Tribal engagement strategies in the Islamic Emirate of Azawad

ABSTRACT

The Tuareg uprising in northern Mali in 2012 culminated in the establishment of the Islamic Emirate of Azawad, a proto-state run jointly by Tuareg Islamists in Ansar Dine and foreign jihadists with connections to al-Qaida in the Islamic Maghreb (AQIM). Ruling over a vast territory where jihadist ideology had little traction, the jihadists had to navigate a complex field of tribal politics to stay in power. This article explores jihadist tribal engagement strategies in Azawad during the occupation of northern Mali (2012–2013) and discusses how the relationships with tribal chiefs and traditional authorities impinged on the state-building efforts. Based on novel primary sources and local media reports, this article shows that the jihadists sought above all to avoid a tribal uprising and, therefore, pursued a relatively lenient policy towards the tribal chiefs. The comparatively weak position of the jihadists gave the chiefs some bargaining power in the new political settlement. However, collaboration with the jihadist rebels increased over time as their territorial control became more consolidated.

KEYWORDS:

• Rebel governance
• jihadism
• Mali
• AQIM
• Ansar Dine
• tribe

Introduction

‘States are not created from one day to the next. Lots of factors are needed for it to succeed […] and gaining the loyalties of the powerful tribes is an important one’ (Bin Ladin 2010, 7–8). Usama bin Ladin imparted this advice on state building to the leader of al-Qaida in the Islamic Maghreb (AQIM), Abdelmalek Droukdel, in 2010. Two years later, the latter forwarded the same message to his sub-commanders (Droukdel 2014, 6), who had just conquered northern Mali and established the self-declared ‘Islamic Emirate of Azawad’. Following the advice of Bin Ladin, and cognisant of the precariousness of their own position, the jihadists in Azawad made cultivating good relations with the tribal chiefs a key policy concern.

The purpose of this article is to explore jihadist tribal engagement strategies in the Islamic Emirate of Azawad (2012–2013), and explain how the jihadists’ interactions with kinship groups in northern Mali impinged on their state-building efforts. I argue that the leaders of the so-called ‘Islamic Emirate of Azawad’ (henceforth ‘the Emirate’) had learnt from previous jihadist state-building ventures and were determined not to provoke a tribal uprising against them, as had happened with al-Qaida in Iraq in 2006 (Cigar 2014, 7–15). Owing to this fact, and to the relatively weak position of the jihadists in the region, opposition to jihadist rule was not met with violent reprisals. Instead, the jihadists treated tribal chiefs with respect, tolerating some dissent while actively trying to persuade them to support the jihadist state project. Eventually, the jihadists managed to secure the support, or at least the acquiescence, of several tribal chiefs in northern Mali.

The past decade has seen increased scholarly interest in the field of rebel governance,¹ and a number of works have been dedicated to the study of jihadist rebel governance in particular.² Most jihadist rebel governance projects have emerged in places where tribes are an important fixture of social and political organisation. However, the literature has only recently started to engage seriously with the interactions between rebel governments and kinship groups.³ The fact that tribal engagement was a major concern for the jihadists in Azawad should alert us to the importance of studying this phenomenon for understanding of the Emirate. Furthermore, it can provide new insights into the general phenomenon of how rebel governments and kinship groups influence each other.

The case at hand highlights two aspects of interactions between jihadist rebel governance and tribes that are under-explored in the literature. First, the Islamic Emirate of Azawad showcases an example of cross-organisational learning. As argued in Whiteside and Elallame (2020), al-Qaida in Iraq (AQI)’s mishandling of tribal relations cost them dearly, forcing its successor group ISIS to adopt a far more fine-tuned tribal engagement policy. This article shows that AQIM also tried to incorporate the lessons from AQI in its tribal engagement policy. Second, most other studies have dealt with situations where the rebel governors are relatively powerful vis-à-vis the tribes, as for example the governance projects of ISIS⁴ and al-Shabaab.⁵ In these cases, the rebel governors have been able to use violence or threats thereof to impose their will on tribal chiefs. The Islamic Emirate of Azawad, on the other hand, presents a case where the position of the rebel governors was far too weak to take hostile action against non-cooperative tribal chiefs, forcing them to adapt their strategy accordingly. In places where the traditional authority of tribal chiefs and religious scholars had greater legitimacy, the jihadists were not able to impose the totality of their political programme, corresponding to the patterns explored by Ana Arjona in Rebelocracy (2016).⁶

While the literature on governance in the Islamic Emirate of Azawad itself remains in its infancy, the past two years have seen an increase in academic studies on this case. Notable examples include Bouhlel and Guichaoua (2021), which investigates differences in the use of violence by jihadist rebels in Kidal and Gao and Skretting (2022), which provides an empirically grounded discussion of governance practices in the Emirate. Another important work for the purposes of this article is Svensson and Finnbogason (2021), which discusses patterns of civil resistance against jihadist rule in Azawad. Previous studies, including Bøås and Torheim (2013) and Raineri and Strazzari (2015), have shown how the emergence of a trans-Saharan smuggling economy led to heightened tensions between tribal interests in northern Mali and contributed to the eventual outbreak of the 2012 rebellion. Overall, however, the influence of tribal interests on the Islamic Emirate of Azawad from a governance perspective remains understudied, and the present study seeks to fill some of that lacuna.

This article relies primarily on written primary and secondary source materials that provide insights into the tribal engagement strategies of the jihadists in Azawad, as well as examples of their praxis. The primary sources include a number of internal letters between commanders in AQIM, between leaders in AQIM and AQC (al-Qaida’s Central organisation), as well as publicly available interviews with jihadist leaders involved in the Emirate, detailing these organisations’ strategies on state building and tribal engagement. These materials include an October 2010 letter from Usama bin Ladin to AQIM-leader Abdelmalik Droukdel retrieved in Bin Ladin’s compound in Abbottabad in 2011 (Bin Ladin 2010), a July 2012 letter from Droukdel to the Shura Council of AQIM in the Sahara and Ansar Dine (Droukdel 2014), and a 176-page interview conducted with a leader identified as Abd al-Aziz Habib, conducted by Zakariyya Bughrara in 2014.⁷ The secondary sources comprise local media reporting in Arabic and French, reports from NGOs, as well as publicly available legal documents with reports and testimonies from the time period in question. The article focuses primarily on Ansar Dine and AQIM’s dealings with Tuareg and Arab tribes in Kidal and Timbuktu, where sources are more abundant than for MUJAO’s occupation of Gao.

The article proceeds as follows: First, I provide a brief discussion of the term ‘tribe’ and its significance in northern Mali, leading up to the 2012 Tuareg rebellion that culminated in the establishment of the Islamic Emirate of Azawad. Next, I explore AQIM’s tribal engagement strategies as laid out in the organisation’s internal documents. Thereafter, I discuss how the jihadists’ tribal engagement worked in practice, showing that they were somewhat naive about the prospects of gaining support for their project. Faced with open opposition from the chiefs of prominent tribes, the jihadists opted for accommodation rather than suppression. In the subsequent section, I show that the jihadists eventually managed to secure some support from tribal chiefs, partly through accommodating their demands, and partly because competing rebel groups that acted as conduits for opposition were marginalised. I conclude by providing a preliminary assessment of the tribal engagement strategy employed in Azawad, and its results.

Background: Tuareg tribes in Northern Mali from independence to rebellion

The vast, sparsely populated territory of northern Mali, comprising the administrative regions of Kidal, Timbuktu, and Gao, is home to the country’s Tuareg population. Although the Tuareg only account for an estimated 1.7% of Mali’s population (’Mali’ 2021), they constitute 86% of the population in Kidal, as well as a significant minority in Timbuktu and Gao, with an estimated 29% and 28%, respectively (Les régions maliennes de Gao, Kidal et Tombouctou: Perspectives nationales et régionales 2015, 13). Tribalism – here defined as a state of affairs where groups bound by inward loyalty based around a (real or imagined) common lineage are a fixture of social and political organisation⁸ – remains a distinct feature of Tuareg society. However, the precise meaning of what a ‘tribe’ is, as well as its social and political significance, has undergone profound changes over the past century. This section will give a brief discussion of the terminology used to describe the phenomenon and an outline of the ‘tribal’ landscape in Mali in the events leading up to the 2012 Tuareg rebellion.

The term ‘tribe’ in northern Mali defies straightforward definition. The Tuareg for kinship group is tewsit, which refers to a single group or clan sharing the same lineage. However, it can also refer to a bundle of clans sharing a more distant ancestor – which can tentatively be termed a ‘tribe’. These ‘tribes’ were incorporated into larger political units, the tribal federation or ettebel, led by a tribal chief or amenokal (Lecocq 2010, 13). Traditionally, this tribal system has been marked by a complex internal hierarchy of status and dependency, which has been likened to a class or ‘caste’ system (Lecocq 2010, 4). The main delineations in the system were those between the imushagh or noble tribes, with dependent groups of imghad, who were considered free but subservient to a noble tribe, as well as inadan (craftsmen) and iklan (slaves).⁹ However, while aspects of this ‘traditional’ picture retain some importance, social and political convulsions over the past 100 years have rendered today’s ‘tribal’ system almost completely detached from its pre-colonial origins.

The boundaries of what constitutes a ‘tribe’, as well as the terminology used to describe the phenomenon, were profoundly influenced by both the French colonial administration and the Malian state’s efforts to administer ‘tribal’ affairs in the north. The French sought to rationalise administration among the nomads by imposing the terms tribu (hence the use of the term ‘tribe’) and fraction as, respectively, the tribal (i.e. tewsit) unit and subunit (Lecocq 2010, 10–12). These terms designated administrative units, which were organised and reorganised to suit the needs of the colonial power, thereby partly detaching the tewsit from its original kinship-based significance. In the post-colonial Malian state, the tribu was eventually abolished as an administrative unit, while the tribal chiefs were informally, and later formally, kept in place as local leaders. Further state-imposed reorderings of the tribal system followed.¹⁰ The fluid and contested nature of the ‘tribal’ phenomenon among the Tuareg in northern Mali therefore makes it difficult to find a useful terminology, but for the purposes of this article, I will rely on the terms clan, tribe, and federation to denote different levels of ‘tribal’ organisation.¹¹

The colonial experience and its aftermath also brought profound changes to the internal, hierarchical structure of Tuareg society. Slavery was outlawed in French Sudan in 1905, leading to the gradual emancipation of Tuareg slaves from the 1940s onward. In the post-colonial era, the Malian state, inspired by Marxist ideology, began a conscious policy to undermine the hierarchy by empowering non-noble clans to break what was termed the ‘feudal’ power of the Tuareg nobles (Molenaar et al. 2019, 39). At the same time, opportunities for social advancement presented themselves to Tuaregs who were previously considered socially inferior through the expansion of the modern state and education system, as well as new opportunities for economic gains through the development of a lucrative trans-Saharan smuggling economy (Raineri and Strazzari 2015). As such, the actual socio-economic class distinctions between noble and non-noble Tuareg clans gradually eroded, opening up a space for previously subaltern groups to challenge the remainders of the old hierarchy.

Although these developments had served to undermine the tribal structuring of Tuareg society, a succession of rebellions and ensuing political instability in the north gave tribal loyalties renewed importance in social and political organisation. While a first, short-lived Tuareg rebellion had broken out in 1963, a much larger, second rebellion followed from 1990 to 1996, in which a number of Tuareg groups fought the Malian government for different goals, ranging from self-determination to full independence (Mali: Avoiding Escalation 2012, 3). As the conflict dragged on, the rebel movement eventually splintered into warring factions increasingly demarcated by tribal affiliations (Lecocq 2010, 316–317). With the peace agreement signed after the rebellion, a process of democratisation and decentralisation of administration in northern Mali was initiated. As it turned out, the practical consequence was that more power fell into the hands of the tribal chiefs (Lecocq 2010, 379–380). Furthermore, the weakening of direct state control in the north gave more space for inter-tribal rivalries to play out, both on the elite level between different clans in the noble Ifoghas tribe in Kidal, and between this elite and an emergent, self-conscious, and increasingly powerful class of imghad Tuaregs.¹² Contestations over elected offices, not to mention the fight for control over the smuggling business, contributed to heightened tensions between various tribal groups (Bøås and Torheim 2013). These fault lines again turned violent when another Ifoghas-led and ostensibly pro-separatist rebellion broke out in 2006, which was eventually quashed by a Malian-sponsored imghad militia led by ex-rebel El-Hajj Gamou. In sum, although successive rebel leaders had raised the flag of a national, Tuareg identity, the end result was a refocusing of tribal identities.

Conflicts along tribal lines likewise played a role in the lead-up to the events of 2012, when another rebellion instigated by pro-independence Tuaregs was eventually hijacked by jihadist groups. The seeds of this hitherto last rebellion were sown in 2010–2011 when Tuareg activists began openly advocating for northern independence, just as well-armed Tuareg fighters from Muammar Gaddafi’s army began returning en masse to Mali as the Libyan regime was poised to fall. The Malian government made efforts to welcome the returnees and facilitate their socio-political integration, but with mixed results. Many of the imghad returnees were integrated into the aforementioned El-Hajj Gamou’s pro-Malian militia. However, most returnees from tribal groupings such as the Ifoghas, Chamanamass, and Idnane regrouped in Zakak in the Kidal Region, where they entered into discussions with pro-independence activists and local power brokers to plan a new revolt (Diarra 2012; ’Tuareg, Mali and a Post-Gadaffi Sahel: Rising Risks to Oil Exploration and Mining Operations’ 2011). The assembled politicians and fighters created MNLA (Mouvement national de libération de l’Azawad), a secular separatist movement that would take the lead role in the rebellion (Mali: Avoiding Escalation 2012, 10).

Shortly afterwards, Iyad ag Ghali, a veteran of the 1990–1996 and 2006–2009 rebellions, established another Tuareg group poised to challenge the Malian state, but with the explicit intention of implementing the shariʿa – Ansar Dine. A sizeable number of Ifoghas joined the group, including several high-ranking nobles with considerable clout and political experience. As Alexander Thurston suggests, their decision to join may have been influenced by the fact that MNLA, even though it was led by the Ifoghas Bilal Ag Achérif, appeared to be dominated by Idnane and imghad fighters, with whom the Ifoghas had been embroiled in bitter conflicts in previous rebellions (Thurston 2020, 130–132). In any case, the presence of prominent Ifoghas politicians in Ansar Dine softened its public image, making it appear less extremist and thereby facilitating its cooperation with MNLA. However, among the early recruits to Ansar Dine were also Tuareg members of AQIM, a jihadist group originating in Algeria, which had maintained a presence in northern Mali since the early 2000s (Skretting 2020). While AQIM’s membership remained mostly non-Malian, its commanders had become tightly integrated into the local political and social landscape through business connections and marriage alliances (Bøås and Elin Torheim 2013, 1287). Even though their involvement was denied by MNLA at the time, AQIM’s southern brigades would fight side-by-side with Ansar Dine, and by extension MNLA, in the rebellion that broke out in January 2012.

By April 2012, all the major cities of the north had been captured by the rebels, but the alliance between MNLA and the jihadists soon began to fray. In June, MNLA was ousted from the provincial capitals of Kidal, Timbuktu, and Gao by their erstwhile allies, who declared Azawad an ‘Islamic Emirate’. Iyad ag Ghali was named governor of Kidal and the overall emir (leader), ruling the Emirate in close cooperation with AQIM, whose commander Abu Zayd became governor of Timbuktu (Bughrara 2014, 13). A third jihadist group, MUJAO (Mouvement pour l’unicité et le jihad en Afrique de l’Ouest), had seized the city of Gao. While MNLA and other militias held on to some minor towns for a little longer,¹³ the jihadists had for all intents and purposes captured northern Mali by July 2012. As state-appointed administrators fled the north, tribal chiefs and other traditional authorities became the primary mediators between the local population and the new, jihadist masters (Molenaar et al. 2019, 43), who in turn sought out ways to win their support.

AQIM’s tribal engagement strategy

The sudden fall of northern Mali to the jihadists prompted AQIM to formulate a strategy for governing the territory and dealing with the local tribes. They were not wholly unprepared, as strategies on tribal engagement had already been an important theme in AQIM correspondence with AQC (al-Qaida Central) prior to 2012. Internal communication within AQIM and Ansar Dine dated shortly after the establishment of the Emirate shows that a tribal engagement policy was being elaborated, based in part on the previous discussions between AQIM and AQC. Successful engagement with the local tribes was considered a key factor for the budding state to succeed. This section will discuss the main contents of these internal strategic discussions.

The backdrop to the discussions on tribal engagement between AQIM and AQC was formed by the downfall of al-Qaida in Iraq (AQI). Despite enjoying some initial support in their fight against the US occupation, AQI gradually alienated the Sunni tribes in the al-Anbar province by trampling on local customs and brutally retaliating against tribesmen suspected of collaborating with the Americans. In 2006, tribal militias known as the ṣaḥwa, with support from the US forces, rose up against AQI and dealt the group a crushing blow (Cigar 2014, 7–15). The lesson from Iraq did not go unheeded in the rest of the al-Qaida organisation, and the importance of avoiding a replay of the Iraqi scenario became the overriding concern in AQIM-AQC discussions on tribal relations in the years that followed.

Thus, in Usama bin Ladin’s treatment of the subject to AQIM-leader Droukdel, a letter dated 17 October 2010, the late al-Qaida leader’s recommendations revolve primarily around what actions to avoid, rather than a positive programme for winning tribal support. The most important thing to avoid, Bin Ladin argued, was provoking the tribes to blood vengeance (thaʾr) (Bin Ladin 2010, 5–6). The jihadists, he says, enjoyed support among Iraqi tribes after the US invasion, and he singles out the jihadists’ killing of tribesmen in Anbar as the main factor in turning tribal sentiments against them. Commenting on the failure in Iraq, Bin Ladin notes that ‘killing one member of a tribe is sufficient to provoke blood vengeance in those circumstances, so what then if a hundred are killed?’ (Bin Ladin 2010, 4). To avoid this scenario, Bin Ladin admonishes AQIM not to attack members of tribes except in cases of direct self-defence (Bin Ladin 2010, 4).

However, Bin Ladin sees the tribes not only as a danger to be avoided but also as a constituency to be courted. Gaining the loyalty of the tribes, he argues, is a key element that has to be achieved to establish an Islamic state (Bin Ladin 2010, 7–8). He states that ‘if the mujahideen deal well with the tribes, the tribes will for the most part side with them’ (Bin Ladin 2010, 6). Exactly how one deals ‘well’ with the tribes, however, is left open to question. Bin Ladin is far less explicit on his plans for gaining tribal loyalties than his strategies for avoiding tribal conflict.

When Ansar Dine and AQIM’s commanders in the Sahara had captured northern Mali in 2012, Droukdel found it necessary to formulate a more detailed programme to win tribal support for the jihadist state in the process of being established. The outline of this programme can be found in a letter from Droukdel to the Shura Council of AQIM in the Sahara and Ansar Dine, dated 20 July 2012, where the AQIM leader lays out a series of policy recommendations for the Emirate going forward (Droukdel 2014). Like Bin Ladin, Droukdel’s primary concern is to avoid a new ṣaḥwa. However, Droukdel, in contrast to Bin Ladin, harbours no illusions of how difficult it will be to navigate the tribal landscape, warning that ‘you are moving in a veritable mine field of tribalism and vendettas and conspiracies, corruption, plots and deception, so you have to be clever and alert’ (Droukdel 2014, 9).

Undaunted, Droukdel proposes a tribal engagement strategy that is above all centred on including members of the various tribes in governance, which, he argues, would serve two important aims. First, it was a practical necessity – since the burden of governing a country far exceeded what the mujahideen could realistically do on their own, they would have to rely on the capabilities of local people (Droukdel 2014, 12). While exploiting the competencies of tribe members was important for practical reasons, the second, more pressing concern was to keep the tribes happy through tribal representation.

Discussing the criteria for choosing future government ministers, Droukdel argues that ‘Competency is important, but tribalism is also important and we absolutely have to take that into account, so let us […] ensure that we include some of the tribal chiefs and, if they can be found, competent people, from every large tribe’ (Droukdel 2014, 13). Interestingly, Droukdel states that ‘loyalty to Islam and the shariʿa’ would be a requirement for those ministries that were especially important to the jihadists,¹⁴ but not necessarily for others. In fact, Droukdel wanted the government of the Emirate to include not only tribal representatives in governance but also representatives from as many constituencies he could find, which also encompassed other political factions, as well as all of Azawad’s ethnic groups. This would serve a dual purpose; first, it would ensure that all major societal groups would have a stake in the new state project, and second, it would prevent the Emirate from being branded ‘jihadist’, with the inevitable negative repercussions that would follow (Droukdel 2014, 12–13).

However, Droukdel does not only wish to gain the acquiescence of the tribes but aims to gain their active support, with all the advantages that might confer. As he states:

As for tribalism, if it is properly directed and controlled by the rules of the shariʿa, then even if it does not confer any benefits, it will not do any harm. However, these tribal animosities, even if they are harmful, their benefits, when properly directed and organised, are greater than we can imagine. (Droukdel 2014, 9)

To achieve this goal, Droukdel advises his subordinates ‘to mobilise tribes with their different orientations, and the local movements, [to work] towards big, shared goals’. Being aware that salafi-jihadist ideology in itself had little traction among the tribal chiefs and was ill-suited to forming cohesion in the broad coalition that he sought to build, Droukdel suggests a more general unifying goal could be ‘fighting against injustice’. Droukdel sees this ‘fight against injustice’ as a common denominator between themselves and the tribes but suggests that the struggle should not be couched in explicitly salafi-jihadist ideological terms, but as a struggle against the injustice and humiliation that the Malian state has inflicted upon Azawad. Droukdel’s idea is therefore to appeal to ‘tribal reasons and nationalist passions’ to unify the north against the Malian state, under jihadist tutelage (Droukdel 2014, 9).

In sum, while Bin Ladin’s advice on tribal engagement above all encouraged the mujahideen not to attack and antagonise the tribesmen, Droukdel also formulated a positive programme for gaining tribal support. This programme emphasised broad tribal representation in government, exploiting nationalist sentiments rather than salafi-jihadism to formulate a unifying goal, and directing tribal animosities towards fighting a common enemy. However, Droukdel’s theorising displayed a profound lack of understanding of how diverse and contested the ‘tribal’ landscape in northern Mali actually was. As shown in the previous section, opposition to the Malian state and support for Azawadi independence was by no means uniform among Tuaregs, to say nothing of other ethnic groups in northern Mali. In other words, there was no single ‘tribal’ interest to which the jihadists could appeal, and salafi-jihadism, as Droukdel had correctly identified, enjoyed little popular support. This problem became clear for everyone to see as soon as the jihadists in Azawad attempted to put theory into practice.

Resistance to jihadist governance: the case of Kidal

When the jihadists had established an Islamic Emirate and ousted their former allies in MNLA from all of Azawad’s major cities in June 2012, they were faced with the daunting task of governing the vast areas they had conquered, and their swift implementation of a fundamentalist interpretation of shari’a law sparked immediate resistance. Previous research has shown that civil resistance against jihadist rule was widespread in the Emirate (Svensson and Finnbogason 2021). This section shows that tribal chiefs played an important role in these efforts, particularly in Kidal, which limited the extent to which a jihadist political order could be established.

Demonstrations against Ansar Dine’s rule erupted in Kidal on 5–6 June, just days after the proposed power-sharing agreement with MNLA fell apart. The demonstrators, mostly youths and women, converged outside of the residence of the Ifoghas tribal chief and amenokal of the Kel Adagh tribal federation, Intallah Ag Taher, requesting his support for their cause (Adam 2012). The Islamic police brutally dispersed the demonstration, whereupon Ansar Dine clashed with MNLA, the former accusing the latter of having stirred the populace against them. Reports indicate that these conflicts were partly fought over tribal boundaries, pitting primarily Ifoghas Ansar Dine members against MNLA members from the Idnane and Taghat Melet tribes (Africa Radio 2012; France 24 2012b). In the end, however, while Ansar Dine gained the upper hand in Kidal militarily, the demonstrators and MNLA won the support of the amenokal. Two days after the demonstrations, Intallah Ag Taher issued an open letter to ‘all those with whom I share bonds of kinship, my friends the tribal chiefs and the ʿulamaʾ (religious scholars)’, asking them to leave Ansar Dine immediately (Keita 2012). Intallah denounced Ansar Dine for its connections and ideological affinity with AQIM. He further criticised the group for imposing religious precepts incompatible with local customs (Keita 2012).

In an effort to respond to the local community and dignitaries in the aftermath of the popular outrage and the harsh criticisms from the amenokal, Ansar Dine leader Iyad ag Ghali convened a conference of more than 100 notables of the Kidal area, i.e. tribal chiefs and religious scholars, on 17–19 June (Mali Actu 2012a, 2012b). However, reports from the conference suggest that Ag Ghali did not attempt to accommodate the criticisms Ansar Dine faced, but instead tried to get the assembly on board with his radical political programme. Rather than walking back on the implementation of his brand of the shari’a in Kidal, Ag Ghali appealed to the assembly to support implementing the shari’a, not only in Azawad but in all of Mali. This was a repudiation, not only of his erstwhile ally MNLA’s secularism but, more importantly, of its goal of making Azawad an independent state. Moreover, Ag Ghali did not repudiate alleged jihadist ties but seemingly confirmed them by calling for a jihad in Mali and throughout the world (Mali Actu 2012b).¹⁵ Iyad ag Ghali received a cold reception; neither the religious scholars nor the tribal chiefs were enthused about his political programme or his plan for holy war but insisted on following the Islam practised by their forefathers (Mali Actu 2012b).

Ansar Dine’s initial overtures therefore did not win them the active support of the tribal chiefs in Kidal, and not even their acquiescence. Intallah Ag Taher continued to speak openly against the group and used his clout as the amenokal of the Kel Adagh tribal federation to rally opposition against them. In September, Intallah Ag Taher and Cheick Baba Ould Sidi El Mokhtar, the chief of the Arab Kunta tribe, convened a meeting of Tuareg and Arab tribes from all of northern Mali to discuss the situation. Ag Taher and Cheick Baba Ould Sidi El Mokhtar secured the support of several tribal chiefs on a statement calling for all foreign members of armed groups to leave Azawad, and for Azawadi members of Ansar Dine to abandon the group and join MNLA (Sahara Media 2012a). Representatives of Ansar Dine reportedly attended the meeting, entered into discussions with the tribal chiefs, and treated the amenokal with the utmost respect (Al-Fajr (Algeria) 2012). Similar meetings were held again in October, where Ag Taher and Cheick Baba Ould Sidi El Mokhtar were joined by the chief of the Idnane tribe, Youssouf Ag Cheick, in demanding that Ansar Dine ask forgiveness for their crimes against the Azawadi people and fight alongside MNLA against the non-Azawadi ‘terrorists’ in AQIM and MUJAO (Ag Mouchallatte 2012).

When the traditional authorities – tribal chiefs and qadis¹⁶ – could not be persuaded to come on board with his programme, Iyad ag Ghali’s opportunities to effect significant political and judicial changes in Kidal remained limited. Jihadist governance in Kidal therefore turned out to be quite light-handed in comparison with Timbuktu and Gao. Several arguably strict policies such as the imposition of the veil, the illegality of cigarettes and alcohol, and the ban on eating publicly during Ramadan were enforced. Corporal punishments, however, were not, albeit with a few exceptions.¹⁷ Prior to Ansar Dine’s takeover, disputes between various clans had been settled by a qadi, who made rulings based on Islamic law (Dakouo 2009). The leading qadi in Kidal before 2012, Ahmad Ould Baana, remained the leader of the council of qadis during Ansar Dine’s control of the city (Dakouo 2009; Bughrara 2014, 74), and the council continued to adhere to a relatively moderate interpretation of the shariʿa under Ansar Dine’s rule, as they had done in the past (Molenaar et al. 2019, 103).

Bouhlel and Guichaoua argue that this absence of shariʿa -mandated corporal punishments in Kidal stems from the fact that Iyad ag Ghali ‘deliberately tapped a local reservoir of jurisprudential resources to minimize violence against civilians’ (Bouhlel and Guichaoua 2021, 16). In my view, a more likely explanation for the discrepancy is that significant local resistance prevented Iyad ag Ghali from imposing the full extent of his political programme. Crucially, this resistance found active spokesmen among tribal chiefs and religious authorities. Acts of civilian mobilisation against jihadist rule were just as prevalent in Timbuktu and Gao (Svensson and Finnbogason 2021, 580–583), but these cities did not have a similar mobilisation of traditional authorities.¹⁸ As for Iyad ag Ghali’s intentions, his initial overtures to the notables of Kidal do not in any way suggest that a ‘pragmatic’ or ‘moderate’ implementation of shari’a law, or indeed minimisation of violence against civilians, was a key concern. It is also known that Iyad ag Ghali personally ratified some of the most egregious crimes against humanity that occurred in Timbuktu under his nominal deputies in that city, including corporal punishments and the destruction of UNESCO world heritage sites.¹⁹

Moreover, whether because the traditional authorities’ rejection of jihadism had influenced them or because their preferences already aligned with theirs, the Ifoghas politicians who held the reins of the Ansar Dine faction in the Kidal also seemed to support relatively moderate policies. They distanced themselves from the hard-line, AQIM-dominated faction in Timbuktu, with Alghabass ag Intallah, the son of the amenokal and himself an Ansar Dine commander, reportedly asking AQIM’s representative in Kidal to leave the city (Thurston 2020, 133–134). Iyad ag Ghali, on the other hand, appeared ideologically and politically closer to the Timbuktu faction. The Ifoghas faction’s eventual breaking away from Ansar Dine when the French military intervention began in January 2013 shows that the ideological split between the two sides was significant, and perhaps also suggests that many of the Ifoghas saw Ansar Dine primarily as a vehicle for their continued political domination in Kidal, rather than as an ideological project for which they were ready to fight to the end.

In summary, Iyad ag Ghali sought the active support of tribal chiefs and religious scholars for a radical programme that envisaged implementing the shari’a in all of Mali through jihad, but was firmly rebuffed. Jihadist governance in Kidal was met with significant civil resistance, and unlike in Gao and Timbuktu, the traditional authorities of Kidal, both tribal and religious, provided legitimacy to this resistance and gave it their vocal support. Consequently, Kidal experienced a ‘lighter’ form of jihadist governance, which was more in tune with local sensibilities and customs, and which saw a significant continuation of personnel pre- and post Ansar Dine’s takeover.

Pragmatic support from tribal chiefs

While Iyad ag Ghali’s initial overtures to the tribal chiefs in Kidal were rebuffed, throughout the latter part of 2012, reports emerged indicating that the jihadists were having more success at securing nominal support or acquiescence from other tribal chiefs. As will be shown in this section, this development can be explained by two main factors. First, the jihadists appear to have refrained from excessive meddling in their internal affairs and adopted a policy that was more accommodating of local customs. Second, taking its cue from Stathis Kalyvas’ theory of control and collaboration in civil war, this section argues that collaboration with the incumbent jihadist rebel regime increased as MNLA ceased being a fighting force that could mount an effective or realistic opposition to the jihadists (Kalyvas 2006, chap. 5).

In a 2014-interview, a leading figure in the Islamic Emirate identified as Abd al-Aziz Habib explains that from the moment they captured the cities of Azawad, the mujahideen engaged constructively with local leaders and tribal dignitaries, as per the instructions of Droukdel. They reportedly met with them to ‘study how the new administration would work, and to seek their aid’, and the response was allegedly positive, ‘especially in the region of Kidal’, with many of them joining the side of the jihadists (Bughrara 2014, 27). Accurate or not, it is interesting to note that the aspect of cooperation with tribal dignitaries is something Habib chooses to highlight. As seen, the initial response from the most prominent tribal chiefs in Kidal towards the jihadists was not uniformly welcoming, but there is some independent evidence for increased support, or at least acquiescence, from other tribal chiefs towards the jihadists at a later stage.

What is clear, however, is that already shortly after the cities of the north had fallen to the combined forces of MNLA and Ansar Dine and their jihadist allies in early April, trickles of MNLA fighters had begun defecting to Ansar Dine (France 24 2012a). MNLA’s conduct when entering Timbuktu and Gao – which had included widespread looting, as well as instances of rape and wanton killing (Mali: War Crimes by Northern Rebels 2012) – had lost the group popular support, especially among the non-Tuareg populations. Ansar Dine appeared to many as the more organised, and as time passed, military stronger party (France 24 2012a). As MNLA was chased from the major cities of the north in June, even more MNLA fighters reportedly joined the winning side (Illicit Trafficking and Instability in Mali: Past, Present and Future 2014, 16). As the months passed and the prospect of an MNLA comeback grew less likely, more high-profile defections happened. In September, prominent qadis of the Kunta tribe publicly abandoned MNLA for Ansar Dine, including Hamdi bin Muhammad Lemine al-Kunti (Sahara Media 2012b), who became the leader of the council of qadis in Kidal (Bughrara 2014, 74). At this point, MNLA still commanded some military force and held on to some towns in Kidal, as well as the city of Ménaka in the province of Gao. However, on 19 November, MNLA lost Ménaka, its last piece of territory of any significance, to MUJAO after two days of heavy fighting (BBC News 2012). As Kalyvas notes, regardless of the political preferences non-combatants may have, collaboration with rebel governors tends to increase with increased territorial control and weakening of competing groups (Kalyvas 2006, 118–132) – a fact which the jihadists appear to have seized upon.

After the fall of Ménaka, the jihadists embarked on a more ambitious project to reach out to tribal chiefs, as well as to religious scholars and other local notables. These efforts are seen most clearly in their organising of various ‘courses’ (ar. dawrāt), which took the form of ideological and military training camps, coupled with community outreach. Three such courses were organised in early December 2012, one for each of the provinces of Azawad – Timbuktu, Kidal, and Gao.²⁰ As could be expected, these courses and outreach programmes were touted as major successes in jihadist propaganda, but even non-jihadist media sources report growing acceptance from local constituents in connection with these events. At the first of these courses, which was held in Aryao west of Timbuktu, one media source indicates that 32 tribal chiefs, accounting for ‘the majority, if not all’ of the tribes of the area swore allegiance (bayʿa) to Iyad ag Ghali (Sahara Media 2012c).²¹ At the course in Ménaka held five days later, several tribal chiefs representing ‘Kel-Essouk, Imachaghane, Daoussak, Ichelene-Harène and Imghad’ Tuaregs reportedly also swore allegiance to the group (Ahmed 2012b). Furthermore, the course in Aryao also saw the drafting of a nine-point memorandum by religious scholars drawn from all across Azawad, calling for the implementation of the shari‘a and the killing of anyone who supported the G5-force that would assist Mali in the upcoming offensive to drive the jihadists out of the north (Al-Hadath al-Azawadi 2012).

While these events appear on the surface to have been major victories for the jihadists, on closer inspection, neither the ‘allegiance’ sworn by the tribal chiefs nor the statements by the religious scholars amounted to an embrace of the jihadists’ agenda. The tribal chiefs at Aryao, according to a local news source, stressed their allegiance to a traditional form of Islam conducive to peace and social cohesion, and not to the ‘Islam of AQIM’ (Ahmed 2012b). Similarly, the tribal chiefs at Ménaka had rallied to Ansar Dine, reportedly on the condition that their attachment to the Islam practised by their forefathers would be respected, and that questions regarding interpretations of the shariʿa would be left to their own religious scholars (Ahmed 2012b). As for the memorandum drafted by the scholars at Aryao, the condemnation of those who cooperated with the G5-force could likely find support both among jihadists and non-jihadist Azawadi separatists. Neither does the rest of the memorandum set out strict parameters for the implementation of shari’a law or stipulate the use of controversial punishments, such as corporal punishments for the ḥudūd.²² In fact, the only infraction it explicitly bans is the use of magical amulets (Al-Hadath al-Azawadi 2012). The rest of the memorandum touches upon institutionalising religious learning in Azawad, setting stricter requirements for the collection of the zakāt, and regulating the conduct of the Islamic Police, possibly to rein in previous abuses (Al-Hadath al-Azawadi 2012).

Overall, there is little evidence that the jihadists could or would interfere in what was regarded as the internal affairs of a tribe. Abd al-Aziz Habib’s testimony mentions only one anecdote of direct interference, where the jihadists forced a wealthy tribesman to pay zakāt, to be distributed to poor people within his tribe as determined by the tribal chief (Bughrara 2014, 59). The sources say nothing of forced military recruitment of young people within tribes, and although persons with various tribal affiliations joined the jihadists on an individual basis,²³ there is no evidence of tribal chiefs actively recruiting fighters from among their own constituents. Finally, while certain marriage alliances had been made between jihadist commanders and tribal groups prior to the rebellion,²⁴ there is little to suggest this practice was regularised during the time of the Emirate.²⁵

In sum, while several tribal chiefs appear to have entered into somewhat more formalised agreements with the jihadists in 2012, this development should be interpreted first and foremost as a pragmatic move to adapt to a new political reality. As pointed out by Molenaar et al., the fact that government officials fled areas under jihadist control left tribal chiefs and qadis as the primary mediators between the local population and the jihadist rulers (Molenaar et al. 2019, 43), and they likely had to pay some form of homage to the new, self-declared overlords in order to keep representing their communities’ interests. As seen, however, the tribal chiefs retained agency vis-à-vis the jihadists and pushed back against the implementation of their ideological agenda. It is furthermore interesting to note how the increase in tribal chiefs’ outward display of cooperation with the jihadists corresponded to a weakening of MNLA’s position. As long as MNLA still commanded some fighting force, the chiefs of the Ifoghas, Kunta, and Idnane tribes could couple their own opposition to Ansar Dine with calls for their constituents to leave the group and join MNLA. However, as MNLA’s defeat at Ménaka rendered the group all but irrelevant as a fighting force, it made little sense for tribal chiefs to continue supporting MNLA as an alternative to Ansar Dine.

Assessing the tribal engagement strategy

Building on the account above, this section will make a preliminary assessment of the tribal engagement strategy employed in the Emirate. As seen in the internal discussions in AQIM, the strategic goals were, first, to avoid a tribal uprising, and second, to eventually secure the active support of the tribal chiefs, if not for jihadist ideology, then for the state project itself.

With regard to the first and most important point, the Emirate managed to avoid provoking a ṣaḥwa. While tribal chiefs, notably the leaders of the Ifoghas, Kunta, and Idnane tribes, mounted open opposition to the jihadist project, it remained non-violent. Moreover, as stipulated in the internal strategy discussions, the jihadists did not quell this opposition through violence or intimidation. Instead, they sought to ease tensions by meeting with the tribal chiefs and treating them with respect. Furthermore, the degree to which the jihadists would meddle in the affairs of tribal chiefs or impose certain jurisprudential choices on them appeared fairly limited – which might otherwise have led to resentment or violent opposition – in contrast to the relatively harsh jihadist governance imposed within the cities of Timbuktu and Gao, where traditional authorities had a weaker and less institutionalised position. This relatively lenient policy towards the tribal chiefs reflects Bin Ladin’s, and later Droukdel’s, exhortations to avoid making the same mistakes that led to the ṣaḥwa in Iraq. While other factors likely contributed to shaping this policy, for example the fact that Iyad ag Ghali originated from the area and had a better understanding of the political and tribal landscape, jihadist governance projects in other parts of the world have shown that also leaders native to a region can adopt fairly repressive policies against tribal chiefs.²⁶ Whatever their impact might have been, it is clear that Droukdel’s recommendations were received in Azawad, and (at least nominally) adopted by AQIM and Ansar Dine.²⁷

While the negative programme to avoid an uprising succeeded, the positive programme to secure the active support of the tribal chiefs was mostly a failure. Although the jihadists allegedly tried to gain support by involving tribal chiefs and other dignitaries in their project, the most ambitious projects in this regard appear not even to have been initiated. For example, Droukdel’s vision of having an ‘inclusive government’ with representatives from all the major tribal and ethnic groups did not materialise – an official government as such was not even announced. Droukdel’s plan had hinged on his – probably unrealistic – suggestion for sharing power with MNLA and other non-jihadist political actors. As Iyad ag Ghali and other jihadist leaders in the Emirate may have recognised, the interests of the different parties were too divergent to be brought under the same umbrella.

Droukdel’s advice to ‘unite the tribes around simple goals’, such as ‘fighting injustice’ and the oppression of the Malian state, seems not to have been taken into consideration. Instead of speaking in these general terms, Iyad ag Ghali attempted from the outset to garner support for a particular and explicitly salafi-jihadist programme. It should be noted here that although Droukdel’s programme showed an awareness of the difficulties of gaining tribal support and advocated for a flexible, pragmatic approach, it displayed a glaring lack of familiarity with the tribal landscape of northern Mali. Descriptions or discussions of individual tribes and their members, their histories, political interests, and so forth, were absent. As seen, support for an independent Azawad was far from uniform in northern Mali, even among the Tuareg. As in the previous rebellions, the armed groups, whether jihadist or not, were not only political entities but also served as conduits for inter-tribal rivalries. Droukdel’s treatment of the ‘tribes’ as a unitary block with common interests underlies the naïveté of wanting to include ‘all the major tribes’ in government and hoping to unite them all around a common goal.

As shown, the one tribal constituent in which the jihadists managed to attract significant support was the Ifoghas, and the inclusion of Ifoghas nobles in Ansar Dine is the case that conforms most readily to Droukdel’s vision of relying on sympathetic locals to govern the budding state. However, this Kidal-based faction of Ansar Dine appeared less committed to jihadist ideology than its counterpart in Timbuktu was, and the limits of their support for the jihadist project were made plain after the French intervention in northern Mali. In January 2013, the jihadists made a sudden offensive against the city of Konna in the Mopti Region of central Mali.²⁸ Its rapid fall prompted fears that the jihadists would continue their march all the way to Bamako, and just a day later, France launched a military intervention at the request of the Malian government to beat back the jihadists and put an end to their control of the north (BBC News 2013).

Soon afterwards, senior Ifoghas Ansar Dine members left the group to form a new group, MIA (Mouvement islamique de l’Azawad) under the leadership of Alghabass Ag Intallah (Thurston 2020, 139–140). Thus, the jihadists lost their most prominent tribal constituent almost immediately. The rapidity with which all the territory in the north fell to the French intervention suggests that the tribal chiefs who had nominally sworn allegiance to Iyad ag Ghali followed suit.²⁹ They clearly had little interest in fighting for the losing side, and for an ideology they did not profess.

Conclusion

The jihadist groups that conquered northern Mali in 2012 had to make a quick transition from roaming rebels to responsible governors of a territory the size of France. They were faced with the difficult task of navigating a complex landscape of independent-minded tribes whom both the French colonial administrators and the post-colonial Malian state had struggled to keep in check. Whereas AQI’s antagonisation of the Sunni tribes in Iraq had led to its downfall, the al-Qaida-affiliated groups ruling the Islamic Emirate of Azawad succeeded in avoiding a tribal uprising in their newly conquered territories. Rather than using force to intimidate tribal chiefs into compliance, the leaders of the Emirate took a pragmatic approach, tolerating some dissent and respecting traditional authorities.

While the negative programme – not to provoke armed resistance – succeeded, Droukdel’s proposed positive programme to gain support among the tribal chiefs did not. Unable or unwilling to turn the Emirate into a (proto-)state with a national rather than purely salafi-jihadist raison-d’être, the jihadists could at most gain acquiescence. Consistent with previous findings in the rebel governance literature (Arjona 2016; Kalyvas 2006), collaboration with the jihadist incumbents increased as their control over the territory increased, but the relatively strong position of certain tribal chiefs gave them bargaining power vis-à-vis the jihadists trying to impose a new political reality.

Finally, while the fact that the Emirate was brought down by a foreign intervention rather than internal strife could partly be seen as a success for the tribal engagement strategy employed, the fact that the tribal chiefs abandoned their reported allegiance to Iyad ag Ghali in favour of collaborating with an ‘infidel’ intervention force shows that their involvement with the jihadist rebels was first and foremost pragmatic. In the end, the tribal chiefs looked out for their own interests and for those of their constituents, not for those of the jihadists.

Acknowledgements

The author would like to thank Brynjar Lia, Michael Skjelderup, and his other colleagues in the Rebel Governance in the Middle East project for very valuable feedback on earlier drafts of this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was funded by the Norwegian Ministry of Foreign Affairs (Utenriksdepartementet)

Notes on contributors

Vidar B. Skretting

Vidar B. Skretting is a PhD candidate at FFI (Norwegian Defence Research Establishment) and the University of Oslo.

Notes

1. See for instance Mampilly (2011), Arjona, Kasfir, and Mampilly (2015).

2. For instance Lia (2015), Lia (2017), Aarseth (2018), Cook, Haid, and Trauthig (2020).

3. Recent examples include Collombier and Roy (2018); Skjelderup (2020); Whiteside and Elallame (2020).

4. See Brynjar Lia’s article in this collection (2022, this issue).

5. See Michael Weddegjerde Skjelderup’s article in this collection (2022, this issue).

6. Arjona finds that stronger and more legitimate pre-existing institutions allow for collective civilian actions against unpopular measures taken by rebel governors, resulting in what she terms an aliocracy, as opposed to a rebelocracy, where rebels are able to intervene broadly in civilian affairs. See Arjona (2016), 3.

7. The letters from Usama bin Ladin’s Abbottabad compound were captured in the US raid that killed the al-Qaida leader in 2011. These materials were released in their entirety on CIA’s website in November 2017. The letter from Droukdel to the Shura Council of AQIM in the Sahara was found by journalist Rukmini Callimachi in 2013 and published online by Associated Press with an English translation (Callimachi 2013), but this version is incomplete and lacks several pages. This article will refer to the original Arabic version, which was published online with an introduction by the al-Qaida-affiliated media group Nukhbat al-iʿlām al-islāmī in 2014 (Droukdel 2014). The interview with Abd Al-Aziz Habib was published in a six-part series on www.marsadpress.net in 2013–2014, and in book-form later in 2014 (Bughrara 2014).

8. This refers to the ‘minimal’ definition of a tribe employed in Collombier and Roy (2018), 5.

9. The finer distinctions in this system, as well as what the terms mean in practice, are subject to debate. See Lecocq (2010), 5–6.

10. See Molenaar et al. (2019), chapter 2 for a more detailed history of these developments.

11. For example, an individual such as Iyad ag Ghali can be said to belong to the Iriyaken clan within the Ifoghas tribe, which again forms part of the Kel Adagh federation.

12. While the imghad were traditionally vassals of a particular tribe, a unitary imghad identity has begun to emerge since the 1970s that cuts across tribe and clan divisions. See Thurston (2018), 23.

13. Notably, the town of Douentza in Mopti, which was held by the Ganda Izo militia until it was captured by the jihadists on 1 September, and Ménaka in the Gao Region, held by MNLA until 19 November. See Ahmed (2012a); BBC News (2012).

14. These ministries would include the army, the media, the judiciary, and the ministry of daʿwa, Islamic affairs, and education. See Droukdel (2014), 13.

15. It should be noted that these overtures by Ag Ghali did not represent a departure from Droukdel’s strategy, since they were made before Droukdel drafted the above-mentioned tribal engagement strategy. If anything, the drafting of the strategy may have been prompted by Iyad ag Ghali’s failure in Kidal.

16. Muslim religious judge.

17. Notably, the stoning of an unmarried couple in Aguelhok in the Kidal Region. See Al-Jazeera (2012).

18. For example, on the relative importance of traditional authorities in Gao in this period, see Raineri and Strazzari (2015), p. 262–263.

19. See ’Decision on the Confirmation of Charges Against Ahmad Al Faqi Al Mahdi’ (2016), 23 and ’Version Publique Expurgée: Rectificatif à La Décision Relative à La Confirmation Des Charges Portées Contre Al Hassan Ag Abdoul Aziz Ag Mohamed Ag Mahmoud’ (2019), 200.

20. Abd al-Aziz Habib also mentions two courses taking place in August and October. See Bughrara (2014), 88–89.

21. Judicial documents in the International Criminal Court case against Ansar Dine Ahmad al-Faqi al-Mahdi provide independent confirmation of this episode (’Version publique expurgée du « Document présentant les conclusions factuelles et juridiques du Bureau du Procureur au soutien du Chef d’accusation dans l’affaire contre Ahmad AL FAQI AL MAHDI »’ 2015, 16).

22. ḥudūd (sg. ḥadd) refers to a punishment fixed in the Quran and hadith for crimes considered to be against the rights of God, such as theft, adultery, and apostasy. See Esposito (2003).

23. Reports on the tribal backgrounds of individual members of the jihadist groups in Bughrara (2014), while not exhaustive, show that several Tuareg and Arab tribes were represented.

24. AQIM-commanders Mokhtar Belmokhtar and Nabil Makhlufi married into the Arab Bérabiche and Kunta tribes, respectively. See Salem (2014), 56–59.

25. There were, however, numerous examples of forced marriages to jihadist cadres in Timbuktu and Gao during the occupation, but there does not seem to have been a component of tribal politics as such in these cases. See ‘Version Publique Expurgée: Rectificatif à La Décision Relative à La Confirmation Des Charges Portées Contre Al Hassan Ag Abdoul Aziz Ag Mohamed Ag Mahmoud’ (2019), ‘Mali Sexual Abuse Survivors Seek Justice – Mali’ (2021).

26. See for example the cases of ISIS and al-Shabab, in Brynjar Lia (2022, this issue) and Michael Weddegjerde Skjelderup’s (2022, this issue) articles in this collection.

27. Droukdel’s programme, which also includes recommendations on other matters related to the governance of the Emirate, was sent to the commanders of the Emirate some time after it had been established. Other internal documents describe that the recommendations were received and adopted by the Shura Councils of AQIM in the Sahara and Ansar Dine (‘Al-Qaida Papers: Internal Correspondence Between the Shura Council of AQIM and Mokhtar Belmokhtar’ 2012, 11). This happened no later than a few weeks after they were drafted, given that Timbuktu governor Abu Zayd published new codes of conduct for the Islamic Police in that city, clearly influenced by Droukdel’s recommendations, already on 15 August 2012 (‘Version Publique Expurgée: Rectificatif à La Décision Relative à La Confirmation Des Charges Portées Contre Al Hassan Ag Abdoul Aziz Ag Mohamed Ag Mahmoud’ 2019, 62–63).

28. As argued by Skretting (2022), the motivation for this offensive was in all likelihood not the conquest of Mali, but a pre-emptive strike to secure strategic ground before an expected Malian and African Union-led intervention force that was scheduled to be launched.

29. The intervention once again opened up the space for contestations within Tuareg society in the north, and as usual, these contestations had a tribal dimension to them. The main delineation came to be between the pro-separatist CMA (Coordination des mouvements de l’Azawad), a coalition of the Ifoghas-led HCUA (Haut conseil pour l’unité de l’Azawad, the successor to MIA), MNLA and MAA (Mouvement arabe de l’Azawad), and the pro-government Plateform, led by the aforementioned imghad commander El-Hajj Gamou and his group GATIA (Groupe autodéfense touareg Imghad et alliés) (Thurston 2018, 27–28). Ansar Dine is today a constituent group of the al-Qaida faction known as JNIM (Jamāʿat nuṣrat al-Islām wa-l-muslimīn). Although not holding territory in an official capacity as it did in 2012–2013, the group has regained significant influence in Kidal. For example, Ansar Dine reportedly levies tax on gold produced in Kidali gold mines (Getting a Grip on Central Sahel's Gold Rush 2019, 7).