the region, relationships, not identity, are crucial determinants of emergent (dis)order. As such, a relational approach is useful precisely because it helps make sense of fluid and shifting identities. JNIM may act as a terrorist group one minute, and as a community protector or political proxy the next. Over time, patterns of relationality emerge that can be analysed and used to predict future behaviour as well as provide the basis for engagement by actors, such as UN stabilisation missions, seeking to transform these dynamics.

UN endogeneity in the Mali conflict system
A relational approach considers MINUSMA as a central part of, rather than external to, Mali's conflict dynamics. In this, MINUSMA's state-centric mandate is crucial, given that the mission's core priorities are assisting the state to displace armed groups, build its institutional capacity, and extend its presence to the peripheries of Mali. In this, MINUSMA and the Malian state have a clear relationship of mutualism: the UN provides logistical support to the army and stabilisation funding to the government, and in return is able to effectuate its mandate.
However, in addition to this strong and direct relationship with the Malian state, MINUSMA also has mutualistic relationships with other key actors. For instance, it is engaged in coordination and informationsharing with the regional and French counterterrorism forces. 24 Association with such forces can negatively impact public perceptions of the UN and its impartiality (Sharland and Novosoloff 2019; Zimmerman, forthcoming), 25 but these relationships also offer advantages. These include the emergency reinforcement offered by the French forces 26 or the strengthening of its own partnership with the Malian government due to triangular cooperation with the anti-terrorist actions of the G5 Sahel. Furthermore, MINUSMA's involvement in the eventual demobilisation and reintegration of combatants sees the UN indirectly supporting both pro-and anti-government elements. MINUSMA also engages directly with traditional authorities through local conflict resolution and reconciliation initiatives as well as some targeted stabilisation programming. 27 Lastly, MINUSMA is ultimately seen by groups like JNIM as the epitome of haram. In their worldview, the UN is the product of an unjust global order that JNIM and other similar groups do not recognise as legitimate. Furthermore, they see the UN's stabilisation missions as 'window dressing' for the counterterrorist agenda and neocolonial interests of Western governments (Chivvis 2015). MINUSMA is therefore pitched into a predatory-amensal relation with violent extremists. The mission has been sent to help the Malian state overcome violent extremist groups and the UN has become part of the raison d'être for groups like JNIM, which outright reject the credibility of the UN. At the same time, continued Western presence in Mali provides fodder for extremists wishing to build on anti-Western sentiment and recruit further into their ranks.
Even a very brief examination of the ways the UN participates in Mali's power networks underscores a crucial point that is often overlooked by UN stabilisation approaches: change does not take place in a linear fashion as ungoverned space gradually falls under the control of state institutions. Instead, UN operations participate in and influence a broad range of relations between different powerbrokers, often falling into the same kinds of symbiotic dynamics we see between conflict actors. This section has illustrated how a relational approach can add value in understanding conflict dynamics in Mali. It highlighted the risks inherent in reductive analytics such as the state/non-state binary when conflict dynamics are the emergent outcome of a highly networked set of actors. However, current approaches to conflict analysis and the stabilisation orthodoxy fail to reflect this relational production of (dis)order. Targeting any one actor in isolation without recognising its interdependencies with others (e.g., tackling the 'terrorists' militarily but not disrupting their involvement in smuggling operations) is like squeezing a balloon. In the absence of a course correction, MINUSMA and the rest of the UN to a large extent are unlikely to address the root causes of conflict and sustainable dividends will remain elusive.

The DRC
Conflict dynamics in the DRC are extraordinarily complex, characterised by highly fluid population movements, rapidly evolving armed group formation, and the most lucrative illicit flows of natural resources in the world. From an international perspective, there is a strong tendency to consider the vast landscape of eastern DRC as an ungoverned periphery, bereft of effective state institutions and subjected to the violence and corruption of local actors (Autessere 2010;Neethling 2014). The international intervener thus understands the core goal for the DRC to be one of extension of state authority from the centre to the periphery, to shape, clear, hold, and build 'islands of stability' that will eventually expand into an archipelago (Barrera 2015). The UN's stabilisation mandate in the DRC reflects this conceptualisation, where the mission is required 'to consolidate State authority in the territory freed from armed groups' (UNSC 2010: Art. 12(p)).
Far from being ungoverned, eastern DRC is in fact one of the richest examples of overlapping spheres of authority and informal governance, where a range of local actors (including armed groups) play multiple influential roles, frequently linked to national-and regional-level powerbrokers (de Waal 2009). In stark contrast to the state/nonstate, governed/ungoverned, and legal/illegal binaries, eastern DRC is better understood in terms of the relationships between these players, and the ways in which influence in the political marketplace is allocated. This section looks at the major trends driving armed group activity in eastern DRC, which provide a context for understanding the symbiotic relationships between some of the most important conflict actors in the country.

Proliferation/fragmentation
The past decade has witnessed a dramatic increase in the number of armed groups operating in eastern DRC, from roughly 20 in 2008 to estimates of more than 70 as of 2015 (Stearns and Vogel 2015). The rise in the overall number is partially driven by the fragmentation of existing groups (particularly the Democratic Forces for the Liberation of Rwanda [FDLR]) but also a continued formation of community-based self-protection alliances such as the Hutu-based Nyatura in the Kivus.

Land and displacement
Much of the fiercest fighting in the past decade has been fuelled by disputes over land tenure and rights of autochthony between the major ethnic groups in eastern DRC (Autessere 2008). While land issues have a much longer history in the DRC, the massive influx of Rwandaphones in 1994 and more recent flows of refugees from Burundi have created new tensions in many areas. In addition, internal migrations -such as large-scale population movements to more resource-rich areas in Northern Lubero and Ituri -have driven inter-ethnic tensions and create high levels of uncertainty over traditional power structures. The pace of such displacement has also grown recently. More than one million people have been newly displaced since the start of 2017 (Norwegian Refugee Council 2017). Day and Hunt: UN Stabilisation Operations and the Problem of Non-Linear Change Art. 2, page 12 of 23

Politicisation of violence
The role of armed groups in eastern DRC cannot be separated from the political trajectory of the country since the end of the Second Congolese War (2002), when provincial governments and political parties were allowed for the first time (Stearns, Verweijen and Eriksson Baaz 2013). Nascent national parties were formed largely on the basis of community and/or ethnic identity, and often secured influence by control of natural resources in eastern DRC. This means that many of the armed groups today, while focused on extracting rents from local communities, are also part of a provincial and national power constellation linked to puppeteers in Kinshasa and other neighbouring capitals. The 2015 découpage process -in which many provinces were broken into smaller administrative units -offered new opportunities for power grabs at the local level (European Commission, Belgian Development Cooperation and UNDP 2010).
The constitutional crisis in the DRC over the failed presidential elections drove President Joseph Kabila's popularity to unprecedented new lows, and has led to a willingness of political actors to provoke violence in eastern DRC (ICG 2017). The surge in violence in some parts of the country was directly linked to the widespread dissatisfaction with the electoral process, with some newly formed armed groups overtly in opposition to the Kabila government (UNSC 2017b). Military operations against armed groups can often be seen through this political lens, as the state security forces attempt to weaken and divide the opposition.

Regional webs
While direct regional interference in the DRC may have become less intense than 15 years ago -when both Uganda and Rwanda maintained troops on Congolese soil -the country's neighbours continued to play an influential role in the political marketplace. In some instances, the link is overt, such as Kigali's role vis-à-vis the M23 rebel group, or the massive quantities of illicit gold trafficked across Ugandan territory every year. But so-called 'foreign' groups like the FDLR have also developed more of a Congolese identity and strong domestic relations, making their roles within the DRC more important than their external sources of support.
These trends combine to create a highly fluid conflict context for the DRC today, one where the relationships between the key powerbrokers are often difficult to fully disentangle, and where the role of the state is often intermixed with a range of non-state actors. However, by looking more closely at some specific cases, it is possible to see symbiosis at play, and to understand how a relational approach could illuminate unseen opportunities for greater stability in the DRC.

Symbiotic relationships in eastern DRC: The case of the Allied Democratic Forces
Armed groups form a key element of the conflict ecology of eastern DRC, and always exist in some form of symbiotic relationship with their surroundings (UNSC 2018b). 28 Adopting a relational approach to the Allied Democratic Forces (ADF) -as just one of many potential nodes in the broader power dynamics of eastern DRC -it is possible to see all three forms of symbiosis at work.
The ADF was founded in 1995 as a Ugandan group based in the Beni territory of eastern DRC, with the goal of ousting Ugandan President Yoweri Museveni and establishing an Islamic state in Uganda. A highly secretive organisation by design, the ADF has kept its internal structure essentially unknown to the outer world for decades, and even its general membership is not apprised of the broader ideological objectives of the group. Over the past 20 years, the ADF has operated in the Beni area, perpetrating some of the most brutal attacks on civilian populations anywhere in DRC and killing hundreds in the past three years alone (UNSC 2018b). In recent years, it has been listed as the highest-priority armed group for the UN and has been the target for the bulk of MONUSCO's offensive military force, which has a mandate to 'neutralize [the ADF] once and for all' (UN News 2015: para. 1).
While there is strong evidence that the ADF was indeed behind much of the most serious violence, the tendency to treat it as a monolithic entity ignores the many ways in which it has become part of the socio-economic milieu of the broader Beni area. Over the past 20 years, the group has become more like a mafia organisation, capable of controlling financial flows, political power, security players, and traditional chieftancies, all through a complex network of local actors. It is therefore more useful and accurate to think of the ADF in terms of how it controls and benefits from influence through different incarnations and identities, rather than its direct risk to civilians.
First, the ADF appears to act in the security realm as a sort of 'police' for a broader range of militant activity in the Beni area. As the most deeply established Beni-based armed group with the ability to back up its position with force, the ADF regulates many of the smaller groups in the area. It does so in part through collaboration with elements of the Congolese army (FARDC), which at a local level has had direct and widely known links with the ADF leadership (see e.g. Congo Research Group 2017; Moloo 2016). FARDC units in the Beni area have at times been permeated with ADF sympathizers -there are even anecdotal reports of Congolese troops leading MONUSCO away from core ADF locations during UN operations -and individuals within the FARDC have reportedly provided some logistical support to the group in the past. High-level elected officials have also been reportedly aligned with the ADF, which has allowed the group to exercise even more control over other militias in the area. 29 The ADF also participates in the highly localised intercommunal dynamics around land tenure and local authority in the Beni area. Over the past 20 years, the gradual migration of the Nande and other ethnic groups has created tensions with those communities that consider themselves the original inhabitants. The ADF has capitalised on this tension, allowing its name to be associated with some of the local groups that are contesting the rights of so-called 'settler' communities. While unconfirmed, some experts believe that a series of so-called ADF attacks on primarily Nande communities in the first half of 2016 was carried out by other armed groups, using the ADF 'brand' to mask their actions (Congo Research Group 2016a).
Relying on its deep set of local alliances, the ADF is an active political player in Beni and was a significant driver of opposition to President Kabila. Some of the ADF's main allies in this are former members of the Popular Congolese Army (APC), the military wing of the political opposition group Rally for Congolese Democracy (RCD-KML). In recent years, the exiled RCD-KML leader, Mbusa Nyamwisi, has reportedly fomented insecurity in the Beni area, using ADF and aligned groups to carry out attacks alongside former APC elements (Congo Research Group, 2016c).
This brief look at the complex web of relationships involving the ADF demonstrates forms of symbiosis in play. Superficially, the ADF and the Congolese state have a confrontational, predatory relationship: the FARDC has been tasked with the destruction of the ADF, while the ADF leaches lucrative natural resources via the illicit timber and gold markets, depriving the state of revenue (UNSC 2015). The extent to which the ADF can undermine the state and continue to control areas in broader Beni is a direct loss to Kinshasa. But there are also gains for each side, allowing the relationship to be understood more as predatory amensalism: The ADF's vicious attacks on civilians offer the government a useful rationale for targeting opposition areas, deploying troops into resource-rich areas, and acquiring material support from the UN; such targeting in turn fuels resentment in the opposition community and may drive recruitment into the ADF and greater popularity for Mbusa Nyamwisi.
Equally, however, the ADF and elements within the Congolese state have a more cooperative relationship based on what could be called parasitic commensalism. Collaboration between the ADF and elements of the FARDC in controlling the illicit timber trade shows a high degree of co-dependence; in fact, it may be complicity by rogue FARDC elements in the illegal market that permits the ADF its degree of capture of resources. Similarly, ADF alliances with the political elite of the Beni area means they are part of both the licit and illicit power structures simultaneously, though the official state line from Kinshasa remains that the ADF is a terrorist enemy that should be destroyed. In discussions with UN officials in eastern DRC, some spoke of the challenge of working with two Congolese armies: one which was wholeheartedly set on destroying the ADF, and the other composed of individual FARDC soldiers who appeared complicit in ADF activities. 30 The mafia-like nature of the ADF means the group also has a parasitic, or even a partially mutualistic relationship with many of the local armed groups and leaders in the area. The ADF gains power by selling its 'brand' and support to these groups, while in turn the groups are permitted to remain active in ADF-influenced areas. The mutual relationship is perhaps most evident in the ties between the ADF and opposition-affiliated ex-APC members. The ADF brand allows them to undertake raids under a helpful nom de guerre, while the ADF benefits from the widespread sense that their influence is far broader than just its core membership. By informally policing armed group activity in this way, the ADF may in fact have a limited stabilising effect, creating predictability and quasi-order among the many groups and factions in Beni.
The ADF is but one example of an incredibly wide variety of symbiotic relationships in eastern DRC. In some, an armed group provides a much-needed good to the public, entering into an illegal, but tolerated, set of relationships with the local authorities to facilitate delivery (e.g., the FDLR's illicit charcoal trade in North Kivu) (Dranginis 2016). In others, a militia will be formed to protect the political and security interests of an ethnic group, thus creating a mutually supporting relationship between a civilian community and an armed group. 31 Common across these is the fact that armed activity is but one of multiple power relationships, often traversing local, national, and regional boundaries, where single groups or individuals often maintain more than one identity.

The UN as a node in the power network of eastern DRC
The relational approach refuses the notion of ' outside' influences and treats all actors in terms of their relationships within a given marketplace. This holds true of the UN as well, which is at times an influential player in eastern DRC. In fact, MONUSCO is deeply implicated in the symbiotic relationships described above, particularly through its stabilisation-related activities.
Returning to the example of the ADF, MONUSCO plays various roles that intersect with the underlying power dynamics in Beni. On its face, MONUSCO is mandated to directly support the state in neutralising the ADF and establishing basic security in the Beni area, entering into a form of mutual symbiosis with the state (including through its US$40 million of stabilisation funding) (UNSC 2014). In these contexts, the UN is an extension and amplification of the state, increasing its monopolising potential in the security realm, helping it to eradicate enemies, and boosting state capacity to deliver basic services to the population. This parastatal role for MONUSCO is often the dominant one in the eyes of the population. Indeed, in recent surveys, more than 60 per cent of the Congolese population thought negatively of the mission's partnership with the Congolese army (Congo Research Group 2016b). But the overt partnership between MONUSCO and the state tends to obscure the more complex roles of and relationships between the mission and key actors in the Beni area. For example, the mission is an employer of large numbers of Congolese citizens, thus entering into a form of mutual symbiosis with the local community. Likewise, MONUSCO engages in a wide range of intercommunal reconciliation efforts, sometimes empowering traditional leaders to resolve conflicts within their own authority and/or bringing political and civil society actors together to address underlying tensions. While formally mandated to support the state, some of these relationships posit the UN in opposition to the state, particularly in upholding human rights in the face of abusive state behaviour.
A final set of relationships concerns the large amounts of money and resources that flow under the stabilisation rubric from MONUSCO to a multitude of Congolese actors. The mission has more than US$40 million in stabilisation funds, which are disbursed under a nationally owned plan. The mission's success -and consequent ability to raise additional funds to maintain stabilisation programming and staff -is measured in large part by its ability to spend these funds on programming. This creates a mutually symbiotic relationship between the institutions (and indeed at times individuals) receiving the funding and the UN, one where progress in the establishment of state institutions is a benchmark of success.
This section has touched on some of the kinds of power dynamics in eastern DRC that can be illuminated via a relational approach. Rather than seeing armed group activity in stark state/non-state terms, it defines armed groups by their relations with other key actors, and by the roles they adopt in these relations. As such, an armed group may simultaneously be a predator, regulator of goods, and provider of protection. Equally, the state may manifest as community protector or predator, as trafficker and counter-trafficker, and as a source of justice while also perpetrating serious human rights violations. More important than these titles is the underlying set of relations that emerges between these actors, and how the basic delivery of governance arises via the networks of power that stretch across eastern DRC. As such, models based on a concept of ' clearing out' armed groups and replacing them with state capacities, may well have negative externalities that could undermine the UN's objectives (Hunt 2017b).

Conclusion
For the UN's peace operations, stabilisation is an attempt to bring order to perceived chaos, to combat what the West sees as the inherent forces of entropy in post-colonial, conflictprone states. The tool for this order is a very particular formation of the state, one largely defined in opposition to the many non-state entities that coalesce in conflict settings. The success criteria for stabilisation missions follow this logic: missions like those in Mali and the DRC will draw down and exit when the state capacity is sufficient to provide basic security and services to the population.
However, as illustrated by the cases of Mali and the DRC, power abhors a vacuum, and governance modalities exist in even the most distant peripheries. Polycentric governance rejects the notions of peripheries and ungoverned spaces and looks instead at relations of power between different actors, whether in a national capital, a foreign country or a rural village. Indeed, Mali and eastern DRC demonstrate the impossibility of compartmentalising groups into centre/periphery taxonomies. The armed groups in Mali and the DRC are simultaneously locally driven, nationally linked, and supported by a range of actors well beyond the countries' national boundaries. A jihadi in northern Mali may be a terrorist to Western donors, but a de facto policeman to the local villagers; a militia leader in eastern DRC might be a spoiler of a peace process in the eyes of the UN, but his group may be the sole source of charcoal to hundreds of villagers. And these identities may be fluid; political leaders can be criminals, criminals can be militia leaders, individuals can contain multitudes.
Ultimately, a relational approach implores interveners to question at least four main assumptions at the heart of the UN's current approach to stabilisation: (1) that the state is the sole legitimate or viable source of governance in conflict settings; (2) that the distinction between state and non-state is clear and fixed; (3) that as institutions are strengthened, risk and instability decline proportionally; and (4) that change occurs in a linear fashion that can be measured in terms of a reduction in violence and the expansion of state capacity.
In fact, this article has offered evidence that all four assumptions are at least partially false: (1) a multitude of non-state actors are seen as crucial, often legitimate providers of governance in conflict settings; (2) the identity of individuals and groups can shift and take on new relationships depending on context; (3) often, the introduction of major institution-building projects has appeared to have no real impact on the levels of violence; and (4) change tends to happen in fits and starts, not as gradual, forward progress.
A relational lens will not solve the wicked problems at the heart of modern conflict or international efforts to stabilise fragile settings. But it offers the distinct advantage of avoiding these dubious assumptions and focuses the intervener on the underlying power dynamics and networked governance providers in a way that avoids reductive, binary approaches to stabilisation. As such it provides the impetus to craft a more reflective, realistic, and analytical strategic approach to the actors involved on the ground. It also has the advantage of positing a far less idealistic end-state for stabilisation missions -eastern DRC will almost certainly never be transformed from lone islands into an uninterrupted terrain of stability under the control of an effective, legitimate state. There is no realistic 30-year plan for Mali that results in state institutions exercising effective control over the full territory of the country. In major conflicts around the world, transformative change from decades of predatory institutions, criminal power networks, and strongman approaches to politics, will not take place within the temporal (or in our view conceptual) framework envisaged by UNSC mandates. 32 Perhaps that is not such a bad thing. Perhaps the Weberian state is a vessel too porous to contain the enormous resources that have been poured into it over the past 30 years. Indeed, now may be the time to acknowledge that the so-called 'third wave of democratization' (Huntington 1991) has begun to fall back on the shore, and that a new conceptualisation of order and disorder is needed. This may require a fundamental rethinking of the principles underlying Western intervention in conflict and postconflict settings. If nothing else, it is certainly time to re-examine the Manichean ideology at the heart of UN stabilisation missions, if they are to be relevant for today's wars.

Notes
1 The UNSC has explicitly authorised four 'stabilisation' missions: first in Haiti (2004), then the DRC (2010), Mali (2013), and the CAR (2014). 2 For example, in the DRC, the UN mission's exit strategy is defined as 'the threat posed by armed groups to civilians has been reduced to a level that can be effectively managed by the Congolese State' (MONUSCO 2016: 9-14). 3 UN peace operations, for example, divide their work between civil affairs (focused on local dynamics) and political affairs (focused on national actors), often with little interaction between the two sections. 4 It is worth noting that the High-level Independent Panel on Peace Operations pushed for more people-centred approaches. However, this is the aspirational aim of an independent panel, not a manifest reality in UN peace operations. 5 The authors are grateful to one of the anonymous reviewers for help in formulating this insight. 6 For a comprehensive analysis of Western doctrines on stabilisation, see, for instance, De Spiegeleire et al. (2014). 7 The phrase 'hot stabilisation' is described in Curran and Holtom (2015: 4). 8 See https://minusma.unmissions.org/ en/stabilization-recovery. 9 Albeit as an adjustment to the mission mandate in 2016, not as originally mandated. 10 Gorur (2016: 5, 12 and 21-25) explicitly advocates for the notion of 'legitimate authority,' not necessarily the government of the day. 11 This analytical framework derives from the biological concept of symbiosis rather than the more abstract or political idea that symbiotic relationships are simply mutually beneficial. 12 Tuareg secessionist insurgencies have occurred periodically since Malian independence in 1962. See, for instance, Manning (1998). 13 Azawad is the name given to the Saharan/Sahelian regions of northern Mali (approximately 60 per cent of Malian territory) claimed by Tuareg separatists. 14 For a more detailed account of the different actors and factors driving conflict and peace in Mali, see Boutellis and Zahar (2017: 21-37