Deliberative Peacebuilding in Somaliland from 1991 to 2005

Abstract

Given the legitimation problem with vertical (state vs. society) and horizontal (modernity vs. tradition) inequalities/differences as historical and cultural causes of conflict, peacebuilding in the post-colonial, post-conflict context requires deliberation to engage in the legitimation problem. Although deliberation has gained academic attention as a means of addressing the ‘legitimation crisis’ in Western liberal democracies (Rawls 1993, Habermas 1996), its application to contemporary peacebuilding remains under-researched. This paper thus aims to theorise ‘deliberative peacebuilding’, applying this to empirical inquiry into the case of Somaliland from 1991 to 2005, and exploring its findings.

Introduction

This paper offers a theoretical and empirical inquiry into deliberative peacebuilding. As colonial rule to construct modern statehood in traditional stateless societies resulted in the legitimation problem with vertical (state vs. society) and horizontal (modernity vs. tradition) inequalities/differences, the post-colonial state often exacerbated these problems as the historical and cultural causes of conflict. Building peace in the post-colonial, post-conflict context will require agencies in conflict to deliberate radicalised inequalities/differences. However, the conventional research and practice of international peacebuilding has highlighted the structural causes of conflict, such as autocracy and poverty, and measures for the so-called (neo)liberal statebuilding which promotes democratisation, economic growth, and consolidation of the state (UN 1992, OECD 2005, World Bank 2011).

Although deliberation gained academic attention when addressing the ‘legitimation crisis’ in
relation to increasing inequalities/differences in Western liberal democracies in the late 1960s and 1970s (Habermas 1976, 1996, Rawls 1993), its application to contemporary peacebuilding remains under-researched. Barnett (2006) is one of the few scholars to have attempted to do so, yet his view is state/elite-oriented in a republican tradition. In Building a Republican Peace: Stabilizing States after War, Barnett (2006) frames deliberation as an institutional means to protect the state from political despotism and legitimise it in ‘the use of the republican principles of deliberation, constitutionalism, and representation to help states recovering from war foster stability and legitimacy’ (2006: 89). Subsequently, Barnett et al (2009) employ this to explore the cause of the prevailing violence in Afghanistan despite the then-President Karzai’s effort for political deliberation with Afghan warlords. Whereas Barnett et al. (2009) relate the cause of this violence to the conflicts of interest between despotic elites, Mac Ginty (2011) links it to the resistance of local non-elites and their exclusion from political deliberation (2011: 108). Their contentious causes (the conflicts of interest between elites vis-a-vis the resistance of non-elites) suggest that deliberation should be embracive of agencies from the modern political to the traditional societal realms, and that it should be re-conceptualised as a socio-political means to address the ‘legitimation problem’ with inequalities/differences for peacebuilding in the post-colonial, post-conflict context in a critical tradition.

In this perspective, this paper will theorise deliberative peacebuilding, apply this to an empirical inquiry, and explore findings from it. For empirical inquiry, this paper will choose the case of Somaliland, a self-declared independent state in Somalia. Whereas British rule resulted in the legitimation problem with vertical and horizontal inequalities/differences, the post-colonial state of Somalia exacerbated them, forcing North-Western Somalis to take up armed resistance in the late 1980s. After the regime’s collapse in 1991, Somaliland mobilised internal forces to deliberate inherited inequalities/differences, and concluded the first electoral cycle in 2005 in an orderly fashion despite the repetition of violent conflict and ceasefire in between. Although the review of the Somaliland case from 1991 to 2005 is brief and selective, given the detailed analysis which I have made elsewhere (Nakagawa 2016) and due to spatial limitations here, the evidence indicates the overall applicability of deliberative peacebuilding to the case of Somaliland, as well as the empirical and theoretical findings, including the actual causes of repeated conflicts and ceasefires and peace thereafter in Somaliland, and the normative merits and practical challenges of deliberative peacebuilding. This paper thus firstly theorises deliberative peacebuilding, secondly reviews deliberative peacebuilding in Somaliland, and thirdly discusses its findings.

Theorising deliberative peacebuilding

This first section theorises deliberative peacebuilding. It firstly unpacks the legitimation problem with vertical and horizontal inequalities/differences in the post-colonial context, secondly re- conceptualises a post-colonial form of deliberation to address the legitimation problem, and thirdly presents a matrix to measure the progress of deliberative peacebuilding given the nexus of power dynamics between deliberating agencies and their choice for approaches to deliberation. The legitimation problem is related to its post-colonial historicity and culture. Western colonisation, which sought to build modern states in traditional stateless societies in the non-Western pre-colonial context, had socio-political ramifications. In order to effectively control indigenous populations, colonisers often adopted indirect rule in which they employed traditional leaders (e.g. kings, chiefs, elders) as their intermediaries to reach the societal majority and introduced a hierarchy between colonisers and colonised (Bayart 1993, Mamdani 1996: 17-19). This asymmetry resulted in a vertical form of inequality/difference between the colonial state and the colonised society at a polity level. In parallel, colonisers often trained indigenous youths as local bureaucrats and soldiers to economise the running of the modern colonial state (Young 1988). Local education and employment, however, created a modern stratum in the traditional segments of society, and generated a horizontal form of inequality/difference between modernity and tradition at an agency level (Bayart 1993). The colonial subjects, especially the emerging modern elites, were, however, not always passive, but often active in collaborating with or resisting the colonisers to meet their interests (Bayart 1993). Despite contextual variations, the state and modernity, and the society and tradition, became increasingly interactive, overlapping, coexisting, yet conflictive as colonisation progressed.

The end of colonisation often reproduced, deformed, and radicalised the legitimation problem with vertical and horizontal inequalities/differences. The modern political elites who took over the post- independent state often imitated colonisers distancing themselves from the traditional societal majority and exploiting the inherited governance system (Bayart 1993, Mamdani 1996). Rampant corruption, for example, indicates that inequalities/differences were reproduced and deformed as
‘traditional-cum-modern’ political elites and ‘modern-cum-traditional’ societal non-elites exploited economic rent and societal relationality in post-colonial politics (Lemarchand 1988). The escalating crisis of the post-colonial state often led political elites to coerce the societal majority (Ninsin 1988, UNDP 1994), yet also prompted the latter to resist the former through such means as coups and insurgencies (Clapham 1998). As a result, the state/modernity-society/tradition divide became increasingly conflictive and precarious (Kohli et al. 1985, Rothchild et al. 1988, Chabal et al. 1999). Figure 1 illustrates the formation of vertical and horizontal inequalities/differences in spatial and temporal transition in the (post)colonial polity. Colonisation introduced the state into the stateless society, creating the state-society interface where colonisers employed local intermediaries (kings, chiefs, elders) in indirect rule (the left figure). As colonisation progressed, a modern stratum emerged in the traditional society due to the colonial education and employment of local youths (the central figure). The post-colonial state replaced colonisers with local political elites as the new ruler, but often kept the state-society interface for its own interests, despite contextual variety (Englebert 2002) (the right figure). In the meantime, vertical and horizontal inequalities/differences were formed, deformed, and radicalised, multiplying ‘modernities’ and ‘traditions’ across the post- colonial polity. Peacebuilding in the post-colonial context will thus require deliberating the legitimation problem with inherited inequalities/differences as the historical and cultural causes of conflict.

In turn, deliberation has gained academic attention as a means to address the ‘legitimation crisis’ and restore political legitimacy given the increasing economic inequality/difference and its socio- political consequences of societal unrest and authoritarian response in Western liberal democracies in the late 1960s and 1970s (Habermas 1976). Although the Oxford English Dictionary (Soanes et al. 2005) defines deliberation as ‘long and careful consideration or discussion’, political philosophers have viewed it not as a mere discussion, but as a means to address inequality/difference and legitimise politics, including, for Rawls (1993), as a political means for elites to rationalise their differences and achieve an interest-free consensus in a liberal tradition, and, for Habermas (1996), as a societal means for non-elites to formulate a consensus for self- determination in a critical tradition. Whereas both regard deliberation as a means to reach consensus over inequality/difference, they differ over: 1) where deliberation takes place; 2) who deliberates for whom; and 3) how deliberation is institutionalised. While Rawls (1993) argues that: 1) deliberation takes place in the state; 2) rational and reasonable citizens deliberate public good on behalf of the Other; and 3) deliberation is institutionalised by a fair procedure, Habermas (1996) contends that: 1) deliberation takes place in the society; 2) argumentative citizens formulate a societal will through the communicative action; and 3) deliberation is institutionalised in a procedural democracy to convey the societal will to the state. Despite their differences, both Rawls and Habermas situate deliberation in the Western polity, where: 1) the state and society are divided in the Weberian sense; 2) the citizenry is individualistic based upon capitalist norms and values; and 3) democratic procedure is well established. This Western assumption is, however, largely different to the post-colonial reality. As noted above, the post-colonial polity often maintains the spatial dualism between the state and society, turning it to a new socio-political space where political and societal forces compete (Migdal 1994: 27) and negotiate (Cornwall et al. 2007: 1). As agencies are largely collective yet divided along vertical and horizontal lines of inequalities/differences,
including tribal groups and kinship networks, the formal channels of communication between the state and society remain ineffective and dysfunctional. Moreover, despite post-colonisation, former colonial powers have kept their material and ideological asymmetry, affiliating their post-colonial
 subjects with international governance systems (Clapham 1998, Jabri 2013).

Accordingly, the endogenous and exogenous structure in which deliberation is situated in the post- colonial context can be characterised as follows: 1) the state and society partially overlap; 2) the citizenry is organised in a collective fashion but divided by inequalities/differences; and 3) the
polity remains ineffective in its attempts at official communication and subject to external intervention. This socio-political reality frames a post-colonial form of deliberation. Ferme (1999), for instance, observes deliberation to take place in ‘secret settings’ in Sierra Leone, where local chiefs and elders interact with political elites to form a socio-political will in their rural houses, and then normalise it in their kinship structure (Ferme 1999: 164, 174). Hashim et al. (2004) similarly see deliberation in the Zakat system in Nigeria, where local emirs and chiefs negotiate with state officials for the modification of (neo)liberal policies in view of the policy gap between the international norm for economic growth and the local need for social justice (Hashim et al. 2004: 248, 251). These ethnographic studies suggest that: 1) deliberation actively takes place in the state- society interface; 2) political/modern and societal/traditional forces interact, argue, and sometimes reach consensus, yet often fall into tension; 3) the institutionalisation of deliberation requires regular interaction between deliberating agencies; and 4) external intervention remains influential. Figure 2 illustrates the endogenous and exogenous structure in which a post-colonial form of deliberation is situated to take place in the state-society interface, where reflexive agencies representing the modern state and the traditional society (dis)engage with the vertical and horizontal inequalities/differences, subject to external intervention.

In practicing deliberation, arguments have been made as to which approaches deliberating agencies should take to address inequality/difference in view of inter-subjective relations, according to: 1) a rational approach; 2) an agonistic approach; 3) a hybrid approach; and 4) an approach to agreeing to disagree. Firstly, Rawls (1993) and Habermas (1996) have viewed deliberation as achieving a rational consensus by either the Rawlsian overlapping or the Habermasian aggregating methods of reasoning. Yet the rational approach to assimilating the Other, including traditional minorities to modern majorities, has raised the pluralist concern about undermining cultural heterogeneity given the irreducible difference in tradition and culture (Lyotard 1984). Taylor (1991, 1994), for example, underlines the right of minorities to maintain their socio-cultural differences and representation in political deliberation. With their Kantian belief in coexistence, liberal pluralists have explored institutional measures to nurture cosmopolitan and national imagination/citizenry (Parekh 2000, Kymlicka 2001). Their promotion for multiculturalism, however, has remained uncritical of the ideological imposition of the rationality of the state over society (Benhabib 2002).

Secondly, a critical contention has emerged to highlight the capacity of minorities to contest and replace a rational consensus with an agonistic consensus. In contrast to the ‘old’ pluralism to accommodate inequality/difference, Connolly (1995) advocates a ‘new’ pluralism to problematise rationalisation as modernising tradition and culture in society, and critically rework and negotiate it.

Given Nietzsche’s worldview and Foucault’s understanding of power, he advocates an inclusive and equal form of agonistic argumentation that allows minorities to ‘form “chains of equivalence” or collective assemblages across differences constructively’ (Connolly 2004: 167). In this connection, Mouffe (1999), from a Marxist perspective, regards agonisation as a means for exploring the conflictive consensus which falls outside the rational consensus based upon modern and liberal dominance. As opposed to Rawlsian and Habermasian reasoning, she promotes societal contestation to replace the non-negotiable rational consensus with an ever-renegotiable conflictual consensus, so as to address social injustice and exclusion, which are structurally entrenched in capitalist, late-modern society (Mouffe 1999: 755, 2000: 104-105).

Thirdly, another contention emerges to see deliberation as a means for articulating a hybrid consensus, instead of seeking a choice-based consensus, either rational or agonistic. Given the permeable nature of difference, Bhabha (1998) proposes opening up space for equivocal, hybrid argumentation where top-down (modern and political) and bottom-up (traditional and societal) forces explore the ‘in-betweenness’ of their inequalities/differences and displace them (1998: 31). Yet his positive view of hybridity has been criticised for its fragility in the face of asymmetry. Said (1978), for instance, problematises persistent asymmetry in hybridity, interrogating ‘Orientalism’ not as an ‘organic’ discourse, but as an ‘intentional’ one due to the asymmetry between the West and the Orient, which allows the former to silence the latter in an asymmetrical form of hybridisation (1978: 3). Benhabib (2002), from a constructivist perspective, applies an organic concept of hybridity to deliberating a hybrid consensus which falls outside the choice-based consensus (2002: 8-11, 19-20).

Fourthly, besides the scholarly search for a consensus over inequality/difference, a new contention
 has evolved, seeing deliberation as transforming difference into disagreement if the difference is
 deeply protracted. In the Rawlsian tradition, Gutmann et al. (1996) view deliberation not as a means

to seek a rational consensus, but to turn difference to moral disagreement. ‘When citizens or their representatives disagree morally, they should continue to reason together to reach mutually acceptable decisions’ (Gutmann et al. 1996: 1). In turn, from a critical standpoint, Connolly (1991) acknowledges the civic right for non-consensus or dissensus, underlining ‘agonistic respect’ as the political ethos of ‘a civic virtue that allows people to honour different final sources’ (1991: xxvi) and trusting the agonised parties to engage with the difference/disagreement constructively (1991: xxvii, 64). Despite their differences in aim (either agreeing to disagree or engaging with non- consensus or dissensus) and manner (either a rational or an agonistic way), their advocacy for deliberation as a means for transforming difference into disagreement can be a critical alternative to seeking ‘the’ consensus (either rational, agonistic, or hybrid), or political relativism.

In order to construct an analytical framework for deliberative peacebuilding, the above four approaches to deliberation will be interactive with the four dimensions of inter-subjective power relations (‘power-over’, ‘power-against’, ‘power-to’, and ‘power-with’) in deforming and transforming conflict. I have argued elsewhere that conflict can be transformed if power in quantity (equal vs. unequal) and quality (coercive vs. cooperative) is fairly and justly distributed between agencies in conflict in order to make their relationship legitimate and ‘peaceful’ (Nakagawa 2017b). Yet this argument does not indicate how a fair and just distribution of power makes human relationships legitimate and ‘peaceful’. Interacting with this, the above theory of deliberation suggests that a fair and just distribution of power can lead agencies in conflict to successfully deliberate the legitimation problem with vertical and horizontal inequalities/differences. First, an unequal and coercive dimension of power (power-over) allows the powerful modern/political to exercise power ‘over’ the weak traditional/societal in making a rational consensus yet aggrieving the weak. Second, if power is balanced, an equal yet coercive dimension of power (power-against) enables the traditional/societal to be ‘against’ the modern/political in reaching an agonistic consensus whilst nevertheless making their relationship precarious. Third, an unequal yet cooperative dimension of power (power-to) allows the powerful modern/political ‘to’ co-opt the weak traditional/societal, making a hybrid consensus yet leaving the asymmetry unaddressed. Fourth, if power is balanced, an equal and cooperative dimension of power (power-with) enables the traditional/societal to cooperate ‘with’ the modern/political in equal terms and engage in their disagreement with civic respect, either rational or agonistic. This will lead the state and society to make a social contract and establish political legitimacy. Accordingly, it can be argued that conflict can be transformed if the quantitative (power-over vs. power-against) and qualitative (power-to vs. power-with) dimensions of power enable deliberating agencies to engage with disagreement over their inequalities/differences, establish political legitimation and societal consent, and thus make their relationship legitimate and ‘peaceful’ (in other word, making a ‘deliberative political order’, or a political order through successful deliberation). On the other hand, conflict can be deformed if the powerful impose their consensus on the powerless, causing political delegitimisation and societal dissent, and making their relationship illegitimate and ‘unpeaceful’ (in other word, making a ‘deliberative political disorder’, or a political disorder through failed deliberation). Figure 3 lays out this analytical framework to measure the trajectory of deliberative political (dis)order in deliberative peacebuilding, given the nexus between the four dimensions of power and approaches to deliberation, and their socio-political consequences to making a deliberative political (dis)order. The indication of this timeline allows this two-by-two matrix to trace a dynamic, non-linear transition of political (de)legitimation over time.

This second section applies the above framework to examine deliberative peacebuilding in Somaliland, reviewing the legitimation problem before 1991 and the deliberative practice from 1991 to 2005 to form a deliberative political order from 1991 to 1993, reform it from 1993 to 1997, and transform it from 1997 to 2005. Long before colonisation began, Somali communities, which had been stateless and without a centralised political authority (Herbst 2000), maintained societal order based upon customary law (xeer) made by elders representing their respective clan families at various levels of elder meetings (shir at the lower and Guurti at the upper levels). As decision-making was participative and consensual, despite its male domination, Lewis (1961) calls this deliberative form of governance ‘pastoral democracy’, in contrast to Western liberal democracy. Britain’s indirect rule since the late 19th century had, however, reconstituted the pre-colonial governance system, introducing modern statehood, educating youths, urbanising pastoral life, and forming vertical and horizontal inequalities/differences. Since the independent Somaliland merged with its Southern counterpart in 1960, the united Somalia had failed to establish constitutional democracy, allowing the Barre regime to emerge in 1969 and exacerbate the legitimation problem thereafter. Defeat in the Ogaden (Ethiopian-Somali) War of 1977-1978 led the regime to return to divide and rule in clan politics, unleashing protracted inequalities/differences and aggrieving North- Western Somalis to the point where they organised the SNM (Somali National Movement) for armed resistance in the 1980s (Samatar 1988). The collapse of the Somali regime in 1991 left radicalised inequalities/differences nationwide (Nakagawa 2016).

The deliberative agencies in Somaliland (dis)engaged with these inherited inequalities/differences, forming, reforming, and transforming a deliberative political order from 1991 to 2005. The first phase (forming a deliberative political order) reveals the emergence of agonistic argumentation in society in response to failed deliberation in state-led rationalisation from 1991 to 1993. Although the regime’s collapse in Mogadishu allowed the SNM to take over the self-declared independent state, the challenging post-conflict setting prevented the SNM from establishing modern statehood in Somaliland’s war-torn society. Budgetary constraints, for example, made it difficult for the SNM to deliver services and demobilise clan militias. The rational demand of modernist civilians in the executive, made to traditional soldiers in the army to hand over the clan-controlled tax bases, the Hargeisa airport and the Berbera port, radicalised their horizontal inequalities/differences in 1992 (Farah 1999). While SNM elites failed in their deliberation nationally, it was societal non-elites, such as local elders and religious leaders, who restored their authority to deliberate horizontal inequality/difference at inter-/intra-clan meetings (shirs) locally (Farah et al. 1993, SCPD 1999). As the UN had focused on enforcing peace in Mogadishu, the overall absence of external intervention allowed traditional leaders to organise more than twenty inter-clan meetings from 1991 to 1993 (Jimcaale 2005, Bradbury 2008) and conclude their agonistic dialogues for a political settlement in the Boroma Conference of 1993. These societal leaders took key decisions, such as the adopting of two charters for peace and governance as national xeers (customary laws), replacing the SNM’s incumbent president with a veteran civilian politician, Mohamed Egal, and integrating the Guurti, the society’s maximum shir, into the state system as the Council of Elders, the state’s legislature, in order to institutionalise a deliberative space for socio-political contestation (Interpeace 2008). This Boroma process turned a deliberative approach from rationalisation to agonisation, enabling the agencies in conflict to deliberate vertical and horizontal inequalities/differences, and ceasing violence nationwide thereafter (Phase I in Figure 4).

The second phase (reforming a deliberative political order) underlines both societal resilience despite the Guurti’s failure and its limitation in socio-political agonisation before the consolidated government from 1993 to 1997. The newly-inaugurated President Egal managed budgetary constraints yet faced contestations from the ousted former President Tuur and his clan constituency. As the former had improved public functions, the latter were aggrieved by their limited access to state power and resources. Given this increasing tension, UNOSOM (United Nations Operations in Somalia) aided Tuur and his allies to rebuild a centralised state in Somalia, inviting them to high politics in Mogadishu and challenging the secessionist move of Somaliland (Bryden et al. 2000). President Egal’s dissensus on UN intervention, however, radicalised the difference between the government and the opposition from 1994 to 1995. Although violence spread across Somaliland, the Guurti was largely inactive in mediation, being co-opted by the government (Bradbury 2008: 121, Renders 2012: 142). The dysfunctional channels of communication between the government and the opposition led to a stalemate until another initiative for agonistic dialogue emerged from society. It was, among others, the Peace Committee of Somaliland, a diaspora civic group, and local elders who partly echoed the Peace Committee and stood up to mediate the radicalised difference (Abdi 1996, Bryden et al. 2000). In the meantime, the departure of UNOSOM in 1995 delegitimised the opposition and hindered their access to resources to keep fighting against the government. These changed dynamics enabled President Egal to intervene in succeeding initiatives of the Peace Committee and local elders, forcing them to discontinue their activities, and instead convene a national conference in the name of the state in Hargeisa from 1996 to 1997 to fight back in agonistic argumentation for his political survival (Abdi 1996, Bryden 2003). Although the political settlement resulted in power-sharing between the government and opposition, the state barely managed to establish a truce between the antagonistic agencies (Bradbury 2008: 126) (Phase II in Figure 4).

The third phase (transforming a deliberative political order) shows the society-led, home-grown democratisation which enabled the deliberating agencies to transform ‘old’ inequalities/differences into disagreement whilst forming ‘new’ inequalities/differences from 1997 to 2005. The power- sharing arrangement was costly, encouraging the enlarged state to approach resilient societal agencies, including capitalists, such as businessmen and the diaspora, who were emerging from economic recovery, and elders running local entities, for funding and taxation. In response, these societal forces formed a deliberative coalition, demanding that the state transform the hybrid polity into a constitutional democracy to ensure their politico-economic and socio-cultural rights (SCPD 1999, Jimcaale 2005). This society-led, home-grown democratisation, however, faced challenges soon thereafter. The constituent assembly reached a stalemate between the modernist executive and the traditionalist legislature. It was external threats, including the formation of Puntland in 1998 and the Transitional National Government (TNG) in Mogadishu in 2000, which pushed the divided politicians to reach a consensus on the draft constitution for a public referendum in 2001 (Bradbury 2008). A rushed move towards a multi-party system after the overwhelming vote for the new constitution, however, exacerbated the entrenched modernist-traditionalist divide in the state (Ibrahim 2007). Yet the sudden death of President Egal in 2002 prompted the antagonistic politicians to set aside their differences before the national crisis (Bradbury et al. 2003). As the carefully-designed electoral system yielded political parties as a new space for political deliberation (J.A.H.I. 2015), Somaliland concluded its first cycle of local and national elections from 2002 to 2005 without a security breakdown. The evidence, such as the orderly-managed close presidential race in 2003 and inter-party cooperation on armed confrontation with Puntland in 2004 (APD
2006a, Renders 2012, Hoehne 2015), suggests that political/modern and societal/traditional forces have engaged in their inequalities/differences with agonistic respect of the democratic rules of the game (Phase III in Figure 4).

Yet challenges remain. Among others, the electoral regime has led to new inequalities/differences between the winners and losers in society (Nakagawa 2017a). The ‘monetisation of elections’ has alienated capitalists and the elders of major clans on the winners’ side, and women, youths, and minorities on the losers’ side (Verjee et al. 2015). While the costly elections have allowed wealthy capitalists to patronise political parties, elders from the major clans have mobilised their clan affiliates for electoral campaigns in order to make the most of their strong clan constituencies (APD 2006, Progressio 2006, NAGAAD 2007, Yusuf 2010). Their emergence, however, undermined and constrained women, youths, and minorities from running for elections. As capitalists and elders are divided due to their different socio-political values, politicians need to engage their differences on the winners’ side, yet often leave the grievances of losers unaddressed (Ibrahim 2010, A.O.S. 2015). Despite the increasing social injustice, other societal forces are largely reactive or inactive. As CSOs remain weak or uncritical, due to their strong association with the winners in funding and operations, religious leaders mostly stay away from politics, while Islamists are diverse, from the radical to the moderate, in how they respond to societal grievances (ICG 2003, M.A. 2015). Moreover, the growing foreign aid for democratisation and socio-political modernisation has undermined the local ownership of elections and the traditional segment in the society, causing

tensions not only between donors and locals, but also between modernity and tradition (Hammond
 et al. 2011, Moe 2013, Verjee et al. 2015). These side-effects increase the risk of deforming the
deliberative political order in the future. Figure 4 traces the trajectory of deliberative peacebuilding through the key political events in Somaliland from 1991 to 2005, reflecting the shift of power dimensions in the deliberative space from the ‘power-over’ exercised by the SNM administration to the ‘power-against’ shown by the local societal agencies in the Boroma and Hargeisa processes, and more recently to ‘power-with’ between the state and society in bottom-up democratisation. This upper-curve progression suggests the reflexive capacity of agencies in conflict to choose the best possible approach to deliberating their inequalities/differences given the changing power relations in the deliberative space. As opposed to theoretical attempts to seek ‘the’ approach to deliberation, approaches taken by deliberative agencies in the post-colonial, post-conflict context of Somaliland are plural and conditional on state-society relations to transform rationalisation into agonisation, and more recently into engaging with disagreement over time. Without direct UN intervention for peacebuilding, socio-political agencies in Somaliland required ten years to adopt the tailor-made Constitution and eleven years to hold the first elections after the end of the civil war. In between, however, the precarious balance and dynamics between state and society led socio-political agencies to repeatedly resume violent conflict. This reality refutes any romanticisation of an emancipatory approach to peacebuilding (Hoehne 2013).

Discussions on empirical and theoretical findings from the case of Somaliland

This third section discusses both empirical and theoretical findings from deliberative peacebuilding
 in Somaliland. First, empirical findings include the actual causes of the 1991 and 1994 conflicts, the  
1993 and 1997 ceasefires, and peace after the Hargeisa Conference, according to the deliberative framework. The cause of the 1991 conflict in Somaliland seems to be the ‘malaise of modernity’ (Taylor 1991). Whereas modernists regard the rational approach to assimilating tradition and
culture into modernity as securing social progress (Rawls 1993, Habermas 1996), pluralists associate it with socio-cultural exclusion leading to political delegitimation (Lyotard 1984, Taylor 1991). In Somaliland, the evidence supports the pluralist interpretation. The political attempt of the modern/civilian faction in the SNM administration to rationalise the clan-centric concerns of the traditional/soldier faction over as to who should manage the key tax bases, either the government or clan militias, excluded the latter from executive decision-making and radicalised them and their clan constituency. The ineffective channel of communication between these two factions prolonged political violence and led to a protracted stalemate thereafter.

The 1993 ceasefire can be seen as the outcome of successful agonisation through the Boroma process. Following the failure of the state in deliberation, societal intervention in the political settlement of the Boroma process proved the capacity of societal non-elites to agonise and interact with political elites and led to the former making a social contract with the state. As widely argued (Farah et al. 1993, SCPD 1999, Boege et al. 2008, Bradbury 2008, Renders 2012, Hoehne 2013), local elders who led a series of inter-clan meetings across the territory were instrumental in establishing the Guurti, the Council of Elders, in the state system as the deliberative space where the modern/political and the traditional/societal could negotiate regularly. As a result, the government and clan elders agreed on how to manage the ports as state resources. Although this institutional innovation institutionalised the agonistic form of peace/order after the Boroma Conference, it failed to sustain it soon thereafter.

The cause of the 1994 conflict seems to have been the ‘insecurity of external modernity’ (Jabri
 2007: 158). As UNOSOM remotely intervened in socio-political deliberation in Somaliland, its
 material power and international legitimacy changed the dynamics between deliberating agencies, exacerbating discursive differences between the federalist position of the UN and the secessionist one of Somaliland, linking them to political positions of competing clans, especially between those supporting President Egal and former President Tuur, and forming identity/difference which Connolly (1991) identifies as the key cause of contemporary conflict. As the degraded qualitative power dimension made agonisation increasingly asymmetrical and antagonistic, the national elites led by President Egal chose not to agonise identity/difference, but to radicalise it for their own political survival (Bryden 2003). This evidence supports the critical contention that asymmetry in contemporary peacebuilding can allow international peacebuilders to impose a global consensus of (neo)liberal statebuilding on national/local agencies, or to ‘other’ them, leading to insecurity (Newman et al. 2006, Jabri 2007), and that an asymmetrical form of agonisation can enable powerholders to securitise or politicise inequalities/differences and antagonise agonism (Connolly 1991: 68, Stewart 2002). This stands in contrast to an essentialist view of agonisation, in which antagonism turns inevitably to agonism (Mouffe 1999: 758). Moreover, the ‘insecurity of external modernity’ in view of a hidden/discursive form of power exercised through international-national interactions contrasts with other claims about the causes of the 1994 conflict, such as the dilemma of hybrid polity in the co-optation of Guurti (Hoehne 2013) and the greed of President Egal (Balthasar 2013), which have been made in view of a visible/behavioural form of power through national-local interactions. This indicates that agonisation and hybridisation in a post-colonial form of deliberation should address the multiple layers of power (not only local and national, but also international), and the multiple forms of power (not only visible/behavioural, but also hidden/discursive) (Gramsci 1971, Foucault 1980), as well as the multiple natures of identity (not only constructed and instrumental, but also primordial) in tradition and culture (Geertz 1973, Laitin 1986, Benhabib 2002).

The 1997 ceasefire can be linked to the successful restoration of agonisation. As the Guurti failed in
 facilitating mediation, reversing the antagonised agonism required the (re)emergence of societal agencies and initiatives for deliberation (e.g. the Peace Committee, the local elders, and the Hargeisa Conference), as well as the exogenous and endogenous change in power dynamics (e.g. the departure of UNOSOM and the defeat of the opposition) in order to lead deliberating agencies to agonise identity/difference. This evidence supports Ramsbotham’s contention that, if
identity/difference is radicalised, continuous agonisation will be the only way to lead the antagonised contestants to explore discursive ambiguity between the radicalised and the agonised difference/disagreement (Ramsbotham 2010: 94, 104-108). The agencies, such as the Peace Committee of Somaliland and local elders, kept agonistic dialogues going, although neither their initiatives nor the Hargeisa Conference resolved their identities/differences.

In this connection, the cause of peace after the Hargeisa Conference is related to home-grown democratisation that structurally changed power dynamics and enabled agencies to transform identity/difference into disagreement. Although democratisation in Somaliland seemed to be a by- product of socio-political deliberation on taxation (SCPD 1999, Jimcaale 2005), as Tilly (1990) observed it similarly happening in Western Europe (Eubank 2012), the customised electoral system played the key role in improving the deliberative quality of power and institutionalising it. Given his own experience as the Prime Minister of Somalia ousted by a military coup in the fragmented democracy of the late 1960s, President Egal looked towards the electoral system as a way of preventing political fragmentation given the complex clan dynamics in Somaliland (J.A.H.I. 2015). Accordingly, he proposed restricting the number of political parties to three, given the three major sub-clans in the Isaaq clan family (the Habar-Awal, the Habar-Jalo, and the Habar-Yunis) (J.A.H.I. 2015), following the electoral system adopted in the 1960s for proportional representation in local multi-member constituencies (Krennerich 2003, Jama 2009). This legal modification relaxed restrictions on the creation of political organisations which could join local elections. At the same time, however, it allowed only three of them to qualify as national parties entitled to run in the presidential and parliamentary elections, provided they either acquired more than 20% of the votes cast in all regions or the highest number of votes cast in local elections (Jama 2009: 94-96). Those which failed to meet these criteria would be merged into those which qualified (Jama 2009: 56). Although this forced dissolution and merger of parties undermined ideological cohesion within them and constrained voters’ political choice (Yusuf 2010: 20), it also yielded benefits in the political context of Somaliland. First, limiting political parties enhanced inter-party interaction in political deliberation (Jama 2009). Second, it also increased intra-party interaction between party members from different societal/clan backgrounds, and led senior leaders to pursue inter-clan power-sharing within parties (Lindeman et al. 2003). The fluid inter-subjective contact across parties prevented inter-clan/regional inequality/difference from radicalising, as the contact hypothesis assumes (Allport 1954). Third, as more than 70% of voters were considered clan constituents (Yusuf 2010: 18), electing local representatives at the national elections forged clanism (Nakagawa 2016: 287), yet also held parties accountable for the vertical inequality/difference to their local/clan constituency (APD 2006: 36). Accordingly, the electoral system in Somaliland has led political parties to pursue increasing inter-/intra-party contact, sharing power within parties and improving accountability to their local/clan constituency, and has thus enabled them to re-engage with the vertical and horizontal inequalities/differences in the medium term. The electoral regime has, however, shaped ‘new’ inequalities/differences between the winners and losers in society.
Also, foreign aid for democratisation and modernisation has exacerbated the modern/political- traditional/societal divide. These side-effects of democratisation have increased the risk of deforming the emerging inequalities/differences in the longer term. Besides empirical findings, normative findings include the theoretical plausibility of deliberative peacebuilding which underlines the nexus between deliberation, political legitimacy, societal consent, and socio-political order. While the mainstream strand upholds a ‘capacity-based’ or ‘coercion-based’ approach to peacebuilding, highlighting the capacity of the state in liberalising the political economy and containing war-torn societies (UN 1992, OECD 2005, World Bank 2011, Nakagawa 2017b), the practice of (neo)liberal statebuilding has often resulted in the legitimacy gap, leading war-torn societies to political, economic, and societal crises (Nakagawa 2017a). By contrast, a critical strand of peacebuilding proposes structural measures to legitimise politics, underlining fairness and justice, such as addressing distributive justice (Pugh et al. 2008) and human security (Futamura et al. 2010, Tadjbakhsh 2011) in an economic domain, and empowering civil society (Lederach 1997, Paffenholtz 2014) and promoting customary governance and restorative justice (Boege et al. 2008) in a socio-political domain. Sympathising with the latter, however, deliberative peacebuilding presupposes a hermeneutic act of human agency to address the legitimation problem with inequalities/differences in a post-colonial, post-conflict context, exploring the traditional/horizontal forms of legitimacy in a society (Holsti 1996, Englebert 2000) vis-à-vis the modern/vertical forms of legitimacy in the state (Hobbes 1968, Weber 1991), given the former’s effect on cultural cohesion. The above review of the Somaliland case in forming, reforming, and transforming a deliberative political order supports this ‘legitimacy-based’ or
‘justice-based’ approach to peacebuilding, underlining the re-distribution of the quantitative and qualitative dimensions of power for successful deliberation, although a single case study of Somaliland limits its theoretical generalisability. Partly interacting with this critical thinking, policy-makers and practitioners have turned their attention to the legitimacy-based approach to peacebuilding, given the recent discussions on state- society relations and social contract in fragile contexts (DFID 2010, UNDP 2012). These debates, however, remain prescriptive in the capacity-based approach to understanding how the state responds to society in order to improve a vertical relationship, masking horizontal variation in society which is often invisible to, and hidden from, the eyes of externals. For example, whereas DFID (2010) underlines state-society relations in peacebuilding yet limits the concept of legitimacy in the capacity of the state for service delivery to meet societal expectations, UNDP (2012) contends for legitimacy based upon social contract yet attempts to secure it in making an effective state (for service delivery) and a resilient society (to be independent from the state). This state- centred, capacity-based approach will continue to leave the vertical inequality/difference unaddressed. In turn, practitioners, including INGOs, have argued for political legitimacy as a by- product of socio-political participation and negotiation (Papagianni 2008, Interpeace 2010). Although societal contestation addresses vertical inequality/difference, their Western ontological and epistemological positions would undermine the non-Western Other in research and practice, and increase the risk of disengagement in horizontal inequality/difference. The empirical evidence in Somaliland supports this concern, in that donors’ modern approach has undermined the traditional segment in society and exacerbated the segmental divide (Hammond et al. 2011, Moe 2013, Phillips 2013). Policymakers and practitioners should thus broaden their understanding of legitimacy to encompass multi-dimensional (not only power-over, but also power-against, power-to, and power-with), multi-layered (local, national, and global), and multi-formed (visible and invisible/hidden) perspectives of power, and keep their policy and practice relevant to addressing the socio-political realities and challenges in the post-colonial, post-conflict context.

Conclusion

Given the legitimation problem with vertical and horizontal inequalities/differences as the historical and cultural causes of conflict in the post-colonial context, this paper highlights deliberation as an agential act to address them, theorising deliberative peacebuilding given the nexus between dimensions of power and approaches to deliberation. The case of Somaliland suggests the overall applicability of deliberative peacebuilding, and identifies the empirical causes of conflict, ceasefire, and peace, and the theoretical plausibility of a legitimacy-based or justice-based approach to deliberative peacebuilding vis-a-vis a capacity-based or coercion-based approach to (neo)liberal statebuilding. Deliberative peacebuilding brings both theoretical merits and practical challenges. Amongst its theoretical merits, deliberative peacebuilding is emancipatory. Its concept of power is multi-dimensional, allowing the traditional/societal weak to address their positions to the modern/political powerholders, despite the risk of their precarious balance leading to insecurity. Deliberative peacebuilding is also practical. Its reflection on post-colonial historicity and culture makes deliberative peacebuilding more context-specific and thus realistic than (neo)liberal statebuilding, which normalises the Western experience of modernisation and state formation, even though deliberation is originally conceptualised in view of Western liberal democracies (Rawls 1993, Habermas 1996). In this connection, deliberative peacebuilding is locally-owned. While (neo)liberal statebuilding relies on external expertise and resources, deliberative peacebuilding draws on the skills and capacity which are locally available, and thus cohesive, legitimate, low-cost, and sustainable. This local turn, however, does not do away with any positive role for external intervention in empowering the weak and enabling them to address their inequalities/differences and grievances. Finally, deliberative peacebuilding suggests a medium/long-term vision. Its comprehensive and longitudinal approach to transforming conflict into peace as deliberative political order allows researchers, policy-makers, and practitioners to map out a dynamic and complex trajectory in a non-linear transition, and identify the remaining gap in the medium/long term, contrasting with (neo)liberal statebuilding, which attempts to solve problems through a linear transition in the short term (UN 1992).

On the other hand, the practical challenges in deliberative peacebuilding include endogenous and exogenous contradictions, as the Somaliland case suggests. Endogenous contradictions are likely to emerge when deliberation is undertaken without acknowledging power dynamics in a post-colonial, post-conflict context. Successful deliberation requires deliberative space to meet the quantitative and qualitative dimensions of power and leads deliberative agencies to engage in the inequalities/differences between the powerful modern/political and the powerless traditional/societal. Yet, rational argumentation is likely to empower the modern/political but disempower the traditional/societal, and increases the risk of the former not tolerating or engaging

with the latter, instead ‘othering’ and regulating it, as the process of the 1991 conflict indicates. However, as the latter are not passively ‘out there’, but capable of resisting the former, the ineffective official channel for deliberation between them in the post-colonial, post-conflict context is likely to radicalise their inequalities/differences and lead to stalemate. Given this, the state-society interface will be instrumental in creating a space where deliberative agencies interact and re-engage with their inequalities/differences, as the 1993 ceasefire shows. In turn, exogenous contradictions are likely to emerge when the national/local agencies encounter international peacebuilders who uphold the global consensus on (neo)liberal statebuilding. The asymmetry between them will allow the latter to impose the globally legitimate consensus on the former, yet increases the risk of Western modernity ‘othering’ non-Western tradition and culture, as the 1994 conflict demonstrates. Given this ‘insecurity of external modernity’ (Jabri 2007: 158), radical critiques promote measures to empower national/local contestants and enable them to critically enunciate, iterate, and challenge the interveners through agonisation (Shinko 2008) and hybridisation (Richmond 2005). In reality, however, it is unlikely that the uncritical interveners hear and engage with the intervened on equal terms (Pouligny 2006, Autesserre 2014). Even if the international interveners engage with the national modern/political elites, the asymmetry in the national/local variation will allow the national modern/political elites to ‘other’ the local traditional/societal non-elites, turning to endogenous contradictions thereafter.

By  Yoshi Nakagawa