Defining the language of rebellion

Before considering insights from the literature and building the typology, it is important to clarify key terms. Throughout the paper, we use ‘transition,’ ‘succession,’ and ‘leadership change’ interchangeably to refer to the process of one leader replacing another. These terms are neutral and do not embed any presumptions about the reasons driving this process. We also use ‘insurgent,’ ‘insurgencies,’ rebels,’ and ‘non-state armed actors’ without meaningful distinction; like Mampilly, we use these terms to refer to armed factions that use violence to challenge the state in pursuit of political goals.Footnote ² We do, however, distinguish between group/organisation, which we treat as synonyms, and movement.Footnote ³ Although some authors to draw a clear line between groups or organisations and movements, others use them interchangeably; there is no consensus on the difference across much of the existing literature, and the terms are often undefined.Footnote ⁴

These terminological boundaries are often blurred. For instance, some actors, such as Ação Libertadora Nacional (ALN) in Brazil, employ organisational language – particularly of a military nature – even when the structures they supposedly reference exist only on paper. Other examples, such as the Sudan People’s Liberation Movement/Army (SPLM/Army) or the Provisional Irish Republican Army (PIRA) and Sinn Fein, typically refer to their political wing as a movement and their military wing as an army, even when there is significant overlap in personnel, as was the case with PIRA and Sinn Fein. However, sections of the literature on non-state armed actors have distinguished between groups and movements, with the latter referring to a much broader category of organisations involved in campaigns of rebellion, who may coalesce or fragment in various ways.Footnote ⁵ The leadership of coalitions and alliances is a more complex issue, and although many of the factors discussed here remain relevant, focusing on relationships transcending groups falls outside this paper’s scope. To avoid confusion, we observe these terminological boundaries here.

Leadership and transition in non-institutionalised groups

A further distinction that requires recognition is the degree of institutionalisation within groups. There can be considerable variation in how groups operate, and they can be organised along a spectrum of bureaucratisation and institutionalisation. At one end, groups can institutionalise to such a degree that they closely resemble the states they challenge.Footnote ⁶ This helps them coordinate beyond personal networks, efficiently manage resources, provide collective goods, and demonstrate the viability of their projects.Footnote ⁷ In such circumstances, leadership – and consequently transition – can be shaped by the institutional mechanisms and procedures that have been established to regulate it. In Weberian terms, rebel leaders can become bureaucratic figureheads whose authority predominantly stems from occupying an institutional position endowed with legal power.Footnote ⁸ We focus our attention on the other end of the spectrum.

Some insurgent organisations make only limited use of formal institutions, procedures, and practices. In such circumstances, rebel leaders are more akin to Weber’s patrimonial or charismatic types, which conceive of leadership in terms of social relationships.Footnote ⁹ Relations within groups are often socially rooted, with blurred hierarchies within groups and blurred boundaries between groups and surrounding communities.Footnote ¹⁰ Without robust institutional mechanisms, leaders represent a group and receive ‘status, prestige, and power’ to further collective goals.Footnote ¹¹ We can understand these relationships by recognising their negotiated nature. Followers do not unquestionably follow their superiors’ demands but instead have their own interests and influence leadership and group processes through the exercise of agency.Footnote ¹² Since leaders lack strong institutional mechanisms to compel compliance, the exercise of leadership becomes one of continual negotiation as different parties seek outcomes favouring their interests. When either party feels their interests are no longer being met, they may terminate the relationship.Footnote ¹³ Relationships are inherently unstable and highly dynamic.Footnote ¹⁴ The use of terms such as ‘organisation’ and ‘group’ should not, therefore, be read as presupposing rigidly defined or static boundaries.Footnote ¹⁵ Leadership transition warrants special attention as a period when these relations are opened up for re-evaluation and negotiations are particularly intense.Footnote ¹⁶ As Schlichte notes, ‘the interdependencies that constitute them [armed groups] change when any of the other relations is changing, for example when new resources become available, when somebody involved in a figuration loses a capacity, or when new agents enter a figuration’.Footnote ¹⁷ The transition process is complete when a critical mass of group participants accepts the terms of the new relationship and thus the claim to leadership.Footnote ¹⁸

Examining organisations at the less bureaucratised end of the spectrum distinguishes this paper from much of the growing literature on rebel governance, which has tended to focus on larger-scale groups capable of contesting and exercising territorial control. This literature has made significant contributions to understanding how rebel groups operate, challenging the prevailing wisdom that war zones are de facto chaotic and the state’s absence or limited presence automatically translates into anarchy.Footnote ¹⁹ Yet it largely neglects the many groups around the world at the lower end of the spectrum of institutionalisation, which are not yet or are no longer capable of contesting territorial control – and the degree of institutionalisation can significantly impact the nature of coordination between actors.Footnote ²⁰ These groups are often labelled ‘terrorist’ or something similar, a label we reject for much the same reason as Mampilly eschews the term ‘guerrilla’ – namely, that it refers to only one of the tactics they use.Footnote ²¹

At the same time, this paper fits within the broader organisational shift in the study of civil wars that this work on rebel governance reflects. Parkinson demonstrates how militant organisations are far from closed systems; instead, individuals’ roles within them are shaped by their position within social networks.Footnote ²² Staniland similarly emphasises the importance of social networks to the subsequent organisation of insurgency.Footnote ²³ In addition to governance writ large, authors have examined questions of rebel diplomacy, alliance formation, legitimisation, and financing.Footnote ²⁴ In their review of this emerging subfield, Parkinson and Zaks draw on organisational sociology to show how organisations can be conceived of as a ‘collection of roles, linked by relations, which produce behaviours, to work towards goals within a given context’.Footnote ²⁵ We contribute to this work by offering a rich, contextualised examination of how these four dimensions – roles, relations, behaviours, and goals – interact during the transition process, focusing on the hitherto-neglected less institutionalised end of the spectrum of organisations.

Transition: A neglected topic

Having established definitional and conceptual parameters for the study, we can now turn to the literature to identify cross-disciplinary insights into leadership and use these to sketch an initial typology of transition tasks that new leaders must tackle. Given the recognised importance of leadership in general, what is remarkable is how limited theoretical development specific to transition has been to date. Political scientists and International Relations scholars have examined how transition in state leadership affects foreign-policy preferences and levels of domestic repression and whether leadership change can be predicted.Footnote ²⁶ These studies examine highly institutionalised and overt actors, where rules are defined, and much of the transition process is externally accessible – offering limited insights into the type of groups under discussion here.

Of more obvious relevance, the issue of forced transition, i.e. leadership removal or ‘decapitation,’ has received substantial attention, mainly from within terrorism studies. Here, findings have been decidedly mixed. David and Hafez and Hatfield argue that targeted assassinations have a minimal effect on overall levels of insurgent violence.Footnote ²⁷ By contrast, Kaplan, Mintz, Mishal, and Samban contend that assassinations increased recruitment to organisations and led to an increase in attacks.Footnote ²⁸ Authors have sought to explain apparent contradictions in impact through reference to a variety of factors: the nature of the leader’s removal (with arrest more disruptive than killing); the level of bureaucratisation and institutionalisation within a group; whether a group is hierarchical or not; the availability of a clear successor; the ideology of the group; and the nature of the support enjoyed by the leader.Footnote ²⁹ This literature, however, has focused predominantly on the degree to which leadership removal impacts operational effectiveness and levels of violence rather than on internal processes of how insurgent groups navigate the transition process. Wu, Carleton, and Davies offer a rare attempt to unpack this process, but they do so by anticipating the outcomes of a single case, AQ, after bin Laden’s death.Footnote ³⁰ Bakke, Cunningham, and Seymour and Mosinger propose leadership decapitation as one of the factors affecting the balance of power within organisations and movements – alongside changing internal demographics, ideology, strategic disagreements, disciplinary mechanisms, rivalries, broader counter-insurgent strategies, battlefield losses, patron preferences, and negotiations – but neither article explores this in depth.Footnote ³¹ Within much of the literature on terrorism, civil wars, and social movements, discussions of transition are buried in empirical research and occur on the margins of studies focused elsewhere. Much richer material can therefore be found by looking beyond transition, specifically to cross-disciplinary research into leadership in general – and it is from this work that we can begin to identify the tasks and roles that leaders must perform.

Key transition tasks: Establishing leadership

First, and perhaps most obviously, we see that leaders must claim and hold on to power. Leadership, after all, is an integral albeit implicit component of numerous theories offered by scholars to explain insurgent behaviour and operation. Gurr, for example, argues that ‘leadership is essential if political violence is to succeed in the face of organised resistance’, while Schlichte characterises it as a ‘functional necessity’ when organisations reach a certain size or need to exercise territorial control.Footnote ³² Staniland and Lyons see leadership as shaping internal group dynamics, which, in turn, affects their cohesiveness, interactions with external actors, and ability to achieve desired results.Footnote ³³ Nor is recognition of the importance of leadership limited to violent groups: empirical studies have examined movements ranging from the transnational jihadism of AQ to rural activism in Nicaragua.Footnote ³⁴ Christia contends that, at the most fundamental level, leaders’ behaviour is shaped by the primary aim of holding on to power in evolving political landscapes.Footnote ³⁵

To achieve this goal, leaders need to neutralise any challenges posed by potential rivals, ensuring obedience and punishing defection.Footnote ³⁶ Rebel actors often fight the existing authorities and each other, and ‘horizontal fragmentation makes the leadership of armed groups prone to factional infighting and coups’.Footnote ³⁷ Tamm sees internal coups originating from changes in the distribution of resources between leaders and their rivals, which affects the balance of power.Footnote ³⁸ As a consequence, it can be argued that new leaders need to ensure a preferential distribution of resources in their favour. Fears of a coup can lead to paranoid behaviours and internecine strife that prioritise individual security over organisational continuity, resulting in a divergence between individual and group interests. Sendero Luminoso’s Abimael Gúzman, for example, surrounded himself with female deputies who were neither prepared for nor seen as capable of replacing him.Footnote ³⁹ Since the transition period is one where new leaders’ power has, by definition, not yet been consolidated, the likelihood of challenge emerging at this time is potentially higher than it is for established leaders – precisely the reason that scholars such as Mosinger list leadership decapitation among their factors influencing the balance of power within organisations.Footnote ⁴⁰ Inhibited information flows may only exacerbate the problem: in clandestine groups with limited institutionalisation, the absence of a clear successor can create a power vacuum, encouraging the emergence of impostors and rival claimants to power.

Key transition tasks: Maintaining organisations and networks

Existing literature also draws our attention to the organisational dimensions of rebellion, including the role of networks. Gurr claims that creating institutional infrastructure is necessary for sustaining political violence – and even smaller and weakly institutionalised groups require some form of organisation and internal processes to regulate activity and survive.Footnote ⁴¹ Staniland argues that the networks around insurgent leaders form the basis of institutionalisation, providing ‘information, trust, and shared political meanings that organizers can use to create new armed groups’.Footnote ⁴² Schlichte contends that there are two pathways to institutionalising power: creating bureaucratic structures and formalising relationships, or creating patrimonial structures that build on client networks.Footnote ⁴³ Both horizontal and vertical relationships impact the trajectories of resistance.Footnote ⁴⁴ In considering transition, this compels us to assess the impact of the disruption to networks caused by the removal of leaders. As already noted, relationships within armed groups are highly dynamic, and a change in leadership can lead to their reconfiguration.Footnote ⁴⁵ For instance, the departing leader may be a crucial node in existing networks, and individuals’ structural positions may influence the choice of successor.Footnote ⁴⁶ Alternatively, removing a leader could break the link between different parts of the network, with the circumstances surrounding that removal potentially exacerbating its negative impact. For example, when Turkish Hizbollah leader Huseyin Velioğlu was killed in a special operation in March 1999, over 600,000 pages of internal documents were found that provided critical intelligence leading to the organisation’s demise.Footnote ⁴⁷ Individuals who join pre-existing insurgencies may not be able to draw on previous networks in the same way as those present at their creation. They may also belong to different networks within the organisation or affect the balance of power when they join.Footnote ⁴⁸ Overall, new leaders face the challenge of reconstructing and reconfiguring the networks that form the organisational basis of rebellion.

Additional transition tasks

These two tasks – claiming the leadership and renegotiating relationships – prompt questions about how leaders meet objectives, which, in turn, leads to the identification of further tasks. An important topic of debate around leadership concerns sources of authority. In political science, there have been efforts to rehabilitate the ‘Great Men’ theory of leadership, drawing attention to the importance of leaders’ personalities, qualities, and abilities in shaping events – even if there is now greater recognition of the importance of context.Footnote ⁴⁹ In the literature on terrorism and criminal groups, we also find repeated references to leaders’ individual qualities and characteristics.Footnote ⁵⁰ In such discussions, the elusive concept of charismatic leadership is never far away. Often overlapping with this is a recognition that leaders can play an important role in articulating an ideology and providing the ‘doctrinal justification for violence’.Footnote ⁵¹ Crenshaw insists that leaders may possess ‘more complex and differentiated belief structures than followers’ and may derive their authority either from their ability to articulate those beliefs or the relevance of their background to the belief system.Footnote ⁵² She also suggests that second-generation leaders may be less ideological than organisation founders, pointing to a diminishing role for ideology. Price contends that ‘values-based’ organisations enjoy a recruitment advantage over ‘profit-based’ ones, while Walter argues that organisations that adhere to an extreme ideology enjoy an advantage over moderate ones.Footnote ⁵³ Of course, the role of ideology in political violence is hotly contested.Footnote ⁵⁴ Yet even when we do not frame the discussion in explicitly ideological terms, there is widespread recognition that leaders are responsible for establishing expectations and demonstrating a commitment to a cause, generating positive social identities and neutralising threats to them, and defining and prioritising objectives.Footnote ⁵⁵

Translating these diverse observations into concrete tasks shows that leaders play an important role in inspiring followers and providing direction to organisations. Indeed, Brenner sees satisfaction of the social identity needs of followers as a critical source of insurgent legitimacy and leadership authority.Footnote ⁵⁶ All leaders must perform these tasks, but they become particularly pressing for new ones.

Weinstein identifies further essential duty of leaders: responding to shocks.Footnote ⁵⁷ The loss of a previous leader can be one of the critical shocks to which new leaders must respond, with a failure to respond adequately risking demotivating followers and engendering a sense that the organisation has lost direction.

Leadership beyond transition

This discussion cannot claim to be a comprehensive account of all dimensions of rebel leadership. Indeed, discussions of the topic often lay scattered across work focused on other questions, making any claim to comprehensiveness questionable. There are also tasks that leaders must perform in the normal course of their duties that are not an obvious priority in the transition period. For example, in his discussion of leadership tasks, Melucci includes the acquisition and distribution of resources.Footnote ⁵⁸ While this is undoubtedly important for all groups, it is not immediately apparent that it should be particularly pressing during transition. If anything, changes in resource availability are more likely to precede and even cause the transition process by disrupting the balance of power within groups – although the importance that should be attributed to resources in leadership authority is contested.Footnote ⁵⁹ However, the question of what causes leadership change is different to that guiding this paper, and one can convincingly argue that other tasks assume much greater urgency during the transition process itself. This brief review of various literatures has enabled us to identify the tasks that appear particularly important during leadership transition within the broader life cycle of groups. We can therefore use this as the starting point for our typology, but we must first turn to our chosen cases and method to consolidate and refine the typology.

Introducing the cases

The insurgency in Russia’s North Caucasus traces its origins to the Chechen wars of 1994–6 and 1999–2002. Following the Soviet Union’s collapse, tensions between a re-emerging Russia and an independence-seeking Chechen Republic of Ichkeria (Chechenskaya Respublika Ichkeria, ChRI) led to war. The ChRI secured a largely pyrrhic victory but faced extensive social, economic, and political problems, including a proliferation of paramilitary groups and a failure to settle Chechnya’s legal status.Footnote ⁶⁰ The interwar period was marked by widespread instability and criminality, and Russia, under the leadership of Vladimir Putin, ultimately renewed the war.Footnote ⁶¹ Over time, in a campaign marked by considerable brutality, Russia drove the rebels from the cities and installed a pro-Kremlin administration in Chechnya.Footnote ⁶²

Meanwhile, the conflict increasingly spread to neighbouring republics due to a deliberate rebel strategy and the preferences of non-Chechen actors. An ideological shift accompanied this process of regionalisation, as a radical faction challenged for, and ultimately assumed, complete control over the insurgency. These dual processes culminated in the decision, in October 2007, to abolish the ChRI and replace it with an explicitly jihadist yet largely notional polity, the Caucasus Emirate (Imarat Kavkaz, IK). It was a watershed moment in the evolution of conflict, which formalised the victory of the insurgency’s Islamist wing over nationalist separatists in its historic Chechen core and changed the rationale for armed resistance. Under the banner of the IK, the locus of the conflict shifted first to Kabardino-Balkaria and then ultimately to Dagestan, and low-level insurgency continued to plague the region. However, by late 2016, the twin pressures of sustained security-service operations and the draw of the conflict in Syria and Iraq had led to the collapse of the insurgency.

The North Caucasus insurgency, and the IK in particular, provide an excellent opportunity for examining leadership transition. Long-serving ChRI president Aslan Maskhadov’s ability to resist the Islamist wing was undermined by the systematic elimination of his allies. Moreover, he enjoyed a legitimacy that his successors could not claim, and his death arguably signalled the end of the ChRI, clearing the way for the insurgency’s ideological transformation.Footnote ⁶³ The deaths of IK leaders occasioned further ideological shifts – albeit within the framework of jihadism – and hampered the insurgency’s ability to respond to emerging domestic and institutional challenges.Footnote ⁶⁴ As such, the IK provides an excellent example of an insurgent group that was not highly institutionalised and could no longer contest territorial control but posed a serious, sustained, and violent challenge to state authority.Footnote ⁶⁵ For much of its existence until IS penetrated the region in 2014, the IK enjoyed an organisational monopoly across the North Caucasus. Still, it always functioned within a broader jihadist movement that comprised North Caucasian jihadism as the inner circle and the global jihadist movement as the outer one. In other words, although the IK was the only recognisable organisation in the region, jihadism in and linked to the North Caucasus was never reducible solely to the IK.

Formally, the group was organised on strictly hierarchical grounds. The basic organisational unit was the jama’at, or city or neighbourhood cell. Multiple jama’ats coalesced to form a sector, and sectors were arranged into vilayyats, or provinces, that largely mirrored the republics of the North Caucasus. At the apex of the organisation stood a leadership that united the different vilayyats under a central command. Each level of the hierarchy was led by an emir who reported to and co-constituted the leadership of the level above them; thus, the vilayyat-level emirs collectively formed a leadership shura, or council, that was responsible for decision-making under the authority of the central IK leader. Leadership was, therefore, a mixture of the collective and the individual: although the emir officially enjoyed final decision-making authority, he was supported by some or all of a naib (deputy), qadi (Shari’ah judge), military emir, and the other members of the shura, as well as occasional individuals with no formal rank.

Lower-ranked members of the leadership enjoyed considerable latitude in their decision-making and operations, especially because of communication and coordination difficulties experienced in a highly hostile operating environment. To a degree, the constituent vilayyats of the IK functioned as distinct and mostly autonomous groups within a broader structure – which facilitates the generalisability of some of the conclusions drawn to the question of leadership transition in insurgencies – but they nevertheless accepted the central IK emir’s overarching claim of authority and strategic direction. In other words, the IK displayed some but not all of the characteristics of a movement, as defined in parts of the existing literature on insurgencies, and it often used the language of a movement, but definitionally it remained closer to an organisation.

On paper, this organisational structure meant that transition was a relatively simple affair. If no formal successor had been designated at the top of the organisation, the central shura decided on a replacement while, at all other levels, the emirs from the level above issued an order appointing the new leader. Groups confirmed acceptance of a leadership claim through a pledge of bayah, or loyalty, to the new leader. As seen in the third section, the reality was considerably different, illustrating how group participants could exercise agency to define the terms of a new relationship. Negotiations among shura members could be intense, and local groups sometimes appointed their own leaders, even in defiance of the preferences of their superiors.

Moreover, communication between the different tiers was often difficult or impossible; notional seniors often lacked means of compulsion, making them reliant on the consent of subordinates, and cross-vilayyat cooperation was limited. Some scholars have gone so far as to question the very existence of ‘a unifying actor, the Emirate, which could act to coordinate actions and activities across the region’.Footnote ⁶⁶ While concerns about the command-and-control capabilities of the leadership are valid, this tends to overlook the significance of a broad acceptance of an overarching ideology and claim of authority and the way the existence of the IK shaped perceptions and, consequently, behaviours. Nevertheless, there was undoubtedly a large gap between the highly bureaucratised and militarised structure on paper and the exercise of leadership in practice. The IK also faced a high-capacity opponent that applied considerable counter-insurgency pressure, meaning leadership transition was a regular occurrence within the group. The IK thus provides ample opportunity to examine how relationships are renegotiated and how actors exercise agency outside of institutional channels during the transition process.

Analysing leadership transition in practice

A case-study approach allows for the intensive study of chosen cases, creating opportunities for insights available through other methods.Footnote ⁶⁷ Our cases are four sequences of leadership transition in the IK era – one at the apex of the group, the other three from within the IK’s vilayyats, as shown in Table 1.

Table 1. Leadership sequences in the IK. Years refer to occupation of the highest leadership position in their groups. The first-named leader is considered only as a predecessor, in order to establish the context of transition.

To assess transition, we collected data from various sources. Throughout its lifespan, the IK maintained a relatively stable online presence of ‘official’ and semi-official websites representing the central organisation and its regional branches.Footnote ⁶⁸ Through a combination of daily monitoring of these resources over many years and targeted supplementary searches of video-sharing websites or online repositories, we compiled a unique corpus of leadership communiqués and other insurgency-produced material. For each leader, we concentrated on communiqués issued in the last year of a predecessor’s tenure and the first year of the new leader, enabling us to establish the context for transition and examine the critical period of flux and uncertainty while still being sufficiently narrow to focus on transition itself. Some leaders took time to announce their succession, while others did not survive this period. In total, we drew on 89 leadership statements in this analysis. We supplemented this with additional information from the corpus, reporting by a broad range of local and federal media outlets during the time period, official and security-service statements, and court proceedings – much of it, again, obtained through routine monitoring of the insurgency that helps build a richer picture than would have been possible through post hoc searches.

We analysed this material using NVivo 11, a tool designed to aid rigorous qualitative analysis. Social movement studies’ theory of framing and the methods it has given rise to influenced our approach.Footnote ⁶⁹ We used the findings from the literature to generate a set of codes covering essential tasks. We then applied those codes to a sample of the material and adjusted the coding schema accordingly. This led to the merger of some codes, where the practical performance of tasks overlapped considerably and made meaningful disaggregation impossible, and the elimination of others, where performance of a task occurred only in an isolated case. Initially, we developed a series of sub-codes to reflect differences within tasks; however, we decided that this level of granularity did not offer clear analytic advantages, whereas a minimalist typology focuses attention on what is critical during the transition process. One challenge lay with how we treated counter-insurgent activities, which posed obstacles to the successful navigation of the transition process and established the context in which insurgent leaders performed their tasks. Initially, we coded these separately, but once we narrowed it down to those that were consistent across groups and could thus be considered ‘key’ challenges facing insurgent leaders, we found that these closely mirrored our already-identified tasks. Consequently, we merged the discussion of these measures into the discussion of the core tasks, helping maintain the focus on the internal processes of movements. Only one task – that of counter-insurgents imposing cumulative losses on groups – was not adequately reflected in the existing codes; consequently, we created the new task of survival to reflect this particular issue.

A non-sequential typology of tasks facing new insurgent leaders

The outcome of this iterative process of refining a literature-derived set of leadership tasks and empirical application to our dataset was a typology of five key, non-sequential tasks, as illustrated in Table 2.

Table 2. A typology of tasks that new insurgent leaders must perform.

Recognising and addressing methodological limitations

Before considering the empirical application of the typology, it is necessary to recognise its limitations and pre-empt some of the criticisms that may be levelled at it because our typology attempts to address some of the issues that have plagued discussions of insurgent leadership. Staniland, for example, recognises charisma and instances where leadership changes have impacted groups in his empirical discussion, but neither individuals’ qualities nor issues arising from their replacement influence the theoretical debate.Footnote ⁷⁰ Rational-action models are commonplace but afford limited space to individual characteristics.Footnote ⁷¹ These widely cited approaches tend to downplay agency and fail to explain the impact of leadership change where institutional structures or networks are unaffected. Capturing these factors, however, is methodologically challenging, and the search for specific traits can quickly become circular: leaders can be considered successful because they are charismatic and charismatic because they are successful, while charisma can be seen as inherent to the individual and conferred by others.Footnote ⁷² That which can be conferred can also be denied, reinforcing the importance of followers’ agency. Elsewhere, the search for qualities is often biased towards Western perspectives on leadership, and ‘Great Men’ theories can be as gender-biased as the label suggests. Some leaders may not have discernible characteristics or formal organisational positions but instead be ‘someone who is perceived as a leader’ who draws their authority from networks and relationships.Footnote ⁷³

Meanwhile, there is a widespread tendency to concentrate on leadership par excellence, cherry-picking those leaders who are regarded as pioneers or exceptionally charismatic, competent, or successful. This prejudices the debate by removing from consideration ‘mediocre’ or ‘bad’ leaders – who may be just as relevant to considerations of transition as ‘good’ ones. By focusing exclusively on the characteristics of successful leaders, we may ignore the many who possess the same qualities who do not become leaders and those who become leaders despite not possessing them and fail to explain why some leaders cannot transfer their authority from one environment to another. Examining processes allows us to move beyond the limitations of characteristic-focused explanations and approach leadership as a highly contextualised phenomenon.

Nevertheless, we face several challenges when evaluating transition in clandestine groups. First and foremost, data availability is critical. We have focused on our chosen sequences because daily monitoring allowed us to gather exceptionally rich data; consequently, we possess a high degree of confidence in the comprehensiveness of that data. Nevertheless, the information environment was highly constrained, and for many aspects of insurgent life in the North Caucasus, the only sources of information were the insurgency or the security services. Only a handful of media outlets conduct original reporting, and these outlets face significant obstacles. Researchers often have limited access to group participants for practical, ethical, and legal reasons, making it difficult to ascertain rank-and-file views in particular. When disputes emerge over leadership, the parties may have unequal access to channels to publicise their claims. Equally, leaders may choose not to communicate extensively, performing their tasks away from the public eye. Some leaders in our sequences were prolific; others maintained a low profile. We see similar practices elsewhere. Early Taliban leaders, for example, were relatively ‘invisible’, heading an ‘anti-charismatic movement’;Footnote ⁷⁴ later factions ‘found themselves hiding, without lines of communication to one another’.Footnote ⁷⁵ Where there are no alternative sources of information, or the interests of the parties align in keeping something from public view, important parts of the transition process may be invisible to the researcher.

Using insurgency-produced material as a primary, albeit not exclusive, data source carries obvious risks: they are hardly neutral observers. In some cases, the information in communiqués reflects actual processes; in others, there will be efforts to manage internal and external perceptions. However, researchers cannot neatly disentangle these aspects of communiqués without additional sources. Leaders may also ‘communicate’ in other ways, including through the use of violence. However, we defend the use of this material on three grounds. First, any study needs to be manageable and bounded. Including insurgent behaviour as a form of communication risks diluting the focus on transition and raises critical challenges about attribution. We do not deny the importance of behavioural aspects of leadership, but we exclude them on practical grounds. Second, insurgent leaders often ‘leak’ far more information than they intend, including around internal disputes. Popular perceptions of insurgent ‘propaganda’ can see it as highly scripted and choreographed, such as that produced by IS. However, the overwhelming majority of communiqués considered here do not fall into this category, and many statements were rambling, ad hoc affairs. As such, they often revealed information that was unfavourable to the group. The IK’s operating environment exacerbated this and forced more information out into the open: intense security-service pressure made coordination extremely difficult, and insurgents used online statements to communicate with each other and external audiences. These particulars likely vary by group, so the persuasiveness of this point will change according to the groups with which researchers are familiar – yet this emphasises the importance of studying groups in a highly contextualised manner. It also leads directly to our third and final point: we should be careful not to base understandings of insurgency exclusively on groups that are relatively accessible and to prioritise methodological purity to such a degree that some groups cannot be studied. For example, the conflicts in Colombia and Northern Ireland are among the most researched in the world and have had an outsized impact on theoretical development. But limited information availability is not just a factor affecting the researcher – it characterises the environment in which the insurgents themselves operate. There may, therefore, be critical differences between insurgencies in information-rich and information-poor environments. We risk drawing the wrong conclusions if we transplant theories developed around the study of insurgencies that are different in fundamental ways. In the North Caucasus, additional sources of information cannot be found with just a little more effort; often, these sources do not exist, will never be publicly available, or can only be obtained through significant ethical violations.

A final limitation of the study is that the nature of the conflict in the North Caucasus limits our discussion to one type of transition: that prompted by the death of the incumbent. Russian counter-insurgency was of the take-no-prisoners kind, with Ingushetian emir Ali Taziyev the only senior IK leader ever to be arrested and tried. This is significant since multiple scholars find the nature of removal affects its impact.Footnote ⁷⁶ We will reflect further on this in the discussion below and the conclusion. However, here it is worth noting that this study is informed by an interpretivist perspective that rejects positivist presumption that underlying laws and rules can form the basis for generalisation.Footnote ⁷⁷ Any case is a priori constrained in its ability to tell us something definite about another context. Equally, the benefits from considering a broader range of cases would be offset by the loss of insights from detailed consideration of a single set of cases. In line with Lincoln and Guba’s interpretivist solution to the positivist dilemma, we aim not for generalisation but for the development of ‘working hypotheses’, which recognise that any situation has its own unique characteristics and that transferability depends on a detailed understanding of the contexts from and to which ideas are being transferred. From this perspective, any version of our typology would require adaptation when applied to other contexts. Scholars unpersuaded by particular dimensions of the typology and the application to other groups they study should therefore embrace the need for adjustment as a feature rather than a weakness of the typology.

Task one: Lay claim to the leadership

Perhaps the most obvious task facing aspiring leaders is to stake their claim to the leadership. Yet this task is fraught with difficulties. Success in making a claim can depend on group structure and the relative strength of rival claimants, and on the ability of other actors to impose their preferences. Transition appears to be quickest in hierarchical groups with clear chains of succession. In Kabardino-Balkaria, the local insurgency’s founding leader, Anzor Astemirov, was a charismatic ideologue with considerable personal authority and theological credentials. He had overseen the transition from an overt local social movement to a clandestine one aligned with regional insurgent actors.Footnote ⁷⁸ When Astemirov died, Asker Dzhappuyev could leverage his role as naib to stake his claim, despite lacking Astemirov’s social standing and intellectual background.Footnote ⁷⁹ His claim did not encounter public dissent and was strengthened by the recent death of another naib who would have had an equal claim.Footnote ⁸⁰ In the insurgency’s senior leadership, Dokka Umarov had risen through the ranks and occupied the ChRI vice presidency before assuming the overall presidency, much like his predecessor, Abdul-Khalim Sadulayev. In both cases, however, the acquiescence of influential figures within a much stronger movement was critical to the movement’s acceptance of their claim.Footnote ⁸¹

Influential or senior figures within a group may attempt to impose their leadership candidates on nominally subordinate factions. However, group members can exercise agency in choosing whether to accept this decision. As IK leader, Umarov often issued edicts designating vilayyat emirs. In some instances, these followed quickly after the death of a previous leader, ensuring a quick transition.Footnote ⁸² In other cases, the appointments appeared to represent little more than a confirmation of operational realities. Rustam Asilderov, for example, had previously served as a sector commander and a naib to the leader of the Dagestani insurgency; he used the title of acting emir from April 2012 but was not formally confirmed by Umarov until August.Footnote ⁸³ The Dagestani insurgency also furnished several examples of local groups rejecting their leader’s attempts to impose a candidate, such as the dispute in 2014 over the leadership of the Aukhovskiy jama’at.Footnote ⁸⁴ Subordinates can also exercise their agency by accepting or rejecting the claims of those who seek to challenge existing leaders and bring about transition by a coup. This was most clearly seen in 2010 when a group of Chechen emirs sought to oust Umarov but found their challenge rejected by the Kabardino-Balkarian and Dagestani emirs, whose support was required to confirm their claim.Footnote ⁸⁵

A leadership claim appears most precarious when there is no obvious candidate or external actor to impose preferences. In this case, negotiations can be drawn out. When Umarov died in September 2013, he had not designated a successor, and the ensuing shura discussions dragged on for six months because of communication difficulties. As multiple leaders testified, physically convening a shura was often problematic for the IK. The death of a leader makes closer coordination a more compelling task than might be the case at other times, increasing the chances of senior figures attempting to meet. This illustrates how counter-insurgency measures can shape the transition process. On this occasion, the Chechen authorities tried to intervene more directly, publishing intercepted recordings of part of the negotiations.Footnote ⁸⁶ Elsewhere, the security services in Kabardino-Balkaria were able to eliminate members of the leadership cadre in successive operations in April 2011 and September 2012 when they attempted to convene a shura.Footnote ⁸⁷ In Dagestan, Asilderov claimed that two months of heavy fighting disrupted efforts to convene a shura to replace Vagapov.Footnote ⁸⁸ When the meeting finally took place, it deliberately did not include all senior leaders ‘for safety reasons’.Footnote ⁸⁹ The difficulties of meeting in person created a reliance on video and audio messages exchanged via courier and the Internet, which increased the opportunities for interception and detection. This was illustrative of the so-called terrorist’s dilemma identified by Shapiro, highlighting the difficulties of balancing operational security with the need for operational control.Footnote ⁹⁰

Uncertainty and vulnerabilities created by the lack of a clear successor raise the question of why existing leaders do not always designate replacements. At least two factors are important. First, any identified successors become a logical target for counter-insurgency operations, creating an external vulnerability. Second, heirs are potentially also an internal vulnerability and threat to the current leader. For example, the 2010 leadership challenge to Umarov centred around his recently appointed heir; it is probably no coincidence that having defeated the challenge, Umarov did not again nominate a successor.

Task two: Renegotiate relationships and (re-)establish networks

If leadership is conceived of as a negotiated relationship between leaders and followers, the emergence of a new leader represents an opportunity to renegotiate that relationship. This can involve deciding whether to continue the relationship or establish new terms, as followers seek to ensure that new leaders adequately represent their interests. Following the deaths in relatively quick succession of Maskhadov and Sadulayev, both Astemirov and Vagapov corresponded with Umarov regarding the ideological direction of the group. They made their support dependent on a formal Islamisation of that ideology, and their influence was a critically important factor in Umarov’s decision to proclaim the IK.Footnote ⁹¹ Astemirov and Vagapov could leverage their theological training abroad and independent support bases within their local networks to push this agenda. Umarov, by contrast, lacked any such background and consistently prioritised group unity over ideological rigour.Footnote ⁹² Thus, individual personalities and social capital were significant factors in maintaining group relationships through the transition period.

When Asilderov challenged Umarov’s successor, Aliaskhab Kebekov, for leadership of the group – as Asilderov sought to align the group with the rising IS – Kebekov and his close ally (and ultimate successor) Magomed Suleymanov responded in the opposite way, prioritising ideological purity over unity. Kebekov and Suleymanov based their leadership on their theological rather than military credentials; neither had evident networks beyond their republic’s boundaries nor demonstrated the willingness to compromise that might have helped build and maintain relationships. In the eyes of those who favoured supporting IS, neither was seen to represent their interests. Accordingly, they switched their support to Asilderov.Footnote ⁹³ By contrast, the failure of the leadership challenge to Umarov in 2010 shows how failed negotiations can result in the collapse of a leadership bid. This also illustrates how the renegotiation of relationships can precede or even cause the actual transition, something most obviously seen in the case of internal coups. For instance, Mosinger argues that ‘leadership disputes result from a shifting balance of loyalties within a rebel organisation’, with network changes affecting individual leaders’ power bases.Footnote ⁹⁴

Additionally, departing leaders may have linked nodes to other parts of the network, leaving new leaders to face the challenge of re-establishing links essential to a group’s function. Asilderov, for example, reported that the Kabardino-Balkarian insurgency lost contact with the central leadership following the deaths of Sadulayev and influential field commander Shamil Basayev.Footnote ⁹⁵ Negotiations over the ideological direction of the insurgency could only start once those contacts had been re-established.

When considering the nature of authority in jihadist groups and how this could differ from other ideological types, it is worth highlighting the role of bayah. This doctrinally mandated pledge of allegiance provides two opportunities to exercise agency. First, new leaders of factions nominally subordinate to a broader group can stake their claim to the leadership by pledging bayah on behalf of the group they seek to lead, thereby creating something of a fait accompli: superiors can only ignore or override pledges in a public manner, raising the costs of doing so. Second, subordinates can signal their acceptance of a new leader’s claim by pledging bayah to them, and the claim’s legitimacy remains in flux until they do so. Similar mechanisms for exercising agency and creating facts on the ground may exist in other types of groups. Sadulayev (2005), for example, cemented his control of the ChRI by deploying constitutional mechanisms to seize control of all aspects of the insurgent hierarchy.Footnote ⁹⁶ However, in at least one regard, bayah is unique: since it is offered to individuals rather than groups, allegiances and alliances do not automatically transfer from one leader to the next. In other words, leadership transition represents a moment of flux in the life cycle of jihadist groups, when relationships are automatically opened for renegotiation.

The process of (re)negotiating relationships and (re-)establishing networks provides further clues about how state opponents can seek to intervene during transition. Distrust and fears of betrayal and infiltration are widespread within illegal, clandestine groups. When previous relationships are weak or missing, these problems can become aggravated, often deliberately. Security services may withhold information about the removal of a leader while capitalising on the intelligence. They may also claim that removal has compromised the networks to which the previous leader belonged. The sole detention of a senior IK leader, Taziyev, occurred days – possibly weeks or months – before it was publicly reported; by that point, he had already been transferred outside the region for interrogation.Footnote ⁹⁷ Yevkurov then claimed that Taziyev would reveal everything, citing examples of other insurgents being ‘loquacious’ in detention and calling on rebels to surrender before it was too late.Footnote ⁹⁸

Meanwhile, the IK–IS schism spilt into the public domain as Kebekov was still consolidating his control over the group. The first pledge of allegiance to IS had some suspicious features that suggested security service involvement.Footnote ⁹⁹ Therefore, the pledge can be interpreted as part of an effort to flush internal disputes into the open, which other leaders later acknowledged dragged on for some time. Insurgent leaders demonstrated an acute awareness of the dangers of authorities capitalising on internal disputes. Suleymanov, for example, warned leaders not to record accusations about other members of the insurgency, arguing ‘if they fall into the hands of the infidels, they will use them for their own purposes’.Footnote ¹⁰⁰

Task three: Demonstrate organisational continuity and inspire participation

A third task facing new insurgent leaders is to demonstrate the ongoing viability of the group and inspire people to keep following it. For all the flaws of the literature on charisma and the methodological difficulties entailed in trying to account for something as intangible as inspiration, they nevertheless attempt to capture something important. Senior figures within the IK made heavy use of the concept of fard al-ayn, an obligation incumbent on all Muslims, in trying to mobilise audiences, while new leaders made similar recourse to this strategy.Footnote ¹⁰¹

Although the challenge of inspiring followers is not unique to the transition process, groups are also attuned to the demotivational impact of leadership losses and recognise the need to counter it. Rhetorically, this can take the form of assertions that the struggle will continue. Dzhappuyev, for example, reported that ‘we will continue in the same direction and we have taken up the banner from the hands of our martyrs’.Footnote ¹⁰² Martyrdom narratives were commonplace, attempting to portray the loss of individuals as the fulfilment of a goal rather than defeat. There were repeated efforts to portray killed leaders as having fulfilled their deal with Allah and thus to be worthy of emulation. These rhetorical strategies are not unique to leadership losses but form part of the response to any setback, with martyrdom remaining a significant theme on IK websites.Footnote ¹⁰³ They are, moreover, just as often articulated by other voices within the group as by new leaders themselves. Following the death of both leaders and rank-and-file fighters, multiple actors across the IK relied on the same Qur’anic verse about how the Prophet’s followers did not lose heart when believers fell in battle.Footnote ¹⁰⁴

Demonstrations of group continuity can also manifest themselves through action. Since several of the leaders examined here did not issue statements, this may be an even more common strategy – although statements can be used to advertise and justify attacks. The Dagestani general leadership portrayed a series of attacks as a response by the vilayyat’s new leader – Magomed Vagapov, a former sector commander who simultaneously assumed the role of IK qadi – to the authorities’ assertion that his predecessor’s death signalled the forthcoming defeat of the insurgency.Footnote ¹⁰⁵ Ingushetian president Yunus-bek Yevkurov and local Dagestani newspaper Chernovik interpreted increased violence following the removal of senior rebel leaders as attempts by remaining insurgents to demonstrate their endurance and leadership claims.Footnote ¹⁰⁶ In Kabardino-Balkaria, Dzhappuyev’s first statements documented several attacks and claimed that some were intended to demonstrate the authorities’ impotence.Footnote ¹⁰⁷ Violence in the republic did indeed spike dramatically: A senior official reported a 400 per cent increase in attacks on law-enforcement personnel in the first eight months of Dzhappuyev’s tenure compared to the previous year.Footnote ¹⁰⁸ Such spikes, however, are not always clearly attributable to leadership strategy. Rhetorically, Dzhappuyev articulated many of the same ‘internal brakes’ – understood as ‘mechanisms through which members of the groups themselves contribute to establish and maintain limits upon their own violence’ – as Astemirov and appeared to seek to restrain violence.Footnote ¹⁰⁹ However, he lacked Astemirov’s authority within the group, and the security services played an important role in escalating violence.Footnote ¹¹⁰ How patterns of insurgent activity vary during transition periods falls outside the scope of this paper but requires further investigation.

Task four: Establish organisational priorities

Embedded in our definition of insurgency is the pursuit of political goals. Logically, therefore, we might expect the leadership to play a significant role in establishing these goals and the general direction and priorities of the group. In reality, however, the performance of this task varied considerably, and it assumed importance only in certain circumstances. Several leaders under consideration did not issue any statements. This cannot be explained solely by the length of their tenure: some leaders survived for only a few months but issued multiple statements, while others were in post for much longer but rarely communicated publicly. Establishing priorities can be considered a collective function, and a leader can delegate the task to others within the group. Taziyev, for example, was a reticent speaker but was accompanied for most of his leadership by Aleksandr Tikhomirov (Said Buryatskiy) – a charismatic convert to Islam who became a leading ideologue despite never occupying a formal role within the IK. Other leaders and insurgent actors, such as the official websites, can also perform the function, leaving leaders free to concentrate on other tasks. In Dagestan, the websites were far more active than vilayyat emirs in communicating a group ideology. These examples highlight the importance of viewing leadership as embedded in broader group activity, challenging the idea that disseminating an ideology is crucial for leaders, thereby affecting who can assume the role.Footnote ¹¹¹

Expectations appear to be an essential factor in determining when the performance of this task assumes importance. When a movement or its legitimacy is seen as closely linked to an individual leader, there may be question marks over the group’s orientation following that leader’s death – and consequently, a greater likelihood of statements asserting group goals and priorities. Similarly, if there is intense competition within a movement over its ideological direction, there may be an expectation that the new leader will clarify their position. ChRI leaders Sadulayev and Umarov faced significant contestation within the group, and both issued statements shortly after assuming leadership of the insurgency that reaffirmed existing policy.Footnote ¹¹² Both, however, went on to oversee significant changes in group ideology, demonstrating the negotiated nature of leadership. In this way, new leaders cannot, from the outset, assert radically different goals unless they already have sufficient support. Kebekov, by contrast, did not initially face significant competition over the direction of the group, and there were no expectations of a dramatic shift. His lengthy early statements neither reaffirmed existing goals nor articulated new ones, despite him being a more articulate and intellectually oriented leader than Umarov. In circumstances that do not compel them, expectations of bold ideological statements are often not realised. If neither party desires nor needs to renegotiate the ideological terms of the relationship, additional explanations of priorities may be superfluous to its members.

Task five: Survive

The task of surviving is so self-evident that it is often not mentioned – indeed, in clandestine groups, it is a task that leaders and activists must perform at all times. Yet there are good reasons for explicitly highlighting it in discussions of transition. The performance of the preceding tasks is self-evidently dependent on success at this one. Moreover, it often requires opening up channels of communication or engaging in activities that increase vulnerabilities.Footnote ¹¹³ Chernovik, for example, reported that security services could locate Vagapov because he went online to upload a video and communicate with subordinates.Footnote ¹¹⁴ The more compelling the task within a broader group context, the more willing leaders are likely to undertake it – and a fundamental justification for focusing on transition rather than leadership writ large is that this period is one where these tasks assume greater importance. Critical to this is the recognition that insurgencies do not operate in a vacuum but face opponents continually interested in their destruction. These opponents will – or should – be particularly alert during the transition period to activities and vulnerabilities they can exploit. The targeted removal of important figures can generate new intelligence on potential successors: the security services claimed they could trace Astemirov after documents recording his travel routes were found following the death of his deputy.Footnote ¹¹⁵ Although such claims are impossible to verify, Astemirov himself provided supporting evidence, recording a final will and testament because he had to ‘move around a lot now, travel to attend meetings with different people, restore the system which our brother Omar had organised’.Footnote ¹¹⁶ Arrests, rather than detention, are also likely to produce valuable intelligence that can undermine the performance of other tasks.

On the other hand, it is worth considering the consequences of failure to perform this task: coming so soon after the loss of a previous leader, it can be incredibly, even critically, damaging to morale and undermine the broader group. In other words, repeated leadership losses can significantly impact the performance of task three. A shortcoming of several studies of leadership losses is their treatment of losses in isolation from one another rather than considering the effect of successive losses and high leadership turnover. Vagapov not only assumed the role of Dagestani emir but also replaced Astemirov as IK qadi. He had arguably developed a charismatic bond and strong standing with local insurgents and could claim considerable social capital, having obtained theological and military training abroad.

Although Vagapov may have had the same potential as Astemirov to exercise cross-faction influence, he ultimately lacked the opportunity – lasting only a few months in the role before he, too, was killed. The IK lost two individuals who demonstrated the ability – rare within the group – to combine military and theological leadership. This occurred when the group suffered the death of several highly influential figures, and although it survived for several more years, it arguably never recovered from this. Indeed, parts of the IK were prone to leadership by default, whereby someone assumed the leadership simply by surviving long enough to be promoted. In Kabardino-Balkaria, all five of Astemirov’s successors were part of his original network; when the group sustained losses, the network contracted, and the insurgency failed to attract enough recruits to compensate. Whereas Astemirov enjoyed considerable social capital and was highly articulate, one of his successors, Alim Zankishiyev, was relatively unknown and struggled to form coherent sentences.Footnote ¹¹⁷ The final leader in the sequence was a former sector emir who had allegedly been removed from his role for passivity.Footnote ¹¹⁸ Ultimately, however, he returned to the leadership because of the lack of candidates with more than a few weeks or months of experience within the insurgency.