They are from within us: CVE brokerage in South-central Somalia

ABSTRACT This article explores how societal actors in Somalia take part in a transnational politics of countering/preventing violent extremism (CVE/PVE) through a political sociological approach to militarisation. We argue that the transnational politics of CVE represents an extension of global militarism by some states, institutions, donors and brokers. CVE works to adapt global militarism and to reconfigure the global-local relationships that sustain it. We explore the roles and influence of local ‘CVE brokers’ in deradicalisation efforts in South-central Somalia. They inadvertently merge the counter-terrorism approach to Somali people, values and territory with non-military means. We show that their key practices – co-ordination, translation and alignment – advance, but also disrupt, alter and transform CVE policy objectives.


Introduction
This article explores how societal actors in Somalia take part in a transnational politics of countering/preventing violent extremism (CVE/PVE) through a political sociological approach to militarism and militarisation. With the wave of reinvigorated IR research on global militarism in the post-9/11 security environment, 1 prominent works have drawn on these concepts in regard to security governance, interventions and war to discuss the roles of the great powers in justifying the sale or use of military hardware and discursive power in pursuit of counter-terrorism activities in Africa. 2 Specific policy domains and segments of populations have been conceived as being militarised by global and African actors. 3 In much of this literature, the purchase of the concept militarisation is to take stock of the process of incremental naturalisation or internalisation of beliefs, practices, discourses and objects connected to war preparation and political violence. Militarisation's conceptual centre is a recognition that military and civilian spheres overlap and interconnect, and that militarisation is a dynamic and contested societal, historical and cultural process. 4 To explore how a transnational politics of CVE is being negotiated, shaped and contested in South-central Somalia, we treat militarism as entailing the dynamic contestation, co-operation and resistance to those social relations, institutions and values relating to war and its preparation. 5 We draw on Lutz' definition of militarisation as, 'an intensification of the labour and resources allocated to military purposes, including the shaping of other institutions in synchrony with military goals'. 6 This approach offers a way to broach competing bundles of actors, modes, practices and instruments, at different levels of analysis, and across power hierarchies. 7 We argue that the politics of CVE has enabled interveners, UN agencies and development partners to combine a wider spectrum of state-building, stabilisation and developmental practices other than those more directly coercive and weaponised options that dominated interventions in Somalia between 1991 and 2012. One of the justifications behind this is to lessen the force of counter-terrorism policies by leading interveners, spearheaded by US and Ethiopia, by adding preventive and peace-building rationales (demilitarisation) to more militarised activities. 8 CVE represents an extension of global militarism by some states, institutions, donors and brokers. CVE works to adapt global militarism and to reconfigure the global-local relationships that sustain it.
Whereas the core of CVE aims at preventing individuals from becoming radicalised or joining a radical group, deradicalisation is a process-based strategy by which you attempt to get someone that is radicalised to exit radical groups or change radical mindset. 9 Narrow deradicalisation can be conceived as aiming for physical and behavioural changes, similar to 'disengagement', whereas comprehensive deradicalisation can be thought of as addressing ideological and psychological aspects that affect an individual's radical world-view. 10 The article explores the roles and influence of local 'CVE brokers' in deradicalisation efforts in South-central Somalia. The concept of brokerage has been usefully applied to the study of state-building, civil wars and development, where it has nuanced the understanding of the influence and empowerment of diverse local elites and power-holders under conditions of state absence, insecurity and power transitions. 11 The concept highlights their strategic roles acting as intermediaries in local/global social fields. Brokers make strategic and sometimes opportunistic use of social and financial capital that, in our case, CVE networks bring. In line with a relational approach to brokerage, our focus is on their social roles as sometimes mobilisers, sometimes informal social workers, and sometimes political or religious entrepreneurs. This approach makes the concept useful for understanding (non-state) modes of governance in areas of limited statehood. We argue that their key practicesco-ordination, translation and alignmentcan advance, but also disrupt, alter, or transform CVE policy objectives. 12 When this happens, local militarised politics may change as well. Drawing on qualitative interviews and focus group discussions (FGDs) with CVE brokers in South-central Somalia, we show how they partake in the militarisation of social relations, institutions and values. Indeed, brokerage involves a degree of local empowerment by defying the categories 'radicalised' and 'extremist'. While implementation of CVE is said to soften the excessive security focus of other global interventionist agendas for Somalia, it simultaneously forms part of global militarism enacted locally. CVE brokers inadvertently merge the counter-terrorism approach to Somali people, values and territory with non-military means. 13 The politics of CVE in Somalia today is dependent on local brokers. Firstly, because of prevailing insecurity and the presence of multiple non-state armed groups, third party diplomatic initiatives or other types of external actors' presence and reach are severely restrained. Secondly, the underpinning logics of CVE emphasise community resilience, and that local partners lead in implementation efforts. 14 We draw on the concept of semi-territoriality to further specify the particularities of CVE brokerage in South-central Somalia. Under conditions of semi-territoriality, insurgent organisations face a superior enemy (often the state or its allies) that can defeat them in open battle. However, due to the negligence/lack of resources/strategic considerations, their enemy fails to establish a more permanent presence amongst the local population, and thus fails to provide protection and governance locally. The government and its allies only maintain sporadic control during limited periods through patrols and campaigns, leaving the population at the mercy of insurgents and their sanctions against unwanted behaviour after counter-insurgency forces withdraw. Insurgents can apply sanctions to control the local population and become the de facto provider of security, governance and some modicum of justice. 15 The local population hedges itself against such potential sanctions implemented by insurgents during government and their allies' absence. They thus often comply with the insurgents' commands even during periods of limited government presence. Social entities (families, clans, religion and businesses) must interact with them to ensure a form of stability and predictability in everyday life. 16 These groups even at times integrate into insurgent structures by acquiescing to intermarriages, providing recruits or paying business taxes to mitigate Harakat al-Shabaab (henceforth 'al-Shabaab') violence.
The mechanisms of semi-territoriality in South-central Somalia allow us to understand the context of CVE efforts in Somalia outside Mogadishu. Insurgents govern territory, even without holding territory in the military sense. They control the population in areas supposed to have been 'freed' (areas cleansed of direct al-Shabaab institutions of governance and larger permanent bases). This means that factors influencing deradicalisation and CVE processes go far beyond any one individual's ideology. CVE takes on a strong communitarian character because of the possibilities of sanctions against families, clans, religion and businesses, including potential loss of income for friends and family working for/with the insurgents inside semi-territorial zones. 17 Semi-territoriality describes the conditions within which militarism interacts, and the local context that influences the interaction between brokers and the prevailing social order. It is from such areas that many of the participants in Somalia's deradicalisation efforts originate. 18 This article draws on qualitative data gathering in Mogadishu and Baidoa, Southcentral Somalia. We spoke to youth leaders, women leaders, traditional elders, religious leaders, members of parliament (MPs), police officers, business community representatives, think tanks, local civil society actors, global donors and international organisation officials. The FGDs were predominantly held in the Somali language and translated subsequently. Follow-up individual interviews were then held with one representative from each category. Further interviews were conducted with representatives from the UN mission in Somalia (UNSOM), International Organisation of Migration (IOM) officials, politicians and ministers from the interim South West Administration (ISWA) and the Somali National Army (SNA). The non-state actors that self-identified as having CVE relevant roles are involved in grassroots state-building, reconciliation and trauma-healing, local governance, gender equality, vocational training for former insurgents and disengaged fighters, human rights, legal assistance, inter alia. 19 The organisation of the article is as follows. First, we discuss the conceptual and operational intersections between the fields of counter-insurgency, demobilisation, disarmament and reintegration (DDR) and CVE. Secondly, we explain how the CVE lens is being adopted and practiced by interveners and CVE brokers in Somalia. Thirdly, we show how CVE brokers such as civil society organisations, religious and traditional leaders, and other societal actors contest, alter and transform CVE security practices.

Global militarism and the politics of CVE
Since the fall of Siad Barre's regime in 1991, African and Western actors have portrayed Somalia as one of the world's most corrupt and dangerous countries. 20 Political violence as a result of internal lawlessness, competition over authority, the presence of diverse warring parties and regional geo-politics produce insecurities perceived as threats to international stability, notably terrorism, piracy and transnational criminality. 21 As a result, Somalia has been the site of several interventions of different kinds. Our objective here is to sketch out the way that three policy fieldscounter-insurgency, DDR and CVE intersect with and co-produce forms of global militarism. This occurs in the context of the broader consensus of the security-development nexus in the post-9/11 security landscape, and the gradual securitisation of underdevelopment and fragile states. 22 In the process, we find that militarist logics and practices are carried over into civic/ democratic sectors and spaces, such as democracy-promotion programmes, security sector reform projects and development assistance. 23 This is not to say that the approach to militarism adopted here may not also involve some analytical drawbacks. Firstly, work on militarism has been perceived as 'normative' and wedded to an anti-militarist posture while some argue that all militarism is not unequivocally 'bad'. 24 Secondly, militarisation complements and needs to be seen in relation to the more elastic concept of security that Frowd and Sandor have argued is better suited to study the full range of forms of violence and control thereof, including symbolic forms of violence, in particular contexts. 25 Thirdly, across different societies and historical epochs, social relations are unevenly vulnerable/resilient to more militarist relations, institutions and values. To date, the literature on militarism lacks fine-grained explanations of how and why some communities or domains transform or reverse militarist agendas more so than others, whether by violent or non-violent strategies. Through our focus on specific societal actors -CVE brokerswe foreground the significance of the interaction and imbrication of local and global actors, norms, technologies and agendas. 26

Counter-insurgency and local power-wielders
Stabilisation approaches for Somalia have combined counter-insurgency strategy with peacekeeping and state-building engagements. 27 We focus on how global militarism is at work in the way stabilisation and counter-insurgency frameworks draw together a large variety of intervening actors, far from all of them military. Somalia has become one of the key theatres for testing 'innovative, low-cost, and small-footprint approaches' drawing on counter-insurgency thinking. 28 Central to these approaches is that key interveners enlist regional and local forces to fight the insurgents more effectively, and to begin to govern the territory more responsibly. In the process, contemporary counter-insurgency has produced 'humanitarian soldiers' 29combining legitimating discourses and also operational tools from both military and peace-building communities. 30 As a manifestation of global militarismhumanitarian soldiers are actors from across civil sectors (political and peace-building) that share with their military planner counterparts the influential view that 'hearts and minds' campaigns help pacify target populations. 31 Combining combat with peace and reconstruction projects and peace-building efforts deepens militarisation. In regard to Somali territories, Wiuff Moe finds that peace-building in the process, having been reframed and made useful to soldiers, becomes subjugated to military tasks and objectives. 32 Pacification, in classic counter-insurgency doctrine (i.e. the coercion-driven process whose ultimate goal is the creation of a lasting peace), becomes an oxymoron from this perspective. 33 Recent work on counter-insurgency warfare indicates that such operations rely on effective coercive capacity, population-centric and hybrid tactics, in addition to the advancement of rule of law and security sector reform. 34 In Somalia, counter-insurgency has over time engaged a widening array of local actors and institutions. Replete with references to 'local ownership', 'resilience' and 'cultural sensitivity', counterinsurgency activities occupy a grey zone making it harder to distinguish and assess separately war and peace-building practices. These activities also push official intervention actors, such as the UN, the African Union Mission in Somalia (AMISOM), regional counter-insurgency forces and a range of donors in the Somali context, to reinterpret and stretch principles and purposes of peace operations. This has included involvement from an expansive set of actors, including elders/leaders, civil society, but also militias, clan-based armed forces, state security forces and private security/military companies (PSCs/PMCs) that enact militarised roles, sometimes through direct coercion, while at other times as governance providers or development-oriented activities and projects.
Thus, counter-insurgency as implemented in Somalia has incorporated a widening and transnational set of actors. These global-local relations gradually influence the character of global militarism. Before we show how this occurs more precisely through the transnational politics of CVE, we first discuss how the interlinkages between DDR and CVE are significant for understanding CVE's expansion in Somalia.

Integration of DDR and CVE -DDR as counter-insurgency
In Somalia, DDR has been an ongoing activity since the end of the conflicts in Somaliland and Puntland. 35 By transitioning former combatants into civilian life, DDR fundamentally aims to re-establish the state's monopoly of violence and the control of force in a postconflict phase. 36 It has therefore been seen as a precondition for ending armed conflict, securing stabilisation, and for initiating a process leading to sustainable development. The DDR policy field is reconfiguring its tools, policies and practices to address the diversity of armed actors in contemporary conflicts. Non-state armed actors regarded by most states and international organisations as 'terrorist groups' or extremists are understood as major threats to international stability in the context of the global 'war on terror'. 37 The scale, complexity and scope of DDR work has significantly grown. As a specific technical field, DDR has come to stretch across a continuum from a strictly minimalist focus on improving security, towards a maximalist understanding of DDR as a wide package of practices for boosting development and reconstruction at the nexus of the security and development agendas. 38 In this way, DDR is no longer a strictly postconflict activity but takes place during periods of ongoing violence. Additionally, DDR rarely follows a scripted and sequenced movement proceeding from disarmament to demobilisation to reintegration, but has become a more flexible and context-specific model. 39 Moreover, the UN occasionally includes deradicalisation components into their DDR programming. An increasing number of non-UN entities such as states independently of UN support, regional organisations, civil society actors and PSCs/ PMCs also conduct DDR programmes, sometimes combined with deradicalisation elements.
Hence, the DDR and CVE fields have become linked at the operational level. The 2015 report 'UN DDR in an era of Violent Extremism', for example, advances a process of integrating policy guidance between DDR and CVE under the term 'Demobilisation and Disengagement from Violent Extremism' (DDVE). 40 In the case of Somalia, these combinations are referred to by UN officials as 'DDR-like' programmes. 41 Expectations have been mounting, notably from Somali government representatives, for UN DDR to encourage defections from violent extremist movements, and to engage with cases of radicalised combatants and foreign terrorist fighters. 42 Observers such as Vanda Felbab-Brown have pointed to ethical, legal and political dilemmas arising in situations such as Somalia where UN-supported DDR activities coincide with ongoing military operations, counter-terrorism activity, partial territorial control by one or several jihadist organisations, as well as weak host state buy-in and capacity. 43 In contexts of ongoing armed conflict, one concern is that when state entities partner up in providing UN-supported DDR work the activities become vulnerable to manipulation by various state entities and other implementing partners. 44 Through their increasingly combined operational roles and policy practices, UN DDR and CVE practitioners have been drawn into a militarised politics where DDR shares central characteristics with counter-insurgency. DDR adjusts to 'non-permissive' contexts, which entails incorporating the supposed virtues of flexibility, and empowering national, local and private security partners that have gained privileged positions within the wider stabilisation policy community. Moreover, for some senior UN officials and CVE experts, the UN is called forth to safeguard its CVE co-ordination role as a means of restraining major powers from resorting to proxy military interventions and bilateral security and counter-terrorism assistance. 45 The Somali National Strategy and Action Plan for Preventing and Countering Violent Extremism 46 has been synchronised with the global CVE approach. Supported by the European Union (EU), UN and other donors, it conforms to the call made at the UN Secretary-General's launch of a UN Plan of Action to Prevent Violent Extremism in 2016 for all member states to consider developing a national plan of action. 47 The need for prevention is linked most closely to those forms of jihadist militancy that currently threaten tolerant, pluralist, democratic political order, and 'as and when conducive to terrorism'. 48 The documents point to the insufficiency of security-driven counter-terrorism measures that have dominated the agenda in the post 9/11 context, seeking instead to unite efforts behind strategic prevention and countering of violent extremism. 49

Expansion of deradicalisation and CVE in Somalia
DDR programmes constitute the largest portion of deradicalisation efforts in Somalia. Participation in these programmes has become a way for civil society actors, regional states and the Federal Government to tap into international funds. In 2012, the Federal Government of Somalia (FGS) developed the 'National Program for the Treatment and Handling of Disengaged Combatants and Youth at Risk in Somalia' as a part of a DDR agenda supported by the UN and the donor community. The purpose was to provide support to low risk disengaged al-Shabaab combatants to reintegrate them back into the community. Four facilities were initially established, most of which were developed in partnership with external actors, in most cases with the UN Mission in Somalia's (UNSOM) Rule of Law and Security Institutions (ROLSI) office as the main collaborating partner. A notable exception was the centre in Mogadishu, named the Serendi Centre, run by a Somali actor partnering with the Danish deradicalisation expert, Michael Taarnby. At the outset, the centre was funded by the Norwegian government, a new actor in this arena in Somalia, while the British government, and later Danish government, continued the funding. The Serendi Centre was heavily criticised by Somali government leaders, local elders and religious leaders, as well as DDR experts for its lack of transparency and human rights violations, with original partners distancing themselves from the day-to-day operations. Human rights actors raised concerns about the treatment of inmates, especially those that were minors. In 2015, a new contractor was brought in, improvements were made and the British government continued the funding. Some problems remain, including its continued poor reputation. This centre illustrates how one of the key concepts in the war on terrorderadicalisationhas established itself in Somalia. In the past, this term had been used for activities trying prevent individuals from re-entering criminal activities. In this sense, the new usages of the concept constitute a deepening of global militarism, whereby this new deradicalisation agenda, heavily influenced by European and North American efforts in the war on terror, was being implemented in Somalia.
A second facility in Baidoa was funded by the German Government, and managed by the International Organisation for Migration (IOM). This facility slowly expanded and grew in importance. Covering the Bay-Bakool area, the centre was viewed by locals as open and beneficial to surrounding community than the Serendi Centre. Smaller deradialisation centres were also established in Belet Weyne (funded by the UK Government and managed by Adam Smith International (ASI)) and Kismayo; the latter is funded by the regional state in question, but has since also received funding from Germany. Importantly, the activities in these centres combine DDR with CVE, adjusting their activities for disengaged al-Shabaab members. They are an illustration of the type of 'DDR-like' work that has drawn the DDR and CVE policy communities closer together.
These are examples of how global militarism, through its inclusion of DDR as counterinsurgency, influences the CVE agenda. 50 The most frequently used deradicalisation institution in Mogadishu was the National Intelligence and Security Agency (NISA), which directly enrolled defectors into its rank, including members from the Serendi facility. This practice created mistrust amongst average Somalis in Mogadishu. The latter often understand the NISA to be an infiltrated body since the same individuals who had been members of al-Shabaab just months prior had now taken on government service positions. 51 The NISA continues to play an important role in screening the participants for the Baidoa and Mogadishu centres. Only candidates that the NISA deems to be receptive and less dangerous are allowed to participate in these centres. The selection of candidates, like NISA's processes of integrating former al-Shabaab members into their own ranks, has been severely lacking in transparency.
Additionally, in Baidoa and Kismayo, NISA access was restricted. The total length of stay from entry to exit was set to three months, compared with the 12 months stay in Serendi. 52 In the case of Serendi, an exit board, consisting of NISA intelligence officials, officials of the Defectors' Rehabilitation Programme and Serendi Centre management representatives was established. In order to be allowed to exit, several conditions must be met including medical approval and NISA certification that the reinsertion area is safe. 53 Yet, only Baidoa and Kismayo included community programmes after exit, such as cultural and sporting events, or 'youth at risk' projects. 54 Such community programmes narrowed the linkage between the global and the local in relation to militarism. They were part of a tendency to expand the meaning of counter-insurgency, therefore a core aspect of global militarism, and to widen the scope of brokers that could take up roles in the implementation phase.

Co-ordination practices of CVE brokers
Over time, new actors took on roles in CVE and deradicalisation activities and programmes, especially in terms of their co-ordinating practices. Co-ordination practices involve strategic efforts by brokers that aim to co-ordinate activities and actors in a way that empowers and enhances the social status of the broker vis-à-vis all parties involved. 55 Elders, religious leaders (sheikhs) and civil society groups have become increasingly integrated into Somalia's burgeoning CVE agenda. Elders, for example, have served as an informal early warning mechanism in Mogadishu, channelling vulnerable youths into newly initiated CVE programmes. In Baidoa, elders and traditional leaders have served a similar purpose. In the words of one Elder: Presently, there is a system in which we are working in Baidoa to maintain the rehabilitation of the youth, backed up with elders; we have agreed that every elder should go back to his clan to mobilise the youth groups in government-held and remote areas in order to work with the existing system. 56 Elders and sheikhs, as representatives of civil community life, were here drawn into the CVE sector by government officials, with the end goal of defeating al-Shabaab. The Somali Ministry of Justice has populated a list of religious leaders it deems suitable for CVE work. Formalising the CVE roles of particular sheikhs, however, involves delicate political manoeuvring. First, since sheikhs might come from a combination of religious backgrounds -Sufi, Shafi and Quietist/non-political Wahhabismtheir different theological leanings and associated social positions make their selection by Government officials very political, as the choice of sheikhs for CVE activities may upset a tenuous balance that exists between these groups. Secondly, drawing on religious leaders can imply that religion/ideology plays a significant role for radicalisation, something which is not uniformly accepted by CVE practitioners. 57 As militarisation involves an intensification of the labour and resources allocated to military purposes, including the shaping of other institutions in synchrony with military goals, elders and sheikhs have undoubtedly been drawn into militarisation's ambit.
The introduction of traditional leaders into official CVE activities, however, is not necessarily all that novel. Known by other terms, traditional leaders have arguably engaged in informal, community-oriented CVE-related work since al-Shabaab initial expansion. Their work, for example with 'traditional dispute settlement', did not previously have a significant transnational quality, nor were they as strongly influenced by global militarist trends. Elders and sheikhs had to involve themselves in new, formal CVE activities in part because dispute settlement was traditionally within their purview, and the most serious disputes had taken on a transnationalised jihadist shape. They also do so because extending state structures and institutions still only had limited power and legitimacy in many regions of Somalia. 58 Thus, the words 'deradicalisation' and 'CVE' are relatively new in this context, while some of the roles that they refer to are not. These terms were brought to Somalia as a part of the wider war on terror and the gradual militarisation of the traditional societal structures.
As funds from the UN and other donors increased for the business of CVE from 2012 and onwards, more local actors entered this sector. Because of the purported preventive and developmental roles that facilitate the objectives of effective CVE work, Somali civil society actors have become visible for their community activities, thereby drawing these organisations into the transnational politics of CVE. Unlike the sheiks, civil society actors had not previously conducted deradicalisation. Rather, their focus was on human rights work and youth or women's empowerment. After 2013, non-state CVE actors grew significantly in numbers. In addition to their traditional human rights work, the Elman Human Rights Centre, for example, became involved in reintegrating children, which included monitoring the conduct of the armed forces. The 'Centre for Community Awareness' (CCA) and 'New Horizons' in Mogadishu became other important actors. The same is the case with the 'Horn of Africa Voluntary Youth Committee' (HAVOYOCO), founded in 1992. Today it operates both in Hargeisa and Mogadishu, as well as in Ethiopia (in Addis Ababa and JigJiga), and has become a CVE-related project implementer, even though the nature of its work is essentially the same as it has been in the past (providing development, education and job training). 59 The CVE agenda has thus drawn a variety of actors into wider attempts to deal with and ultimately defeat al-Shabaab. While it is important to note that not all CVE initiatives will involve all types of interlocutors, their growing inclusion in CVE-related activities indicates a drawing in of societal actors into global militarised politics. As a result, several Somali institutions like the Elman Human Rights and Serendi centres are adjusting their activities to synchronise with military goals.
Such a widened scope has the potential to enhance the deradicalisation, and therefore counter-insurgency process, but also introduces new challenges. The newfound CVE brokers often attribute the label CVE to already existing activities to prevent recruitment to local militias, to access alternative livelihoods or to reintegrate former combatants or radical Islamists. Because there has been no system in place in Somalia to ensure a uniform CVE approach, diverse actors have ample space to appropriate, alter and transform CVE brokerage according to their beliefs, interests and immediate context. 60 The CVE agenda, therefore, empowers some actors more than others, notably the religious leaders on the 'approved' list, the clan elders most regularly consulted, certain government ministries and the civil society actors that receive grants.
Drawing together these activities with the aim of stopping al-Shabaab attributes military values to development-oriented youth empowerment schemes. In Somalia, this uneasy transversal function of CVE is made necessary by a situation where counterterrorism operations are sharing the scene with a government amnesty programme for defected (moderate) al-Shabaab leaders/fighters, and other leniency measures directed at some high-risk or strategically important defectors. 61

CVE brokerage and semi-territoriality
In South-central Somalia, the SNA, special/regional police forces, AMISOM and the Ethiopian Army have insufficiently secured the local population. 62 They remain concentrated around their base areas except for limited patrols and well-planned offensive operations. Local communities, therefore, know that their primary and regular interactions will continue to be with al-Shabaab militants. As a result, this ensures that local communities maintain a form of loyalty to the insurgent group due to their fear of reprisals. The al-Shabaab presence and ability to collect taxes, zakat, is permanent enough to secure stable incomes for recruitsmeeting their potential opportunism, desires for income and the need for security. 63 As such, al-Shabaab is able to operate relatively freely in the Somali countryside. Moreover, SNA levels of corruption are high (implementing illegal checkpoints and protection rackets), and the Ethiopian National Defence Forces (ENDF) are perceived by the majority of those interviewed as a conflict actor (and sometimes as an imperialist presence) in the Bay/Bakool region, making loyalty to al-Shabaab more likely. 64 Under such a regime of fear, CVE brokerage is both unavoidable and dangerous. Rural social institutions, rural women's groups, elders and religious leaders risk being sanctioned by al-Shabaab for their roles as brokers. One CVE actor concluded that even if al-Shabaab's ideology is waning, the fear among ordinary people and the collusion of business circles and politicians with al-Shabaab remain daunting. 65 Insecurity also hinders defections from al-Shabaab and severs loyalty towards government institutions. A former al-Shabaab member, for example, explained how he hesitated to defect for a long period because he feared that he would be tortured or shot depending on who, at what point in time, got a hold of him. 66 As another respondent explained, 'one day a special police officer, and the next an al-Shabaab operative, then back to being a beneficiary of a vocational programme. But when facing unemployment, he carries out a small job for al-Shabaab'. 67 In this sense, the original militarist agenda outlined by the international community and Somali government actors is significantly altered by brokers. CVE instead becomes a tool to gain power, options and funding rather than a counter-terrorism strategy.

Translation practices
The inclusion of local elders and local religious leaders in CVE activities has provided donors and UN agencies with bridges between them and government actors, as well as to rural areas where government authority has limited influence. For their part, brokers act as translators, or bridges, which means that through their actions they produce relations and realities that are irreducible to the interests of the bridged parts. 68 Importantly to our analysis, as elders and religious leaders began to engage in CVE brokerage, this gradually transformed related Somali societal structures. Traditionally, religious leaders were supposed to act as neutral arbiters in conflict. Similarly, Somali clan leaders have traditionally held mediator roles more so than outright warlord roles. In one sense, their role has become hybridised, and merged with modern roles they traditionally never had. 69 Respondents raise the risk that elders and sheikhs become 'just like other politicians', meaning corrupt and politicised elites in the country. Certain elders and sheikhs play roles in labelling individuals as 'extremists' and 'radicals'. Their involvement in programmes to rehabilitate defectors means that they partially serve the interests of external CVE actors (understood by the authors as serving global militarist agendas). 70 These external interests do not always align with community welfare. Militarisation leads to a partial alteration of their traditional roles, and can undermine their legitimacy. Conversely, such a hybrid position may also yield the possibility for using their newly-attained legitimacy in ways that can reduce violent action, and increase an acclaim from local communities. Thus, the involvement of traditional leaders in CVE work illustrates tensions and contradictions when their roles become entangled with local militarised politics: in some cases, this militarisation can present new problems, but can also solve some older problems, which as a result transcends the simple view of the militarisation phenomena being associated with either 'good' or 'bad'. The fact remains that in the semi-territorial areas of Somalia, where the government and other structures are simply too weak to provide consistent protection to local communities, the involvement of elders and sheiks remains crucial since they can assess and mitigate community risks coming from individual defections, and decipher possible sanctions from al-Shabaab that might result.
Brokerage by Somali local actors has also enabled international actors to support CVE in remote spaces far from Mogadishu, despite their own inability of having staff present. While local actors have channelled new resources and terms to conduct many of the same functions as before, CVE efforts nevertheless require them to also place themselves in relation to a counter-terrorism-inspired language that brands their interlocutors and programme beneficiaries as extremists, radicalised and defectors. Thus, as Somali NGOs have gained a newfound status, and can access a well-funded transnational CVE agenda, they nevertheless partake in extending CVE's importance at the local level. By so doing, local brokers become imbricated in global forms of militarism associated with the fight against violent extremism. In the process, they also transform its content and operations. CVE brokers inadvertently lessen the coercive aspects of the counter-insurgency approach to Somali people, values and territory, yet their non-military means still retain linkages with an excessive, and inherently exclusionist security focus.

Alignment practices
CVE brokers often become social mobilisers or quasi social workers. They mediate between conflicting interests and aim to bring the actors connected in and through brokerage 'in line'. 71 The nature of community recruitment by al-Shabaab has created additional layers of division and social challenges with Somali society, such as broken families (single-headed households). As one respondent put it, al-Shabaab has 'killed our social institutions'. 72 In the place of local government, CVE brokers have thus directed their efforts at families and community: 'al-Shabaab is everywhere, so we have to begin with families'. 73 The sentiment is equally shared by former al-Shabaab members. For example, a former al-Shabaab fighter stated, 'if family accepts me, maybe others can'. 74 It is notable that many CVE actors interviewed have found that defectors principally return to family settings, especially to parents, and in some cases wives or husbands. If these are missing, the next available kinship relation is activated. One female CVE broker explained how her organisation performed functions akin to social workers, health workers and legal aides by visiting families and gaining the confidence of women and children. 75 With this trust, they work to raise awareness among mothers in single households on factors that can prevent radicalisation. In this way, CVE operates as a form of feminism, yet female CVE brokers express how they would have liked to be empowered as development and community actors in the first place. 76 This form of brokerage, therefore, attempts to redefine (and civilianise) militarism locally. The women CVE brokers in part accept to work under the CVE umbrella because they hope that programmes and incentives for women might replace the informal/shadow support roles that radical networks enjoy. 77 Interlocutors spoke at length about their developmental activities, education or vocational programmes, community engagement (inter-faith, inter-clan, and female empowerment activities), illustrations of how militarism embeds itself into these spheres. One respondent explained that the work they do is 'positive' since it aims to rebuild a cohesive community. 78 Their intent is to channel CVE funding towards 'righteous' community life, both in terms of social behaviours and religious beliefs. 79 Therefore, the locally known cases of loopholes in screening processes (for instance, cases brought to the Baidoa Centre of young people who are not actual defectors) are viewed pragmatically. For them, this sort of opportunism is easier to understand when you consider the insecurity in the region and the history of clan politics, including the marginalisation of some groups.
Of course, al-Shabaab is not the only factor creating insecurity in Somalia. Clan conflicts, and manifold grievances from the era preceding al-Shabaab's ascendancy, where human rights violations occurred in battles between Somali warlords, often following sub-clan cleavages, have contributed to a variety of other conflicts. 80 Somalia's historical context leads many CVE brokers to explain the reservations they have with terms like 'deradicalisation', 'extremist', 'defector', all of which tend to stigmatise and attribute blame onto large segments of the community. Others articulate that the communities do not need specific interventions to 'deradicalise', since the large majority of people do not benefit from such categories. They would explain that due to the absence of functioning state institutions, non-state actors perform governance, development, social or humanitarian roles. In South-central Somalia, involvement or support for al-Shabaab covers a wide range of activities, including routine and non-radical acts such as paying or collecting zakat, collecting supplies, living as an infiltrator in the community to gather intelligence on competing groups, or accepting services provided by al-Shabaab such as justice provision or financial aid to run a market stand that doubles as tax collection point. In this context, such minor supporting acts are carried out because people fear for their lives and property, which ultimately informs how community members are considered; in ways that do not centre on 'radical' or 'extreme' beliefs or actions.
This suggests that brokers have reservations, and hold ambiguous attitudes regarding their own position and involvement in the transnational politics of CVE. It is worth recalling that their brokerage link is itself a result of the war on terror's introduction to Somalia, and the military agenda to defeat al-Shabaab, which ushered in CVE activities to the region. Global militarism can 'do work' through brokers, underlining its fleeting and succumbs to local forces that adapt its dominant practices to suit local context. CVE brokers saw it as their role to plan activities that better recognise the many causes of engagement with al-Shabaab in Somalia, including economic reasons, the need for security, and the need to appease or collaborate with al-Shabaab for mobility, security or business operations. 81 Somali brokers' simultaneous adoption and adaptation of global militarism, therefore, is preferable to many community members over the available alternatives, for example continued al-Shabaab dominance and the militants' ability to attract former recruits back to their ranks. 82 CVE brokers have a significant level of influence and independence to carry out CVE work in South-central Somalia. The development of CVE brokerage may lead to shortterm empowerment of some actors, but could also produce longer-term changes in Somali societal structures. For example, in the recent past, sub-clans have attempted to manipulate al-Shabaab, and other powerful militias to gain advantages in local clan-based conflicts. Sub-clans are known to have infiltrated al-Shabaab to gain its support. Local conflicts stir up the need for al-Shabaab connections in order to gain the advantage of powerful allies. Of course, al-Shabaab does not always master this game. But they can nevertheless benefit from inter and intra clan insecurity. Al-Shabaab has often taken advantage of, or has manipulated inter-clan conflicts to their benefit. When the presence of fewer CVE brokers among marginalised groups, for instance, intensifies clan politics over the resources that are introduced to the CVE scene, this could easily sow the seeds of future patterns of armed conflict, and further anchor informal militarised fighting strategies between communities. 83

Conclusion
This article examines the interplay of global militarised politics and local brokerage by showing that CVE brokers leverage and transform expanding views of what counts as CVE. Somali NGOs have gained a newfound social status by their access to transnational networks of funding and influence. Such access provides benefits such as vocational training, grants and contracts and networks with global organisations. However, their participation in a form of counter-insurgency against their communities is transforming and potentially delegitimising the sources of their social power.
Brokers partake in extending CVE's importance at the local level and by so doing become imbricated in global forms of militarism. In the process, they also transform its content and operations. In some cases, CVE becomes a pretext for doing what NGOs did before, for example programmes with focus on vocational training, youth integration etc. The CVE and deradicalisation scene in Somalia is a telling example of what occurs when CVE activities draw together bundles of military, developmental, political and humanitarian actors. CVE networks assign new rules of the game for global-local CVE cooperation, set standards of expertise for this work, and label a diverse set of actors as 'extremist' or 'radicalised' groups and as CVE actors, respectively. They blur distinctions between pacifist and military forms of interventions to counter support for terrorism and the often-undefined category of extremist groups. We find that as a result of local brokerage, global militarism is widened and altered in south-central Somalia.
The politics of CVE in Somalia remains associated with the broader effort to defeat al-Qaeda's local affiliate, al-Shabaab. Yet, the social and political focus of CVE makes the term more palatable to development and humanitarian actors than counter-terrorism. 84 CVE co-ordination and information-sharing is a form of interaction that nurtures and constructs assumptions about (de)radicalisation processes. This process reinforces connections between poverty and insecurity, and redraws the lines between civilian oversight and military force. Notes 55 respondents. During the course of interpreting the interviews, we have paid close attention to the fact that respondents take risks when sharing information with Gelot who, as an 'outsider' in many people's minds, becomes associated with al-Shabaab's enemies. This means, for example, that a larger proportion of the respondents she has met with belong to a well-connected transnational Somali elite, and that IHRO would meet with a wider set of actors, thus at times obtaining different responses. Close collaboration with district-level authorities and community-level actors has been key to our field research, we direct heartfelt thanks to the Baidoa district commissioner, and all the respondents from Bay/Bakool area. We have ensured strict confidentiality of all respondents. 20. Fergusson 49. UN General Assembly, 'Plan of Action', para. 5-6. This is a sentiment with broad support among donors and Somalia's international partners, which the High-Level Partnership