Counterinsurgency as order-making: refining the concepts of insurgency and counterinsurgency in light of the Somali civil war

ABSTRACT In this paper, we argue that the current insurgency and counterinsurgency discourse is dominated by concepts that are too narrow and too isolated from the wider civil war literature within which insurgency and counterinsurgency occur. Rather than accounting for the complex political processes and wide range of forces and actors that shape conflict dynamics, the dominant insurgency and counterinsurgency debate tends to reduce highly messy contexts to a competition between the often false dichotomy of insurgents and counterinsurgents, usually understood as the state versus one or more non-state violent actors. In order to understand civil war contexts like South-Central Somalia, we argue that orthodox reductionist concepts and assumptions underpinning the dominant insurgency and counterinsurgency discourse provide limited value. Building on recent critical literature, the paper proposes a refined conceptualization. Instead of understanding insurgency and counterinsurgency as peculiar forms of war, strategies, or sets of guerilla tactics, we follow Jaqueline Hazelton’s line of thought, suggesting that insurgency and counterinsurgency are mere elements of a broader process of violent order-making. Thus, insurgency and counterinsurgency are, in our view, comprehensive processes of organized challenge to and consolidation of established political order within the context of civil war.


Introduction
Civil war 1 is a complex, messy, and intensely political phenomenon whereby established orders are challenged and new ones produced. 2 The scholarly undertaking to understand, explain, and at times predict such messes is difficult at best, yet the prevalence and impact of the former, current, and future civil wars compel us to do our best. In this quest, the academic literature on insurgency and counterinsurgency which quickly lost popularity after US failures in Iraq and Afghanistan, has, nevertheless, provided vital insights into and knowledge about why insurgencies emerge, how they endure and how they end 3 , and how governments succeed or fail in defeating them. 4 However, the insurgency and counterinsurgency discourses are dominated by concepts that are too narrow and too isolated from the burgeoning literature on the wider phenomenon of civil war within which insurgency and counterinsurgency occur. 5 Rather than accounting for the complex political processes and wide range of forces and actors that shape conflict dynamics, the dominant insurgency and counterinsurgency debate tends to reduce highly messy contexts to a competition between the often false dichotomy of insurgents and counterinsurgents, usually understood as the state versus one or more non-state violent actor. In reality, governments and violent non-state actors are only part of a conglomerate of other actors, institutions, and forces shaping conflict dynamics. The recognized governments themselves may even be the ones challenging order exercised by a 'non-state' armed group, 6 or they may cooperate with one or more non-state challengers in some areas while fighting them in others. 7 Who is the challenger ('the insurgent') and who is the ones consolidating established order ('the counterinsurgent') may fluctuate considerably throughout the same conflict over time. In civil war contexts like South-Central Somalia, recognized state authority has been either absent or at best limited, and de facto dominance exercised by militant Islamist groups. We argue that orthodox reductionist concepts and assumptions underpinning the dominant insurgency and counterinsurgency discourse provide limited value in order to understand these kinds of conflicts.
Building on recent critical literature on insurgency and counterinsurgency 8 , the paper proposes a refined conceptualization of both terms and points out how these may increase our understanding of civil war dynamics. Instead of conceptualizing insurgency and counterinsurgency as peculiar forms of war, strategies, or sets of guerilla tactics, we follow Jaqueline Hazelton's line of thought 9 and understand insurgency and counterinsurgency as elements of violent order-making. Thus, insurgency and counterinsurgency are, in our view, comprehensive processes of organized challenge to and consolidation of established political order within the context of civil war. By emphasizing the politics of order-making, the dynamics of violence is first and foremost seen through the lens of domestic power competition, whereby the intensity of violence and who is 'insurgent' and 'counterinsurgent' largely follow the dynamics of elite competition and the effects of elite bargaining on the ground. The paper's arguments draw on a review of scholarly insurgency and counterinsurgency literature in combination with insights acquired from qualitative studies of the Somali civil war, a case largely overlooked by the wider insurgency and counterinsurgency discourse. 10 Our analysis of the Somali conflict relies extensively on the authors' long-time engagement with Somalia, including the main author's previous fieldwork in southern Somalia and Kenya, as well as recent interviews with key Somali politicians conducted by the co-author. The first part of the paper discusses existing conceptualizations of insurgency and counterinsurgency and suggests a refined approach. The second part applies the refined conceptualization to an analysis of the Somali civil war and demonstrates how these concepts make us better suited to comprehend key features of the highly complex civil war context compared to orthodox conceptualizations.

Method
Our attempt to understand and analyze the complexities of ordermaking in the violent Somali conflict will naturally be influenced by our own notions and assumptions about groups, dynamics, and mechanisms, thereby adding to the inherent difficulties in conducting empirical research within conflict and post-conflict zones. Narratives from politicians, militia leaders, foot soldiers, and 'ordinary citizens' about incidents and processes often vary considerably, even among those who were directly involved, depending on factors, such as an individual's political position, clan affiliation, authority, and gender. Yet, while we carry with us our own preconceptions and biases, we believe our long-time engagement with the Somali conflict, including more than a decade of work and field work in Somalia and Kenya, helps us, at least to some degree, to navigate this messy landscape.
The paper draws on scholarly debates on insurgency and counterinsurgency as well as on the growing literature on violence and order-making within the context of civil war. Our analysis of insurgency, counterinsurgency, and order-making in southern Somalia is informed by a combination of written media reports, NGO/GO reports and semistructured interviews by both authors: 45 interviews conducted by the main author in Kismayo, the capital city of Jubaland between 2018 and 2020 11 , and 10 interviews conducted by the co-author between November 2022 and January 2023. 12 Due to the enduring instability and uncertainty on the ground in southern Somalia, every interview subject has been anonymized in order not to inflict unnecessary risk to their safety and wellbeing.

Conceptual debates on insurgency and counterinsurgency
There is an abundance of literature explaining how dominant conceptualizations of insurgency and counterinsurgency spring out of the thinking and writings of colonial officers, especially their experiences of the twentiethcentury anti-colonial struggles and the rise of independence movements inspired by left-wing ideologies. This line of thinking, which later came to influence US and Western counterinsurgency discourse after 9/11, posited that insurgency was a peculiar kind of warfare, different from 'conventional' war, i.e. continental war in Europe between industrial armies 13 , notions which Steven Metz refers to as the 'orthodox' conceptualizations of insurgency and counterinsurgency. 14 The orthodox view of insurgency and counterinsurgency has been extensively criticized since its 'renaissance' in the mid-2000s. According to critics, an obvious problem was the small number of cases, such as the British experience in the Malayan Emergency and the French decolonialization struggles in Indochina and Algeria, on which counterinsurgency theory was built. 15 Another, and far more profound, critique rests on the ideological and theoretical assumptions underpinning the orthodox notions, which, as demonstrated by several critics, are largely dominated by modernization theory. 16 Such notions reflect ideas about civil society and politics born out of the European Enlightenment, where a government should be 'of the people', with 'the people' defined by citizenship of a nation. A legitimate government is one that seeks the best interests of a majority of its citizens. Whenever this does not happen, conflict and violence will occur. Hence, insurgency is a symptom of 'bad governance' and the cure is to increase a government's legitimacy through various modernization projects in the form of political, social, and economic reforms. 17 The paradigm of modernity was so profoundly taken for granted within the orthodox view that any opposition to counterinsurgency as a modernizing project was, according to Metz, 'attributed to evil or at least misguided people. The solution was to eradicate the evil and educate the misguided'. 18 While the orthodox paradigm of insurgency and counterinsurgency is currently highly contested, the two concepts are still commonly treated as binaries, where counterinsurgency is largely seen as the response to insurgency. Any definitions of the concept of 'counterinsurgency' are therefore often based on how one defines 'insurgency'. 19 No commonly accepted definition of 'insurgency' currently exists in the scholarly literature. However, the debates tend to revolve around definitions of insurgency as either a peculiar form of war, as propagated by the orthodox paradigm, a strategy, or as mere operational techniques or tactics. 20 While state actors, as Daniel Whittingham and Stuart Mitchell point out, may apply insurgent methods 21 , there is a clear tendency to identify insurgency with the weaker party of the conflict and counterinsurgency with the stronger, often set up as a dichotomy between the non-state challenger and the incumbent state regime. 22 This duality is well illustrated by David H Ucko in his recent book, The Insurgent's Dilemma, where he bluntly defines 'insurgency' as 'a politico-military campaign against the writ of the state'. 23 M L R Smith and David Martin Jones are highly critical of the dominant tendency to treat insurgency as a specific form of war. According to them, by categorizing insurgency as a subset of war based on the observations of non-state violent actors using guerilla tactics within an intra-state war, one runs the risks of losing sight of the inherent complexities and political realities of the specific conflict at hand, including the motivation, interests, and opportunities of the opponent, subsequently hampering any efforts to formulate a suitable counterstrategy. If one perceives insurgency as a set of symptoms of a distinctive kind of war, and counterinsurgency as a general cure, one may tend to facilitate inflexible, decontextualized, and destrategized policy responses. 24 According to Smith and Martin, the concept of insurgency should thus not be understood as a type or subset of war, but rather as a mere synonym for guerilla tactics, i.e. what can be described as asymmetric methods, such as hit-and-run attacks, ambushes, and so on where direct and sustained confrontations with the enemy are avoided. 25 If insurgency is a faulty concept, or at best a category of descriptive tactics, what then is the meaning of the concept of counterinsurgency, perhaps besides a set of techniques or tactics to counter guerilla methods? According to critical perspectives, orthodox counterinsurgency conceptualization is nothing but a (misguided) narrative, a story produced by cherrypicking historic examples of 'counterinsurgent victories' and tailored to suit US and Western policy discourse by seductive ideas like 'hearts and minds', 'armed social work' and 'restrained level of violence'. 26 While the modernizing, technical counterinsurgency discourse may sound plausible and tempting, promising a universal panacea for insurgencies regardless of context, the politics of war, according to Smith and Jones, 'is discounted in favor of the application of the rationalist technique of an instrument to guide conduct'. 27 An instrument which ultimately may as well fuel rather than defuse conflict if you do not understand the political context wherein you are operating. 28

Counterinsurgency as violent order-making
Would we be better off dropping the concept of counterinsurgency altogether? Reflecting upon this possibility, Ucko points out that if this was the case, a precondition would be that the notion of 'conventional' warfare, to which counterinsurgency thinking was in many ways a counterreaction, is not allowed to dictate dominant perceptions of war. Instead, war should be seen as a highly complex and intensely political phenomenon which often hosts both 'irregular' and 'regular', or 'conventional', features. Furthermore, he adds that the many lessons learned from past and current counterinsurgencies, and discussions that follow in their wake, are of broader value for understanding violent conflict and should therefore not be forgotten. 29 In a sense Hazelton, picks up the baton from Ucko and other critics 30 in her recent treatment of counterinsurgency. 31 However, she does not conceptualize counterinsurgency as 'mere' war but treats it more broadly as primarily a domestic political process of violent state-building. While international actors and states may intervene in intra-state conflicts, political order largely arises from domestic elite efforts to come out on top in violent political rivalry. Contrary to assumptions of the orthodox counterinsurgency discourse, Hazelton stresses that elite groups rule to protect their own interests, not those of the populace. Once elites have determined through violent competition which of them will dominate the rest, and at what cost to which actors, then political stability will follow as long as the elite bargain holds. 32 Hazelton's conceptualization of counterinsurgency as domestic statebuilding has several theoretical implications. First, 'insurgency' is first and foremost to be understood as an element of violent domestic state-building processes in the form of the actors striving to seize or transform the established or dominant political order. Second, it does not distinguish between the weak and the strong or reduce counterinsurgency to a specific type of strategy or tactic which, as previously noted, may fluctuate throughout the same conflict. Third, it emphasizes the political nature of violent struggle. How the established authorities decide to manage challenges to existing order largely depends on the dynamics on the ground. For example, it may be that authorities are focused on other, more pressing issues than on fighting the insurgents, such as accommodating local elite groups, if insurgent activity is not seen as a vital threat to the dominant order. It may even be the case that the authorities care little about insurgent violence and atrocities against civilian communities as long as the mechanisms upholding their power position remain in place. In what Ucko 33 terms rural 'localized insurgencies', the violent challenger may thrive in the periphery, controlling considerable territories and populations, as long as it keeps below a certain threshold and is not seen as a direct threat of the incumbent state's elites. Or the opposite may be the case, i.e. that the established authorities are willing to crush the insurgents by any means possible if they feel sufficiently threatened, as in the case of the Tamil Tigers (LTTE) in Sri Lanka. 34

Insurgency, counterinsurgency, and civil war
Hazelton's armed domestic state building and Ucko's 'localized insurgents' operate within a context of civil war, understood in broad terms as a 'situation where a government faces a violent challenge by one or more rebel actors over some articulated goal of political change'. 35 We strongly disagree with Kaushik Roy's 36 separation of the concepts of insurgency and civil war as the effects of insurgents' use of guerilla (irregular) tactics versus conventional (regular) tactics as the key variable. Yet, we follow his argument that an emerging insurgency may (or may not) escalate to the level of civil war. 37 Leaving aside the debate on civil war threshold and how to quantiatitively demarcate it, 38 Stathis Kalyvas makes a useful distinction between an emerging insurgency or rebellion, i.e. armed challenge to the state, on the one side and civil war on the other, where insurgency is one of several potential causes for civil war, alongside, for example mass protest, military coup, or intercommunal violence. 39 Hence, while there can be an (emerging) insurgency without civil war, no civil war exists without an insurgency: when civil war is in effect, dominant order (often the state) is always challenged by one or more insurgent groups. Who is the challenger and who is the opponent may, however, fluctuate throughout the conflict, as in southern Somalia.
Based on this understanding of insurgency and civil war, one can study insurgency as an independent phenomenon from civil war, for example origins of an insurgency, the causes of insurgency etc. However, given a broad understanding of civil war, most visible insurgencies and hence counterinsurgencies operate within the larger context and complexities of civil war. This realization urges research on insurgency and counterinsurgency to closely interact with and benefit from the extensive civil war literature and rich debates on political violence more broadly. For example, where orthodox counterinsurgency discourse treats 'the population' as an aggregated and oversimplified category, which are to be won over by coercion and/or provision of governance and socioeconomic reforms, there is a quickly growing body of literature within civil war studies demonstrating the multifaceted roles of civilian communities. 40 Recent insights from this literature explores, for example, how the quality and efficacy of local institutions may extensively influence relations between civilian communities and armed groups and increase civilian agency and leverage. 41 Recent civil war literature also suggests that the dichotomy between nonstate and state may be unclear -and even unnecessary for meaningful analysis. There may well be cases where the internationally recognized government is actually the one that challenges the established order in territories ruled by 'non-state' authorities. Who the challenger and defender of established order is may fluctuate widely throughout the conflict. 42 Following this line of thought, we define insurgency and counterinsurgency as the comprehensive process of organized challenge to and consolidation of established political order within the context of civil war. In this perspective, studies of insurgency will not necessarily be focusing on the non-state challenger(s) of recognized states but rather on the broad processes, both violent and non-violent, that challenge dominant order. Likewise, studying counterinsurgency implicates exploring the actors, institutions, and forces striving to safe-guard the status quo within the wider context of civil war. Recognized governments, non-state armed groups, non-combatants, criminal organizations, intervening organizations, and states, etc. all play into these processes in various ways, some spurring change, others striving to defend established order. Instead of focusing solely on how a government with or without external support moves on to defeat its non-state challenger-(s), in line with orthodox views, studies of counterinsurgency now become a broad approach to grasping how established forms of order in civil conflicts may or may not withstand armed challenge, including political and military strategies applied by the dominant players, be it states or non-state groups.

The case: civil war and the construction of state authority in Somalia
Conceptualization of insurgency and counterinsurgency as violent domestic order-making highlights the complex political processes surrounding elite competition, where outright fighting between established authorities and one or more opponent may be only one, and perhaps less important, manifestation of the ongoing contest for dominance. 43 In southern Somalia, where state authority has been more or less totally absent since the downfall of the dictator Siyad Barre in January 1991, lasting authority has not been the rule. 44 Authority and order have been continually co-produced and reproduced by a wide variety of actors and institutions contesting dominance. 45 The ones consolidating established order (the 'counterinsurgents') and the ones challenging it (the 'insurgents') have varied extensively over time within a 'mosaic of power' 46 , or what Ucko 47 terms 'heterarchical governance', denoting a multiplicity of political power structures coexisting within one national space, and whereby the strategy and tactics applied by the actors have changed according to developments in the conflict dynamics.
Despite multiple, and often rivaling, internationally sponsored peace and reconciliation conferences, one cannot arguably speak about any sort of state authority in southern Somalia 48 until the Djibouti process in 2000, the tenth internationally sponsored attempt, when the Transitional National Government (TNG) was created. Gathering some of the Mogadishu-based factions, several civil society groups and various religious groups, the TNG was, nevertheless, a short-lived venture, soon challenged by a competing process in Eldoret, Kenya in 2002.
After two years of negotiations, the Eldoret initiative culminated in the empowerment of long-standing faction leaders, headed by the notorious militia leader from Puntland, Abdullahi Yusuf, and the establishment of an internationally recognized Transitional Federal Government (TFG). 49 On the same grounds as its predecessor, the TNG, it is fair to label the TFG a 'paper state', constituting primarily a specific segment of elite groups that exercised limited authority and focused on external funding and expected spoils rather than on administering the state. 50 While there are few reasons to believe that TFG was inherently better suited to consolidate state authority in Somalia than the TNG, it had one major advantage: the rising fear of Islamism and terrorism in the wake of 9/11. Having been listed by US as a potential 'breeding ground for terrorism', renewed US and international interest emerged in establishing a government in Somalia that could keep 'malign' forces in check. In contrast to the peace conferences held in the first years following the UN withdrawal in March 1995, which had primarily been driven by neighboring states and the Arab League, the Eldoret process had broadbased backing from the regional organizations as well as from the US, the UN, and the EU. 51 However, the TFG and an additional US-sponsored counterterrorism alliance of Mogadishu-based faction leaders failed to hamper the emergence of a powerful counterforce consisting of business leaders, powerful clan factions, civil society, and a loose alliance of Islamist groups which came to be known collectively as the Islamic Courts Union (ICU). After a few months of intense fighting in Mogadishu, the ICU defeated all opposition and by September dominated most territories in South-Central Somalia. 52 Alarmed by Islamist takeover of South-Central Somalia, Ethiopian forces with US support intervened in December 2006 and quickly defeated the poorly organized, trained, and equipped Islamist forces. However, despite protecting the TFG in Mogadishu and dominating the general military terrain, the Ethiopian and TFG forces were facing mounting resistance, spearheaded by a loose alliance of the more radical Islamist elements from the previous ICU. When the Ethiopian forces finally withdrew in January 2009, the Islamist alliance already dominated large parts of the South-Central territories and pushed for final takeover of Mogadishu. Had it not been for the quite small contingent of well-trained and armed African Union (AU) soldiers who remained in the city after the Ethiopian withdrawal, the Islamists would most likely have conquered the capital. Clinging on to a small enclave in the city center of Mogadishu, however, the TFG was unable to exercise any meaningful authority throughout Somalia. The Islamist alliance, increasingly dominated by al-Shabaab ('the youth'), on the other hand, exercised de facto authority 53 in South-Central Somalia. 54 Through their quick installation of a functioning, though brutal, justice system and a rudimentary governance system and their pragmatic approach to cooperating with traditional authorities, the Islamists established relative peace and stability 55 , though, not surprisingly, without any kind of international recognition.
However, Al-Shabaab's Islamist 'state' came under increased military pressure in 2011 and 2012. First, the Ugandan and Burundian AU forces under the African Union Mission in Somalia (AMISOM) went on the offensive in Mogadishu, conquering most of the capital. Second, in October 2011, the Kenya Defence Forces (KDF) intervened in southern Somalia, pushing to conquer the strategic city of Kismayo and establish a buffer zone along the Somali -Kenyan border. Third, a month later, in November 2011, Ethiopian forces re-entered Central Somalia and opened up a third front, prompting al-Shabaab to retreat from major cities. From having approximately 8,000 troops in August 2010, with parts of the Ethiopian forces finally under its umbrella, AMISOM numbered around 22,000 troops by January 2014. 56 Parallel with and partly due to tactical military progress by AMISOM, renewed political initiatives were fostered. After extensive international pressure, monitoring and financial support, a new federal parliament was selected and recognized in late August 2012, followed by the election of a federal president in September. However, despite establishment of new institutions, increased international political interest and financial support, and articulated visions for a federal system, the Somali Federal Government (SFG) was still largely a 'paper state'. Security was contingent on the presence and capabilities of AMISOM, and SFG's authority was extremely limited, even within Mogadishu, not to mention in other regions of Somalia, where political processes for the establishment of federal member states were already ongoing with limited influence from Mogadishu. 57

'State', 'non-state', and elite competition in Jubaland
Within a context of profound fragmentation of the nation-state, state actors and non-state armed groups often engage in complex overlapping authority relationships, not only over means of violence, but also over authority and governance. 58 According to Malmvig, 59 the 'non-state' prefix may therefore in many ways be a misnomer, obscuring insights into how closely interwoven non-state armed groups are with state or state-like power. Instead of perceiving a civil war situation as a zero-sum game between the state and contesting non-state actors, as in orthodox counterinsurgency discourse, the nationstate may host a wide variety of power-constellations and forms of order shaping both patterns of governance and violence. 60 The southern provinces of Lower Juba, Middle Juba, and Gedo, the region later to become the Federal Member State of Jubaland, have witnessed an ongoing contest between competing elite groups in the form of clan-based militia alliances since the outbreak of the civil war in the late 1980s. Alliances of dominant elite groups defending a certain form of order have on several occasions soon found themselves toppled and transformed into the challenger. Especially the strategic deep-sea harbor in Kismayo, the current capital city of the Federal Memberstate of Jubaland, has been a highly contested area for the rent-seeking elites. 61 Hence, there are, for example, few reasons to treat the internationally recognized TFG-installed administration in Kismayo of early 2007 as more of a 'state' than its soon to be successor, a militia alliance dominated by the infamous faction leader Abdikadir Adan Shire 'Barre Hiiraale' or the subsequent Islamist alliance of 2008. Indeed, both Barre Hiiraale's militia rule and the Islamist administration could be said to have exercised more authority than the local representatives of the Mogadishu-based Transitional Federal Government, which had severely limited powers to dictate politics in the southernmost provinces. 62 The case of Jubaland's current President, Ahmed Mohamed Islam 'Madobe', is illustrative in showing how blurred the lines between recognized 'state' and 'non-state insurgent' may be within the fragmented context of 'heterarchical governance'. 63 In 2006, Madobe, then commander of a local Islamist group called Mu'askar Ras Kamboni ('the Ras Kamboni camp'), often referred to as the Ras Kamboni Brigade (RKB), fought alongside the infamous al-Shabaab and shared power with the group in Kismayo under the umbrella of the Islamic Courts Union. When Madobe was wounded in battle near the Kenyan border in January 2007 and brought to Ethiopia for treatment, he was at best seen as a spoiler of and challenger to recognized rule by international actors. 64 However, Madobe, originally from the Somali region in Ethiopia, but with ties to prominent clan groups in southern Somalia, was perceived by several elite figures in Mogadishu as a potential candidate to undermine al-Shabaab influence in the south. After a deal was reached between the newly installed TFG president, Sheikh Sharif Sheikh Ahmed, a former fellow ICU leader, Madobe and the Ethiopian government, Madobe was released and went to Mogadishu. However, he quickly left Mogadishu and resumed his activities with al-Shabaab and Mu'askar Ras Kamboni in Kismayo, which ruled most of southern Somalia at the time. 65 Cooperation with al-Shabaab soon soured. Striving to prevent al-Shabaab's tightening grip on power in Kismayo at the expense of Mu'askar Ras Kamboni, Madobe and several Ras Kamboni commanders fueled emerging rivalry, which culminated in a brief battle in the streets of Kismayo in the fall of 2009. A battle Madobe lost. 66 Following his defeat, Madobe and a faction of the RKB escaped tothe border areas of Ethiopia, 67 his position radically changed. At that time, Kenya backed the so-called Azania project of TFG's defense minister Mohamed Abid Mohamed 'Gandhi', which aimed to establish an autonomous member state in Somalia's southern provinces. By training and equipping between 2,000 and 3,000 Somali fighters, Kenya hoped to create a buffer zone along its border to prevent spillover, including large numbers of refugees, from al-Shabaab-dominated areas. 68 However, Gandhi's project did not fare as well as planned. Not least, it failed to foster the expected support from several elite groups of the local clans. According to 'Mohamed', representing one of the clan groups in a meeting with Gandhi in Dolow in early 2011, several of the clan representatives were critical of Gandhi's proposed power-sharing model and felt their clan groups would not acquire sufficient power. 69 There were also doubts as to whether Gandhi, an academic and politician, would be able to mobilize sufficient military capabilities to defeat al-Shabaab on the battlefield. When Ahmed Madobe, as an already battle-hardened commander with clan ties to the dominant Ogadeen/Mohamed Zubeyr clan group, displayed successful resistance against al-Shabaab forces along the border, he emerged as a promising alternative. According to 'Ismail', a clan fellow of Ahmed Madobe, Ethiopian officials assessed Madobe to be a suitable choice for a proxy in southern Somalia and successfully convinced Kenya to embrace him and his experienced and well-established Ras Kamboni Brigade. 70 When Ahmed Madobe was finally elected first president of Jubaland in May 2013, a position later approved by the federal government despite initial vocal resistance, and by the international community in August the same year, this was without doubt contingent on Kenyan political and military support. 71 Ahmed Madobe's career and trajectory to power and presidency is intriguing. In just a few years, he rose from the position of marginal Islamist commander to senior leader within the Islamic Courts Union, then insurgent on the run, political prisoner in Ethiopia, recognized politician, defeated militia commander on the run, Kenyan proxy and, finally president of Jubaland. In such fluid and shifting contexts of competition and ordermaking, orthodox categories of 'state' versus 'non-state' and 'insurgents' versus 'counterinsurgents' as applied in traditional insurgency and counterinsurgency discourse provide limited meaning if seen as contingent on weak versus strong, or as a specific subset of war, strategy or set of tactics, all of which fluctuated throughout the southern Somali conflict. 72 On the other hand, through the lens of Hazelton 73 as broader processes of order-making and state-building, Madobe's transformation from local Islamist leader to president of a Federal Members state may not be surprising if seen in the light of elite competition and bargaining in southern Somalia where negotiations and alliances tend to be shaped by patronage networks shaped by clan affiliation. 74 According to Christopher Day and William Reno,75 counterinsurgency in African states tends to play out as extensions of patronage-based practices where co-optation of insurgent leaders and other elites from segments of the population supportive of the insurgents may be the preferred strategy. Hence, while Ahmed Madobe was a leading militant Islamist and a threat to the elite order in Mogadishu, he was also, potentially, an important ally, who, if co-opted, could strengthen the dominant order.

External intervention and local order-making
International recognition -or lack thereof -can have huge discursive and potentially political, military, and economic implications for the contesting actors. 76 For the federal government in Mogadishu, as for Ahmed Madobe and his Jubaland government, enjoying recognition and support from international and regional actors have been essential for their political and military survival in the ongoing elite contest for dominance. 77 As Robinson and Matisek point out, 78 there is little reason to believe that neither the still fragmented national nor the provincial forces would stand a chance against al-Shabaab's centralized and battle-hardened fighting force had it not been for the military support of AMISOM/ATMIS 79 or other neighboring forces. In Jubaland, after several failed attempts, it was not until Kenya launched a fullscale military intervention into southern Somalia in October 2011 to back Ahmed Madobe and his Ras Kamboni Brigade that local militias stood any chance to challenge al-Shabaab dominance in the south. 80 Yet even Kenya and Ethiopia, both profoundly engaged politically and militarily in Somalia and with long experience of assisting local military forces on the ground in southern Somalia, are often unable to direct and shape political bargaining and micro-dynamics within Jubaland or between the member state and the federal government. The same counts for the US security force assistance in Somalia. 81 According to a West Point workshop on US counterterrorism in Africa, 'successive U.S. administrations have resorted to collaborating with non-democratic African regimes, whose buyin for U.S. counterterrorism programming has often been either minimal or who have used the assistance from the United States for personal or regimerelated purposes'. 82 An expanding body of literature on counterinsurgency and intervening states' use of host governments or 'sub-state' and 'non-state' actors as agents to promote their security interests shows that it is remarkably difficult even for a powerful principal like the US to control and direct the actions and behavior of the minor, host agent when interests diverge. 83 Continued al-Shabaab attacks across the border into Kenya, high-profile attacks in Mogadishu, and lack of progress by the AU, US, and government forces in degrading al-Shabaab activity in southern Somalia may seem frustrating in a policy discourse driven by orthodox counterinsurgency and counterterrorism ideas. And it may be easy to lament the inefficacy and corruption of local partners. 84 While such claims could in many instances be justified, lack of local initiatives and military progress against al-Shabaab may make perfect sense in the view of counterinsurgency as domestic ordermaking. 85 Following Day and Reno's point on African counterinsurgency as endogenous to regime (patronage) politics 86 and Ucko's 'limited statehood' logic,87 as long as al-Shabaab is not deemed an immediate and existential threat to the established elites' power position, the local authorities are best served by focusing their attention and resources elsewhere. For Ahmed Madobe and his government of local elites, it may be far more pressing to forge and maintain local alliances and keep the federal government in Mogadishu at bay than to fight al-Shabaab through a costly military campaign. According to 'Abukar', a seasoned politician from Jubaland, al-Shabaab is indeed the immediate military challenge to the local government within its province. However, the survival of the current elite constellation in Kismayo is first and foremost a political concern. 88 As long as Jubaland continues to enjoy military support from Kenya, the Jubaland government's primary concern is at least threefold. One, to meet the interests of its established power base, which centers on support from an alliance of Ogadeen clan elites and a number of other key clan elites in the south. 89 Two, to negotiate and partly counter elements from opposing elite groups, especially those of the large Marehan clan group in the Gedo province, the northernmost province of Jubaland. 90 Three, to counter the federal government and central elite groups in Mogadishu, which, not surprisingly, strive to project their central powers throughout the entire federal state and, not least, to expand their revenue base to include Kismayo harbor. 91 A well-established narrative among senior Jubaland politicians is that Ahmed Madobe has few incentives to focus on the military campaign against al-Shabaab. If Madobe pushes al-Shabaab further out of its remaining strongholds, for example in the Middle Juba province, he will lose more politically then he will gain, despite tactical successes against al-Shabaab on the battlefield. Elite groups in newly conquered areas, hosting a wide range of non-Ogadeen clan groups as well as historically traditional intra-Ogadeen rivals, may strengthen or alter established alliances unfavorable to the current Jubaland government. 92 Hence, for Ahmed Madobe and his elite alliance, counterinsurgency is first and foremost exercised through political maneuvering and cooptation, not primarily through the gun nor the popular support of 'the people'.

Conclusion
Our refined conceptualization of insurgency and counterinsurgency as processes of violent opposition to and consolidation of domestic established order, based on critical literature from the insurgency and counterinsurgency discourse, treats insurgency and counterinsurgency as broad processes related to domestic order-making, civil war, and state-building rather than as a peculiar kind of war, a specific type of strategy or sets of tactics or techniques, as has been the case in orthodox conceptualizations. By applying this conceptualisation to the complexity of the Somali civil war, the present paper demonstrates that both insurgency and counterinsurgency as broad processes of order-making provide analytical value in understanding the political dynamics shaping military efforts in southern Somalia. From this perspective, insurgency and counterinsurgency are first and foremost political, in this case translated as a contest between domestic elites, with or without external support, for dictating what should be the dominant order. (Relative) stability is thus produced through either elite agreements or (sufficient) superiority by one group of elites and suppression of others. As long as established order is not existentially threatened, dominant elite groups may lack the incentive to divert efforts to rout out violent challengers, such as al-Shabaab.
Any fruitful conceptualization of insurgency and counterinsurgency should acknowledge the complexities of (civil) war and conflict, and provide meaning and analytical and explanatory power in its engagement with the extensive and multidisciplinary research on civil war, peace and conflict, and state-building. If so, the currently not so popular discipline of counterinsurgency may survive and bring with it valuable lessons learned from the wars of the past two decades, knowledge that is vital for the years and the wars to come.  58-59. 48. In contrast to southern Somalia, factions in the north managed their own peace process and declared independence for the former British territories of Somaliland already in 1991. However, Somaliland has so far not received international recognition for its claims.