Naked Bodies and Collective Action: Repertoires of Protest in Uganda’s Militarised, Authoritarian Regime

ABSTRACT How can citizens living under increasingly militarized and authoritarian regimes exercise political voice? Using an in-depth case study of naked protest in modern day Uganda, this article finds that naked bodies allow citizens to employ three types of overlapping power to confront a militarized authoritarian state: biopower, symbolic power, and cosmological power. The study illustrates one way in which citizens seek to engage militarized regimes—and in doing so, how political voice takes particular forms with limited capacity to instigate broader political claim-making that might be associated with country- or region-wide political action.


Introduction
Over the last few decades, military forces have adopted an increasingly prominent position in many African governments, playing a key role in wider statebuilding processes and social transformations (Day, Khisa, and Reno forthcoming). A notable outcome is the militarisation of not just state institutions, but also society and the economywith profound implications for the expression of political voice. Indeed, authoritarian rulers have long leveraged military force to limit collective political action; and civil society actors have responded with a variety of creative approaches. Recent scholarship, for example, has examined how citizens leverage new technologies to pursue social and political organisation, open avenues for collective action, and protect themselves against unaccountable and violent state repression (Tufekci 2017). However, authoritarian regimes have adapted quickly, using these technologies to surveil their populations, tightening control over social media, andshould those means failreverting to violent repression (Roberts 2018, Gohdes forthcoming). How can citizens living under increasingly militarised state institutions exercise political voice? To answer this question, this article focuses on an ancient form of protest: the baring of naked bodies. Public nakedness has been documented as a means of protest worldwide, in vastly different cultural contexts, from India (Misri 2011) to Canada (Makarova 2013), and in colonial settings to modern-day. This longstanding approach to challenge authority has remained marginal to studies of popular protestand although it may be used infrequently, its conditions of possibility tell us about the available avenues for collective opposition to violent and repressive rule.
This article offers an in-depth study of collective naked protest under one such regime: modern-day Uganda. The article examines a case in which Ugandan citizens used public nakedness as an alternative and complementary repertoire of protest when interfacing with the militarised Ugandan state. 1 The article finds that naked bodies allow citizens to employ three types of overlapping and entwined power to confront the much more powerful state in this militarised public sphere: biopower, symbolic power and cosmological power. The study illustrates one way in which citizens seek to engage militarised regimesand in doing so, how political voice is pushed into particular forms with limited capacity to instigate broader political claimmaking that might be associated with country-or region-wide political action.
The article first introduces the case of Uganda, briefly outlining how the regime has militarised both state and society, curtailing citizens' ability to exercise political voice. It then outlines the argument in relation to existing literature on public protest and collective action under authoritarian militarised regimes. The article focuses on the collective public use of nakedness as a form of protest, recognising that this may be different fromwhile still related toindividual uses. The third section offers a case study of one such naked protest in Uganda, and examines the three types of power in action. The article concludes with a broader discussion of what naked protest can tell us about prospects for collective action under such repressive regimes, including why and how militarisation of the state shapes the exercise of political voice.
The article is based on fieldwork conducted in 2018 on naked protest in northern Uganda using focus group and individual interviews, as well as the combined experiences of the four co-authors, who include two Acholi Ugandans who have worked as researchers in the north for a cumulative 18 years, and two foreign academics from the fields of anthropology and political science, with specialisations in gender and security, respectively. The insights from Kerali and Abonga, along with the ethnographic work (conducted by Porter since 2008) and qualitative fieldwork (conducted by Tapscott since 2014) allowed us to set local attitudes and interactions expressed during the focus group discussions and interviews in their broader historical, cosmological, and political context. An ethnographic approach, as Jackson observes, demands 'not merely an imaginative participation in the life of the other, but a practical and social involvement in the various activities, both ritual and mundane, that contextualise and condition the other's worldview' (Jackson 2002, p. 264). Indeed, many of the themes which emerged had been observed in people's lives over years and decades. Additionally, being able to pursue and discuss insights from the qualitative research with people with whom we have longstanding relationships fostered open communication and trust and gave us insight into the evolving narratives of those involved. The triangulation of methods and data further allowed us to crosscheck our ideas and interpretations.
This article thus combines interdisciplinary, foreign, and local perspectives to the lived experiences of ordinary Ugandans. Much of the empirical material is based on the context of naked protest amongst the Acholi people of northern Uganda. This region bears the scars of a 20-year-long civil war, which inevitably continues to shape residents' views of the ruling regime and contribute to tensions around civic representation and rights. However, similar instances of naked protest have occurred across the country, indicating that this type of protest reflects broader dynamics between the state and civil society that are not limited to the north. Our ethnographic interpretive approach combined with other qualitative material thus benefits from observing social processes unfold over time, it is attuned to silences, actions and gestures and is not overly or exclusively reliant on narrative evidence (Clifford and Marcus 1986). This better enables us to analyse contextual associations withand meanings ofnaked protest that can offer insight into naked protest as a phenomenon in Uganda, while the overall findings can help understand the nature of political voice and dynamics of claim-making in militarised regimes more generally.

Civil-Military Relations in Contemporary Uganda
Uganda is in many ways a paradigmatically militarised state. In 1986, the National Resistance Army led by now-President Yoweri Museveni overthrew the government and took over the country. In the subsequent years, the National Resistance Movement (NRM) regime has merged state institutions with NRM party and military power, producing a kind of governing assemblage that serves the interests of the ruling regime. This arrangement finds its legacies in colonial intervention, when state security organs were primarily used to pacify the population in pursuit of economic development (Oloka-Onyango 1990, p. 7). Museveni has further cultivated this approach over the past three decades, overseeing the militarisation of state security forces, including the police, as well as state institutions (Kagoro 2015). Today, soldiers hold important positions in the legislature and executive, and the government continues to prosecute civilians in military court despite the condemnation of human rights organisations (Burnett 2011, Kafeero 2019. In the security sector, Museveni's divide and rule strategy has produced an 'alphabet soup' of formal and informal organisations (Mwenda 2007, p. 32-33). While these organisations are nominally distinct, they often are deployed interchangeably to police civil society, limiting their accountability. Moreover, individuals frequently move amongst the armed services, both overtly and covertly. As a result, many Ugandans express the view that the different branches of state security are interchangeable elements working in service of Museveni and the NRM regime, in spite of their different coloured uniforms.
Under the NRM regime, society has also been militarised. For the first 17 years of his rule, Museveni implemented a 'no party' system. While purportedly fostering meritocratic political competition untainted by sectarianism, in practice, the NRM conducted itself as a political party while outlawing all others. The arrangement meant that Ugandans were de facto considered part of the NRM, linking Ugandan citizenship to the NRM party (Carbone 2008). For years under Museveni, government appointees and elected officials, along with ordinary citizens, were encouraged to participate in a security training called chaka mchaka to become NRM cadres (Lanken Verma 2012). Much of the state's neopatrimonial system of resource distribution occurs through security organs. For example, before the 2016 elections, the regime recruited tens of thousands of 'Crime Preventers'nominally community police organised to 'help with the elections'who were promised and received benefits via the police and Ministry of Internal Affairs such as bicycles, motorcycles, low-interest loans, and short-term employment (Tapscott 2016, Kagoro 2019. The regime also recently placed its agricultural extension programme and its national ID project under the auspices of the military, and has appointed military officers throughout many of the country's ministries and agencies, including to the State House Anti-Corruption Unit; the Department of Citizenship and Immigration; and as liaisons to the Uganda National Roads Authority and the Uganda Revenue Authority (Kafeero 2019).
The NRM regime has been described as embracing a type of 'military masculinity' in which hegemonic and military masculinity is prized above all others (Dolan 2011, Tapscott 2018. This version of masculinity is based on competitive performance in a hierarchical structure, in which manhood is associated with warriorhood, obedience, and discipline. Kagoro has described a 'warriorised field of power' in Uganda, in which citizens are accustomed to a melding of the military and politics, wherein they attach a high sociopolitical value to the military and 'its associated ethos such as uniform, gun, rank, training, and participation in war appear to have become a source of symbolic capital' (Kagoro 2015, p. 123). Thus, power is equated with the military and an ideal-type masculinity. The military masculinity that pervades the public sphere simultaneously restricts the forms of public protest that resonate with the public, and makes naked bodies particularly poignant, as shall be illustrated further in this article.
The NRM regime has long attempted to control public gatherings. In 2008, Opposition MP Alice Alaso noted in Parliament that the heavy hand of Uganda's security services, and extensive militarisation of public spaces, effectively suppressed political voice: [I]t is practically impossible for . . . any political leader . . . to address or consult with his or her people without permission from the Inspector General of Police . . . you effectively have Police as state machinery watching over you all the time . . . as people in the Opposition we have no idea whether we will always be granted permission to hold rallies or to freely assemble and demonstrate because we do not know what the Inspector General of Police will think next time. All we know is that he only unleashes teargas. (Uganda Hansard 2008) As Alaso notes, the police unpredictably approved some rallies and banned others. Many attempts to organise publicly are intercepted and dispersed. The regime frequently paints its detractorsand particularly those that seek to organise collective political protestsas terrorists, criminals, or bad elements. Discussing ongoing political protests in the aftermath of the 2011 presidential elections, the former Kampala Metropolitan Police boss, Andrew Kaweesi said, 'The rallies have become concerts for political opponents to incite violence, name call, abuse the President and instigate a sense of hatred among the public based on sectarian tendencies' (Anguyo and Tatyamisa 2013).
Public protests are commonly shut down when the police arrive in riot gear and deploy teargas, and either rubber or live bullets. Teargas has been 'a major tool of keeping demonstrators at bay, countrywide' (Mubiru and Masaba 2018). In 2018, the Uganda Police Force reportedly allocated 44 billion shillings (approximately 12 million USD) to purchase teargas, becoming so common that it is practically satirised: 'Tear gas' is now an everyday term in Kampala . . . Residents exchange advice on prevailing winds prior to planning their journeys through town . . . And Uganda's Finance Minister even cited the government's decision to import rather than manufacture tear gas as a reason for the poor performance of the Ugandan shilling. (Monteith 2016) An illustrative example of how the regime handles public protest can be found in the Walk-to-Work protests in 2011. Opposition politicians organised Ugandans to walk to work, rather than take motorised forms of transportation, to protest the hyperinflation of food and gas prices, which many believed resulted from the irresponsible fiscal choices by the NRM regime (Perrot 2014). Hundreds of protesters were arrested and detained for their participation in Walk-to-Work, many sustained injuries, at least nine were killed by state security personnel, and the major opposition figure, Kizza Besigye, was shot in the hand with a rubber bullet and later pepper sprayed (Perrot 2014, p. 426).
Opposition leaders, arrested on charges of treason and sedition, were held at Luzira Prison for 2 months before receiving bail (Lukwago 2011). Perrot notes that these protests may have been particularly threatening to the newly reelected regime because they initiated a 'transformation of post-electoral modes of protest by efficiently developing mobilization through direct action and civil disobedience' (Perrot 2014, p. 427).
In response to this popular protest, the regime introduced several new measures, chief amongst them the Public Order Management Act (POMA), which prohibits public gatherings without notifying the police 3-15 days beforehand. Before, during, and after the 2016 elections, the police used POMA to justify violent dispersal of events organised by opposition political parties, as well as the arrest and detention of participants (Human Rights Watch 2015). Presidential candidates were required to travel with a convoy of heavily armed police officers, which is commonly understood as a means for the government to surveil their whereabouts and activities, rather than for their protection. On Election Day in 2016, social media was shut down and cell phones were nominally banned from polling stations. Museveni explained that these measures were designed to prevent rumour mongering; while others said they were intended to prevent reporting on election fraud and to stymie opposition organisation on polling day. 2 During the President's inauguration in May 2016, media were banned from covering protests (Paul 2016). These regulations and their violent enforcement are a mode of intimidation, making police and military violence more difficult to document and report.
The unpredictable use of unaccountable and excessive force tips civilmilitary relations in favour of the militarised and masculinised state (see also Tapscott 2018). In such an environment, public protest is extremely riskycitizens know that they could easily suffer physical harm and be brought before the court as a criminal. As noted by Branch and Mampilly, in addition to a socially and politically fragmented polity, 'protest politics [in Uganda] would have to deal with a highly militarised regime, one that depended upon a militarised police force and, in the extreme, upon the military itself to maintain ordera regime that did not hesitate to use violence against the population' (Branch and Mampilly 2015, p. 122). Indeed, state authorities such as the police and military, working in tandem with party stalwarts, can repeatedly illustrate their ability to crush individual and collective claims that civilians attempt to make on state authorities, thereby offering the regime a platform to showcase its strength compared to ordinary civilians. This means that collective public action can produce the perhaps counterintuitive effect of reinforcing state authority by highlighting the regime's access to overwhelming force, its willingness to deploy it against ordinary citizens, and the absence of accountability for resultant loss of life, damage to property, or physical harms.

Repertoires of Collective Protest under a Militarised Regime
It is perhaps intuitive that political voice is one of the first things to suffer under militarised and authoritarian regimes. Scholars have long documented how authoritarian regimes use unpredictable and excessive violence to produce fearful citizens, unable to freely exercise political voice (Greenblatt 1980, Wedeen 1999. Protest in such a scenario is extremely challenging, though it has much to tell us about regime type and statesociety relations (Tilly 2010). In their recent book on protest in Africa, Branch and Mampilly note that most protests fail to catalyse political change, in part because of the prevalence of state violence. It is difficult to overcome state violence while maintaining non-violent and inclusive strategies. Moreover, the potential for state violence can make collective political action a 'near-suicidal strategy' (Branch and Mampilly 2015, p. 9). Authoritarian regimes repress voice through the exercise of violence, but also through restrictive policies like Public Order Management Acts, public media campaigns that paint members of the opposition as rebels, terrorists, or enemies of the state, and the use of surveillance and threats to intimidate opposition. 3 In response to such oppression, scholars have identified alternative registersor 'repertoires', following Tilly (2010)that citizens employ to express political voice, such as humour (Wedeen 1999), art (Chaffee 1993, Adams 2002, or rudeness (Summers 2006). Building on a long history theorising the practice and effects of non-violent collective action, recent scholarship has noted that non-violent means have historically been twice as effective at instigating political change as violence for several reasons: non-violent action bolsters the legitimacy of the protesters compared with their target, and makes it more difficult for the target to justify counter-attacks (Stephan and Chenoweth 2008). Moreover, protest and other forms of civil disobedience can reveal that the regime is unable to maintain order. As Zunes explains, this is particularly risky for authoritarian systems that often derive legitimacy from their ability to maintain order. When protests and acts of civil disobedience reveal that the regime has only partial control, erstwhile supporters often seek alternative leadership (Zunes 1994, p. 404). Thus, non-violent forms of protest are potentially useful in the face of a militarised state, which regularly employs violence to suppress opposition. 4 The Naked Body as a Site of Resistance in Uganda and Beyond: Bio, Symbolic and Cosmological Power Scholars have studied naked protest as an approach that draws on norms ranging from Judeo-Christian notions of sacrifice (Barcan 2002), a shared humanity (Sutton 2007), or using the object of dispute as the vehicle of protest (Tyler 2013). Ugandan scholar Sylvia Tamale notes the potential of naked protest to achieve emancipatory ends, writing that it is 'always effective in that it draws attention to the issue under dispute . . . [and can] undermine[] the foundations of the hegemony of repressive regimes' (Tamale 2016, p. 7). Naked protest has been used widely in Uganda, often in relation to land. For example, in 2012, women in Amuru District stripped to protest the potential sale of land to Madhvani Group (Lawino 2012); on several occasions in 2014 and 2015, women in Soroti stripped to protest Soroti University's encroachment onto their land (Onyango and Emwamu 2015); and in 2018, women in Bulambuli District stripped to protest government-planned relocation after a landslide (Kirinya 2018). In our interviews, respondents reported non-land-related protests as well, for example during the LRA war, 5 when Acholi women stripped to demand that leaders end the conflict; or in 2012 when women in Kampala stripped to their bras to protest police harassment (NTV 2012). We argue that collective naked protest allows citizens to project three types of overlapping and entwined power to challenge the often overwhelming force of the state in this highly militarised public sphere: biopower; symbolic power and cosmological power. Before describing each type of power, we examine how nakedness is used in moments of protest amongst the Acholi people of northern Uganda. Focusing on northern Uganda allows us to elaborate context-specific details that help elucidate the types of power, allowing us to develop a framework that can help think about naked protest in militarised contexts more broadly.
In contemporary Acholi culture, it is believed that public nakedness is 'gwok', often roughly translated as 'obscene'. Gwok refers to an occurrence that is incongruent with the normal order of thingswhether social, natural, or cosmological. Such an 'out of place occurrence' can be itself a disturbance of order, or an indication of such, and if unrectified, may prove dangerous (Victor 2018, p. 68). Intentional public nakedness, accompanied with curses and gestures directed at a perceived offender, creates a powerful curse that causes misfortune and even death. While the nakedness of adult men and women can produce this curse, for reasons discussed below, the potential power of the exposed female body is often seen as more powerful (and common) in acts of protest. 6 Amongst Acholi, nakedness in social conflict occurs mainly at an individual-level rather than collectively, and entails the baring of a breast or genitalia (male or female) and pointing it at the offending party while cursing them. Such acts are typically associated with the elderly, who often direct naked curses at a younger person to whom the elder is related or responsible for in some way, while uttering a curse, such as: 'I am the one who gave you life, if you are the one who stole this money then you will continue to steal your entire life and you will not be able to stop'. The most immediate purpose of the curse in such cases is to sever relationships and punish or harm the person to whom the curse is directed.
Such occurrences are not new. Naked curses, for example, appear in anthropological work of the 1950s and 60s (Hirsch-Foster 1954, Girling 1960. In interviews, all of our informants recalled instances when they had either been cursed themselves or had witnessed naked cursing in their families or amongst their neighbours. They noted the subsequent downfalls, calamities, sicknesses, and sometimes deaths of the people who were cursed. Respondents agreed that the naked curse is generally not intended to change the actions of its recipient, as it is cataclysmic when used; nonetheless, the curse has the potential to effect long-term social relations. For instance, one man reported that he had caused the death of his own son by a naked curse. 7 He explained: 'There is nothing that he [my son] can do when he is dead, but it [the curse and death] is a lesson to the other sons that if we elders tell you something, you must obey or face consequences' (Authors' interview, Gulu, August 2018). Further illustrating this view, many criticised a well-known academic when she stripped publicly to protest her working conditions. 'She had other means . . . she has dignity and is educated, she could have fought legally' (Authors' interview, Gulu, August 2018, see also Tamale 2016). In many of the instances that were recounted as examples of interpersonal use of naked curses, the gerontocratic authority of elders was being ignored or had often been eroded by circumstance, and the naked curse was deployed to remind the younger generation that elders have an irrevocable power that youth ignore at their peril. In enacting the curse, the elders reasserted authority in a social order gone awry.
While individual and interpersonal uses of naked protest differ from collective uses, which we discuss in the remainder of the article, they offer important insights into the idioms of nakedness. It appears that collective uses of nakedness draw their power in part from the reference to the more familiar and common individual uses. In focusing on an act of collective naked protest used against the Ugandan regime, this article seeks to elaborate how citizens employ this repertoire of protest to make claims on an often violent and unaccountable regime. It elaborates three types of power that give naked protest meaning: (1) Biopower: The body is a key locus of governance, and thus an important resource for protest or disobedience. Indeed, as Zunes (1994) notes, at the heart of governance is the control of bodies. The practice of biopolitics is the control of bodies at population-scale, producing legible populations that selfpolice and self-mould in response to public policies (Foucault 2010). In this sense, sovereign power is produced through the bodies of the governed population. Therefore, refusing to be a disciplined member of society can be a deep challenge to sovereign authority. This logic helps explain certain types of protest that appear to employ anti-social uses of the body, such as hunger strikes, dirty protests, or naked protest. Bodies offer a democratic and practical site of protest particularly for the marginalised, as everyone has access to one. Indeed, Cahill asks that we 'understand the body as a central site for both the production of power and the possibility of resistance' (Cahill 2001, p. 4).
In a biopolitical register, the power of naked protest derives from the body as a political object, manipulated and used to subvert political ends of the sovereign. Purposeful public nakednesswhich reveals bodies as defenceless, unprotected, and uncontrollablecommunicates an underlying message of the state's failure to control its subjects, and the breakdown of selfpolicing or self-discipline. As Hansen and Stepputat explain, state power is realised in opposition to bare bodies, upon which it can inscribe itself. The body is then transformed into 'an object of collective projections of the plebeian crowd whose presence was essential to these performances of sovereignty' (Hansen and Stepputat 2001, p. 11). The voluntary destruction of, or rendering vulnerable, one's own body can thus have a transformative effect, consolidating leadership amongst protesters. Nonetheless, the power of this may be limited. While subjects can use bodily resistance to challenge the sovereign's arbitrary power, Mbembe shows how, in the postcolony, such actionsparticularly those that draw on obscenitycan reproduce the logic of that very power. Mbembe adds that Strictly speaking, this process does not increase either the depth of people's subordination or their levels of resistance; it simply produces a situation of disempowerment [impouvoir] for both the ruled and the rulers . . . Though it may demystify the commandement [sovereign's command] or even erode its supposed legitimacy, it does not do violence to the commandement's material base. At best, it creates pockets of indiscipline on which the commandement may stub its toe, though otherwise it glides unperturbed over them. (Mbembe 1992, p. 15) Bodily protest is thus a viable means to trouble potentially overwhelming and arbitrary power. Yet, in its usage, it can also reinscribe the very logic of that power, thereby making it simultaneously an act of power and futility.
The works of Foucault and Mbembe reveal that acts of naked protest are biopolitical encounters between the subject and the sovereign. However, as illustrated in the remainder of this article, these acts can also be viewed as an expression of political ends that, in some instances, challenge the sovereign. Thus, while the body is evidently a site of governance and biopower, it is also important to consider other registers through which bodily protest exerts powerin this case, through symbolic and cosmological power.
(2) Symbolic power: By the symbolic power of naked protest, we mean power that derives from the body as a gendered symbol. Here we turn to Butler, for whom the body is inscribed with, and in turn iterates, socially produced gendered scripts. It is this co-productive dynamic that makes the naked body such a powerful symbolic tool. Because bodies are produced as 'male' or 'female' through certain performative acts, acts that fail to adhere to the resultant gendered order are disturbing, putting into relief and troubling the constructed nature of social order (Butler 1990). This helps explain why naked protest is gwok and a powerful tool to create and express social unrest. As Tamale notes, the embodied female subject is inscribed with a gendered script of 'subordination, passivity, and sexuality'; her nakedness is then 'read by the public as shameful and embarrassing' (Tamale 2016, p. vi, p. 29), lending naked female bodies symbolic power in protest (Barcan 2002, Sutton 2007, Tyler 2013, Tamale 2016, Ebila and Tripp 2017. The gendered scripts disrupted by naked bodies include localised beliefsfor example, in Uganda, that mothers can produce life and thus can take it away, thereby suggesting that the targets of the performative demonstrations of nakedness are 'dead to society' (Tripp cited in Guyson 2016). This helps explain why the nakedness of womenespecially of elderly motherscan be particularly shocking, as even more so than that of young women and men it is meant to be covered at all times (Tamale 2016).
This symbolism of biological and social reproduction as well as potential inherent power to give and take away life were salient amongst our respondents. It is also noteworthy that naked protest in Ugandawhether individual or collectiveis more common amongst mothers who are no longer of reproductive age. There is an assumption that their bodies will no longer elicit sexual desire from onlookers, and thus their nakedness is sheltered from accusations of sexual impropriety. Elderly mothers that employ the symbolic power of their bodies in protest thereby pull the terms of debate for such protests onto moral ground. The state is doubly at a loss here. First, it undermines the Ugandan state's assumed political and economic registers (Scott 1976, Tapscott 2016. Second, it restricts the state's ability to deploy violence, since embedded in the very masculinity of the militarised state is the widely shared belief that it is not palatable to use violence against vulnerable female (maternal) bodies (Zunes 1994, Chenoweth andCunningham 2013, Boyd forthcoming).
(3) Cosmological power: The power of naked protest derives from the cosmological register of the naked 'curse'. We use cosmological to refer to culturally shared understandings of forces in the universe (sacred and profane), within a dynamic worldview inhabited by both seen and unseen purposive powers that permeate the cosmos (Abramson and Holbraad 2014). Scholarship on cosmological power (including witchcraft and sorcery) suggests that the personified use of such power in many contexts is understood to offer explanations for general misfortune (Evans-Pritchard 1976). Far from stagnant, these practices are often adapted to interpret and respond to modern social threats (Allen 2007, Allen andReid 2015).
Writing in particular about witchcraft as one medium of wielding such power, Stewart and Strathern note that it is intimately associated with the body as a relational object: Witchcraft is seen as a power belonging to persons through their bodies or spirits, giving them an ability . . . to kill, harm, or inflict sickness on those whom they intend to weaken. Characteristically, the witch is seen as a kind of cannibal, eating the victim's life-force as a way of self-augmentation. (Stewart and Strathen 2004, p. 6) The power of witchcraft rests in the body, weakening others through its exercise, thereby building on the relational aspects of symbolic power discussed above. Though the naked curse is not viewed as witchcraft nor its maker as a witch, its power analogously draws on embodied cosmological power and the intentions of the person who uses it. As noted by Finnström and p'Bitek, 'in Acholi thought, powers to heal are also potential powers to harm, depending on shifting contexts . . ., the blessing is one side of the famous coin, of which the other is the curse' (Finnström 2008, p. 209; following p 'Bitek 1971, p. 146; cited in O'Byrne 2015, p. 37). Thus, individuals' personal experiences of cursing, and its effects, gives the naked curse weight as a collective enterprise.
Cosmological power is thus important to the use of naked protest. While symbolic power emphasises the gendered nature of this form of protest, cosmological power makes naked protest dangerous and allows the effects of the protestthe possibility of suffering due to the curseto endure beyond the moment of collective action, even in situations when the protest itself is brutally suppressed. This makes naked protest particularly valuable in contexts where protest is often shut down quickly and violently.
The power of the naked curse has been widely observed in African contexts. As Turner writes: [T]he [naked] curse is invoked only under the most extreme circumstances. Before it is even threatened, women usually take a formal vow to honour the enormity of its symbolism. We all come into the world through the vagina. By exposing the vagina, the women are saying: 'We are hereby taking back the life we gave you,' [. . .] It's about bringing forth life and denying life through social ostracism, which is a kind of social execution. Men who are exposed are viewed as dead. (Turner in Ekine 2008, in Tyler 2013 The above quote highlights both symbolic power (through the symbolism of the exposed vagina) and cosmological power (through the assumed lethal power of the curse). The use of naked protest-curse at times provides no remedy for the injustices that inspired it. Its use leaves the protester vulnerable, both to the humiliation of being seen naked and the denial of their claims to moral authority. Moreover, once released, the naked curse is unpredictable: it can be redirected against its maker if its original target is underserving or if words are uttered to reverse the curse. 9 Thus, people generally understand naked protest as a powerful but risky measure of last resort, used when there is no other avenue to makes one's voice heard and one's needs respected. It is most effective when the cause is just and the person who produces it has the moral and relational authority to use it.
Biopower, symbolic power and cosmological power converge in naked protest, creating a powerful and emotive form of collective action thatthough it is still often unable to overcome the physical and material violence of the militarised statecarves out a space for political voice in an otherwise extremely repressive context. This space, however, is narrow and fleeting. It is able to exist because it works in a different register to those of the state; but this in turn allows it to be delimited and framed as a traditional or cultural activity. Moreover, as it is understood as a tool of the vulnerableand indeed displays (literally) naked and raw vulnerability in the face of powerit in some ways highlights the dynamics that protesters seek to upend. The case below further illuminates these dynamics.

Land Conflict and Naked Protest in Apaa, Uganda
In perhaps the best-known case of naked protest in northern Uganda, in 2015, five adult women stripped naked in front of government authorities to protest land alienation. A police officer described the events: [The people] were well organized or prepared for the worst and quiet [ly] carrying placards with writing that they were prepared to die for their land . . . All of a sudden [the community] . . . broke into crying for their land. Even Hon Gilbert Olanya [the area MP] himself was crying as if someone had died . . . The ministers and all of us were surprised. What shocked us the most was when five women jumped out of [the] crowd and started somersaulting while completely naked before the ministers and all of us. 8 According to a journalist, 2 weeks prior to the naked protest, armed forces had raided the area, beaten civilians with sticks and thrown them out of their homes. Several of those present that day reported that the security officers included soldiers, police officers, and even Crime Preventers or cadres dressed in state uniformanyone that could be recruited at short notice. The journalist recounted, 'I saw in the faces of the men, they had become warlike, and women and children looked helpless and hopeless. You feel it in the environment'. He continued, 'On that day, they did not fight, they sat in the road to block the UPDF [military] trucks. Their numbers were overwhelming. People feared because of the brutality of the security forces and the fierceness of the local community'. He explained that although the minister's visit was purportedly a 'courtesy visit', it was widely believed that an eviction was planned, and the trucks and armoured military vehicles that accompanied the minister would be filled with the residents if they cooperated, and with their corpses if they resisted. He estimated that, when the convoy arrived, they were met by 3,000-5,000 residents of Apaa sitting on the road. Elders sat in the front with calabash, water, and oboke olwedo (tree leaves)objects used in ritual blessings or curses. Women of child-bearing years sat behind them with their children. Men sat in the back. The journalist explained, 'Women were [seated in the front] because they [the community] wanted to show that they are not violent. And none of the men stood up. Soldiers surrounded the crowd with machine guns and empty trucks and police and UPDF were moving around them'.
The elderly women at the front of the crowd reportedly undressed and rolled on the ground while crying, begging, and at the same time, cursing the authorities. Respondents recalled that the women said: 'You, government, watched us being killed during the LRA war and now you want to forcefully take the land that we are surviving on so that our clan perishes all, you will collapse soon, die mysterious deaths one by one' (Focus group discussion, Gulu, February 2018). A man who participated in the protests explained: The situation called for them to do something obscene. Because if you see your wife helpless and undressing, you feel like you should really do something obscene. But we decided that people should not be violent . . . Imagine that day if some people tried to beat up a soldier or police officer and kill him, how many people would have lost their lives? We said people cannot see our mother's naked and we are just sitting here, so we cried. (Ebila and Tripp 2017, p. 40) A newspaper reporter wrote: On that day, two government ministers had arrived with surveyors; their plan was to demarcate the land where the women and their community live. But they were shocked by what they saw. As a policeman took pictures, one of the women approached him by rolling on the ground and then raised her leg. He ran away. (Byaruhanga 2015) The journalist further reported that the government officials and everyone who witnessed the eventsapart from the Minister of Internal Affairs, General Aronda Nyakairimawere brought to tears (Authors' interview, Gulu, August 2018). The land demarcation was delayed. Five months later, Aronda died under mysterious circumstances. Amongst some, it was rumoured that his death was caused by the curse that he suffered, offering an indication of its lethal nature (Wilmot 2016). 10 The protest in 2015 built on nearly two decades of ongoing conflict and government attempts to alienate the land. In the 1990s, the Uganda Wildlife Authority (UWA) had proposed establishing a wildlife reserve in Apaa, noting that the land was abandoned because of insecurity, caused by the ongoing LRA war, which had displaced its inhabitants to camps. The government noted that because the population had been displaced, 'limited encumbrances would be encountered during the process of establishing a game reserve' (Serwajja 2014, p. 179). In 2009, the government signed a management concession with the South African owned Lake Albert Safaris for approximately 40 square kilometres of land (Authors' interview, Gulu, November 2015).
After the war, the UWA blocked approximately 1,500 displaced citizens from returning to Apaa, and progressively used violence to force those who had returned to leave (Serwajja 2014, p. 176-77). Between 2010 and 2012, the UWA, the Uganda Police Force and the UPDF evicted an estimated 10,000 residents, razed hundreds of homes, and destroyed property and crops (Lenhart 2013). Another journalist recalled the events of 2012, when the government conducted mass evictions: Over 200 people were injured and one died. It was done under the disguise of UWAthe militaryviolently started moving people, taking them physically and throwing them into the truck to move them. They dumped everyone in Pabbo [IDP camp]. (Authors' interview, Gulu, November 2015) Here, the journalist notes the blending of the UWA with other security forces, specifically identifying the military's participation. In 2012, Olanya and others received a court injunction on the land to prevent further evictions and require community dialogue, and later met Museveni to advocate for residents of Apaa: We met the President in his home. I told him clearly . . . 'Now that the war is over, I am not happy to see my people go back into suffering. I'm taking people back to their land . . . if you don't want to see people there, kill all of them, bury them and take their land. This time we're not going to allow our people to suffer'. (Authors' interview, Gulu, November 2015) Some viewed this naked protest as successful because the authorities temporarily retreated. At the same time, there appears to have been little change in the ongoing conflict. As one MP noted in Parliament in 2016: Last year in September, you [the NRM government] sent the police and soldiers and many people were assaulted [in Apaa], many people are still missing and two people lost their arms during that conflict. I even talked to the Minister of Lands and Housing about that issue, but up to now it has not been sorted out. (Uganda Hansard 2016) Since then there have been numerous attempted evictions by government authorities, huts have been burned and civilians beaten and arrested under mysterious circumstances, several more small-scale instances of naked protest have occurred, and Apaa residents conducted an 'occupation' of the United Nations office in Gulu Town from July to September 2018 to demand action (Ocungi 2018(Ocungi , 2019. At the time of writing, Museveni had commissioned a committee to 'seek a lasting solution to the Apaa land conflict', but many Acholi politicians and Apaa residents have expressed concern over the composition or legitimacy of the effort (Ocungi 2019). 11 The Naked Body and Its Power as a Repertoire of Subversive Protest The case in Apaa illustrates how the three types of power converged in the naked protest and curses of the elderly women. Many residents felt they had no other option at that time, after having been forcibly displaced multiple times with no resolution in sight. From this case we can see that there is something distinct and potent about the conjunction of the use of biopower, gendered symbolism, and cosmological power of the curse enacted through naked protest. These elements were evidently intertwined through various aspects of the protest which we explore in this section.

First, Bio-power
The protesters used their naked bodies to demonstrate that the people of Apaa would not be governed by the Ugandan state. In the face of overwhelming violence, this was one of the few options available to them. As Olanya said: 'if you don't want to see people there, kill all of them, bury them and take their land'. Here, he recounts explicitly telling the President that his constituents would not submit to the state's authority; the state would have to kill Apaa's residents, leaving no one to govern. The journalists' memories of the events further highlight the stark juxtaposition between the naked and vulnerable bodies of elderly women and the uniformed and armoured bodies of the security forces officers.
The journalists also recounted the strategy behind the physical configuration of bodies on the road. The physical stance and organisation of the community members allowed them collectively to present a paradox: on the one hand, their strength (in numbers, justness of cause and cosmological power) and on the other, their vulnerability in the face of the overwhelming military power of the state. By bringing this disjuncture into relief, the community challenged the role of the state as a protector of life and legitimate governor of society. As Tamale reminds us, 'naked bodies make strong political statements that challenge structures of domination or exclusion. They have the capacity to disrupt and, in a spectacular way, turn vulnerability into empowerment' (Tamale 2016, p. 31).

Second, Symbolic Power
As described potently by the journalist quoted above, the naked elderly female bodies were gwok, producing a strong, emotive reaction amongst those present. Males and females, young and old, civilian and military alike shed tears in response to this disturbing scene. The symbolism of the naked elderly female bodyreferred to as the symbolic 'wife' and 'mother' of the communityfurther was able to shift the question of land ownership from legal terrain (and the state's duty to enforce contracts and right to alienate its population) onto moral terrain. Indeed, as a question of land, this case offers particular symbolic dimensions to draw upon. For Acholi, and many Ugandans, land is considered life; the denial of it is a threat to existence. This is particularly the case for the rural poor, who may own nothing of value other than a piece of land.
Thus, a powerful symbolism was derived from the juxtaposition of the women's naked bodies and the land, in service of the claims they sought to make. It is useful to again return to the journalist's description of the event. Through tears, 3 years after the event, the journalist recalled: As the elderly women began to undress, wail, and curse, the soldiers held their guns, and all the police were touching their pistols . . . if you were there, you could do nothing but cry . . . .the naked women came harmlessly to the soldiers . . . I imagine that when the women undressed, it would have been a good moment for Aronda to arrest people, but even the officers were shedding tears. If Aronda had ordered them to arrest people they might have started with him. They had trucks prepared to arrest all these [male] youth, but it was the elderly women now! . . . One of the officers told me, 'The guy [Aronda] is lucky today. If he had ordered me to shoot my mother [the naked protesters], I would have shot him first. I am an Acholi like you'. (Authors' interview, Gulu, August 2018) The journalist's report of the soldier's words suggests that, when confronted with naked protesters, even those participating in the militarised state were forced to wrestle with the moral implications of their actions. Moreover, the naked women literally rolled on the earth, juxtaposing the powerful symbolism of their reproductive power to the earth. Ebila and Tripp (2017) similarly highlight in their analysis of this protest the role of 'symbolic resonance' and the body symbolism of motherhood which was mobilised as a resource to struggle for land rights in the face of repressive authority.

Lastly, Cosmological Power
Cosmological power was clearly an important aspect of the community's decision to employ naked protest. While stripping, the women uttered curses, directing their naked bodies at the government officials they viewed as threats, accompanied by the use of ritual objects like the calabash and leaves from a sacred tree. This drew on familiar cultural idioms which held both great symbolism and cosmological power. In the moment, the women's curse appeared to have frightened the government officials, causing them to physically retreat. Although the conflict continues, the subsequent death of Arondathe only official who reportedly did not shed tears at the protestis held by some as an indication of the potency of the Apaa residents' collective action and curse (see also Wilmot 2016). In this way, the curse imbued the protest with a kind of longevity, allowing it to live on in people's imaginations, memories, and narratives of the conflict beyond the moment of confrontation.
Though naked protest in the case of Apaa has not been successful on conventional measures (that is, it has not fundamentally changed the terms of debate), it appears to have drawn the attention of at least some government officials to consider that the community has undergone immense and unjust suffering. For some, for a short time, this changed the calculation in regards to how to handle the issue at hand. Such approaches have been used elsewhere in Uganda to situate claim-makers through relationality. As Boyd explains: More than anonymous victims, these exposed bodies revealed how rights claims are situational, dependent not on the abstract equality of individuals, but rather on the relationships of difference, dependence, and inequality that have long provided the terms of moral arguments about obligation, justice, and rights in Uganda. (Boyd forthcoming) Nakedness thus uses the body, and its gendered symbolic scripts, to trouble the assumed social order, and, as Boyd elaborates, draw attention to social relations imbued with moral obligation. This helps explain the soldier's comment that rather than obey orders from a senior commander, he would have defended the naked mothers before him. This resonates with other naked protests on the continent, when women use their naked bodies and gendered norms to shame the authorities. Cases have been documented across Africain 1992, Kenyan mothers of political prisoners went on a hunger strike and stripped naked to demand the release of their sons; in 2001 also in Kenya, women stripped to protest land annexation at a nature reserve near River Tana; in 2002, Nigerian women stripped naked to protest environmental degradation by oil companies; in 2003, Liberian women stripped to demand an end to the civil war; in 2016, South African women stripped to protest sexual violence on campus at the University of Rhodes. These cases demand that authorities consider their relational obligation to women and society more broadly.

Conclusion: Implications for Political Voice and Collective Action
Following Branch and Mampilly writing on African uprisings, we seek to place naked protest 'within the broader debate about today's outbreak of protest around the world -. . . [while also] discerning what makes that protest specifically African' (Branch and Mampilly 2015, p. 4). Naked protest allows us to examine repertoires of protest that we observe in politically repressive and militarised states, and further to analyse how their particular social, cultural, and political contexts give them meaning and power. This offers a valuable window onto how ordinary citizens living under increasingly militarised state institutions seek to exercise political voice. Our analysis highlights that citizens use creative techniques that draw on contextual specificities to leverage biopower, symbolic power and cosmological power outside the control of the state. Though these techniques may open a space for public engagement and claimmaking, the space is often small and temporary.
Indeed, although public nakedness can have political effect as a form of demonstration to hold authorities to account, our respondents emphasised that it is not without its risks. In Uganda, citizens employ naked protest in situations they view as desperate, where authorities are violating fundamental rights, particularly in the area of landboth in Apaa and elsewhere. While some of our respondents argued that it shows absolute powerlessness, our analysis suggests that it is better understood as a 'last resort' to an inalienable power of the body. In this way, it is a potential weapon of the weak and vulnerable. At the same time, it also shows how, in this context, the body works not just through the register of biopower, but also gendered symbolism and cosmological power. Naked protest is an assertion that even the most vulnerable actor possesses an undeniable and inherent embodied power which emanates from troubling social order and re-centring local moralities.
In this context, naked protest is a strategy used by marginalisedor 'muted'populations who lack access to dominant modes of communication to express their grievances (Ardener 1983cited in Barcan 2002. This approach is poignant in a context like Uganda, where the public space is heavily militarised and policed. Acts of naked protest can be understood as an attempt to reframe moments of conflict within local narratives of morality and relationality. We found that those who engaged in it at times made this move deliberately and explicitly, as one woman explained, 'If you do it [naked protest-curse] it is a sign that you are giving tradition/ways of the ancestors more prominence' (Focus group discussion, Gulu, February 2018). Yet, as illustrated, the very factors that make it 'traditional' contribute to its power.
Naked protest therefore offers us a window into the strategies that ordinary citizens can use to exercise political voice in contexts increasingly marked by overt and more subtle forms of repression. It illustrates that though authoritarian states seek to dominate and control public space, citizens can still use creative means to trouble the symbolic and discursive infrastructure upon which this power is built. This is important particularly in contexts where the state both relies on repression but has limited capacity to deploy repression, and thus must also induce the population to self-police (Tapscott 2017a). In such contexts, acts that challenge the social order have the potential to reveal this weakness in authoritarian control. However, despite the latent power of such acts, they also reveal the extremely limited space for expression of political voice. It is in this space of tension that naked protest exists, highlighting both its power and its limitations. Notes 1. Here, we do not propose that the NRM regime should be understood purely as a militarised regime. Indeed, scholars have illustrated the hybrid nature of the Ugandan state under the NRM, such that it melds rule of law compliant institutions with arbitrary use of state power (Rubongoya 2007, Tripp 2010, Goodfellow 2014Tapscott 2017b), though it has recently been recognised as increasingly authoritarian (Abrahamsen and Bareebe 2016). However, militarisation influences many aspects of the NRM's hybrid regime, for example, shaping the flow of neopatrimonial resource distribution and building new social networks amongst those who participate in military trainings. It is thus an important factor to understand prospects for citizen claim-making on the state. 2. One key aspect included denying the opposition mobile money transfers that were critical for facilitating polling day logistics, especially for polling agents. Thanks to Moses Khisa for highlighting this point. 3. See also Khisa 2019 for how the War on Terror was instrumentalised to suppress opposition and public protest. 4. While we might question whether naked protest is truly non-violentin particular because its curse is understood to be potentially lethalwe categorise it as such because it offers insight into the different registers of protest that civilians use in this category, as contrasted with violent forms of protest, like rioting, destroying property, or looting. 5. The LRA war refers to a 20-year conflict between the Government of Uganda and the Lord's Resistance Army that devastated the north of the country. For more, see Allen and Vlassenroot (2010), Branch (2011), and Porter (2017). 6. Other explanations for the gendered differential in the use of naked cursing offered by our informants included: naked protest is viewed as a last resort and perhaps men have more alternative avenues they can try first; women are more invested in the proper upbringing of their children, and are thus more aggrieved than men by their failures; women have shorter tempers; and men used naked protest more in the past but 'these days men are taking a break from responsibility' and as a result, the onus of protecting moral order increasingly falls to women. 7. It should be noted that we are less concerned with the empirical status of spirits or spiritual power of curses than the lived experiences of those who are affected by them. We therefore take a basic phenomenological approach to the problems experienced by the people we have interacted with in the course of research. We strive to take things 'as they are' (Jackson 1996), not searching for empirical 'truth' or 'existence' of the spiritual force of a curse, but how they work as social realities, having observed that our respondents do not experience all spiritual phenomena as supernatural, but as arising from both ordinary and extraordinary events, local histories, and social processes (Victor and Porter 2017).
8. The target of the curse enacts the reversal by uttering the words: 'Let it [the harm of the curse] begin with the person who curses' or 'Okwong ki lalam'. 9. Quotes from interviews and fieldwork have been lightly edited for clarity. 10. Aronda's death was formally attributed to a heart attack, though many speculated that he was poisoned by the NRM government. For example, Kizza Besigye wrote on Facebook that the death 'occurred against a background of many suspicious deaths of high-ranking government and military officials . . . [and] a specific warning that his life could be in eminent danger' and advocated for a full investigation (Besigye 2015). 11. For more on the evolving situation in Apaa see 'Apaa Monitor', a watchdog group on Facebook.