Insurgent courts in civil wars: the three pathways of (trans)formation in today’s Syria (2012–2017)

ABSTRACT As part of research on the meso-foundations of conflict, the field of ‘rebel governance’ examines political institutions that regulate the affairs of civilians in wartime as well as their relations with armed actors. Judicial institutions play an important role in this and research has shown that they are widespread among both historical and current insurgencies. However, usually these bodies have been analysed in the context of one hegemonic faction like the Tamil Tigers in Sri Lanka and the Afghan Taliban. What is missing so far is an analysis of different pathways of (trans)formation of rebel courts. As exemplified by the three case studies of judicial institutions in Eastern Ghouta, Idlib and Aleppo, these are shaped by the distribution of power between ‘same-side’ groups, yielding unipolar, bipolar, or multipolar constellations. The analysis is located on the meso or movement level of insurgent social structures, complementing research on the micro and macro levels.


Introduction
It is natural that in any country in which a war occurs, criminals, bandits and gangs proliferate. Someone has to stop them, so it is a normal thing that the rebels set up military courts. How can we stop them if we are not stronger than them? 1 Rebels in civil wars do not only fight, but they also attempt to govern by establishing order and security and by providing essential services. Judicial courts play an important role in this as exemplified by the many examples in the history of insurgent groups. One explanation for the abundance of these bodies through time and space is that judicial institutions are central in exercising political power. Establishing alternative courts is 'the ultimate denial of the right of the state to determine the outcome of disputes and consequently the ultimate denial of that state's legitimacy'. 2 If insurgents manage to set up these institutions and gain the acceptance of the target population, this ultimately expresses not only their desire but also their right to rule.
Under which conditions do rebel courts emerge and (de)stabilize in the context of a multiparty civil war? Existing studies have shown that insurgent courts are a globally relevant phenomenon, ranging from the 'Republican Tribunals' in the Ireland of the 1920s, the courts run by the National Liberation Front in the Algerian War of Independence, the Marxist Naxalites of India, the Sierra Leonean Revolutionary United Front, the FMLN in El Salvador and the Afghan Taliban to the Islamic State's 'Caliphate of Law' in Iraq and Syria after 2012. 3 Often, justice institutions are at least one ingredient in a broader toolbox of political, social, and economic goods provided by insurgents as shown by the emerging field of rebel governance.
However, existing studies have been confined to bodies founded by a hegemonic faction. What is missing so far is an analysis of different pathways of (trans)formation of rebel courts which can exist under conditions of unipolarity or hegemony (i.e. one dominant insurgent group), bipolarity (two dominant groups), and multipolarity (no dominant group). These power constellations were present in the three opposition regions in Syria analysed here: Although only for the duration of 2 years, the Judicial Council in Eastern Ghouta attained a remarkable degree of stability and strength in the besieged enclave compared to judicial institutions in other opposition areas. In contrast, the courts in Aleppo city and its countryside could never fully establish themselves but remained in a state of fragmentation. Finally, the judicial dynamics in Idlib province were characterized by an enduring competition between two bodies, the Islamic Committee(IC) for the Administration of the Liberated Territories and Dar al-Qadaa, which were connected to the Islamist strong weights of the civil war, Harakat Ahrar al-Sham al-Islamiyya and Jabhat al-Nusra (JN), respectively. Although since early 2016, this balance has tipped in favour of the latter, it could only be cemented by a military takeover of most institutions in Idlib in the course of the first half of 2017. In order to analyse these case studies, I conducted a multisited ethnography and an analysis of primary documents.
As I will show, these power constellations of hegemony, bipolarity, and multipolarity in the judicial sector are strongly connected to the internal military situation between groups that fight on the same side of the Syrian civil war. 4 The focus is mostly on structures, institutions, and interactions in this arena and not on the substance of the law applied, which is an important topic in its own right. 5 By analysing institutional dynamics and interactions between armed factions, my analysis is located on the meso or movement level of insurgent social structures, which complements research on the micro and macro levels.

Rebel governance and courts in civil wars
In the 2000s, the scholarly field of 'rebel governance' emerged as an important theme in political science-oriented peace and conflict studies, building up on older research on guerrilla governance in Latin America. 6 Based on this literature, two basic views about it can be distilled. First, why would rebels form governmental structures at all, when it takes away essential resources from their military survival? 7 The opposite, highly instrumental view most prominently represented by Mao Tse-tung sees it as 'natural' since rebels need to win the hearts and minds of the populationand governance is one way to achieve this. Mediating between these rather extreme positions, rebels might also have an incentive to create governmental institutions in order to extract resources in a more reliable way. 8 For armed groups controlling a territory, these institutions might ease the control of the population and free space to use violence elsewhere. 9 As a basic definition, rebel governance refers 'to the organization of civilians within rebel-held territory for a public purpose. These purposes include rebel encouragement of civilian participation, provision of civilian administration, or organization of civilians for significant material gain', with one of these activities being sufficient to call it governance. 10 This broad definition which allows for a low threshold of inclusion might make it difficult to delineate it from other political and social phenomena in civil wars. There is the danger that every interaction between rebels and civilians is considered rebel governance. In this context, clear scope conditions are needed: First, rebels must have some degree of territorial control. 11 Second, civilians must live in the respective area. The third element deals with the 'rebel' part of the definition: the group has to commit violence in order to become rebels which it either continues or credibly threatens its use in territory it controls. 12 One of the most striking features of rebel governance is its variation, for example concerning the range of activities. While some armed groups have concentrated on one or few 'core' areas such as the provision of basic services and security, others have also managed the economy, organized civilian participation, and even regulated private conduct. 13 In order to explain this variety, scholars have looked at pre-war and population attributes, armed group characteristics and war dynamics. Mampilly has argued that the preexistence of state structures and the degree of the state's penetration into society influences the rebels' likelihood to do the same or even better by coopting extant institutions and networks. 14 This only explains the probability of service provision, not its specific nature, however. Concerning Syria, this seems to apply at least partly since regions that were first favoured by the Baath party after it came to power in 1963 in terms of redistribution and since the 1980s were affected by the withdrawal of local state institutions became important centres of opposition administrative and juridical activities. 15

Insurgent courts
In a global perspective, most insurgents practice some level of arbitration and jurisprudence within a broader arsenal of governance. The New People's Army/Communist Party of the Philippines and the Communist Party of Nepal, the LTTE in Sri Lanka, the FMLN in El Salvador, the Taliban in Afghanistan, and the so-called Islamic State represent diverse groups that managed to develop sophisticated structures. 16 Yet, there are few works dealing explicitly with the judicial arena as a subfield of rebel governance. For his classic study on guerrilla movements, Jon Lee Anderson visited several insurgent movements all over the world. He concluded that insurgents establish courts as 'revolutionary rehearsals of the exercise of the power they hope to wield one day on a larger scale'. 17 The book Rebel Law by the barrister and former intelligence officer in the British military Frank Ledwidge explores insurgent courts ranging from those set up by Sinn Féin and the IRA during the Irish War of Independence, the Taliban in Afghanistan, the Union of Islamic Courts in Somalia, courts run by different rebel groups and JN in rebel-held Aleppo to the Islamic State in Iraq and Syria. His main finding is that even in the case of the most ideological groups and the institutions founded by them, what mattered most was that they could establish hegemony, which is a combination of (perceived) procedural fairness, coercive jurisdiction, the ability to enforce judgements; and the elimination of rival courts. 18 Existing works are a fascinating excursion into the world of rebel courts. However, they are largely descriptive and mostly look at the emergence of judicial institutions in the context of a two-party civil war or after a hegemonic group prevailed over its competitors, such as the Taliban in Afghanistan and the LTTE in Sri Lanka. Hence, they did not consider different pathways of emergence and (de)stabilization of insurgent courts in an early or undecided stage of an insurgency. Here, the meso-level structures of insurgencies play an important role.

Meso-level structures of civil wars
Civil wars differ in terms of the number of armed actors involved, both on the side of the incumbent and the opposition. A common distinction is between a two-party and a multiparty civil war. 19 However, I argue that this characterization fails to capture important dynamics of the Syrian war. While some reporting spoke of more than 1000 groups, 20 there were many minor factions included not relevant for the larger conflict dynamics. As will be shown in the subsequent analysis, what was decisive was the concentration or dispersion of power between groups, while the number of actors involved played a secondary role. Hence, we should add the following to the measures of fragmentation: the degree of institutionalization and the distribution of power among groups, 21 which can be concentrated in one or few centres, yielding unipolar or bipolar orders, or dispersed among many, resulting in a multipolar constellation. 22 These concepts emerged in the neorealist discussion in International Relations (IR) on the stability of the international system under different actor and power constellations. While this extensive debate clearly exceeds limits in space, the initial assumption was that multipolar systems are the most stable ones. 23 This was famously questioned by Kenneth Waltz who argued instead that 'the inflexibility of a bipolar world, with the appetite for power of each major competitor at once whetted and checked by the other, may promote a greater stability than flexible balances of power among a larger number of states'. 24 Following the fall of the Soviet Union and the rise of the US as sole superpower, some scholars have argued that unipolarity tends to be more stable than the other systems. 25 Concerning the application to internal dynamics in civil wars, several scholars have argued that armed groups behave like sovereign states in the international system characterized by anarchy. 26 However, with the exception of Christia and Krause, none of these works has explicitly focused on power dynamics, which is the central theoretical argument here. 27 I argue that the relationships between armed groups constitute a level of analysis in its own right, nested within the larger dynamics of the macro level (the fight between an incumbent and opposition factions including transnational and international factors), and the micro foundations of conflict (spatial variation, multiple paths of recruitment, and wartime changes). 28 Going beyond existing studies I show that judicial bodies are also established in the context of a multiparty civil war with continuing 'on-side' fighting, 29 that is where a hegemon has not (yet) emerged. The power constellations in which insurgent courts (trans)formed are an expression not of the macro dynamics but the internal military situation between same-side groups and have important consequences for the stabilization or destabilization of these courts. In multipolar constellations, reoccurring conflict is expected since the (power) relations between groups are in constant flux. Both bipolar and unipolar constellations are assumed to be more stable unless changes in the distribution of power between them are expected or occurring.
3. The Syrian case: from peaceful protests to an Islamist-dominated civil war Following the events in Tunisia, Egypt, and Libya, the Syrian 'revolution' started with peaceful demonstrations which turned violent as a reaction to their brutal suppression and the quick and widespread arming of segments of the protestors, crossing the threshold to a civil war already by mid-2012. 30 2011 and 2012 were mostly characterized by the emergence of many small, local groups which soon coalesced into larger groupings like Jamal Maarouf's Syrian Martyrs' Brigades in the Idlib region of Jabal al Zawiya, Liwa Ahrar al Shamal, and Liwaal-Tawhid in the Aleppo region. Since the end of 2012 we have seen the emergence of super-coalitions on a national scale assembling the dominant armed groups, most notably of which was the Islamic Front. 31 Not coincidentally, Islamist actors of different strands started to become increasingly dominant. While 2014 was mostly characterized by the fight against IS, which was followed by a discursive dissociation of most of the other groups from it, 2015 was marked by two important military alliances, Jaysh al-Fatah conquering Idlib city and the rest of the regime-held areas in Idlib province and the Fatah Halab Operations Room in order to coordinate battles against the regime in the Aleppo region. Islamist and Salafist factions were dominant due to their manpower, internal cohesion, superior command, and control structures and resources. This is also reflected in the courts analysed here as 'hotspots' of Islamist activity. Hence, a short note on the ideology of the relevant actors is in order. While definitions of Salafism usually refer to the direct connection with Islam in its earliest form represented by the Prophet and his first followers, in the context of the Syrian war mainly the restrictive definition of the Islamic faith amenable to sectarian polarization 32 is relevant. The defining feature of Salafi-Jihadism is the embrace of a global agenda. In line with this, Suqour al-Sham in the Jabal al-Zawiya region in Idlib, and Jaysh al-Islam in the Eastern Ghouta qualify as Salafist (although they espouse different versions of it). JN or Jabhat Fatah al-Sham as a (former) al-Qaeda affiliate is seen by most analysts as Salafi-Jihadist. 33 Ahrar al-Sham is a disputed case. While it clearly started out with strong Salafi-Jihadist credentials and connections to Jihadi networks beyond Syrian borders, it more and more moved towards the rebel mainstream, actually avoiding taking a clear stance against al-Nusra until the latter took over most Ahrar al-Sham positions in Idlib in mid-2017. The best characterization, which also seems to apply to Hayat Tahrir al-Sham, a coalition of several groups dominated by Fatah al-Sham and its most recent rebranding, is that these groups are best seen as larger movements including a more hardline-Jihadist camp and a more mainstream, revolutionary one.
Several interviewees, both judicial actors and members of armed groups, confirmed that beyond fighting the judicial arena has been the Syrian insurgency's central field of interaction between armed groups. Hence, I selected the most important joint rebel courts which were located in Eastern Ghouta in the vicinity of Damascus, Idlib province and Aleppo city as well as its western countryside, regions which have been important and early hotbeds of the peaceful and later armed rebellion. The Juridical Council in Eastern Ghouta achieved a near situation of hegemony for approximately 2 years, since Jaysh al-Islam imposed its control ruthlessly. However, the system began to crumble after another group with a similar strength materialized following the merger between the second and the third strongest groups. The council finally split in two geographically separated units. The judicial system in Aleppo remained fragmented and dysfunctional, mirroring the meso-level fragmentation of the insurgent landscape. In Idlib, the IC and Dar al-Qadaa took control over most of the other courts in the governorate resulting in a situation of bipolarity, but the balance of power changed in favour of Nusra since the beginning of 2016 and its dominance was completed by mid-2017. 34

Methods
The article is based on two different, empirical sources. First, a multisited field research conducted between August 2017 and March 2018 in different places in Turkey and second, primary documents published by the relevant actors on social media (Facebook and Twitter). For obvious reasons it is not possible to conduct field research in Syria. Being one of the largest receiving countries of Syrian refugees and enjoying neutral or even cordial relations with most of the mainstream opposition groups, Turkey was the logical destination to conduct the interviews. Albeit this comes with some drawbacks, it represents the second-best and only feasible option for conducting ethnographic research on an ongoing war, which has been marked by intensive regime bombings particularly on Eastern Ghouta during the time spent in the field. 35 In order to ameliorate these shortcomings, I also conducted several interviews with armed opposition actors and activists residing inside Syria. Despite the difficult access due to the nature of the actors interviewed and the sensitive nature of the topic, I managed to carry out deep ethnographic field research not the least because I could conduct most of the interviews in Arabic. 36 The sample consists of current and former members of major armed groups, Syrian journalists and researchers, civil activists, members of local councils, lawyers and former judges as well as religious figures. The unit of analysis is not the collective actors themselves but the interactions between them and the armed groups' behaviour in the judicial field.
In order to analyse these different kinds of data and to establish a link between events and outcomes, I use the approach of longitudinal case studies, which is a combination of process tracing and comparative case studies. 37 For the analysis of relationships between collective actors, tools from linguistics can be used. The sequence Subject (S) -Action (A) -Object (O) and respective modifiers are recognized to be the most basic narrative structure of a text or frame, also known as semantic triplet. This so-called story grammar generates relational data on actors and their interactions. 38 Unipolarity, short-lived: the case of Eastern Ghouta Eastern Ghouta is part of an agricultural belt encircling Damascus. The Syrian regime lost control over the opposition enclavewhich only has a size of about 100 km 2early on in the conflict, and has imposed a siege on it since mid-2013. The area experienced several nerve gas attacks, culminating in a massacre of civilians on 21 August 2013. Due to the proximity to the capital the military pressure remained immanent. Hence, the enclave was left to its internal dynamics characterized by the siege making the defence of its economic lifelines (underground tunnels) as well as its organization and order tantamount. It is in this context that the behaviour of the Islam Brigade and later Army should be seen that managed to assert itself as the dominant military power in the Eastern Ghouta from mid-2013 until early 2016, which was accompanied by its dominance in the Judicial Council. The strong charisma, resolve, and brutality of its leader Zahran Alloush played an important part in the gradual swallowing or extinction of rival factions. It is known from other insurgencies like the Tamil Tigers and the Taliban in Afghanistan that this can only be achieved by the extensive use of violence. 39 Judicial and military unification under pressure After regime forces were expelled from Douma, the biggest city in the Eastern Ghouta in late-2012, chaos and a lack of security prevailed in the area that was worsened by the imposition of a siege starting in April 2013. 40 In order to cope with this, each local faction established its own security and judicial branch which expectably did not improve the situation. In order to unify this disarray of private sharia courts, in June 2014 the most important military groups in the enclave, Jaysh al-Islam, Free Syrian Army (FSA)-linked Faylaq al-Rahman, Ahrar al-Sham, al-Itihad al-Islami Ajnad al-Sham, JN, and a number of smaller factions founded the Judiciary Council (JC) in the Eastern Ghouta. The factions did not intend to run the council themselves but charged a panel of religious scholars to take over this task. 41 However, most of the sheikhs could be seen as representing specific factions. 42 While the leadership echelons were clearly dominated by them, the organizational structure also included practising lawyers. 43 As expected, Jaysh al-Islam tried to translate its strong military position into dominance over the Judicial Council. As the spokesperson of Faylaq al-Rahman told me, Jaysh al-Islam appointed judges belonging to it and enforced itself as executive force. After 6 months we discovered that the independent judicial institutions were not independent anymore, but belonged to Jaysh al-Islam and the members assigned by Faylaq al-Rahman and al-Itihad al-Islami [Ajnad al-Sham] were weakened. Finally, we withdrew, like we did from all other joint institutions. 44 According to the researcher Youssef Sadaki, there was a case where a local council misspent an international loan. Instead of going to court, the parties tried to find a solution before because they were afraid that Jaysh al-Islam would be involved in the decision. This also had practical reasons: the office of the council was in Douma where Jaysh al-Islam was basically the only force in town after outcompeting its main local rival, the Douma Martyrs' Brigade. 45 Predictively, the group's dominant role resulted in problems like the high-profile resignations of Ahmed Abdelaziz Uyyoun (whose presumed links to the former leaders of the Umma Army might have been a reason for his assassination in October 2015) in summer 2014 and Khaled Tafour in July 2015 as a result of their disagreements with the Islam Army and infringements on their authority. 46 Also, Sheikh Abu Rateb Abu Diqqa stepped back after just 2 days in office. 47 The unification of the judicial field was followed by the foundation of the most important military alliance of the Eastern Ghouta, the Unified Military Command (UMC). This was created on 27 August 2014 by five of the enclave's most powerful factions, the Islam Army, Ajnad al-Sham, Failaq al-Rahman, the al-Habib al-Mustafa Brigades, and the Eastern Ghouta branch of Ahrar al-Sham. 48 It was led by Zahran Alloush of the Islam Army while Abu Mohamed al-Fatih from Ajnad al-Sham served as his deputy and Faylaq al-Rahman's Abdel-Nasr Shmeir as field commander. 49 According to the former spokesperson of Jaysh al-Islam, after the local council in Darayya, it was the most successful project in Syria: 'the council succeeded because it was led by one faction and it was institutionalized'. 50 However, this project was not uncontested. Resistance came from two new coalitions partly consisting of old enemies, the Umma Army which emerged from the remnants of the Douma Martyrs' Brigade and Failaq Omar. Alloush reacted with outrage and declared that there could not be 'two heads on the same body', launching a major crackdown and brutally finishing off his old rival from Douma. 51 His allies in the UMC were unwilling to enter, despite their criticism of his singlehanded approach. 52 Based on its alleged links with the Islamic State, Alloush also cracked down on Failaq Omar and pressed the remaining leaders to join the Islam Army in April 2015. 53 Another unruly faction to the unified system of the UMC and the JC was al-Nusra. In contrast to the 1.5 years it stayed in the Shari'a Committee in Aleppo, the Nusra Front split from the Juridical Council in Eastern Ghouta just 1 month after its foundation to run its own sharia courts instead, using similar justifications such as the wanting ability of the council to enforce its rulings. 54 In reaction, the 'general leader of the Eastern Ghouta' Zahran Alloush declared that he would not permit any faction to run a judiciary outside the Judicial Council. 55 The UMC and the JC also called on al-Nusra to shut down its courts and submit to the Judicial Council. 56 But the jihadis did not comply and Alloush did not use military force against them since he was afraid of the repercussion this could have on weaker Jaysh al-Islam brigades in northern Syria where al-Nusra was a major force. 57 In February 2015 the UMC banned the formation of new factions, basically leaving the Islam Army as the dominant force, as well as Ajnad al-Sham, and Failaq al-Rahman in charge, while tolerating the Nusra Front. 58 This meant that all major factions in the region except the Salafi-jihadists had endorsed the Judicial Council. However, a year later in 2016 the power system in Ghouta started to crumble when the third strongest group, Ajnad al-Sham, merged with the second strongest, Faylaq al-Rahman, destroying the relative stability created by Jaysh al-Islam's hegemony. 59 The enclave's demise, and with it that of the Judicial Council began in spring 2016 when Faylaq al-Rahman and Jaysh al-Islam were at each other's throat for the first time. The council did not collapse, but split into two parts. As one respondent explained, The main office of the Council was in Douma, so most people living in the middle section under Faylaq were wanted by Jaysh al-Islam. Because of this, most of them can't go to Douma. So they needed to create their own court. There are still connections between the judges in two areas. But because civilians can't go back and forth, there is a de-facto separation between the courts. 60 This has basically sealed the insignificance of the once powerful institution. Its split along factional lines also shows that finally the military logic prevailed over the judicial, which was not the case in 2014 and 2015.

Bipolarity: the case of Idlib province
The IC for the administration of the liberated territories: from a joint institution to domination by Ahrar al-Sham Before the formation of the IC in Idlib in the beginning of 2014, there was a great number of single courts in the whole province. In 2013 representatives from these courts met and decided to found a joint body, which became the IC with its central court in Binnish. 61 According to Sheikh Hassan Dreim, one of its founding members and former judges, it was formed by six factions, Ahrar al-Sham, Suqour al-Sham, Faylaq al-Sham, Harakat Hazzm, the Syrian Revolutionaries Front (SRF), and Liwa al-Haqq. In the beginning, the founding factions shared the burden of funding and staffing the court's executive force collectively. But approximately 1 year after its foundation, it was dominated by Ahrar al-Sham, resulting from the annihilation of most of the other factions by the Nusra Front. After Suqour al-Sham had merged with al-Ahrar, who remained were the latter and Faylaq al-Sham. 62 According to one member of the group, even before the unification of the judicial field by the IC, '90% of the courts were under the control of the judiciary belonging to al-Ahrar. The other factions were present but for instance in Salkin there were 6 courts belonging to it and two to other factions'. 63 Before the foundation of the IC, there were also joint courts founded by the factions present in one area to regulate civilian affairs. One former member of the Hazzm movement suggested it was also dominant in these: the courts were not joint in the true sense of the word but rather al-Ahrar asked Hazzm and the Syrian Revolutionaries Front if they also wanted to have one judge in the committee which they agreed to in order to have at least one just voice represented. 64 Enduring competition with al-Nusra According to Hassan Dreim, there were two factors that hindered the work of the IC, first the concentration of regime bombings on it and second, the rise of JN. 65 In the early periods, in 2012, 2013, and even 2014 al-Nusra participated in the joint military-judicial institutions. Already in this period it tried to take control over the courts, but by 'soft power', for example, 'by supporting them in financial ways, and place its members or people loyal to it in there'. 66 Concerning the IC, Hassan Dreim describes Nusra's behaviour as follows: a delegation from Nusra came and discussions about them joining the committee began, which took 6 months. [. . .] They don't have judicially trained people; they are a group of amateurs. We even established an institute for them and trained them in judicial procedures. If they have to take over the judiciary, they should at least have the skills to do so. The dialogue continued for 6 months and after that they agreed to enter the judiciary in an organized way. Then suddenly they decided to leave without further discussion. And a representative of Jolani came and said: 'I am an envoy of Mohamed al-Jolani and we reject the former dialogues and form what is called Dar al-Qadaa '. 67 In summer 2014, al-Nusra started to create Dar al-Qadaa courts in different towns in Idlib province and began to compete with the institutions founded by the FSA and al-Ahrar. 68 As Hassan Dreim noted, 'for the citizens the confusion started, where should they go? But most people are simpleminded, they go where they get their right'. 69 That is why people in some regions reportedly turned to the Dar al-Qadaa courts. 70 The conflict between the IC and Dar al-Qadaa intensified. 'The Islamic Committee began to take selfdefensive measures. But Nusra used force against it, and attacked Maarrat Hamra, Kafranbel, and took control over the courts and 14 stations of the Free Police which coordinated with the Islamic Committee.' 71 Sometimes al-Nusra took control over the courts purely by military force, for instance in Maarat al-Nuaman; sometimes by a mixture of military and other measures, for example by founding or infiltrating a civil institution like the local councils. 72 But my interviewees agreed that its military presence in a region is the decisive factor. With the passing of time the Nusra Front less and less relied on its soft power approach adopted initially but supplemented it with violence.
Looking at the larger dynamics in Idlib province, an increasing centralization of the judicial sector took place. Initially there were many smaller and informal institutions that sprang up after the regime withdrew from large areas which were gradually replaced by more official courts founded by Ahrar al-Sham. The stronger FSA-factions like the SRF and the Hazzm movement that emerged since the end of 2013 also often were represented in these institutions. 73 After their foundation, the IC and Dar al-Qadaa increasingly monopolised these courts 74 with the balance of power increasingly favouring the latter since the beginning of 2016. 75 However, in some places a regional division of authority between the two bodies reportedly also emerged, like in Sarmada. 76 To summarize, the era of the IC lasted from the beginning of 2014 until the beginning of 2016. Dar al-Qadaa emerged in the end of 2014 and completed its control over the judicial field in the middle of 2017. 2015 was the year of competition. In the course of 2016, Dar al-Qadaa began to dominate more and more judicial institutions. 77 On 1 August 2017, JN's successor Hayat Tahrir al-Sham took control over the courts of the IC, with the exception of the court in Armanaz which continued its work until early October when it also fell under the control of the group. 78 There is one consistent pattern in the behaviour of al-Nusra: first, it joined or even helped founding shared judicial institutions in Aleppo, Idlib, and Eastern Ghouta in 2012 and 2014, respectively and then gradually withdrew from them starting in spring 2014. How can we explain this? According to Hassan Dreim, in the beginning it tried to infiltrate the courts, but it has a problem it cannot overcome, the lack of qualified people. [. . .] So it appointed amateurs that are corrupt and destroy the reputation of the court. Then it tried to target the heads of the court. When it did not succeed with this, it withdrew. [. . .] You can say it participated in the Islamic Committee to study its destruction from within. When it could not achieve that, it founded Dar al-Qadaa. 79

Multipolarity: the courts and factions in Aleppo city and its countryside
The juridical-factional landscape in Aleppo was characterized by a multiplicity of actors and a strong dispersion of power which prevented any kind of consolidation of the nascent institutions. The pioneer among opposition courts, the Unified Judicial Council (UJC) was founded in October 2012 by a group of defected lawyers, judges, and shari'een of the groups in Aleppo to regulate the relations between civilians and military actors and to protect public as well as private property, a necessity facing opposition activists after controlling a large city for the first time. 80 It was formed after several rounds of consultation with the various factions. All groups in Aleppo, with the exception of al-Nusra and al-Ahrar, signed a declaration to subordinate themselves to the UJC. Liwa al-Tawhid, one of the biggest groups in northern Syria, although it signed the founding declaration, withdrew from the council and became part of the Sharia Committee, 81 which is discussed below. After that, most of the Islamist factions fought the council, although they were consulted in its foundation. The main conflict was that it used codified law, the Unified Arab Code, which the factions rejected. In November 2013, the Council's First Judicial Chamber was attacked by Tajammu Fastaqim Kama Umirt, one of the larger factions in the city, which in the following declared its replacement with a new body called the Sharia Judiciary (Kadaa Shara'i) in Aleppo due to corruption charges. 82 This occurred after the merger between the Sharia Committee in Aleppo and the UJC had failed. Concerning the accusation of corruption, a lawyer from Aleppo said that this was true, 'but it was caused by the factions that falsely accused people in order to confiscate their property'. 83 The Sharia Committee in Aleppo and its Countryside (ASC) was formed through an agreement between the biggest four factions in Aleppo, Liwa al-Tawhid, Ahrar al-Sham, JN, and Suqour al-Sham in December 2012, because of which it was also called 'Committee of Four'. Since it was established by the strongest factions at that time and had generous funding compared to other projects its role was very important initially. 84 Soon after its foundation the Committee also expanded geographically and formed branches in the city, and in the northern and northeastern countryside of Aleppo. It also expanded to Idlib province, to Binnish, Bab al-Hawa, and Sarmada. 85 As one former member of the group told me, 'actually the Committee was Liwa al-Tawhid's idea. The idea was to form a court for both civilian and military affairs. But we were betrayed. When Nusra joined, it took everything'. 86 The Shari'a Committee did not continue its work for long, since it was weakened after the withdrawal of Nusra and ended effectively when the war with Daesh started since all resources were needed in this fight. 87 The Sharia Court in Aleppo and its Countryside emerged from the remnants of the Sharia Committee, the UJC, and other Sharia committees. 'After Fastakeem had taken over the Council and expelled its judges and replaced them with their own it was merged into the Sharia Committee together with many other Sharia courts'. 88 However, the issue of corruption did not improve, it became even worse after the security establishment of the Levant Front, an ostensibly FSA faction, had taken control over the work of the Court by approximately summer 2015. This happened by way of appointing its shari'eenwho managed the security apparatus of the groupas judges. 89 Resulting from the largely negative experiences of the juridical bodies in Aleppo, the Supreme Judicial Council was founded in July 2015 following the unification of a number of already existing courts. Several armed groups, including Nour al-Din al-Zenki and Division 16 promised to recognize the new body as the only judicial power in Aleppo governorate. 90 The Free Police served as the executive force, in addition to the factions present in the region. 91 As of June 2016, the branches of the Supreme Judicial Council, which is still existing, are primarily in the northern and western Aleppo countryside. 92 In September 2016, the Sharia Court in Aleppo merged with it and ceased to exist as a separate entity. 93 All courts in Aleppo except the UJC were connected to specific factions, such as the Sharia Committee to its founding factions, and the Sharia Court to the 'security apparatus' of the Levant Front. 94 The branches of the Supreme Judicial Council in Aleppo province are linked to the factions that are dominant in these areas, for example the court in Azaz is linked to the Levant Front and the court in al-Qasimiya is linked to Zenki. 95 The latter controls a great area in the western countryside of Aleppo and the courts in this region belonged to the court in al-Qasimia. The group also controlled the Salah al-Din neighbourhood in Aleppo and founded a branch there belonging to the Council. 96 That is why there was the impression that Zenki was in control of the whole court. 97 The UJC indirectly relied on Fastakeem or one of its sub-factions since its judicial police was too weak to engage in bigger operations. 98 All courts except the Sharia Committee were too weak to reign in the factions, 99 so they continued to operate private security and judicial institutions which arrested and kept civilians rather than transferring them to the courts as agreed upon earlier. 100 This points to the prevalence of the military logic over the judicial.
The competition between the armed groups also translated into competition between the judicial institutions, which sometimes even ended in violence. For instance, there was an enduring conflict between the Shari'a Committee and later court and the UJC in Aleppo which even resulted in the latter's occupation by the Committee and al-Nusra once. 101 As a lawyer from Aleppo and one of its founding members said, in the beginning nearly all the factions recognized the UJC, but after a while this changed in favour of the Sharia Committee. 102 To conclude, in Aleppo, rather than one court being more or less supported by all factions as in Ghouta or two competing courts as in Idlib, one could almost say that each faction had its own court. This resulted in a highly volatile and instable situation. In line with this, Aleppo experienced the greatest degree of institutional demise and the highest number of violent attacks by armed groups on the judiciary, which were mostly aimed at taking control over these institutions. In line with this situation of multipolarity is also that the Nusra Front was only one perpetrator among many in contrast to its role as the dominant challenger of the IC in Idlib and that of the rogue element in Ghouta being somewhat allowed to exist. Groups using the 'FSA' label were equally involved in attacks on the judiciary in Aleppo. 103

Conclusion
While the literature on guerrilla and insurgent groups has documented that most armed actors form courts and other, justice-related institutions, this paper has extended these insights by showing that there are several pathways of (trans)formation of insurgent courts. These are strongly connected to the power distribution between groups fighting on the same side, yielding a system of unipolarity, bipolarity, or multipolarity. As shown in the analysis, the situation in Eastern Ghouta was characterized by a multiplicity of armed groups, a tightening siege and a deteriorating security situation. Due to the proximity to the capital, the threat of a large regime offensive always loomed as a threat in the background which finally materialized in early 2018. Together with the siege, this imposed a great degree of discipline on the insurgents. In this context, Jaysh al-Islam managed to establish a system of hegemony between mid-2013 and early 2016 by imposing its control ruthlessly.
Between 2014 and mid-2016, the joint military and judicial institutions reached a remarkable level of consolidation in an otherwise hostile context. In contrast to Aleppo, the victims of attacks against the judiciary were mostly persons, not institutions, which confirms this relative stabilization. However, the system was weakened the moment a competitor with a similar strength entered the picture due to the merger between the second (Faylaq al-Rahman) and third strongest groups (Ajnad al-Sham). In the case of the bipolar situation in Idlib, the competition between the IC and Dar al-Qadaa resulted in a less stable situation particularly since 2016 when there was the constant danger of an attack by al-Nusra. However, a certain consolidation of the judicial sector was still achieved. Finally, the situation in Aleppo remained highly fragmented and did not pass the phase of emergence. The constant external military pressure certainly contributed to this, but as I have shown another central factor was the wanting concentration of power in one or few actors. This was also due to the quick demise of groups that have been strong in early times such as Liwa al-Tawhid. As several interview partners indicated, the success attributed to the judicial institutions analysed here was in the same order.
The findings are in line with those IR scholars who argue that unipolarity is the most stable system, followed by bipolarity. Multipolarity has proven most instable. However I would argue that it is not the constellation as such (which can be quite static) but rather dynamic factors such as the degree of institutionalization (which proved quite low due to the overall trajectory of the war) and changes in the distribution of power between groups that explain conflict. IR theories developed to explain the international system have some use in shedding light on the processes analysed here, but need to be supplanted by anthropological work on wartime governance and law that provides better insights into the internal dynamics.
Another theoretical question concerns the links between the external (macro) as well as the internal military situation and the (trans)formation of the insurgent courts. I have argued that primarily the latter is relevant for the dynamics analysed here. However, it would likely be too reductive to see the judicial arena as a mere copy of the distribution of power between military groups. While all courts emerged as part of the factions' organizational structure, their path towards autonomy differed. While the judicial institutions in Aleppo never could emancipate themselves from the military logic, the Judicial Council in 2014 and 2015 could break free to a certain degree, but finally the force of arms prevailed when the internal conflict between Jaysh al-Islam and Faylaq al-Rahman resulted in the split of Eastern Ghouta into two. All interviewees confirmed that the judicial independence of the IC in Idlib expanded over time, as exemplified by the increased role of lawyers, despite its greater reliance on Ahrar al-Sham. Why some judicial bodies could at least partly leave the orbit of armed groups while others did not achieve this should be explored by further research.
From my analysis, a number of questions arise for comparative studies of insurgent courts: Is Syria a special case? Or do different pathways of (trans) formation also exist in other contexts? How do insurgent courts established by actors with a strong Islamist clout differ from institutions founded by other types of insurgents concerning the dynamics analysed here? Are there other pathways of emergence and (de)stabilization not considered here? When asking my interview partners about failure and success of the opposition courts they always referred to Dar al-Adl, the House of Justice in Southern Syria as the best example. Here, Islamist groups were marginal and the dominant FSA groups subordinated themselves to the court, offering a venue for further research and maybe a more hopeful view into Syria's future.  " 137-38. 33. It publicly declared its split from Al-Qaeda in July 2016, which was seen by many analysts as a ruse, however. Recent differences in opinion and armed clashes between members of the two camps have however put this reading into question, pointing to that it was at least by Jolani intended as a an actual split. Al-Tamimi, "Al-Qa`ida Uncoupling"; Al-Tamimi, "The Hay'at Tahrir Al-Sham-Al-Qaeda Dispute I-III"; Lister, "How Al-Qa`ida Lost Control of Its Syrian Affiliate." 34. During 2015 the regime's forces withdrew from the remaining areas in Idlib and the leading opposition coalition, the Army of Conquest (Jaysh al-Fatah) formed its own court in the provincial capital. However, it stopped functioning after a short while following aerial bombardment in which 30 civilians died. As a consequence, people avoided it. 35. The siege and particularly the intensive bombing by the Syrian regime made access to interview partners from or residing in Eastern Ghouta more difficult than access to people from Aleppo or Idlib. 36. For the problems of fieldwork in the context of violent conflict see Nordstrom and Robben, Fieldwork under , 1968-1969," 121. 39. The use of torture was widespread in the detention centres run by Jaysh al-Islam. al-Zarier, "How Does a Jaish Al-Islam Detention Center Differ from a Regime Prison?" Furthermore, its leader Zahran Alloush was accused of kidnapping the "Douma Four", famous human rights activists such as Razan Zeitouneh whose fate is unknown until today. 40. Youssef Sadaki, researcher, Skype, 31 January 2018; Lund, "Into the Tunnels." 41. "Almjls Alqdaʾ Llghouta Alshrqyh." 42. For instance, Khaled Tafour had served as a religious teacher for the leaders of Ajnad al-Sham, Abu Abderrahman Zeinelabidine had links to the Islam Army, and Abu Rateb Abu Diqqa was the father-in-law of Failaq al-Rahman's leader Captain Abdel-Naser Shmeir. Lund, "Into the Tunnels."

Notes
friendship. I am grateful to the Max Planck Institute for Social Anthropology for funding my doctoral research, including my stays in the field.

Disclosure statement
No potential conflict of interest was reported by the author.