Extinction Rebellion: Green activism and the fantasy of leaderlessness in a decentralized movement

This article interrogates the idea of leadership in a decentralized organization, using Extinction Rebellion (XR) as a case study. Through close observation of this environmental movement, we problematize the notion of leaderless organizations and question whether the idea of a hierarchy- and power-free ‘decentralized organization’ is a fantasmatic endeavour. Psychoanalytic analysis of discussions between XR members following two disruptive actions during XR’s International Rebellion in October 2019 reveals that these actions temporarily provoked members to question the attainability and effectiveness of XR’s so-called ‘leaderless autonomous organization’. This article contributes to psychoanalytic studies of leadership and to the social movements literature and argues that the power relations present in any form of organized endeavour must be recognized in order to develop effective and democratic activism.


Introduction
Grassroots activism has grown rapidly in recent years (Bevington, 2009;Gieseler, 2019;Jakobsson and Korolczuk, 2019). Decentralized protest movements, against issues such as environmental collapse, increasing inequality and systematic injustice and violence, often adopt flat, nonhierarchical structures that eschew formal leadership. These decentralized mobilizations are made possible by technology, which now enables so-called spontaneous leaderless movements (Castells, 2013;Ross, 2012). Political engagement through social media has reframed people's agency, providing and supporting alternative routes for oppositional activism (Lievrouw, 2011) on a wide range of issues. Information sharing on commonly used social media platforms has influenced people's discussions and worldviews while expanding their repertoires of action (Earl and Kimport, 2011). The interactivity characterizing social media has also enabled social movements' transformation from hierarchical forms of decision making (Bennett and Segerberg, 2013;Gerbaudo, 2012;Lievrouw, 2011) towards networked, neo-anarchic, horizontal forms without formal leadership (Teixeira, 2018;Western, 2014).
For many movements, being leaderless is a political stance, with roots in autonomist and anarchist politics (Graeber, 2011;Western, 2014). From this viewpoint, the entire notion of leadership is suspect because its ingrained authoritarianism goes against egalitarian principles and democracy through individual empowerment. Western (2014: 675) associates this with their 'disavowal of all leadership (including autonomous forms)'. Such ideas inspire many current 'leaderless' social movements, appealing to numerous diverse groups of citizens around the world who feel defrauded by their political class (Gerbaudo, 2012). However, lack of formal leaders may pose challenges. Spontaneous leaderless movements may successfully achieve clearly defined short-term goals, such as toppling a tyrannical ruler, but often struggle to maintain momentum to achieve long-term objectives (Western, 2014). Some have argued that this is not problematic in itself as many such movements aim to raise awareness rather than implement sustained political changes (Gerbaudo, 2012).
In the last decade, social movements with autonomous structures and no leaders have been effective in spontaneously mobilizing large crowds (Snow and Moss, 2014), although their popularity may wane just as quickly. A recent review of experiences in several countries (Youngs, 2019) summarizes developments in the aftermath of such mobilizations as protesters join mainstream political parties or traditional non-governmental organizations, or move into low-profile community organizing. Protesters may also withdraw from political action altogether, partly to protect themselves from government repression, and may adopt new forms of organization and resistance (Tarrow, 2011;Youngs, 2019). The new 'networked' social movements rely on thoroughly integrated digital connectivity for organizing, publicity and effective communication (Lim, 2012;Van Den Donk et al., 2004). This has purportedly led governments to use political and financial pressure to increase their control over broadcast media and demonize social media (Tufekci, 2014). Lacking organizational depth, leaderless protests are often forced to operate in a polarized environment, unable to create national dissent and ensure their movements' continuity (Mongiello, 2016;Tufekci, 2014).
Overall, it is unclear whether such movements achieve their goals more effectively than those with conventional forms of leadership (Tufekci, 2014;Youngs, 2019). While the literature on social movements has advanced understanding of the organization of these spontaneous movements (e.g. Pilati et al., 2019), few studies have examined this phenomenon from a leadership perspective. Knowledge is lacking on whether the various alternatives termed 'leaderless leadership' mean turning away from the defunct top-down model towards distributed, servant or relational leadership, or whether they are an expression of an anti-leadership trend (Kellerman, 2012;Sutherland et al., 2014).
In this article, we ask what is the role of leaderless leadership in oppositional grass-root initiatives and how this impacts movements' achievement of their goals. Using the example of Extinction Rebellion (XR), a non-violent grassroots organization established to sensitize politicians and the public to the urgent need for action to prevent impending climate disaster, we suggest that the notion of leaderless leadership on its own maybe problematic for achieving lasting social and political transformation. This is because without workable democratic procedures, it is difficult to maintain legitimacy in inclusive movements and cooperative efforts involving multiple organizations (Kwok and Chan, 2017). Conversely, autonomy at the base sometimes makes it difficult to implement coherent strategies across their membership in decentralized movements (Tarrow, 2011). Further, in the absence of clear alternatives to hierarchical leadership, individuals eventually assume or are forced to assume leadership roles that could displace their original goals (Sutherland et al., 2014;Western, 2014). Symbolism and emotions strongly influence how contentious politics are formed and reinforced in collective action (Jaspers, 1998), often centring on the narratives of 'leaders' and 'leaderlessness'. Viewed from the critical organizational leadership perspective, leaderless movements appear to reject the romantic view of leadership (Meindl, 1995) that lionizes the lone individual hero-leader on the one hand, while avoiding common leadership dysfunctions such as narcissism (Kets De Vries, 1991) and abuse of power by individuals (Kets De Vries and Miller, 1984) on the other. Our analytical framing is inspired by critical leadership studies in general, and by psychoanalytic approaches to leadership in particular.
This article contributes primarily to psychoanalytic approaches to leadership (Costas and Taheri, 2012;Driver, 2013;Gabriel, 1999;Western, 2014) by deploying the Lacanian concept of fantasy to elucidate avoidance of leadership in a leaderless social movement. We adopt a psychoanalytic perspective to show how the fantasy of being a 'leaderless organization' enables activists to avoid debate over the nature of leadership in decentralized social movements (Kwok and Chan, 2017;Nepstad and Bob, 2006;Rahmouni Elidrissi and Courpasson, 2019;Western, 2014), and potentially makes democratic politics and social change impossible. We thus argue that such movements' failure to explicitly articulate these issues and fully implement democratic procedures may lead to evasive behaviour, confusion and abdication of responsibility, which may undermine their political and organizational goals. Specifically, we highlight the deployment of fantasy as a way to avoid decisions on forms of democratic power sharing to temporarily neutralize tensions and anxieties within the movement and in its relationships with the outside world. Researchers have begun to employ this idea to show how relying on the fantasmatic notion of a conflict-free future in the fields of social enterprise (Kenny et al., 2020), public services (Ekman, 2013;Fotaki and Hyde, 2015) and financial markets (Komporozos-Athanasiou and Fotaki, 2020) may ultimately bring about counterproductive outcomes. However, this framework has not previously been applied to social movements. Focussing on people's desires, imaginings and projections of the anticipated future as an important driver of both individual and collective political action (Fotaki, 2010;Stavrakakis, 2007) may yield novel insights into leaderlessness in many contemporary social movements.
In the remainder of this article, we briefly review the literature on anti-hierarchical, distributed leadership and other forms of egalitarian leadership and define key characteristics of leaderlessness in the organization of protests. We then present our case study of XR and outline the methods used to analyze chat group conversations in one local organization of XR in the UK, as well as documentary references and two short interviews. Based on analysis of the material, we problematize leaderless leadership through a psychoanalytic lens to delineate the possibilities for democratic and inclusive leadership. We also articulate some further theoretical and practical implications for organizational practice in the context of social movements. apex and followers who cheer and follow them, many social movements have adopted the logic of leaderlessness to counteract authoritarianism and promote radical, democratic, participatory forms of organizing. They have no assigned leaders, and it appears that the people themselves spontaneously come together to demand change. Some of these tendencies can be traced back to autonomist anarchy or feminism, setting individual freedom against the notion of people exercising power over others (Bamyeh, 2012;Western, 2014). Well-known leaderless social movements (Cornell, 2011;Gerbaudo, 2012;Ross, 2012) include the Arab Spring, Occupy in the US, Spain's Indignados movements and Black Lives Matter, as well as XR. Yet, leaderless leadership also drove earlier movements, including Mexico's Zapatista Army of National Liberation (EZLN) which mobilized communities of Chiapas, landless movements in Brazil and the Catholic Left-inspired Plowshares movement (Nepstad and Bob, 2006;Speed, 2008), and continues in many contemporary protests, such as the Gilets Jaunes in France, protests against political-class corruption in Lebanon and price rises in Chile and demands for greater political freedom in Hong Kong. Although not considered in this study, leaderless resistance may also be a strategy of opposition allowing and encouraging individuals or small cells to engage in acts of political violence, or even terrorism (Sageman, 2008), entirely independent of any hierarchy of leadership or network of support (Joose, 2007).
The advent of social media is commonly cited as enabling the emergence of leaderless movements. Nevertheless, people join such movements for a reason. For instance, they are often disillusioned by politicians reneging on their promises, being detached from the concerns of those whom they are supposed to represent and instead engaging in corrupt and self-serving behaviours. This results from a failure of political hierarchical authority (Bamyeh, 2012), often exacerbated by lack of opportunities for young people and worsening of existing inequalities in the developed and developing worlds (Sen, 2002).
Devolved 'leaderless' protests first appeared around 2011, in the Occupy movement and the Arab Spring protests, and spontaneous protests emerged in the aftermath of the global financial crisis (Ancelovici et al., 2016). Mass mobilizations have occurred in democracies and non-democracies and advanced and developing economies alike, becoming a major feature of global politics (Youngs, 2019). In addition to opposing corruption and economic injustices, protesters have taken to the streets to speak out about environmental problems, repression and local issues (Youngs, 2019). In 2019, mass demonstrations in Africa (Algeria and Sudan), the Far East (Hong Kong and Indonesia), the Middle East (Lebanon, Egypt and Iraq), Latin America (Chile and Bolivia) and Europe (France and Spain), to name but a few, adopted various forms of decentralized organization without formal leadership.
A strength of such oppositional movements is their reliance on informal networks to mobilize mass protest, making them difficult to suppress (Pfaff and Yang, 2001), especially when they adopt non-violent forms of action. This also allows for flexibility and adaptability to changing conditions (Welty et al., 2013), given their independence from established opposition parties and other alternative formations (Oikonomakis and Roos, 2016). Decentralized organizations are less vulnerable to the loss of a charismatic leader. Their spontaneity and power to bring large numbers of people rapidly onto the streets, as shown in the recent Black Lives Matter demonstrations, has led to a hypothesis that the future of social movements lies in networked, non-hierarchical forms of organizing (Keating, 2020). At the same time, both critics and sympathizers point to their limited ability to achieve lasting change (Mongiello, 2016), or even to survive as a movement (A'Lee Frost, 2020), partly due to the establishment's backlash against their actions (Tufekci, 2014;Youngs, 2019). Their various forms of collective action have triggered a new and more aggressive repertoire of protest policing (Tarrow, 2011). Furthermore, enthusiasm for the empowering potential of social media (Gerbaudo, 2012;Jackson et al., 2020) to promote social activism is tempered by evidence of data and information manipulation, as well as the media owners' lack of transparency and/or accountability to the public. The ever-present and ubiquitous danger of reproducing conventional or coercive power dynamics is unacknowledged by 'leaderless' movements (Western, 2014). Equally, the various means to counteract these issues and maintain legitimacy with participants (Kwok and Chan, 2017), including mechanisms to dissolve power (Uitermark, 2017), may impede the movements' achievement of their aims.
These developments in the oppositional politics of contention (Tarrow, 2011) somewhat parallel academic questioning of the idea and role of leadership as an all-encompassing solution (Barker, 2002;Currie and Lockett, 2007;Gemmill and Oakley, 1992;Meindl, 1995). For example, Gemmill and Oakley (1992) regard the whole idea of leadership as a socially fabricated myth that functions to reinforce existing social beliefs and structures around the necessity for hierarchy and leaders in organizations. Meindl (1995) forcefully challenges the strong belief that leaders are responsible for the outcomes of organized action, calling this the 'romance of leadership'. Barker (2002: 478) rejects conventional leader-centred theories that juxtapose individual and institutional needs and decries the common bemoaning of 'lack of leadership' as a sign of increasing social despair and learned helplessness. Currie and Lockett's (2007) study of the education sector shows how policy makers rather than leaders transform the context in which leadership takes place. Various notions of relational (Uhl-Bien, 2006), servant (Greenleaf, 1971) and distributed leadership (Bolden, 2011) are also proposed to counter the dominant Messiah leadership discourse (Western, 2008). Some stress the role of 'followership' (Uhl-Bien et al., 2014), or focus on distributed leadership as a unit of analysis (Gronn, 2002), and others offer authentic leadership (Avolio and Gardner, 2005) as a way to achieve greater well-being, happiness, autonomy and ethical responsibility and emancipation (Costas and Taheri, 2012) in organizations.
While all these approaches oppose traditional authoritarian structures and hierarchical followerleader relations, instead advocating a distributed, egalitarian and altruistic leadership style, they often ignore power relations. While recognizing that power is not derived from the leader but is given by the followers (Uhl-Bien and Carsten, 2007), this relational reconfiguration of leadership does not account for how power flows in these relationships, nor does it negate the need for some form of leadership. Equally, power in distributed leadership is supposed to extend beyond the person of the leader (Bolden, 2011;Gronn, 2002) to include followers, but how this is achieved and the implications of shifts in power dynamics are rarely examined. Finally, the idealization of followership may be driven by a desire to control 'the masses' (Ford and Harding, 2018), reproducing the very kind of leaderism it seeks to condemn (Learmonth and Morrell, 2017) by obscuring power relations. In this article, we argue that all these dynamics are present in leaderless organizations as there can be no organized action without power. It is therefore important to understand how nonhierarchical alternatives work in practice and particularly how leadership is practised and perceived in leaderless social movements.
To examine these issues, we take inspiration from psychoanalysis, which considers power as both an intra-psychic (occurring internally within individuals' psyches) and inter-psychic phenomenon in relation to others. This derives from Freud's (1923) psychoanalytic philosophy, which highlights the impact of unconscious drives and early-life attachments on our future actions and patterns of behaviour. Psychoanalysis has long contributed to our understanding of unconscious dimensions of the leadership phenomenon (Simpson et al., 2002;Zaleznik, 1989) and its dysfunctions (Kets De Vries, 1991;Petriglieri and Stein, 2012;Stein, 2007). In the psychoanalytic conception of leadership practice, leaders and followers are bound by deep, unconscious links (Burns, 1987) that permeate all aspects of organizational reality. Leaders are believed to have a unique capacity 'to spin followers' dreams' (Gabriel and Hirschhorn, 1999: 140), thereby enabling the group's inner desires to be expressed. Various psychoanalytic schools of thought have inspired a range of developments in leadership. Working from an object relations theory perspective, Krantz and Gilmore (1990) consider splitting heroic leadership from the technical, problem-solving process of management as a social defence against the uncertainty and turbulence faced by many organizations. Object relations theory also provides startling insights into the genesis of corporate misconduct and decay embedded in narcissistic entanglements between leaders and organizations and leaders and followers (Stein, 2005), as well as executive hubris enacting leaders' fantasy of omnipotence enabled by followers' idealization (Gabriel, 1999). Yet on the whole, psychoanalysis has historically been weak in linking unconscious dynamics to power and empowerment.
These issues are only partly addressed by critical scholars, who draw on the Lacanian theory of subject formation to theorize leadership identities as imaginary constructions that invariably fail because they are reiterating a lack of being (Driver, 2013). This lack is constitutive of the subject, who constantly seeks to cover it up by engaging with various imaginary constructs and fantasies that can never be realized (Fotaki, 2009;Kenny et al., 2020). Fantasies originate in the imaginary realm, but are expressed through symbolic means such as language, social discourses, laws and social norms comprising the big Other (Stavrakakis, 2008). In their Lacanian analysis of authentic leadership, Costas and Taheri (2012) show how such ideas dominate both academic and organizational discourses.
In this article, we draw on interpretations of fantasy as a core aspect of human subjectivity, originating in an imaginary lack in the subject but manifesting itself in the symbolic order through language. According to Lacan, fantasy necessarily emerges as part of the structure of language and related social norms, but is never fully accessible to the subject because it is rooted in the unconscious (Fotaki, 2009). The subjects find themselves covering over fissures in the symbolic order to avoid the anxiety produced by this lack, and fantasy emerges as an important mechanism to achieve this. Nevertheless, the fantasy provides only short-lived relief from the lack by creating a temporary placeholder for the desired state. Western (2014) argues that the discourse of 'leaderlessness' is one such fantasy that temporarily fills the gap caused by the subject's fragmentation and lack of wholeness. However, according to the Lacanian ontology, such fantasy is also a symptom of the lack, and as such it signifies a desire for (unfulfilled) leadership. Lack is therefore a cause of both desire and the 'leaderlessness' symptom, pointing to a desire for 'leaderfull' leadership. This, as we argue, indicates the productive aspect of fantasy which makes possible the imagining of different ways of organizing.
In what follows, we present the case of XR as an example of a new oppositional movement, combining anti-establishment discourse, decentralized organizing through digital networking and the discourse of 'leaderlessness'. We then outline our methods and discuss the findings emerging from analysis of a forum discussion of two contentious incidents and two short interviews.

Case presentation
XR is an environmental mass movement aiming to achieve systemic change through non-violent civil disobedience (Westwell and Bunting, 2020). Since its inception in the UK in May 2018, it has spread globally, replicating a decentralized model in which local groups are formed to coordinate local actions. By December 2020, XR's website (https://rebellion.global/) listed 70 countries in which a total of 1138 XR groups were active. There has so far been very little academic research on XR, so its structure and decision-making processes are relatively unknown. As detailed examination is beyond the focus of our research, we report here on how the movement introduces itself in its multiple websites, to show how this has contributed to the fantasy of a leaderless organization.
XR's global website introduces the movement as 'a decentralised, international and politically non-partisan movement using non-violent direct action and civil disobedience to persuade governments to act justly on the Climate and Ecological Emergency'. XR often cites the inspiration of grassroots movements, such as Occupy, Satyagraha (the non-violent resistance started by Gandhi during India's struggle for independence), the suffragettes and the civil rights movement of the 1960s, as models for the movement's governance and strategies (Taylor and Gayle, 2018). For instance, it encourages its members to engage in 'non-violent' civil disobedience actions, potentially leading to their arrest, arguing that this will help to turn bystanders into upstanders.
The movement also boasts of its grassroots and decentralized nature. Various XR websites claim that power is decentralized in XR as any person or group can organize autonomously and take action in the name and spirit of XR, as long as the action fits with XR's principles and values. XR's global website emphasizes that it is not a decision-making body for XR, claiming that 'In fact, there is no such body … Global Support does not decide whether a group is XR or not, what decides that is whether a group is living and applying the 10 Principles & Values and working towards the 3 core Demands.' XR's global website lists the movement's 10 principles, two of which revolve around issues of power and autonomy, supporting its claim to be a leaderless organization that emphasizes 'autonomy and decentralization': 'We collectively create the structures we need to challenge power. Anyone who follows these core principles and values can take action in the name of Extinction Rebellion' (Principle 10). The damaging role of power is highlighted, and the movement pledges to 'break down hierarchies of power for more equitable participation' (Principle 7). Further details of its governing structure, which is dubbed a 'self-organizing system', are provided in XR's constitution. Various arrangements are made to break traditional hierarchies of power, including distributed authority, decentralized power and transparent decision-making processes. It is argued that this system ensures that 'no individual has power over another', mitigating the negative features of a hierarchical structure. In keeping with its status as a leaderless organization, XR does not list any representatives or leaders. Only in the FAQs section of XR's original website (now relabelled XR UK) do the names of key members of the movement appear, in response to the question 'where did the idea come from?': Extinction Rebellion has come about as a result of academic research and mass engagement. Roger Hallam and Dr Gail Bradbrook … did some thinking and talking … and following discussions with others a movement was born (https://rebellion.earth/the-truth/faqs/). Nevertheless, Roger Hallam and Gail Bradbrook, along with a few other most active members, are often named in the press as leaders of this social movement. Their actions and words spark discussions both within and outside the movement.
In addition to an overarching structure to coordinate national and international events, the movement also enables the creation of informal groups, including 'general meetings' organized at the local level, for example, in boroughs and towns. These local groups initiate new members and facilitate activity groups (e.g. art groups and political groups). Individual members are encouraged to work with their affinity group (AG)-usually those trained together on non-violent direct action (NVDA)to develop their own actions in addition to those promoted officially. The movement also pushes for the creation of 'citizens' assemblies' at the borough/city level to promote its vision of netzero greenhouse gas emissions by 2025. XR UK's website claims that the Citizens' Assembly on Climate and Ecological Justice can bring ordinary people together to investigate, discuss and make recommendations on how to respond to the climate emergency (https://rebellion.earth/the-truth/ demands/). XR encourages its members to respond to the climate emergency in an individual capacity, for example, by lobbying their MP, mobilizing local support for XR's aims and carrying out NVDA. Examples of NVDA listed on XR's website include occupying or chanting during meetings; blockading/sit-downs/swarming outside symbolic buildings, meetings or MPs' offices; vigils/ camps/short hunger strikes with participants outside symbolic buildings or MPs' offices; naked protests; and gluing/locking onto the doors of buildings.
The movement is most visible in national and international action days or 'uprisings', when its activities are intensified, and groups coordinate their work to produce maximum disruption. For instance, in the October 2019 uprising, XR called for an International Rebellion to demand that governments around the world take urgent action to tackle the climate crisis: 'Together, we will peacefully occupy the centres of power and shut them down until Governments act on the Climate and Ecological Emergency' (https://rebellion.earth/event/international-rebellion-begins-7-october-2019/).
Rebellion protests were organized worldwide, gathering in capital cities. The International Rebellion campaign is reported to have drawn 30,000 activists to London alone (Townsend, 2019). For a fortnight, XR members held marches and demonstrations involving civil disobedience, for example, by occupying 11 sites in central London. In the first week, over 1000 XR members were arrested (Jarvis, 2019), yet some activities attracted major support both within and outside the movement. For instance, on 8 October 2019, an XR group of which Matthew Shribman, a scientist, was a member organized for 1000 trees to appear outside the UK parliament, with the first 650 allocated to British MPs. Over 400 MPs collected trees, and some publicly committed to accelerating reforesting in the UK (Jones, 2019). In another example of more disruptive but popular activity, on 12 October 2019, XR held a 'funeral procession' along Oxford Street, which allegedly attracted 20, 000 activists (Jarvis, 2019).
However, other actions have resulted in backlashes, from both members and wider society. In particular, two events on which we focus in this study generated considerable debate about the nature of XR, and specifically its decision-making process and leadership. On 17 October 2019, XR activists targeted rail and underground services heading to the Canary Wharf financial district by climbing onto or gluing themselves to trains at Shadwell, Stratford and Canning Town stations. Although this action proved unpopular among XR members at large, some well-known members went ahead with it. The tube action at Canning Town, a largely working-class neighbourhood, caused particular uproar, with an activist being dragged from the top of a tube train and beaten by angry commuters. Videos of this event went viral and caused many to question the group's decentralized structure allowing individual groups to act on their own.
A second event had taken place 2 days earlier. When the police announced a ban on all XR-related activities, the movement staged a protest at the Department for Transport at 8 a.m. on 15 October 2019, during which Gail Bradbrook stood on top of the building's entrance and 'called on ministers to explain how their continued expansion of roads and airports fitted with a net-zero emissions target' (Taylor, 2019a). Rather than adopting the usual XR tactic of gluing or locking herself to the building, Bradbrook cracked a window pane using a hammer and screwdriver.
We decided to analyze discussions generated by these two examples as they clearly capture the questions examined in this article: how leaderless leadership materializes in practice, and whether it is aharbingerofnewwaysoforganizinginsocialmovementsasamoreeffectivemeanstopromotetheirgoals.

Research design
Our research design used interpretive theory building (Klag and Langley, 2013) to investigate the notion of the leaderless organization by questioning how members of a so-called leaderless organization perceive leadership. We did not start with a firm assumption about leadership, but adopted an exploratory-inductive approach (Bansal and Corley, 2012). The research was motivated by the authors' interest in both decentralized movements and the politics of climate change (Harrison et al., 2019). One author joined XR and participated regularly in XR events in the UK from July to October 2019. He also joined a local XR group that held weekly meetings, and participated in a public online discussion forum. While, as researchers, we sympathized with the movement's goals and ambitions regarding curtailing climate change, we distanced ourselves from full engagement with the group to allow greater reflexivity as part of the activist research (Reedy and King, 2019). We maintained a passive observer role, exercising a non-obtrusive research style (Cole, 1991) while having direct, real-time access to data. We attended rallies and protests, but did not engage in activities that might be considered illegal. We also maintained a non-participant observer role in an online forum. This allowed us to observe how discussions emerged in their natural settings in an internet environment (Nørskov and Rask, 2011). Given our interest in leadership, we focused on discussions within the movement that shed light on how leadership was practised in XR and how activists perceived leadership. While one co-author's regular participation in XR activities helped us to contextualize our analysis, we focus here on discussions that took place in the aftermath of two critical events, which we label 'Canning Town tube action' and ''hammering glass window at Department for Transport', as briefly discussed above. We chose to focus on these two events because they were heavily debated within and outside the XR community, and generated lively discussion of the concept of leadership in the XR activist community.
We analyzed the content of these discussions, which took place in the chat group, to help us understand how members of grassroots organizations perceive the purpose, and even desirability, of leadership in decentralized social movements. Our aim was to better understand the participants' views on whether incidents involving elements of violence were justified, even if not approved by the majority of members; whether these acts signalled a departure from the collective form of leaderless leadership adopted by the movement; and whether they felt that this shift promoted the organization's goals.
We followed guidelines issued by the British Psychological Association (Hewson and Buchanan, 2013) on ethics for internet-mediated research. In particular, before treatment of the online data, it was important to ascertain the likely perceptions of social media users on whether these media are public or private. Based on our exchange with several participants, we concluded that the forum was not seen as a private group. The material's availability in the public domain was reiterated by several members, who emphasized that they were aware that this was a public forum. Nonetheless, some private information (e.g. names) was shared on the forum, so all names and identifying information were deleted from or adjusted in the dataset to maintain participants' anonymity and privacy. The content of the public forum was redacted and stored in a secure Word file (∼13,000-word document).
We conducted a first round of primary coding, using thematic analysis (Attride-Stirling, 2001) to capture how members reacted to each event. Centring around the notions of 'leadership', 'democracy', 'representation', 'decentralized structure', 'goals of the movement', 'non-violent action' and 'accountability', we identified two key themes: who is a leader in a decentralized organization, and how one becomes an accountable leader in a leaderless organization. We further analyzed the findings drawing on our psychoanalytic framework. Specifically, we sought to reveal tensions in how forum participants referred to leadership while claiming to belong to a decentralized organization with no formal leaders. Rather than downplaying inherent contradictions, a Lacanian framing works with and through them (Kenny et al., 2020;Parker, 2005), for example, seeing inconsistencies as manifestations of how affective anxiety emerges from a lack in the symbolic order (Lacan, 2006). In the next section, we present a narrative of how issues of leadership and power were discussed in XR.

Findings: Questions of leadership in a leaderless organization
Event 1 -Hammering glass window at Department for Transport: Who is a leader in a decentralized organization?
On 15 October 2019, almost 10 days after the start of XR's October rebellion, Gail Bradbrook grabbed the national headlines by appearing at the Department of Transport's London office and cracking the glass window of the entrance with a small hammer. Her action was meant as a protest against a call by London police to ban all XR activities in London. While many XR activists had been involved in designing and implementing a range of non-violent civil disobedience activities (e.g. gluing themselves to buildings, blocking roads), this action by a leading character in the movement was seen by many members as a violation of NVDA. This led to an outpouring of criticism, specifically of her assumption of leadership in performing an act that went against the movement's espoused principle of non-violent action. It also brought the issue of authority and representation to the fore, raising questions about whose actions speak for or against the movement. In the public discussion forum, views were divided. Almost half of the members supported Bradbrook's action. Supporting arguments ranged from claiming that cracking glass is not strictly a violent action, to advocating the necessity for more extreme actions given the urgency of the situation, and drawing parallels with earlier resistance movements that used similar tactics, such as the suffragettes: I do agree that doing it may put people off. Just that it is not necessarily violent. Gail was quoting the suffragettes as she did it (participant 8).
I think window smashing (cracking to be more accurate) is a reference to suffragettes breaking windows? And at that time it stirred huge public debate and condemnation. In retrospect, I think how mild those women were just breaking windows to get the right to vote and take part in public life etc. And how massively criticised they were for it. And that they won (participant 5).
It's a shame that Gail's actions have been perceived by many as violent, it's quite the opposite. The violence that is inflicted by BEIS continued inaction is leading to the death of peoples globally. A cracked piece of glass is nothing in comparison (participant 17).

Some participants defended Bradbrook's action by referring to how the movement's founders understood non-violent action:
Window cracking is listed in the nvda training script as one of the examples of a tier 3 non-violent action (along with spray painting). Based on this, it's clear that XR founders and trainers view it as a high risk non-violent action considered by the movement founders as a strategy that could generate more engagement (participant 1).
Pretty sure there are stats showing April's higher risk actions garnered more engagement and donations (participant 7). Others referred to the organization's decentralized structure and its principle of defending Bradbrook's right as a member (rather than a designated leader) of an AG to organize such an action: In nvda training, we talk about how any AG can organize an action as part of XR as long as it's in line with the principles (participant 1).
Yet almost half of the participants in the public forum questioned the action's appropriateness. Those against it were concerned that such actions might make it more difficult to convince the public to join or support XR and to engage with other stakeholders: If people start moving in this direction and treading the line between violence/criminal damage that will put all of us who have already been arrested and/or been a part of it in an extremely difficult situation where the public and our peers/friends/family will see us as advocating this behaviour and assume we share this attitude (participant 2).
Despite divergent views on whether breaking a glass window constituted violence, there was real disagreement on the authority for such acts among the movement's participants. Critics were particularly concerned that it was performed by a member of XR's leading team. They expected leaders to be more tactful and aware of the consequences of their actions for the rest of the group. They thought that leaders should not engage in any activity not supported by the main body of the movement or inconsistent with the organization's values: I'll admit I'm a bit troubled by the co-founder of a non-violent movement causing criminal damage, especially considering what we learn in NVDA training and after a week of 1000s upholding values of peaceful protest (participant 3).
I suppose a potential danger could be other less disciplined rebels getting carried away and not understanding the reasons/method... If this gets endorsed as a non-violent thing to do in XR, it surely needs to be done extremely carefully? I've never cracked a window though so not sure how easy/difficult it is to remain safe when doing so (participant 7). While XR's website does not refer to leaders as such, the movement's founders are clearly seen as occupying this position by both the membership and the media. This leads to a conundrum about how one becomes a leader in a leaderless organization, how such authority is brought about, for example, by becoming a founding and committed member (see , and what are the accountability structures, as explored next.

Event 2 -Canning Town tube action: Becoming an accountable leader in a leaderless organization
On 15 October 2019, an announcement on XR's website advertised a planned action to disrupt London Underground 2 days later. The activists planning the action introduced themselves as 'several XR AGs (autonomous civil disobedience groups) [who] are planning to non-violently disrupt Tube services to highlight the Climate and Ecological Emergency' (Extinction Rebellion, 2019). They proposed that the planned action would alert 'the public to the crisis [climate change] by disrupting their daily lives'. This, they argued, was necessary when 'There's a fire in the house!' However, the majority of XR activists were less than enthusiastic. Many were surprised that the event had been advertised on XR's official website, despite wide-ranging opposition from the main body of the movement, including the unanimous disapproval of XR's political circle. Several participants with whom we interacted were very concerned about the potential consequences, and it was clear from discussions on the public forum that most members had serious reservations. A range of issues was raised in the forum, from concerns for safety to the potentially damaging effect on XR's reputation and loss of public support. For example: I'm a bit worried about the tube action in rush hour. We're surely going to lose people's support if we start doing that. Attacking public transports it's a bit paradoxical as they are sort of greenish options people have to commute (participant 26).
Other public and private polls showed that 70 to 80% of members were against the action (Taylor, 2019b). In the midst of confusion about XR's official position, members learned in the early morning of 17 October 2019 that the tube action had indeed gone ahead in several locations and that in Canning Town it had led to a backlash from commuters. A video that went viral showed XR activists being pulled from the roof of a train and physically assaulted by members of the public who were frustrated about the disruption to their journeys. Ultimately, it was unclear who had endorsed the action.
The action and reactions to it triggered sombre debate among XR's members about this type of disruptive action. Beyond this, several members questioned the appropriateness of the movement's decentralized structure for its governance. Some defended it, suggesting that occasional setbacks are inevitable. They argued that the decentralized structure implies that different members may engage in actions with which not everyone always agrees: Although I am also not a fan of the action, I'd say it is up to us not to let it divide us. The bigger the movement will get, the more likely it is that there will be people/actions with whom/which we disagree. We all have our own limits but XR is a decentralised movement so we cannot impose our own views on others. Maybe we could try and take this action as a learning opportunity rather than a tipping point whichever way it goes? IMHO this would have at least triggered very healthy and needed conversations and arguments that are questioning who we are, how we work, our values, strategies, aims etc so at least that is positive (participant 9).
However, others were more sceptical. Some members questioned what being decentralized means in practice and raised questions about its practicality. Some went further, announcing that they could no longer support XR until a mechanism was established to safeguard against independent actions deemed unacceptable by the majority of members: There should be mechanisms to safeguard against actions that are dangerous to participants, bystanders & the reputation of XR. Until these are in place the movement will remain vulnerable to misrepresentation, extremism and violence. The PR damage done here alone is a significant setback & I personally am not sure I can continue to affiliate with a movement that operates without such agreements in place (participant 22).
There was recognition that the absence of explicit structures specifying decision-making processes and responsibilities was potentially harmful to the movement and might lead to departures from its principles, such as adhering to NVDA: I feel what's missing is some safeguard. We are decentralised but there should be a way to overrule if more than 70% are against. Or if several circles are unanimously against. Something that would only work in extreme disagreement like this. Bit of a shame that it doesn't happen naturally ... Decentralised only work if we consider each other's as equal (participant 12).
A few members drew attention to issues of access and equality within the movement, as shown in the previous quotation. For instance, several members questioned why the group of activists who had undertaken the unpopular action still had access to XR's website: The website is public domain & represents XR as a whole to many viewers. Using it to publicise actions that are so contentious within the movement misrepresents many others & feels problematic for me personally (participant 20).
It's a problem that the resources available (i.e. the website and possibly other things) are unbalanced in this situation... There's not the equality needed (participant 24).
The problem that AGs appear to be disrespecting the views of the majority is seriously compounded when those AGs have privileged access to official XR media channels. That privilege sh[oul]d bring responsibility to respect the movement (participant 25).
Others went further, suggesting that the incident showed that XR has a 'structural problem' and is 'undemocratic'. This allegation led to further discussion about the distribution of power and the role of leadership in the movement. Some associated the problem with members of XR's leadership, who had engaged in controversial and unpopular actions in the past: The problem starts with Roger Hallam and his daft drone stuff that he insisted on doing despite the majority being against it. Where's the democracy? (participant 30) I also struggled with Gail Bradbrook yielding a hammer and breaking glass earlier this week TBH (participant 33). This discussion revealed tensions underlying the organization's commitment to a decentralized structure and its implied leaderlessness, and a desire for effective and democratic leadership. Two other members suggested that the 'decentralized approach' is not only dysfunctional but also problematic in its own right as leaders are not accountable to the rest of the movement since they are not elected by them. They called for an official leadership structure in which leaders could be elected by members: This is why we should actually elect our leaders (participant 30).
I say this with love, but I've never been a fan of the decentralised approach … XR instituting some kind of direct democracy, and possibly a leadership structure, would be a great thing for the movement IMO (participant 34).
What is the point of demanding a citizens assembly if XR is so undemocratic (participant 16).
These members also highlighted that a decentralized structure implies that leaders can claim to be acting in their own capacity as independent members, rather than as members with a particular role and responsibility for XR as a whole. They questioned the notion of decentralized leadership or a 'leader-less' organization, arguing that, in effect, this makes those occupying leadership positions (e.g. founders) untouchable. They cannot be replaced because they do not hold any official position: The big problems come when the notion of decentralization is applied to the leadership. The national office can't be an affinity group that does whatever it wants. It has to be democratically accountable. Publicizing the tube action last night went against the majority view in XR and yet there isn't even a mechanism to re-elect the current leadership (participant 30).
The discussions surrounding the Canning Town tube action provided a temporary space that allowed several members to express their concerns about XR's leadership structure, many of whom had remained silent about the previous episode at the Department for Transport. Nevertheless, calls to reevaluate the leadership structure were gradually replaced with euphoria about a hopeful and better future: 'Love each other and love the earth, and let money be at the service of those two. Beautiful' (participant 47).
Some members suggested that the incident was an accidental 'fu ÃÃ up' and that XR's existing mechanisms (e.g. mandate and polling) could solve the problem if utilized effectively in future. Others tried to assuage group concerns by suggesting that the movement is not perfect so some mistakes are inevitable. This was evident in the responses of two vocal participants (participants 1 and 7) who had clearly supported the movement's co-founder in the previous incident: Frankly, it's impressive that we've got a full year without this happening before (participant 1).
Absolutely. Solidarity with the bravery and sacrifice of rebels this morning, however misplaced we think their action might have been. Let's acknowledge this tension in our decision-making structure, and work together to do better next time (participant 7).
During our exposure to the movement (July-November 2019) both before and after the events described above, we noticed that members continued to embrace the notion of a 'leaderless' organization. They appeared inspired by the movement's decentralized structure and lack of hierarchical leadership. For instance, one member explained, 'I felt invigorated. You know it looked like you could take any action you want! But reflecting back ultimately we did not do anything!' (participant 48).
Members' socialization in the movement, through orientation meetings, training sessions (e.g. 'self-organizing systems workshops') and interactions in the general meetings, also reinforced their attachment to the notion of a decentralized and leaderless structure across the movement. For instance, in the lead-up to the October rebellion, which clearly involved some centralized decision making, vocal members insisted that other members should not talk about XR Headquarters or XR Centre since XR is a decentralized, leaderless organization and has no headquarters. In the next section, we discuss this 'inconsistency' through the lens of a proposed psychoanalytic Lacanian framework as we outline our theoretical contributions and emerging practical implications.

Discussion
In this study, we examine the tension between formal rejection of leadership in a leaderless organization and members' desire for accountable and democratic leadership. This contradiction came to the fore following controversial actions unsupported by the majority. Faced with the reality of decision making in a broad social movement, some members began to question the group's attachment to the ideals and implications of a leaderless organization. In particular, they underscored ways in which power was being practised and simultaneously disavowed. For instance, while the Canning Town tube action was seen as an autonomous action by certain AGs, these members questioned why, unlike many ordinary members of the local group, these 'autonomous groups' had privileged access to XR's website. They felt that this inner group was enacting leadership without being held accountable for its actions.
Arguably, neither action represented 'what decentralization looks like', as a fervent advocate of XR's leaderless model argued, but instead showed how a self-proclaimed leaderless (decentralized) social movement can be captured by leadership by stealth (Sutherland et al., 2014), when actions designed by small groups of members at different levels shape the movement's direction. For instance, while a large proportion of XR activists questioned Gail Bradbrook's breaking of the Department of Transport's window, our observations indicate that this action is now being used in XR's NVDA training as an example of how window breaking can be used as a form of non-violent civil disobedience.
To interpret this apparent contradiction, we turn to Lacanian psychoanalysis. However, rather than focussing on the notion of lack to understand the impossibility of leadership (Driver, 2013), or on imaginary illusions (Western, 2014), we propose the notion of fantasy as a necessary condition for imagining different modes of organizing. This appears to root people's desire for democratic and egalitarian leadership, and alleviate future anxieties, with the risk of disabling what it seeks to achieve. New imaginary significations are important for creating different meanings and modes of understanding that question the entire socio-political order, as philosopher Straume (2020) argues in the case of climate change. While horizontal, networked, leaderless movements may successfully initiate such developments by producing radically new significations and meanings, they are much less successful in bringing about the sought-after change. Put differently, leaderless social movements are an effective vehicle for bringing into view important but neglected issues through mass protests, but less effective for policy changes (Mongiello, 2016). This, we suggest, may be because they often ignore issues of power, and because they must go beyond desiring democratic governance to systematically implementing and continuously practising it.
In this study, we provide an account of how the leaderlessness discourse in organized social protest movements results from a desire for idealized egalitarian power-sharing arrangements that are not always or often actively pursued in everyday organizational practice. Leaderlessness is thus evoked as a way to highlight and exorcise abuses of power and corruption. In this sense, it is an expression of oppositional and contentious politics (Tarrow, 2011), but also enables the imagining of different possibilities for more emancipatory and democratic organization.
Our central contribution is to stress the role of fantasy, which seeks to overcome the incompleteness and fragmentation in the human psyche, as both a productive activity and a limiting condition. Because of its capacity to envision alternatives, fantasy is indispensable in struggling towards a better future when the present is no longer tenable (Kenny et al., 2020), but it may become counterproductive when it ignores real power dynamics and conflicts. As Costas andTaheri (2012: 1211) argue with reference to authentic leadership, 'its categorical emphasis on love, completeness and harmony as a way to replace symbolic authority leaves subjects in a paranoid dependency on an irrevocable fantasy figure akin to the Freudian primal father'. Fantasy may be truly destructive when attachment to the very idea of leaderlessness or followership (Ford and Harding, 2018;Learmonth and Morrell, 2017) takes over organizational reality and becomes an end in itself, as our research shows. Thus, this study's first contribution is to psychoanalytic studies of leadership in organization studies. This novel, and in our view much-needed, input builds on previous elaborations on this topic by other authors (Costas and Taheri, 2012;Western, 2014) engaging with the emancipatory potential of non-hierarchical forms of organizing by putting power dynamics at its centre. While their journeys into the unconscious have significantly advanced our understanding of relations between leaders and followers, their main preoccupation has been with explaining the vicissitudes and perils of leadership originating in the misuse of authority and pathologies of leadership resulting from projections and re-enactions of one's own past (Gabriel, 1999;Kets De Vries, 1991;Stein, 2007). These have been examined in relation to the business sector, but with the exception of Simon Western's recent work, have rarely, if at all, been concerned with social movements. Drawing on Lacanian theory, Western (2014: 675) argues that discarding traditional leadership creates a lack and leaves a gap: Utilizing the signifying term leaderless to celebrate the absence of leadership does not fill this gap, but temporarily covers it over. For a short time the anxiety and desire created by the lack is displaced by the temporary enjoyment gained through the identification of being 'free protestors'. However, this enjoyment is short lived, for to fill the gap will take new and innovative forms of leadership, not the celebration of its absence.
Our second and related contribution is, to elaborate on the intra-psychic processes through which power operates in organizations, which we examine in the context of decentralized protest movements. The intra-psychic dimension concerns the constant interplay between destructive and life-affirming dynamics that Freud called death and love drives (Thanatos and Eros). These manifest themselves intersubjectively in the social domain vis-a-vis others as we form groups and engage in organized action. Lacan stresses the importance of symbolic authority in disallowing fantasmatic strivings from taking over these endeavours 'by forbidding relentless efforts to complete the big Other via imaginary simulacra (fantasy, identification)' (Costas andTaheri, 2012: 1200). This understanding helps us elucidate what happens in a movement when formal hierarchy and traditional leadership are removed without explicitly acknowledging the need to direct organizational actions purposefully. While the leaderless leadership implies the absence of formally designated leaders, this does not mean there is no leadership in such organizations as our findings demonstrate. We suggest that the attachment to the illusory fantasy of 'leaderlessness' disables the alternative organizations' capacity for becoming 'leaderfull' without succumbing to the defunct hierarchical models of leadership. Our study also reveals how such psychic dynamics may hinder social movements from tackling their dysfunctional attachment to fantasmatic ideals that remain in the realm of desires without being embedded in organizational practices.
Interestingly, however, the fantasy of a decentralized, leaderless movement was not unanimously shared in the case study presented here. Some members seemed to be more strongly attached to this fantasy than others. Tension between the idea of a leaderless organization and enactment of leadership became clearer when XR's leaders and members pursued actions that were considered more divisive. In the hours and days following the two disruptive actions discussed earlier, questions were raised about who makes decisions that impact the whole movement, the role of leaders in sanctioning, blessing or hindering an action and whether leaders should follow the majority's point of view or should inspire the main body of the movement to adopt new directions. The reality of decision making and its outcomes in an assumed leaderless context revealed, at least temporarily, both the unattainability of a decentralized organization free from hierarchy and power and its deficiencies and shortcomings in sustaining collective action. Temporary realization that the object of fantasy is unattainable may cause frustration and anxiety (Glynos, 2008). This was evident in the public forum discussion, when several XR members, particularly perhaps those with stronger attachments to the ideal of the leaderless organization, attempted to neutralize tensions and anxieties within the movement by inviting other members to accept these tensions as the imperfections of a decentralized organization. However, this may also impede more thorough re-evaluation of the ideals of the leaderless organization and realization of inclusive forms of leading/organizing a movement, such as task-based temporary forms of leadership.
Affective attachment to the identity of being leaderless is often a barrier to the development and agency of social movements (Western, 2014). On the one hand, it neglects alternative, potentially empowering forms of organizing, such as autonomist leadership. On the other hand, it may lead to leadership by stealth (Sutherland et al., 2014) by a small inner group because power dynamics are always present as part of intra-and inter-psychic mechanisms. This suggests that leaderlessness is an aspirational idea that helps mobilize individuals striving for non-authoritarian forms of organizing only when tensions between power, representation, accountability and purposeful action are recognized. Neglecting these important factors and dynamics could also lead to adverse political outcomes. One of many adverse outcomes of the 'lack of leadership' evident in the Arab Spring uprising, for instance, meant that the protesters ended up with a detrimental political reality they were fighting against in the first place (e.g. Muslim Brotherhood taking over and then, ultimately, the establishment of a military dictatorship). 1 In contrast, the anti-austerity protest movement Podemos originating in Spain has led to the formation of a new political party which is now a minor coalition partner though its ability to influence developments is limited 'but at least it presents new possibilities' (Iglesias, 2020). In other words, we suggest that anti-authoritarian organizations must embed the desire for different ways of enacting politics in their everyday practices, rather than being merely enamoured with it.
Acknowledging the psychic conflicts and dynamics involved in such processes provides muchneeded insights into what drives leaderless social movements. Only when such ever-present tensions and conflicts are transparently acknowledged can decentralized and anti-authoritarian modes of governance become developmental rather than regressive. On a practical level, 'This means working through difficult challenges such as addressing power dynamics and struggling for meaning and ideas, which contrasts with the regressive dynamic of imposing constraints on themselves in regard to taking up leadership in order to maintain the fantasy of being "leaderless"' (Western, 2014: 676). Another way of avoiding this is to recognize the emergent forms of empowering leadership, while correctively acting to curtail the imposition of conventional and regressive top-down forms of leadership. Various strands of anarchist and autonomist thought have proposed and implemented schemes in which leaders do not hold power over others and any formalization of a leadership position is temporary, open to recall and dissolvable at any time. Feminist theory of democratic participation originating in embodiment and connectedness to everyday experiences (Fotaki and Daskalaki, 2020;Vachhani, 2020) can provide another way forward. Simultaneously, we must also stress the importance of authority to contain movements' autodestructive tendencies and leadership stealth (Sutherland et al., 2014) as we highlight how existing power dynamics are obscured by the leaderless discourse.
To sum up, we draw on the notion of fantasy described above, to suggest that power flows are influenced by psychic dynamics, which are ridden with the power struggles within us and their social expressions as we relate to others. We argue for acknowledging the productive role of oppositional protests for opening new possibilities and spaces for alternative significations while stressing the importance of sustained work for building alternative emancipatory and leaderfull organizational practices. This can also help us envisage alternatives and solutions to social issues of inequality and discrimination, while also speaking to theorization of the connection between ethics, dissent and democracy (Vachhani, 2020).

Conclusion
In this article, we problematize the notion of the 'leaderless' organization adopted by many recent social movements, including XR. We acknowledge that social media have changed how social movements are organized, relying on spontaneous and decentralized mobilization. Nonetheless, our research underscores ways in which power is both disavowed and practised in decentralized social movements. We attribute this to attachment to the fantasy of leaderlessness, which could take over the organizational reality and become an end in itself. Our findings show that this may impede more thorough re-evaluation of the ideals of the leaderless organization and realization of inclusive forms of leading/organizing a movement. At the same time, we acknowledge the productive role of fantasy, which sparks action and brings social projects to life (see Dey and Mason, 2017;Kenny et al., 2020). We stress the key role of leadership in this context, not because of its mystery and charisma, which are often ascribed to exceptional individuals, but because all organizations require the capacity to fulfil the objectives that are their raison d'etre. The presented research demonstrates there is rather a desire for leaderfull organization rather than a leaderless leadership. In our view, this means building leaderfull organizations that acknowledge the power dynamics within the movement and vis-a-vis the external environment, and work towards establishing more egalitarian and participatory ways of sharing and distributing roles and responsibilities. In so doing, we highlight the potential of oppositional protests for envisioning alternatives that become part of the political solution that social movements strive to achieve.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.