Thesis for master’s seminar in Peace and Conflict studies, 30 credits Master's Programme in Crisis Management and Peacebuilding, 120 credits
Abstract
For long the discipline of peace studies have investigated causes of war, rather than causes of peace, in an African context. In the northern peripheries of Somalia, a nation ravaged by civil war and conflict, two apparent peace zones have emerged following the complete state collapse of 1991: Somaliland and Puntland. The study explores whether or not these two realities of peace can be defined and characterised as Zones of Peace, or sanctuaries, amidst a civil war. Utilising the analytical tools of Zones of Peace – hitherto applied on conflictual contexts elsewhere but the Horn of Africa – this study suggests that both Somaliland and Puntland are, despite the territorial conflict between them, peace zones granting shelter from the civil war. Suggestively, peace has prevailed in both Puntland and Somaliland due to Somalia’s deteriorated situation, not in spite of it. The study concludes that in order to optimise research concerning Somaliland’s and Puntland’s peace(s), the framework of Zones of Peace can offer in-depth insights on local everyday milieus. The framework partially explains why these local peace(s) has lasted despite lacking external attention and allow for thorough comparison between two homogenous cases. Lastly, both Puntland’s and Somaliland’s inviolability and durability remain unchallenged and rigorous, possibly because of the civil war’s status quo, and since the international community’s foci on south- and central Somalia persists.
Table of content
1. Introduction 1
1.1. Purpose and research questions 2
1.2. Limitations and demarcations 3
1.3. Explaining Zones of Peace and sanctuaries in brief 3
1.4. Previous research on Puntland and Somaliland 4
2. Zones of Peace theory and conceptual framework 8
2.1. State (re-)building where there is no state 10
2.2. Zones of Peace amidst armed conflict 11
2.3. The logics behind the creation, inviolability, and durability of a sanctuary 16
3. Comparative method and guidance of framework 18
3.1. Research design 19
3.2. Case selection 20
3.3. Material 21
3.4. Measuring Zones of Peace underlying logics 22
3.4.1. Answering the research questions on inviolability and durability 26 4. Puntland and Somaliland as sanctuaries amid civil war 28
4.1. Conflict background for context 29
4.2. Puntland 30
4.2.1. Type of sanctuary in Puntland 30
4.2.2. Functions of Puntland peace 31
4.2.3. Puntland’s environment 33
4.2.4. Protection offered by Puntland 35
4.3. Somaliland 37
4.3.1. Type of sanctuary in Somaliland 37
4.3.2. Functions of Somaliland peace 38
4.3.3. Somaliland’s environment 40
4.3.4. Protection offered by Somaliland 42
4.4. Comparison 44
4.4.1. Similarities in characteristics, relationships, and role of patrons 44 4.4.2. Differences in characteristics, relationships, and role of patrons 46
5. Analysis 48
5.1. Impenetrable land(s) of lasting peace(s) 52
6. Conclusions 53
6.1. Thriving in a conflictual context 54
6.2. Discussion on theory guidance and future research 55
References 58
List of abbreviations AIAI – al-Itihaad al-Islamiya
AMISOM – African Union Mission in Somalia AU – African Union
BBC – British Broadcasting Corporation
CTC – Combating Terrorism Center at West Point EU – European Union
FGS – Federal Government of Somalia ICG – International Crisis Group ICU – Islamic Courts Union IS – Islamic State
NGO – Non-governmental organisation NIS – National Intelligence Service
PDRC – Puntland Development Research Center PIS – Puntland Intelligence Agency
PSF – Puntland Security Force SNM – Somali National Movement SSDF – Somali Salvation Democratic Front TFG – Transitional Federal Government TNG – Transitional National Government UN – United Nations
UNITAF – Unified Task Force
UNSOM – United Nations Assistance Mission in Somalia US – United States of America
1 1. Introduction
“Following the melt-down of the state in 1991, Somalia became state collapse’s archetype, being synonymous to the outside world with hunger and lawlessness. UN-sponsored attempts to re-establish central authority went nowhere. Yet while most of the international political attention has gone to “national reconciliation conferences” with warlords, several spontaneous bottom-up processes have begun providing genuine – and increasingly comprehensive – human security to Somali communities.” (Verhoeven 2009: 406)
The results of the Somali civil war, and of the humanitarian and militarised intervention United Nations Assistance Mission in Somalia (UNSOM), have had long-lasting impacts on the Horn of Africa. Following the arguably abashing retreat of United Nations (UN) and United States (US) military personnel in 1995, international peacebuilders and observers have dismissed “the Somali fiasco […] as an outlier – a case of imperial hubris where there was no peace to keep” (Jenkins 2013: 9, emphasis added). Whilst acknowledging that Somalia – officially the Federal Republic of Somalia – has yet to reach a negative peace (see Galtung 1969) for the international community to keep or build upon, the Horn of Africa nonetheless provides scholars with two distinguishable examples of traditional and lasting peace(s): Somaliland and Puntland (see Appendix 1, Figures 1 & 2 for maps).
Although Somalia as a whole is far from nation-wide reconciliation and peace – principally between the internationally recognised Federal Government of Somalia (FGS) and the terrorist group al-Shabaab – two apparent Zones of Peace (ZoP) in the northern parts of the nation seem to have defied monumental odds in shaping lasting security and peace, in lieu of subsiding to the civil war. Somaliland and Puntland, betwixt armed clans and Salafistic jihad groups, present some of the more perplexing observations a student of peace- and conflict may encounter: why peace during war? Specifically, why peace despite a notably long-lasting and violent civil war. The two sovereign Somali regions have ostensibly utilised non-violent conflict resolution, promoted inter-clan reconciliation, and reached significant levels of civilian security against violence within their own territories – effectively creating inner sanctums of peace. Whereas central- and southern Somali clans, warlords, and more problematic civilians, have been involved in a soon to be 30 years long-drawn-out civil war.
Informed by Mitchell’s query, “some form of international interest has helped maintain the inviolability of some local ZoPs as sanctuaries, but is this a general pattern?” (2007: 24, emphasis added), this thesis ponders on whether or not the lack of international attention to the
2 peacebuilding processes in Somaliland and Puntland can nuance alternative insights on causes of peace during war. One puzzling remark is apparent, which previous literature has yet to tackle: can these two empirical observations of peace amid civil war be considered ZoPs, even if external peacebuilders have not substantially intervened nor assisted in their local processes? Existing literature does provide invaluable material on each specific case, certainly so within a larger Somali context (see, especially, Lewis 2008 and Hoehne 2015). This thesis strives to thoroughly investigate why Somaliland and Puntland have achieved and maintained peace despite noticeable spoilers and lack of international assistance, or perhaps more intriguingly: what are their causes of peace? Moreover, what potential lessons there are to be learned for the framework of Zones of Peace in the midst of armed conflicts elsewhere.
1.1. Purpose and research questions
The over-arching aim of this thesis is to contribute to the theoretical queries on ‘why peace’ by analysing the characteristics of Somaliland’s and Puntland’s peace(s). The purpose is specifically to answer why virtually high levels of security and stability have been attained in the regions, as both have factually maintained peace following the Somali civil war(s) of the 1990s and 2000s. Thus, I will derive whether or not these two cases can be considered ZoP and compare them. The first research question is formulated in a descriptive manner and addresses the territories similarities and differences. Whereas the second research question is of a more analytical nature and explores what traits (or lack thereof) these cases of peace amid war rely on in order to maintain their integrity and certitude as sanctuaries:
1. How does Somaliland and Puntland conform to the logics of ZoP, can they be characterised as sanctuaries amid armed conflict? What are their similarities and differences?
2. Why have Somaliland and Puntland remained inviolable and durable examples of peace, in lieu of receding into a state of civil war?
I incorporate a ZoP framework which informs the analysis, thereby adding to the pool of knowledge concerning literature on Zones of Peace. This study compares and evaluates the two self-governing regions and conclude on why they have achieved peace where a considerably persistent civil war has continuously raged since the early 1990s. In addition, this thesis will hopefully provide some material, via the exploration of two hitherto under-studied cases, to the general queries on causes of peace amid civil war in an African context.
3 1.2. Limitations and demarcations
At the time of writing Somaliland is yet to be internationally recognised as a sovereign state. Somaliland is, however, acknowledged as a relatively more stable part of Somalia than other remnants of the nation – hence enjoying recognition in practice if not in theory. Somaliland is thereby included in this thesis analytical endeavour as it provides an exemplary counterpart to neighbouring Puntland, who enjoy de jure independence from Somali central authority. Concerning Puntland there are additionally four federal member states – with a myriad of sub-state administrations – in Somalia (South West State of Somalia, Jubaland, Hirshabelle, and Galmudug (Appendix 1, Figure 2). These are however not included in the comparison merely because they have not attained the same levels of peace as Puntland and Somaliland have upheld since the 1990s – nor have the other federal Somali states achieved the same levels of autonomy as Somaliland and Puntland (Dill 2012: 292-293).
It should be pointed out that this comparative study equalises the local with several localised communities and aggregation of peoples. The ‘local’ is not one particular site but rather encompasses the entirety of both Somalilanders and Puntlanders domains of influence. The regional/outside, in turn, includes the total topography and societal realm of federal Somalia. The external – applied interchangeably with the international and global – involves sovereign states; non-governmental organisations (NGOs); regional organisations; and supranational entities. To clarify, outsider is conceptualised to depict armed actors on all levels (local, regional, or global) and will be analytically treated as such. Ergo, the relationship between the local, outsiders, and global includes the linkage between those residing in the peace zone (residents), external forces interacting – supportive or refutative – with the zone (patrons), and those who are posed as a potential threat or possible ally toward the zone (outsiders).
1.3. Explaining Zones of Peace and sanctuaries in brief
According to Hancock & Mitchell (2007) Zones of Peace are defined as social sanctuaries (geographical or socially construed) wherein people are sheltered from armed conflict based upon accepted rules of conduct and societal guidelines. Examples of ZoPs are observable all over the globe, albeit varying in size and scope of their missions. ZoPs may be extra-societal, i.e. without the geographical territory of those seeking sanctuary; or intra-societal, i.e. in an explicitly demarcated territory adjacent to armed conflict, attempting to protect all residing within its spatial and jurisdictional boundaries. Note that the concept of ZoP is not to be
4 confused with Kacowicz’s (1995: 265) definition, which classifies “Zones of Peace” as a group of states deliberately avoiding warfare with one another. A sanctuary may, to some observers, come across as synonym to ZoP. This is partly true, as both concepts evidently refer to a place which is held inviolable and protects its residents from harm. Sanctuaries are “designed to produce negative peace by removing the threat and use of violence through the agreement of all warring parties” (Hancock 2015: 237). Thereby, a sanctuary per se encompass outsiders and their conditionalities (e.g. agreed upon rules and relations) against the region, whereas a ZoP may solely involve the local. This study applies ZoP and sanctuary interchangeably: conceptually referring to local protection against outside threat of violence or conflict via local-global/national interactions.
1.4. Previous research on Puntland and Somaliland
Previous literature on peacebuilding processes on the Horn of Africa have in general applied its analytical foci on international peacebuilding failures and shortcomings, often critiquing the liberal peace paradigm (see Ahmad 2012; Autesserre 2019; Menkhaus 2007, 2009). Somalia and the nation’s institutional structures are, and has been for nearly three decades, in a state of ruin and anarchy – seemingly unable to gain momentum in any shape or form of state building (Ahmad 2012; Allo 2009; Haldén 2008; Hammond 2013; Menkhaus 2006/07), let alone peacebuilding (Menkhaus 2009; Wiuff Moe 2016; Njeri 2019; Pham 2011; Silvestri 2019). Whilst recognising the validity of these observations, there have been some studies devoted to examining successful attempts at reconciliation, state building, and subsequent peacebuilding, namely in Somaliland (Johnson & Smaker 2014; Schwoebel 2018; Balthasar 2019; Farah & Lewis 1997) and in Puntland (Albrecht 2018; Dill 2012; Helander 2005). This larger field of research is of significance when examining both regions, either separately or comparatively as this study aims to do.
Indubitably, Somaliland and Puntland are cases of successful, local peace(s) (Hoehne 2015; Samantar 2009) wherein both regions have achieved substantial societal reconciliation between former enemies – despite the fact that neither Somaliland nor Puntland have received any particular assistance from the international community (Lewis 2008: 93). Nor have the two regions resolved the ‘Puntland–Somaliland border dispute’ where Somaliland has explicitly seceded away from federal Somalia, and Puntland claims autonomy but continuously endorse a functioning and unanimous Somali state (Hoehne 2015: 15). Therefore, it remains valid to highlight that neither Somaliland nor Puntland are utopian cases of peace wherein no conflicts
5 takes place, but rather emphasise that both regions have achieved peace within their own territories despite the border dispute concerning the sub-regions of Sool, Sanaag, and southern Togdheer (Ayn)1 (Appendix 1, Figure 1). Even more intriguing, both Puntland and Somaliland remain relatively peaceful despite a long-lasting Somali civil war where there is a high presence of potential violators toward each zone (e.g. warlords, rival Somali clans, and al-Shabaab). It has, furthermore, previously been established from a variety of studies that Somaliland is a de facto independent, albeit not internationally recognised, nation-state where Somalilanders effectively govern and maintain peace and security themselves (Farah 2001: 138; Johnson & Smaker 2014: 6-12; Kaplan 2008: 147-149). Puntland, recognised as a de jure autonomous state of federal Somalia, governs their own institutions, territory, and peacebuilding processes, and enjoys self-governance away from a centralised Mogadishu administration without substantial external interferences (Marchal 2010: 10, 25; Johnson & Smaker 2014: 13-17). In addition, the works by Samson (2015) and Lewis (2008), among others, further strengthens this thesis assumption that Somaliland should be considered a sovereign unit of analysis, i.e. as a separate entity away from the state of federal state of Somalia. Whilst simultaneously admitting that the secessionist state has a shared history with Greater Somalia2, it becomes apparent that Somalilanders adhere to their own independent principles, procedures, and agendas – with or without international recognition and support (Nordstrom 2004: 171-172).
Arguably, prior expositions on Somali peacebuilding have approached the northern region in a state-centric manner, including Hoehne’s seminal work Between Somaliland and Puntland (2015) which purpose does not go beyond decoupling state building vis-à-vis peacebuilding, in order to explain Somaliland’s and Puntland’s particularities (2015: 160-162). Hoehne does make an important contribution to the field of study: while much has been written on Somaliland, far less literature is available on Puntland – an issue Hoehne mitigates. Important scholarly editions, among them Hoehne (2015) and Lewis (2008), provide crucial insights on Somali clans in particular and Somali culture in general, which is deemed invaluable for this study. One cannot aptly examine either Puntland or Somaliland without embracing Ioan Lewis social anthropological overview, Understanding Somalia and Somaliland, written in 1978 and revised 2008. Lewis – acknowledging Somaliland’s peacebuilding successes and criticising the
1 These regions claimed their independence in Khatumo state, a proto-state which never received recognition
(Stremlau 2018: 79-80; Hoehne 2007).
2 Greater Somalia is the notion of a united Somalia which encompass all Somali speaking nations in or near the
6 failures of external interventions in central- and southern Somalia – provides the reader with a thorough chronology on the entirety of the region (2008: 123-128). Perhaps more relevant for this thesis intentions, he argues, compellingly, that the international community’s role should be consultative rather than interfering – e.g. offer support to but not dictate Somaliland’s and Puntland’s processes (Lewis 2008: xi).
There are relatively few scholarly attempts to analytically compare the two Somali states, even though both self-governing entities have achieved what the international community has sought for so long in central- and southern Somalia (ergo non-violent conflict resolutions and power-sharing agreements). The much-needed but often over-looked separation of state-centric (top-bottom) approaches and of the local turn in Somali peacebuilding, or bottom-up approaches, is further merited by Little’s (2003: 167) observation on Somali culture, where “radical localization is the norm, and even before the state collapse, most meaningful politics were conducted outside official channels.” Arguably, if linking “most meaningful politics” with local peace processes, analytical endeavours in African peace literature should thread carefully as not to intertwine state building and peacebuilding too much, since it may risk undermining an exposé of the local. One interview study does highlight the need to recognise local agency, specifically women, as peacebuilders in Somaliland and Puntland. Dini (2009: 36) argues that recognition of women’s legitimacy and efforts could complement ongoing practices of peacebuilding in both territories. This observation, among others, make the issue of how Puntland and Somaliland may be considered, and compared as, ZoPs even more relevant.
In one of the more recent literatures, and in sharp contrast to this thesis assumptions, Silvestri (2019: 13-14) argues for a continuation of the liberal peacebuilding paradigm in the republic of Somalia; inclusive and protective of all its citizens, including Somaliland. The author emphasises Somali unity – through assistance from regional and international actors – and maintain that the legitimate government of Somalia, in Mogadishu, ought to respect and adhere to international law and humanitarian values as formulated externally. I do not contest the reasoning, nor do I make any remarks on the valid concerns of the “very real possibility that Somalia’s fragile security progress could be undone by mismanagement of the political situation” (Silvestri 2019: 14). But where Silvestri and others (see Bryden 2004: 31-32; Bryden & Brickhill 2010: 261; Horton 2019: 24) seemingly promotes a more comprehensive re-run of the hitherto unsuccessful concept of a central state apparatus equals national peace, this study aim to explore in-depth the localised, more successful, peace(s) in northern Somalia and what
7 insights they may offer. In another recent body of work, Schwoebel (2018) explores Somaliland and Puntland in-depth, particularly concerning their local state building practices and peacebuilding successes. Schwoebel pinpoints many of the particularities of Somaliland’s inherent difficulties in finding national harmony – considering the state peripheral-central clan dichotomy – and concludes that the legitimacy gained by local actors very much relies on the traditions of Islam, rather than the successes of multiparty democracy (2018: 218-219). This is highly relevant when examining Puntland and Somaliland since Islam is the undisputed major religion and way of life for nearly all Somalis; and has lead both Somaliland and Puntland in a direction where peacemakers arguably “have had to marry a more contractual-based legitimacy with an affinity-based legitimacy in order to persuade some clans to accept state authority” (Hancock & Mitchell 2018: 225).
Lastly, few scholars have offered such insightful observations into peace(s) amid armed conflict as Nordstrom’s inspiring book Shadows of War (2004). Nordstrom signifies Somaliland as a “curious inversion” where peace was attained despite the de facto state being engulfed in a raging civil war. Adding to the critique against the liberal peace paradigm, Nordstrom identifies that the international community for too long has “insist[ed] on speaking of “Somalia” as if it were a state, and the battles of Mogadishu as battles for the control of Somalia” (Nordstrom 2004: 172), rather than recognising the particularities of a self-governing island of peace in the north. Survival is argued to dictate the wants and needs of locals, rather than the amalgamation with a former state apparatus – and if informal agents of peace manages to create sanctuaries against violence, the “state” could become superfluous (2004: 172-173). This is an important observation concerning Puntlanders and Somalilanders everyday milieu, since they would presumably differ: one wants to rebuild a federalist state regime in Mogadishu, whereas the other actively seeks to separate from federal Somali authority. In sum, Somaliland and Puntland are under-explored cases vis-à-vis peacemaking during civil war and despite armed conflict. The studies that have explored either cases’ peacemaking successes are yet to comprehensively compare the two, and at the time of writing no ZoP framework has been applied on either region. The theoretical approach this thesis adopts is far from a simple one, yet highly warranted as both regions seem to uphold security measures and manages to create peace writ large (to borrow Hancock’s (2017: 268) notion that peacebuilding initiatives should boost larger societal benefits). Even in the face of terrorism, climate crises, and localised border contestations, Puntland and Somaliland seemingly prevail in peacemaking in lieu of plunging back into civil war.
8 2. Zones of Peace theory and conceptual framework
“Peace is not a universal concept that can be transposed identically between different contexts of conflict. Rather, unique forms of peace arise when the strategies, institutions and norms of international, largely liberal-democratic peacebuilding interventions collide with the everyday lives of local actors affected by conflict.” (Richmond & Mitchell 2012: 1)
The most widely used conceptualisation of peace resides within Galtung’s (1969) negative and positive peace dichotomy. Negative peace becomes apparent in the realm where violence and/or fear of violence is absent, namely in a protected sanctuary. Positive peace, on the other hand, is true, lasting, and builds upon justice and fairness for all people within the spatial boundary. Peace can thereby – more than the mere absence of war and conflict – either be personal and direct or structural and indirect (see Galtung 1969: 183); and for the sake of conceptualising ZoP, both the negative and the positive spectrum of peace is adhered to and accepted as epistemologically feasible. Moreover, in the context of islands of peace amid armed conflicts, the underlying assumption is that negative peace is initially sought after (e.g. immediate refuge away from violence), but in due course the desired peace is stipulated as positive (e.g. structural and collective safety-nets against violence, such as economic growth or strengthened individual agency).
To reiterate this thesis purpose, and highlight why this undertaking is necessary, there is an echoing lack of ZoP theory application in the context of African peace(s) amidst civil war; a context which is warranted for closer examination. The theory of ZoP was formulated and has been continuously developed in the context of mainly Colombia and the Philippines (Mitchell & Allen Nan 1997; Mouly et al. 2015, 2016; Idler et al. 2015; Garcia 1997). Concerning other regions (e.g. sub-Saharan Africa) where sites of sanctuaries remain understudied, scholars have previously identified the need for a ZoP framework, but retain that they have:
“come across some information about efforts to establish peace zones, weapons free zones or peace corridors in various African countries suffering from various types of intra-state conflicts – for example in Sudan and South Sudan, in Somalia, and Nigeria – but details about these efforts
is difficult to discover.” (George Mason University 2017, emphasis added)
Moreover, Africa is quite often presented in a very dismal, ceaselessly warring, context: adding fundamentally mismatched pieces to the grand puzzle of why peace. Johan Brosché and Kristine Höglund devote their attention to causes of peace, rather than causes of war, in the cases of Botswana, Malawi, and Zambia. Their project aims to offer an explanation to “why peace has
9 prevailed in these three countries, despite being part of the world’s most volatile continent and with their neighbours struck by conflict” (Brosché & Höglund 2017). Parallel to their grand task, this study attempts to answer why peace has been attained in Somaliland and Puntland, despite both regions’ adjacency to a long-standing failure of a state, via the framework of ZoP.
With the notable exceptions of Höglund’s and Brosché’s ongoing research project ‘Causes of Peace’ (see Uppsala Universitet n.d.), and Allouche & Jackson (2019) who identifies six distinct periods of ZoP in Côte d’Ivoire and Sierra Leone in West Africa, contemporary and much needed literature on African ZoPs remain conspicuously absent – where the “logic of non-violence during conflict is scarcely covered in the African academic literature” (2019: 71). Moreover, seeing how the international community have not expressed particular interest or lobbied for large-scale peacebuilding programmes on-site in Somaliland and Puntland (with very few exceptions), an analyst may similar to Hancock urge for “one potential model for post-conflict peacebuilding that has hitherto remained below the radar screens of many in the international peacebuilding community – the local zone of peace” (2017: 260). Assuming that most ZoPs are by and large characterized by non-combatants who formulate their own peacebuilding agendas, claims on agency, and push-back against violence (Hancock 2017: 261), Somaliland and Puntland would at first glance fall neatly within the frames of ZoP. But I digress, this is left for the upcoming analysis.
For analytical clarity, ZoPs are at the onset seeking negative peace but can presumably only succeed in this endeavour if they strive for positive peace, where global and liberal values collide and combine with local and traditional norms (cf. Björkdahl & Höglund 2013). The ZoP framework does acknowledge the importance of agency and autonomy, but it also recognises the importance of external intervention and patron interest; it emphasises the analytical limits of territories and spatial boundaries but need not attribute this to a state-centric lens, necessarily (Mitchell & Hancock 2007: 217-218). According to Richmond & Mitchell (2012: 1) unique types of peace(s) – more often than not hybrid in nature – evolve when liberal and international peacebuilding strategies mismatch and create friction with the local, or the everyday, in a specific context. Highly relevant for this thesis exploration on localised forms of peace (traditional in nature) are key concepts such as (attained or perceived) agency and unity utilised by inhabitants residing within a peaceful sanctuary. Furthermore, Richmond & Mitchell’s query “[h]ow are acceptance, co-optation, resistance and rejection expressed in peace processes?” (2012: 2) presents itself as a viable question for this undertaking in order to nuance the
10 ‘everyday’. The above-mentioned interactions between the global-local (see Björkdahl & Höglund 2013) may aptly be integrated in a deeper analytical node, by coupling encounters and frictions, to examine what challenges a ZoP may encounter – wherein externally and internally initiated peacebuilding processes and conflict resolutions may affect Somalilanders and Puntlanders possibilities to seek or create shelter from violence, and more importantly preserve the inviolability of their sanctuary.
Note that the concepts of liberal and hybrid peace are well cemented and illustrated in contemporary literature in the field of peace- and conflict studies. This thesis does not, however, embark on such a theoretical exercise. A rather substantial amount of studies has coupled Somaliland’s and Puntland’s relative successes in security and peace with state-centred theories, which I argue is too insubstantial for this thesis enterprise. Similarly, studies have attempted to explain and investigate namely Somaliland’s peacebuilding processes by approaching the analysis with a pseudo-hybrid lens (see Schwoebel 2007, 2018). Alas, as warranted as Njeri’s (2019: 39) critique is – claiming that “Somaliland is marginally mentioned within the discourse of hybridity” and where the author “posit this is a result of a state-centric approach to peacebuilding” – I argue that a preferable approach would be that of ZoPs when examining Somaliland and Puntland comparatively. In a sense, state-centric lenses and approaches to peacebuilding have been numerous concerning Somaliland, and to some extent on Puntland. However, the international community has endorsed few peacebuilding missions in northern Somalia. Ditto, theoretical explanations conceptualising hybridity risks rendering buzzwords; not sufficiently explaining Puntlanders and Somalilanders attained levels of peace and simply avoids decoupling peace- and state building altogether (Balthasar 2015: 28-29). To reiterate, a ZoP framework may result in a marginally different theoretical (and plausibly empirical) understanding concerning the general question why peace when war, certainly so in one of many understudied African contexts.
2.1. State (re-)building where there is no state
As previously mentioned, a number of studies have linked the relative successes of peacebuilding in northern Somalia to local and traditional efforts at state building (see Duffield 2014). The correlation and intimate relationship between state building and peacebuilding has largely been assumed as an unquestionable fact in the schools of the social sciences for quite some time. A more nuanced understanding – beyond state ‘failures’ or ‘successes’ – why peace occurs is arguably merited (see Brickhill 2010; Call 2008; Suri 2016; Haldén 2008; Njeri 2019).
11 As Boege et al. highlights, “there is a need to develop alternative non-state-centric approaches to governance, the control of violence, peacebuilding, and development” (2009: 14, emphasis added). This critique is especially striking when investigating how local and traditional methods may develop mechanisms of conflict resolution that does not warrant violence, consequently avoiding relapse into war, without a formal state apparatus supporting the peace zone. Thus, attempts at analysing such pockets of peace may favourably penetrate the veil of state-centred analyses, and rather investigate the particularities and uniqueness of ZoPs (Campbell & Peterson 2015: 340).
It is sensible, however, at this juncture to underscore and acknowledge the importance and actual virtues of state building. The continuous aim of the international community, namely the UN and AU, in bolstering intra-personal peace and reconciliation between competing Somali clans and warlords is not without its merits. Assumedly, neither Somaliland nor Puntland could ever have achieved such levels of security and peace (as to merit them ZoPs which this thesis assesses) without robust – albeit fundamentally defragmented and localised – (in)formal institutions absorbing the brunt of the burden which civil war applies (Zoppi 2015; Wiuff Moe & Simojoki 2013). This thesis undertakes no such analytical task, as it has been done prior to the onset of this study and scholars of social sciences and IR theorists will most likely continue detangling the web of hybrid- and traditional forms of peace in par with state-centric approaches in a similar fashion. It is here assumed that state-centred theories alone cannot explain peacebuilding processes, nor fully promote it, as it merely engages with an international, or state building, discourse rather than combining it with the local and informal. Besides, this combination results, more often than not, in frictions between the global-local (Björkdahl & Höglund 2013: 289) understandings on how to build lasting peace, i.e. the dichotomous relationship between ‘top-down’ and ‘bottom-up’ (Richmond & Mitchell 2012: 4-5).
2.2. Zones of Peace amidst armed conflict
Zones of peace. Indubitably, these three words evoke optimism and hope in most. But what exactly is a ZoP and what makes them safe? In its cradle the concept of ZoPs stems from attempts at civilian resistance against armed violence, where local non-combatants intended to create islands of sanctuary against armed violence within Filipino communities during the turmoil of the 1980s (Hancock 2018: 28; Hancock 2015: 237). In a contemporary context, where civil wars are more common than war between states, these peace zones are virtually
12 always created, implemented, and hopefully sustained in reference to an ongoing armed conflict, as observed by Christopher Mitchell (2007):
“Being “fought over” in recent decades has been reason enough for local people to seek some form of protection from violence by trying to construct a sanctuary. Being the targets of threats, unwilling recruitment, ejection, and killing is a major spur to finding or creating a place where whole communities are safe – even if only temporarily.” (Mitchell 2007: 3)
Inspirational for my analytical endeavour is Idler et al. (2015), where the authors compare varying outcomes of peace initiatives in two Colombian communities, and subsequently categorise eight factors into three sub-groups which they argue should be closely examined in order to analyse a ZoP (2015: 2). Following Allouche & Jackson (2019: 74-77) analytical mode – who operationalise their ZoP analysis on Sierra Leone and Côte d’Ivoire by Idler’s et al. three groups – I shall examine what peace(s) my two cases represent, via the three dimensions concerning the potential for, and continued existence of, ZoPs amid civil war:
1. The first concerns the characteristics of the internal resistance against outside violence, more specifically this includes the structure and cohesions of Somalilanders and Puntlanders communities within the peaceful zone. Informing this dimension are e.g. participation (in decision-making), collective leadership, and internal cohesion vis-à-vis outside, external threats.
2. The second feature to be disentangled is the local’s relationship with armed actors, or the potential violators of the sanctuary. This component is highly merited seeing how potential ‘spoilers’ to peace may be related or well-known to the subjects of a peaceful zone. Informing this dimension is e.g. what defines boundaries and trust in-between residents of a sanctuary, or what potential incentives there are for both outsiders and locals to violate the safe haven.
3. Thirdly, the role of external actors, i.e. peacebuilders and external interveners who do not reside in the sanctuary but impose agendas, for better or worse, upon the peaceful zone. This query is informed mainly by external actors’ peacebuilding programmes, policies, or financial and humanitarian aid objectives (patronage). Even more relevant concerning agency and autonomy: how locals interpret and respond to these global-local interactions and encounters.
A noteworthy assumption is that ZoPs are specifically and territorially defined sanctuaries, both empirical and theoretical (Hancock & Mitchell 2007, 2018; Hancock 2017). Sanctuaries are difficult to establish during civil wars, whilst non-combatants struggle to escape or find sanctuary in other geographical spaces – they may need to create one of their own. Bluntly, the mere prerequisite for a sanctuary denotes that there is a great and valid need for it (Mitchell 2007: 15-16). Generalisations on sanctuaries, or ZoPs, as categories is somewhat problematic, given the myriad of practices, organisations, and circumstantial specificities. Nevertheless,
13 going forward it should be emphasised that four varying, though interconnected, categories need to be taken into consideration: (i) the type of sanctuary, (ii) the functions of the sanctuary, (iii) the environment in which the sanctuary operates, and (iv) what form of protection the sanctuary offers (Mitchell & Hancock 2007: 189). These categories will assist in deducing whether or not my chosen cases can be defined as ZoPs.
A ZoP has one primary function: to protect those residing within it from outside violence, or negative peace as per Galtung’s theorem. However, there are also secondary functions for any ZoP: to promote economic development, evoke participation in societal functions, strengthen inclusion, and to formulate and implement justice- and policy reforms (Mitchell & Hancock 2007: 192-193), more in par with Galtung’s concept of positive peace. It may at first seem paradoxical, as sanctuaries are primarily established to provide protection to those inhabiting the spatial area which it encompasses. Nevertheless, in order to fulfil its primary function of negative peace, sanctuaries often seek to strengthen and improve societal, political, and economic structures – allowing for peace writ large to take hold (see Hancock 2017: 268). In other words, the strive for positive peace can become a means to an end in itself:
“it would be possible for particular ZoPs to be relatively unsuccessful at providing long-term safety for their denizens but highly successful at banishing corruption, improving living conditions, or developing a sense of unity, purpose, and shared identity.” (Mitchell & Hancock 2007: 194)
Notably, Mitchell & Hancock seek, in their widely cited book Zones of Peace (2007), to distinguish what defines the efficiency of a ZoP, or what Mitchell frames as: “the search for common factors that help to increase the chances of sanctuaries being effective in offering protection to those within the spatial and other boundaries of some protected and (hopefully) inviolable zone” (2007: 3). Additionally, the establishment of a ZoP – when defined by territorial boundaries – may be one of the more valuable methods a community or society can adopt when trying to mitigate or remove themselves from a conflict, rather than to resolve it per se. Thus, protracted and complex armed conflicts are more often than not difficult to manage long-term; whereas short-term solutions such as a temporary sanctuary can be enacted fairly immediately (Hancock & Iyer 2007: 29-30), which in turn may provide long-term shelter. Considering that ZoPs may either be locational – i.e. constructed to protect an area – or personal, Mitchell (2007: 1) identifies that a zone of personal protection includes “priests or other “holy” individuals, merchants and traders, doctors and other medical personnel, peasants,
14 women, and children”, ergo prominent persons, categories of people, or inviolable spaces (hospitals, UN compounds, schools, religious buildings). As this thesis foci on macro units of analysis, rather than micro, the personal category becomes irrelevant (though merited for other studies). Rather, the locational or territorial zone of protection are of analytic interest.ZoPs are often established and maintained within an ongoing violent conflict, or in a post-conflict setting. The zones duration may be temporary or permanent, and functioning agents of preserving the sanctuary may be local communities themselves, national and/or regional leadership, or even international peacebuilders and/or experts (Saulich & Werthes 2020: 42).
Worth mentioning is that the creation of a ZoP can be both top-down, or externally generated, or alternatively stem from a locally and collectively formulated need for a safe sanctuary (Mitchell & Allen Nan 1997; Mouly et al. 2015). However, the literature is in accord that the founding process, and subsequent survival, of a ZoP is far more likely to succeed if the ownership and control over the process is bottom-up, rather than top-down (Hancock 2018: 28; Allouche & Jackson 2019: 77; Saulich & Werthes 2020: 32). This is particularly interesting for this study, considering the relative lack of international attention both Puntland and Somaliland have received. Whilst recognising that external actors have “limits […] in local conflict management and prevention” and acknowledging “the fact that local communities are not at the mercy of external forces” (Saulich & Werthes 2020: 34), it remains prudent to assume that external assistance, or patronage, is merited for analysis. Evidently, external and international interventions – or lack thereof – is of import when examining any and all ZoP during armed conflict and war. Militarised, humanitarian interventions may fortify agitation towards the ‘outsiders’ and embolden resistance, as was the case of UNSOM (see Duffey 2000; Kapteijns 2013). In essence, there may be several risks attached when external peacekeepers and -builders seek to intervene in local ZoPs:
“neither peacekeeping nor even outright occupation can necessarily provide long-term security for people living in conflict zones; even further, these external controls on violence often backfire, cementing divisions between groups and fanning the flames of resistance.” (Hancock & Mitchell 2007: xiv)
On the other hand, external interventions – if appropriately implemented with a legitimate mandate – may strengthen a local sanctuary’s inviolability (Mitchell 2007: 22; Mitchell & Hancock 2007: 215-217). External attention to Somaliland and Puntland is arguably needed then, for them to stand a chance at preserving their statuses as relatively peaceful regions:
15 “The initiatives in the north need to be supported. Such external support, however, needs to
recognize the sensitivity of the recovery process. […] For the time being, external assistance must supplement rather than overwhelm the kinds of local grassroots initiatives that already exist.” (Farah 2001:143)
The instinctual urge to dismiss the role of external actors – seeing how they have been more or less absent since Somaliland’s and Puntland’s claims on secession and autonomy – should therefore be halted in its tracks: patronage remains a viable analytical factor in order to evaluate and compare these two cases. Incorporated in this thesis framework is Mitchell & Hancock’s modelling on the effectiveness of outside patronage (Appendix 2, Figure 4). In addition, the authors create a typology of sanctuaries to provide a simpler overview of how they may develop and shift functions over time (Appendix 2, Figure 3). Considering inviolability, ZoPs require non-interference from potential violators, i.e. the enemy, toward the sanctuary. Assuming that interactions between the national-local are even conceivable, the varying functions of a sanctuary can be categorised by analysing each case practices and procedures concerning how they realize positive peace within the zone, in combination with their relationship to the national (Mitchell & Hancock 2007: 196-197).
Noteworthy to reiterate: the local units of analysis are Somalilanders and Puntlanders – be they elites, armed actors, or clients for protection against violence. The outsider refers to national and/or regional neighbours, e.g. Somali clans and Jihadists. Thus, the relationship between the ZoP, local ditto outside armed actors, and patrons become indicative for the ZoPs inviolability and durability. In the introduction I assumed that outsiders may traverse across levels of adherence, whereas the local or global stays within their realm of influence. This may come across as an oversimplification, since both global as well as local actors can and so often do fluctuate between what is to be considered local (communities in the peace zones) and what is global (organisations or NGO offices). Yet, for the purpose of this comparison, Somalilanders and Puntlanders relationship with the outsiders, i.e. other Somalis, neighbouring African nations and UN peacebuilders, is what constitutes the local-outsider and local-global elements. The categories of analytical enquiries from Allouche & Jackson (2019) simultaneously informs both the inviolability and durability of a ZoP, albeit varying in size and scope. For example, the role of external actors in a sanctuary cannot be fully comprehended without extracting some information on how the sanctuary defies or cooperates with a national government. Nor would it be possible to obtain any solid information regarding the stated purpose of a sanctuary without coupling this to an “enemy” endangering the peace, or some other manifest threat to the zone.
16 2.3. The logics behind the creation, inviolability, and durability of a sanctuary
Irrefutably, the logics of a ZoP are to protect those residing within the zone against further or renewed conflict. Seemingly, sanctuaries are often unable to protect their boundaries or jurisdictions via the usage of force, but then how are ZoPs able to protect the denizens of a sanctuary? Crudely, it comes down to the outsider, or armed actors, agendas and motives. Mitchell & Hancock (2007: 203-204) conclude that armed actors may respect the integrity and inviolability of a ZoP if the region adheres to two fundamentals: lacking incentives for violation (i.e. few potential gains) and absence of perceivable threats (i.e. few potential risks). Outsiders are, arguably, more likely to respect the inviolability of a sanctuary if the zone does not harbour or support opponents to the armed actor; perceived as potential or future threats. If the sanctuary is sheltering enemies or potentially high value targets (e.g. wealthy individuals, wounded enemy combatants), and integrating them into the society where outsiders’ enemies, their supporters, and neutral denizens mix – the risks for violation against the ZoPs inviolability and integrity become more significant.
Furthermore, ZoPs does not exist within a vacuum: global actors and organisations play a crucial role in examining apparent peace zones amid war, where the aforementioned global-local interactions, or frictional encounters between external actors and Somalilanders and Puntlanders are produced and/or reproduced. Depicted in Table 1, informed by Björkdahl & Höglund (2013), are the frictional encounters and local’s response to the global-local interactions which may occur – and these will be investigated to nuance the outcomes on Somaliland’s and Puntland’s functions as peace zones. The local response toward global and liberal values envisages how peace is created and primarily governed. The external impact on Somaliland’s and Puntland’s visions of peace informs what outcomes these encounters have had on the peace processes, at large. First, what (potential) external patrons promotes becomes highly relevant: grassroot ownership and individual agency, or the elite’s narratives. Basically, who sways influence over the continuation of the ZoP, and who are excluded? Second, the local response toward potential frictions, which inadvertently affect the outcomes of peace processes, may further dictate the external patron’s iteration at influencing the sanctuary – thus creating a feedback loop (Björkdahl & Höglund 2013: 297-298). Noteworthy is that this feedback loop may (re)create new realities within the sanctuary, and consequently stipulate new issues or challenges for the zones’ inviolability and continued survival (2013: 296), making the frictions between global proposals and local responses of value for the analysis on why peace has lasted.
17 Table 1. Global-local interaction in peacebuilding (Björkdahl & Höglund 2013: 297).
To summarize, there are ten interconnected criteria, or logics, which inform the potential success or failure of a ZoP, i.e. informing the type and function of a sanctuary, environment where it operates, and what protection it offers (Mitchell & Hancock 2007: 212-213, 218). Notably, not every logic has to be fulfilled in order for a ZoP to occur and thrive - but needless to say it alleviates and assist the regions’ inviolability and durability:
1. Levels of internal unity concerning the sanctuary’s purpose, and effectiveness of non-violent conflict resolution mechanisms for dealing with divisions and disagreements within the sanctuary.
2. Scope and variation of development programmes beyond those projects involved in safe-guarding the sanctuary from armed actors (positive peace as means to maintain negative peace).
3. Effectiveness of the leadership and width of preparations for the sanctuary’s continuation if violated. 4. Territorial and spatial demarcation and sets of boundaries.
5. Spatial remoteness from or adjacency to armed conflicts and civil strife. 6. Impartial and neutral conduct from the sanctuary toward outsiders.
7. Legal and/or ethical basis and motivation for the existence of the sanctuary.
8. Threats posed by the sanctuary against local armed actors or outsiders, and presence of valuable resources, notables, or strategic properties that may incentivise violation.
9. External actors’ agenda in providing patronage, protection, or support to the sanctuary. 10. Levels of credible sanctions against violators of the sanctuary.
Hence a sanctuary can be (i) identified, (ii) compared against other zones, and additionally (iii) evaluated based on its properties (inviolability and durability). In sum, this study will analyse what defines Somaliland’s and Puntland’s peace processes internally, the regions relationship with armed actors and outsiders, and to what extent these peace processes have been influenced or devised by external actors. This approach allows for a focus on both the creation of the zones, and additionally on the inviolability and durability of each peace zone. I define ZoPs – or sanctuaries – analytically by four categories (Hancock & Mitchell 2007), and measure in
18 chapter 3.4 their inviolability and durability by cross-referencing Allouche & Jackson (2019) and Björkdahl & Höglund (2013). To recap the research questions, the framework will be utilised in order to (1) determine whether or not Somaliland and Puntland can be characterised as ZoPs – how do they differ and what similarities do they share? And (2) conclude how both regions have attained internal peace despite the Somali civil war – why have they remained durable and inviolable safe-havens?
3. Comparative method and guidance of framework
This thesis is conducted as a qualitative and comparative analysis with a macro lens, informed by a variety of material much due to the inaccessibility of first-hand sources and accounts on site. The query whether or not Somaliland and Puntland can be described as ZoPs is interesting, yet not unproblematic to assess without direct access to each region. The framework is, however, formulated in such a way that a desk study is feasible through analysis of qualitative secondary sources. Interviews, for example, may not necessarily be the best approach, as noted by Höglund & Öberg (2011: 3): surveys and interviews are commonly used techniques for information gathering, yet viable alternatives such as desk studies are readily available. As the topic of interest for this thesis are sanctuaries and their logics, informing terms concerning the theoretical and methodological approach requires rigid conceptualisation and operationalisation – i.e. variables need to be defined and measurable – prior to the analysis (Höglund & Öberg 2011: 185).
Hence, this thesis adopts a theory-guided and comparative approach to decide whether or not the regions may be defined as sanctuaries, and what their respective causes of peace potentially are. Moreover, considering matters of ontology and epistemology, I assume that ZoPs are “an external reality which exists independently of people's beliefs or understanding about it” (Snape & Spencer 2003: 11). Thereby, peace zones may exist with or without third-party acknowledgement and labelling, albeit that these factors assumedly influence the zones inviolability and durability. Alternatively, as a ZoPs existence necessitates some levels of legitimacy, they may aptly be considered constructed external realities: a sanctuary is, after all, not a natural phenomenon but created. I assume that the analysis will provide this thesis with intersubjective results, ergo if “several reports confirm a statement then it can be considered true as a representation of a socially constructed reality” (Snape & Spencer 2003: 14). Therefore, this theory-guided exercise may, primarily, shed some light on the application – or efficiency – of ZoP theory on as mentioned earlier understudied African cases.
19 Prior to the methodological review of this paper, some key acknowledgments have to be made, informed by Bhattacherjee (2012: 94-95), as this comparative case study sets out to: (i) explain why and how Somaliland and Puntland have achieved peace where war rages; (ii) examine multiple units of analysis, i.e. local, national, regional, and global actors in a macro perspective; (iii) perform a multiple case analysis in order to enable comparison; (iv) the sites chosen for this comparative case study are the Puntland and Somaliland regions, sharing as aforementioned numerous similarities; and lastly (v) the material collected is extracted via secondary sources including, but not limited to, previous research, reports, news articles, and policy papers – ergo the technique of document analysis is partly applied.
3.1. Research design
Seeing how this is a theory-guided effort with a comparative disposition, the framework of ZoP remains instructive for formulating a research design and delineate my field of inquiry. The choice of cases allows for comparison within the cultural and state context of Somalia, where pre-existing in-depth knowledge on Somaliland and Puntland is required of the researcher (Höglund 2011: 117). The administrative units and armed actors for analysis are at times fluid, which necessitates caution and analytical transparency on my part (see chapter 1.2). Nonetheless, the scientific rigor required in peace- and conflict studies in general, and on analytically complex cases specifically, is adhered to as envisaged in Table 2. Rihoux & Marx (2013) seminal and informing work implies that potentially explanatory factors (cf. Table 2) may produce an intriguing phenomenon, or in this thesis the outcome of sanctuary and peace in the midst of civil war. This thesis also assumes that varying conditions, i.e. the logics of ZoP, may result in the same outcome which needs to be explained, i.e. the inviolability and durability of a sanctuary. Moreover, “depending on the context, a given condition may very well have a different impact on the outcome” (Rihoux & Marx 2013: 168), indicating that if – hypothetically – Puntland and Somaliland share similar conditions, the outcome of relative peace is not necessarily causal but requires a more nuanced depiction on what their differences and similarities are.
Thereby, the purpose of this comparative study is to conclude on whether or not Somaliland and Puntland conform to the logics of ZoP, and subsequently answer why the regions have maintained peace instead of receding into civil war. As Höglund (2011: 115) denotes: a comparison between two entities may be sufficient for such an endeavour, at least for the single researcher since the specific field of research require vast knowledge and understanding of the
20 local context. The observable phenomena further strengthen the choice of (only) comparing two cases of peace amid conflict, as mentioned earlier: Puntland and Somaliland are two extraordinary discrepancies in an otherwise conflict-ridden context. What is yet to be satisfactory answered beyond state-centred explanations is the why. The research design for this comparative study has three particular issues – beyond the conceptualisation of ZoP in the previous chapter – that need to be addressed prior to the analytical effort: first the selection of cases; second sorting through the material and decipher how it may answer my research questions; and third the operationalisation to enable measurement of potential explanatory factors (Höglund 2011: 116).
3.2. Case selection
In order to analyse and compare Puntland and Somaliland as peaceful phenomena – ostensibly similar yet different – this thesis applies what Höglund (2011: 116) defines as an appropriate frame of comparison. The frame entail, first, that various outcomes may be identified in each case, whereas both peace zones have experienced (roughly) similar levels of violence prior to the sanctuary’s creation. Secondly, case homogeneity must be met: both cases may be compared because they share similarities in the procedures that allows protection and creates peace (2011: 116). Therefore, I attempt to discover linkages and connections between my two cases, further strengthened by Ritchie et al. (2003: 248) who highlights that this is a common approach in qualitative and comparative analyses – as linkages may occur between both highly similar ditto fundamentally different cases. For my purpose, both the thematic, or otherwise observable, similarities and differences are of interest. Notably, when addressing the issue of case selection, I assume that a single case may be examined as numerous events, or what Levy (2008: 2) coins “spatially and temporally bounded set of events”. The theoretical framework adheres to this assumption, as it covers both clearly defined temporal and spatial limitations; the entirety of Puntland’s and Somaliland’s strive toward internal and lasting peace will be analysed following the Somali state collapse of 1991 where Somaliland and Puntland are acknowledged as two Somali administrative entities roughly spatially equal.
Considering that this thesis is theory-guided, rather than inductive or hypothesis-generating, it entails that the analysis becomes idiographic, with the “aim to explain and/or interpret a single historical episode rather than to generalize beyond the data” (Levy 2008: 4). The idiographic lens (1991-) thus becomes applicable regarding Puntland’s and Somaliland’s formation as relative peaceful regions in adjacency to the continued civil war of Somalia in general.
21 However, caution should be raised concerning the potential generalisation of this thesis: there are obvious difficulties in generalising vastly different realities on the ground and produce some concrete statements considering the usage of ZoP theory, if not on sanctuaries differences and similarities. Nevertheless, this comparative approach allows for in-depth insights on each case simultaneously to some levels of generalisation (Rihoux & Marx 2013: 168), albeit more theoretical than empirical per se.
To reiterate, Somaliland and Puntland are apt to be examined via this comparative approach, rather than against other non-African geographical and cultural contexts. Somaliland and Puntland have, albeit varying in scope, largely separated themselves from the rivalry clan politics and violent conflicts of central- and southern Somalia. There are other examples of secessionist elements and autonomous claims in other parts of the country, but these entities have not realized their claims to the same extent as Somaliland and Puntland have – nor are there equally substantial levels of peace and security in the remainder of Somalia. The generalisation that may occur should therefore be framed in regard to the theory of ZoP, but only in parts and with caution on empirical observations of ‘why peace’ elsewhere. In sum, a potentially different approach, considering the usage of the ZoP framework, could have incorporated a comparison between previously identified ZoPs – e.g. local Colombian or Filipino communities – and one or both of the two chosen cases. Alas, this study does not embark on such an investigative journey, as previously denoted: the choice of understanding causes of peace rather than war is underexplored, certainly so concerning sanctuaries against violence amidst civil war in a general African context. Furthermore, the two cases share religious, linguistic, and cultural homogeneity as the populations are predominantly Somali nationals (Ismail 2010; Samatar 2001; Prunier 1996) – making them ideal for comparison.
3.3. Material
The upcoming analysis revolves around a vast array of secondary source material, including news articles; reports by NGOs and other organisations; Somaliland’s and Puntland’s constitutions; and most extensively utilised previous research, published by peer-reviewed scholars who have often conducted on-site studies in either region. Most sources are in electronic print, with the exception of various books, which is a widely used form of empirics gathering in peace research (Dulić 2011: 36). Furthermore, the usage of printed sources, albeit secondary in nature, offers the analysist certain opportunities to perform triangulation on the material, i.e. collecting information and verifying the results from several independent sources
22 (2011: 39). Alas, as marked out by Dulić (2011: 36-37), published information readily available on the internet is not without its problems, even if it alleviates triangulation; the authenticity of a source may be hard to verify. This is taken into consideration, as all sources besides prior research (e.g. documents, reports, and news articles) are treated with caution considering data veracity.
Dozens of peer-reviewed scholarly articles and books compose the majority of material that will be analysed, but some examples of other utilised sources include Interpeace, Pillars of Peace, Puntland Development Research Center (PDRC), UN accounts, International Crisis Group (ICG) reports; agreements and negotiated settlements between Puntland, Somaliland, and Somalia retrieved from UN databases; and news publications from Reuters, the British Broadcasting Corporation (BBC), and the Combating Terrorism Center Sentinel (CTC). In addition, concerning the reliability – or the replicability and preciseness of analysis – and validity – or the accuracy of material being analysed – of this thesis, I acknowledge the fact that a perfect measurement cannot be developed in full (Sundberg & Harbom 2011: 98-100). As noted by Levy (2008: 12), if either compared case evidently are examples of ZoPs – and subsequently follows the set of logics presumed – it would increase the confidence for ZoP theory, however incrementally. I argue, furthermore, that the framework may sufficiently balance between what is considered too ‘narrow’ or too ‘broad’ as it explicitly guides my effort on what is of theoretical interest: the generated assumptions of theory-guiding studies allows for some flexibility, as the framework guides the analysis rather than the empirics.
3.4. Measuring Zones of Peace underlying logics
In order to conclude if Somaliland and Puntland may be typified as ZoPs I will first determine how to characterise them and second unearth their levels of inviolability and durability, where the complete list of analytical questions is provided in Table 2. First, as previously mentioned in the theory section, types of sanctuaries can either be locational or personal (see Appendix 2, Figure 3) but are for this study presumed locational. The usage of ZoP often indicate, as previously mentioned, a geographical aspect of clear delineation (Mitchell 2007: 2), as per case the two Somali regions offer. The locational, or spatial, type of sanctuary can be two-folded: intra- or extra-societal. The intra-societal sanctuary is situated within clarified jurisdictional and geographical boundaries, located in proximity to or amongst the threats (i.e. outsiders) against those that reside within the ZoP. The extra-societal sanctuary is where geographical and jurisdictional boundaries are clearly demarcated yet located at a distance from the threat against
23 those sheltered in the sanctuary (Mitchell & Hancock 2007: 190). Thereby, this thesis will examine whether the two cases are examples of intra- or extra-locational ZoPs. Furthermore, if they exhibit traits as intra-societal sanctuaries, the issue of longevity becomes relevant: is the sanctuary (i) traditional, i.e. has the zone lasted a long period of time and offers protection to a wide range of peoples; or is the zone (ii) constructed, created fairly recently and/or ad hoc to meet specific threats aroused from immediate dangers and hazards, e.g. a nearby battle (Mitchell & Hancock 2007: 191). Furthermore, who founded the sanctuary and how the peace zone is legally or ethically motivated are indicative concerning how the ZoP may be perceived by armed actors and outsiders residing without the sanctuary.
Second, the varying functions of a sanctuary need to be identified. Peace writ large, applied in this thesis synonymously to positive peace, entails peace on a societal-level; contributing to larger, collective, goals that involve each and every one residing within the sanctuary (Chigas & Woodrow 2009; Hancock 2017). Connecting the functions of the sanctuary, Puntland’s and Somaliland’s levels of attained negative peace – and more importantly strive toward peace writ large, or positive peace – needs to be measured by extracting information concerning what makes the regions distinctly peaceful in context of civil war. Enlightened by Abdi (2012: 60-61), the zones levels of democratic participation and claims on monopoly of power informs the over-arching collective vision of peace (i.e. goals which involve the entire zone to maintain inviolability and durability). As noted in chapter 2.2, a ZoP cannot stay durable nor inviolable without a strive towards positive peace, and to measure this I will investigate each zones’ commitments to community-based reconciliation efforts and non-conflictual settlements – both factor’s indicative for the aim to keep a negative peace (Mitchell & Hancock 2007: 192-193). Moreover, the leadership can be either collective or elitist, informative for the zones’ levels of unity and cohesion and subsequently whether or not denizens abide by the formulated rules of conduct for the ZoP. Additionally, ZoPs may combine efforts of local peacebuilding with processes at the national level: indicative for this is whether or not there is mutual recognition of legitimacy, and if there is a functioning relationship with the outsider (e.g. negotiations prior to coercion). Concerning the local response to global values, as noted in chapter 2.3, Table 1, where if the local response is compliance, the sanctuary is characterised by submission to international agencies values and principles. If the response is adoption, the sanctuary willingly adopts global values and functions in order to attain more positive peace. If adapted, the sanctuary intertwines global norms with traditional values, e.g. hybrid governance. Co-opted sanctuaries strategically soften their traditional principles in order to maintain patronage or
