Introduction

Abstract

The introduction to this book defines cosmologies of war as different theories of the world that sees social dynamics through the optics of warfare, showing how the field of policing in Brazil can be understood through the histories of colonial and cultural war. These notions highlight the material and symbolic dimensions of the Brazilian state's war on drugs. I suggest that the Deleuzian concept of war machine and state dynamics is useful in untangling the overlapping tension, symbiosis, and blurring between different agents of power in Rio de Janeiro across binaries such as state and non-state, allowing us to grapple with an empirical reality where the police are both enforcing law and order and operating outside of it and where the state engages in symbiotic relationships to a multiplicity of agents and forces that challenge its sovereignty.

It is late in the afternoon. Thick clouds hover over the hill towering in front of us. Hundreds of houses and small buildings cascade down the hillside. They resemble Lego bricks carelessly stacked on top of each other—like a jigsaw puzzle of bricks and mortar. I am walking in a long row of about 40 black-clad police officers moving through the alleyways and stairs towards the heart of the favela with stealth. They have raised their guns and are ready to retaliate should we come under fire. Some of them whisper quiet orders at each other. They fear an ambush and do not want to announce their presence to the drug traffickers. There are many places for the traffickers to hide and many blind alleys where the police might get trapped. In some places the passages are so narrow you can touch the walls on both sides of the alley if you extend your arms, and so steep that the stairs have been built to scale the hillside crossways. The officers at the front-guard are armed with semi-automatic FAL machine guns. These are war-grade weapons and offer more power and precision than the usual police guns. But above all, the officers tell me, they warrant respect. As we make our way through the neighborhood, they aim their guns at the windows and terraces of the low buildings surrounding us, where the enemy might be hiding (Fig. 1.1).

Fig. 1.1

Police officers patrolling in Mangueira, April 2015

In the last few weeks, the situation in Mangueira has been tense. The favela complex is controlled by Commando Vermelho (CV, the Red Command), one of the three main gangs vying for control over Rio’s favelas. The name is a nod to the socialist rhetoric that the group used to gain legitimacy among residents when they emerged in the 1980s. Initially, they presented themselves as a welfare state for the poor, offering handouts and help to favela residents in exchange for their support. Since then, traficantes (lit. “drug traffickers”; gang members) from different gangs have been fighting each other to gain control of the drug retail in the favelas. Meanwhile, the police have played an ambiguous role, staging spectacular “invasions” in the favelas, confronting the drug traffickers and recurring to torture, executions, and massacres in an all-out war on drugs, while simultaneously extorting fees from the gangs, funneling the money to local politicians, and establishing a symbiotic relation with the favela-based drug economy. This modus operandi has produced levels of armed violence and death that are normally only seen in warzones.

In 2009, the Rio’s Military Police forces began to establish Pacifying Police Units (Unidades da Policia Pacificadora, UPPs) within the favelas in an alleged attempt to “pacify” them: to assert police authority and force out armed drug traffickers. Through the pacification, the police sought to quell the armed violence and gunfights that periodically transformed entire neighborhoods into battlegrounds.¹ Officially, the pacification was also meant to signal the abandonment of war-oriented forms of policing and favor a democratic modernization of the police force inspired by the principles of community policing. Addressing security concerns ahead of the 2016 Olympics, the UPP at Mangueira was inaugurated in 2011, right across Maracanã, the football stadium where the games’ opening and closing ceremonies would take place. In the first years of its existence, it seemed the police’s efforts to repress armed violence had largely succeeded.

But in 2015, while I was doing fieldwork at Rio’s UPPs, police officers were frequently involved in confrontations with traffickers. Increasingly, Mangueira as well as many of the other “pacified” favelas were becoming battlefields in the war between police and traffickers from different gangs. Following a wave of armed violence, police leaders promised to “re-occupy” Mangueira. The Commander in Chief at the UPP reorganized patrol to saturate the favela with police officers and assert territorial control. In areas where the risk of armed confrontations was high, several units patrolled together to offer each other tactical support in combat situations. The re-occupation involved the coordinated deployment of officers at strategic points in the favela, with all the patrol units—some 50 officers—leaving the base simultaneously in a spectacular performance of police power that was meant to discourage attacks.

I walk at the back of the group, right behind the only female officer in a long row of men. We move quickly, and since I am wearing a bulletproof vest with the police’s insignia, I have hung my camera across my shoulder to signal that I am not a police officer. My hope is that by keeping my camera visible I will be seen as a journalist rather than an officer dressed in civilian clothes and that this will give me at least some protection as a non-combatant. But I also feel a slight discomfort at being associated with the Military Police by the people living in Mangueira. The institution is poorly viewed by many favela residents. On this day, the streets are empty and the few people we encounter silently step to the side as we pass. Most of them look down to the ground, pretending that we’re not there. They are used to seeing their neighborhood transformed into scenes from a war movie and don’t want to attract attention from the police. Some, however, gaze at us in defiance. I feel like an intruder or member of an invading force. We climb the stairs to a small open square at the top of the ridge. It is protected by a tall brick wall and a large water tank that services the community. The police consider it a safe place for a short break. They share a bottle of water to quell the thirst after the steep climb.

We continue down one of the main streets; a winding road that runs along the upper part of a narrow gorge covered by forest. To our left, the slope is too steep to accommodate any buildings, allowing for a panorama of the favela. The small brick homes roll over the edge on both sides of the gorge. In the eyes of the police, such open landscapes are dangerous. Deprived of shelter, we become vulnerable targets. Gang members can easily hide in the buildings across the gorge, making us easy targets with no chance to retaliate should we come under fire. To reduce risk, we cross the most exposed stretches of the road one at a time. The only female officer is ordered to escort me across the patch of open road. She tells me that the men see her as a liability for being a woman. One of the men gives me a quick order as we get ready to cross: “Don’t stop moving!” The officers see me as a liability too, forced on them by their superiors. The female officer grabs the handle attached to the back of my bulletproof vest. It’s meant to make it easier to pull wounded officers out of combat situations. With a firm grip on the vest, she forces my head down as we crouch to make ourselves smaller and harder to hit. Then we run.

Cosmologies of War

Arguably, the installment of the UPPs in Rio’s favelas was supposed to end the logic of war that has characterized Rio’s public security policies since the return of democracy in 1985. But rather than ending the war on drugs, the pacification resulted in intensified police presence and the perpetuation of war dynamics in Rio’s favelas, especially from 2012 when the project expanded to the favelas of Rocinha, Complexo do Alemão, and Penha (see Menezes 2018). Coinciding with the multiplication of “old practices” of policing (i.e. repressive and truculent forms) at the UPPs, the prevalence of armed drug traffickers in Rio’s pacified favelas produced a sense of crisis of the project that would eventually lead to its official dismantling in 2018.

The collapse of the UPPs (between 2012 and 2018) was accompanied by a similar unraveling of Brazilian democracy. A succession of corruption scandals, economic crises, political realignments, and moral panic targeted at the Worker Party, which had governed for four consecutive periods, led to the destitution of Dilma Rousseff in 2016 and mock trial of Lula da Silva in 2018, paving the grounds for the Presidency of Jair Bolsonaro. Although most of the police officers I met during my fieldwork in Brazil in 2015 were arduous Bolsonaro supporters, at the time he was still a fringe figure in national politics. To me, the authoritarian worldview that he invoked seemed like a gust from the past, diverging from the image I had of Brazil as a liberal albeit “disjunctive” democracy (Holston 2007) that was going through a process of modernization of its institutions and society.²

In hindsight, my fieldwork took place at a moment of contestation over Brazil’s future that is still unfolding. While Bolsonaro drew on a worldview with a long historical trajectory in Brazil, he must be seen as part of a global (re)emergence of the far right. In Brazil, this movement found its expression in the figure of a former army captain who had built his political career on his unwavering support of the police and military, and whose family has been tied to paramilitary groups—the milícias (militias) dominating Rio’s suburbs. In the aftermath of Bolsonaro’s presidency, academics have tried to understand how a far-right politician and outspoken defender of the military dictatorship could gain so much support among Brazilian voters. Explanations tend to follow four different avenues (Duarte and Martínez-Moreno 2023): some note the growth and strengthening of the paramilitaries and the reorganization of illicit markets in Rio, alongside a macroeconomic restructuring as an important prelude to Bolsonaro’s bid for presidency (Menezes 2018; Grillo 2019; Manso 2020); others have focused on the political turmoil that followed the popular uprising in 2013, plunging Brazil’s political class into a moral crisis from which it has not recovered (Neiburg and Thomaz 2020; Nobre 2020, 2022); a third analytical avenue emphasizes the surge of right-wing and religious discourses in conventional and social media, and the affective materiality of digital media that produces effects of its own (Cesarino 2020, 2022; Salem and Larkins, forthcoming)³; finally, some scholars point to a cosmological reordering largely shaped by conservative and religious moral crusades (de Almeida 2017; Messenberg 2019).

Most analyses, while situating the emergence of Bolsonaro as part of a global resurgence of right-wing authoritarianism, emphasize the political power negotiations on a national level and the fact that he drew on a hardline approach to policing and security that resonates with many Brazilians. Political philosopher Marcos Nobre (2020) warns that calling Bolsonaro stupid or crazy immediately shuts down the possibility of understanding the political crisis that brought him to power. He argues that Bolsonaro’s presidency must be understood as a war against the democratic system and his government, as a government of war, permeated by a military logic (see also Durão 2020). Bolsonaro’s understanding of politics and the state differs from liberal political philosophy in its rejection of any notion of common good or social contract. He understands politics as a process whereby the will of one group—the true Brazilian people—is imposed on the rest (Feltran 2020b). In this sense, Bolsonaro’s understanding of the state is rooted in a dialectic conflict, like that of Marxist philosophy. Similarly, anthropologists have problematized the bias that results when we refrain from studying the subjects that are often referred to as fundamentalist, authoritarian, or fascist—the “repugnant others” according to Susan Harding (1991). Harding shows how, in these cases, social scientists tend to retreat into binary thinking, where the non-modern other is associated with religion, magic thinking, and backwardness, while the modern us is attributed with rationality and civility. These scholars advocate the need to take this “other” seriously, rather than disregarding them as irrational and easily manipulated (Pasieka 2017).

In this book I explore the relation between the cultural meanings and social practices that preceded conservative backlash in Brazil by looking at the exercise of police authority in Rio de Janeiro’s favelas through the notion of cosmologies of war: different theories of the world that understand social dynamics through the optics of warfare. Rather than discussing whether Rio’s police forces are really at war according to the conventional criteria of warfare, I approach war as a cognitive framework or cosmological force that shapes Brazilian social relations, subjectivities, landscapes, economies, and politics (see Grillo 2019). Acknowledging the longstanding configuration of policing as warfare against racialized territories and populations in Brazil, I unpack the moral universe and cosmological order that is produced through the Military Police of Rio de Janeiro’s war on crime and analyze its effects.⁴ I situate my analysis in relation to some of the most recent and ambitious ethnographic attempts at theorizing police power and urban conflict in Brazil and suggest that the Deleuzian concept of war machine and state dynamics offer a powerful theoretical framework through which different but complementary perspectives can be synthesized. Finally, in my conclusion, I show how policing at Rio’s pacified favelas can illuminate wider political currents in Brazil, especially the tension between modernizing projects and the resistance to these.

In the following sections, I will build on the ethnographic work of Jaime Amparo Alves, Graham Denyer Willis, and Gabriel Feltran to show how the cosmologies of war that shape Brasil’s urban conflicts draw on the notion of colonial war as a racialized process of resources extraction, organized around necropolitical forms of governance, and of cultural war as a particular way of understanding the relation between different values and meanings that draws on religious tropes and narratives. These authors have analyzed the dynamics of police violence and urban conflict in São Paulo, a context that differs from Rio de Janeiro in some important respects: Whereas Brazil’s biggest city is characterized by the predominance of a single, centralized criminal organization (the PCC), the situation in Rio is more complex, with a multiplicity of armed groups vying for power (see Hirata and Grillo 2017). However, what concerns me here is to explore how the notions of colonial war and cultural war can be applied in a Brazilian context. I will return to the particularities of Rio de Janeiro in the last section of this chapter.

The Racialized Necropolitics and Predation of Colonial Warfare

With some of the world’s most violent police forces as well as levels of armed conflict that are often compared to recent wars in the Middle East, Brazilian big cities have garnered attention from social scientists that try to grapple with issues of securitization, militarization, urban conflict, state violence, and police terror.⁵ Following in the tradition of Franz Fanon, Jaime Amparo Alves (2018: 12) suggests that Brazilian civil society is a fundamentally anti-black “political space for the heteronormative white male subject of rights.” He reproaches that despite the “affirmative action policies and welfare policies that took millions of black families out of poverty during the Workers’ Party’s government” we still observe “genocidal proportions of violent black death during the same leftist administration.” This, he argues, reveals “the limited impact of the politics of rights in challenging the black structural condition in Brazilian society” (Alves 2018: 14). How is it that despite decades of democratization, market liberalization, and economic growth, the Brazilian state has not managed to establish an egalitarian sociopolitical order in the ideal image of liberal democracy? Instead, the almost forty years of democratization has engendered a political project with a strong totalitarian thrust, emerging from the intestines of extractivism, religious fundamentalism, and militarism. The framework of liberal political philosophy, building on the idea that the urban conflict is the effect of an insufficiently developed democracy (implying that liberal democratic systems are anti-ethical to violence) is unable to offer plausible explanations (Alves 2018).

Challenging Foucauldian and Agambian approaches to sovereignty, the postcolonial scholarship of Jaime Amparo Alves and others who draw on Achille Mbembe’s notion of necropolitics have signaled how a racialized politics of death originating in the country’s colonial history is shaping the Brazilian social order (see, e.g. do Nascimento 2016 [1978]; Gonzalez 1988; Vargas 2012; Salem 2016; de Oliveira 2016; Alves 2018; Saborio 2018). In these analyses, the favelas appear as colonial spaces, embedded within the Brazilian state order.⁶ As a necropolitical terror formation, colonies are places where disciplinary and biopolitical forms of governance converge with the administration of death, in such a way that “vast populations are subjected to conditions of life conferring upon them the status of living dead” (Mbembe 2003: 40). In the colonies, war becomes a mode of governance and is often associated with extractivism and predatory accumulation, such as the narcotics industry, that characterizes war-machine formations (see Bourgois 2018). Brazil’s urban violence, according to this perspective, must be understood as part of an ongoing process of colonization and of the production of an anti-black social order through the killing of black bodies and war within black spaces, but also as the production of spaces for different forms of capital accumulation centered around violence and predation.⁷

The argument put forth by Graham Denyer Willis in The Killing Consensus (2018) is amenable to such analyses. In his analysis of the dynamics of urban violence in São Paulo, where the exercise of sovereign power is shared between the criminal organization Primeiro Comando da Capital (PCC) and the state’s security forces, Willis suggests a shift in emphasis from the sovereign agents that exercise the right to kill, to “the existence of a ‘disposable’ population that states allow to be preyed on under an acknowledged definition and common denomination of deservedness” (Willis 2018: 40). If routine violence appears uncontrolled, he writes, “it is only because the definition of life and death is so expansive”—i.e. the violence that the state allows appears limitless because it encompasses a large segment of the population (Willis 2018: 12).⁸

According to this approach, state and crime are not antagonistic powers but “morally and practically nested, operating in mutually beneficial and symbiotic ways” (Willis 2018: 94). At first glance, this claim seems to be in line with Amparo Alves’ (2018) proposition that the Brazilian urban order is structured around anti-blackness as an organizing principle. The killing consensus that Willis observes attributes a multiplicity of groups with the right to kill as long as those killed are black, poor, criminalized, and therefore imagined as deserving. However, the framework of a killing consensus suggests that criminal organizations such as the PCC are encompassed by or nested within the Brazilian state formation. In the words of Willis, it posits the existence of a single sovereign but a multiplicity of agents exercising the right to kill within a common, overarching framework. Here, Alves offers a radically different reading of the relationship between the PCC and the state. Understanding the war between the PCC and the police as a contest over territorial sovereignty, he argues that gang violence can be seen as the political expression of black criminal agency—i.e. as a strategy of black resistance that explicitly challenges society’s moral norms (Alves 2016, 2018).

The idea of an expansive killing consensus offers a powerful analytic lens from where to gauze the racialized macro dynamics of urban violence. Similarly, the anti-black, necropolitics-perspective effectively underscores the historical continuities of the Brazilian state formation and points to a significant gap in traditional analyses of urban violence in Brazil, which have until recently avoided the issue of race (Vargas 2012). Following Alves’ anti-black approach, the political emergence of the far-right does not appear as a radical break with a former, democratic state order, but should rather be interpreted as a weaponization of existing security discourses and intensification of necropolitical modes of governance that have been active throughout Brazilian history (Ystanes and Salem 2020: 54).

The notion of colonial warfare is good at capturing the historical continuity of anti-black state terror throughout Brazilian history and signals how racial capitalism (Seigel 2018) is organized around practices of predation and extractivism that are underpinned by a logic of extermination. However, the emphasis on the continuity of anti-black state terror shifts the focus from what is new or emergent in Bolsonarismo in relation to the (in important respects) socially inclusive politics of the Workers Party which extended political rights to new, individualized subjects.⁹ In this book, I argue that the conflicts that I observed within the Military Police reflect the larger dynamics of Brazilian modernization processes and reactions to these that also draws on a notion of war—albeit a cultural one.

Furthermore, in Willis and Alves’ different emphasis on the dynamics of criminal violence as consensus-driven on the one hand and a form of resistance on the other, we are alerted to a tension between transgressive and authorized forms of violence in a way that contemplates expressions of violent resistance towards the Brazilian state while simultaneously acknowledging the ways in which the state draws on and directs unauthorized forms of violence. Many Brazilian scholars have grappled with this tension (see Misse 2008; Machado da Silva 2004; Hirata and Grillo 2017; Albernaz 2020), and I suggest that a Deleuzian analytics organized around the notion of war machine and state dynamics can offer a lens that integrates different approaches.

Cultural Warfare and the Religious Battle Between Good and Evil

With regards to the question of what is emergent or new in Brazilian right-wing populism, Gabriel Feltran’s comprehensive analysis in The Entangled City (2020a) offers important insights. Noting how criminal organizations have become important security providers to the people living in São Paulo’s favelas, Feltran examines the production of urban order through the notion of different normative regimes. He explicitly builds his analysis on the debates forwarded by a tradition of Brazilian scholars who have tended towards an interpretation of the armed violence of Brazilian urban contexts as a conflict “between subjects that do not share the same plausible parameters of action and […] do not occupy different subject positions in a common urban order” (Feltran 2020a: Introduction, para. 15). In other words, as a conflict between subjects belonging to different cosmological orders, or as Feltran explains (I am paraphrasing): Within a shared normative regime subjects might disagree on how the state should go about in order to provide security and protection from crime to its population, but in the plural or disjunctive Brazilian order, there are subjects who see the world of organized crime as their main source of security rather than the state.

For these subjects, the liberal framework of citizenship, democracy, and the rule of law are neither useful as frames of analysis nor as universal values to be reached. Rather, Feltran argues, “the notion of a war between sliced-up sub-sections of the population, which no longer make up a single moral or legal community, seems to make more and more sense for understanding contemporary conflicts.” He suggests that evaluating the Brazilian state in relation to a Eurocentric normative framework ignores the fact that for many Brazilians, the world of crime is an important security provider and that we need a theoretical toolkit that grapples with the co-production of social order between criminal organizations and the state.

Through the notion of normative regimes, Feltran argues that Brazilian liberal democracy has been unable to resolve the tension between the seemingly unsurmountable differences in how the state and organized crime have been perceived by different segments of the population. Following democratization in 1985, these differences have been mediated through money. However, Bolsonaro’s incendiary rhetoric has successfully synthesized the two positions in a new, fundamentalist, authoritarian, and, I would add, totalitarian political order. He achieves this through the promise of overcoming social divisions by appealing to moral values in a cosmological reordering of the world according to evangelical religious doctrine, which has appeared as an emergent force in the Brazilian social field.¹⁰ In other words, what Feltran signals is the existence of at least two lines of conflict: One between those who rely on state and non-state providers of security (which are the “normative orders” of his analysis), and another between liberalism and it’s opposite, whether it is referred to as fundamentalism, authoritarianism, religion, or (a patriarchal) culture. Bolsonaro’s synthesis of the former is achieved by appealing to the latter. Feltran’s understanding of the role of Pentecostalism in producing this synthesis is consistent with Kristin Kobes du Mez’ (2023) understanding of U.S. evangelicalism “as a cultural and political movement rather than as a community defined chiefly by its theology” that must be understood as a larger cultural identity and urges us to examine how evangelicalism, conservatism, and militarism bleed into each other through the notion of cultural war.

In the religious sphere, scholars have noted how Catholicism, with its multiple saints, has tended to counteract polarized understandings of the social while Pentecostalism reinstates “a dualism between good and evil in its wake,” dividing “the world in two, allowing for no middle ground” (Mafra and de Paula 2002: 61 cited in Vital da Cunha 2018: 5; see also de Pina Cabral 2007).¹¹ In this context, “‘War’ signifies the ever-present spiritual struggle of the church and its members with the devil and the forces of evil,” and the “sense of a forceful battle waged daily against the devil on behalf of the afflicted” (Kramer 2005: 106–109). In the world of policing such an understanding of human existence as a battle between forces of good and evil can strengthen antagonisms, with police officers distancing themselves from the people they are supposed to protect (Albernaz 2015: 531). For those who support Bolsonaro, “the crisis that is devastating Brazil is […] fundamentally a crisis of values, where leftists, feminists, or gays are dangerous categories. It is the strategy of cultural [wars,] of the moralization of politics and the demonization of opponents transforming them into enemies” (Solano 2020: 216).

Influencing the urban culture of the peripheries of Brazilian big cities, evangelical churches emphasize meritocratic discourses, the strengthening of traditional family values and gender roles, and, importantly, promote the use of metaphors that refer to warfare and struggle between good and evil (Vital da Cunha 2018: 20). Some of these churches have taken it upon themselves to liberate Brazil from the influence of the Devil, for example, through the exorcism of demons but also through purifying violence (de Pina Cabral 2007: 499). In territories associated with drugs, crime, death, and indecency, violence is understood as a manifestation of divine fury, and the suffering it produces is seen to redeem or save people and areas that have been abandoned by God from the forces of evil. This understanding gives meaning to experiences of suffering as necessary stages in the transformation of thugs into men of God (Birman 2012: 137–144). Patricia Birman (2019: 113) sees a connection between the secular warfare I have described with reference to Brazil’s colonial history, and the spiritual warfare that evangelical churches preach. While secular warfare casts “the other” as an enemy to be eliminated, spiritual warfare sees the enemy as a subject to be saved and transformed. These are not incompatible paradigms but rather complementary ones: Thugs can either be transformed through redemptive suffering or eliminated (or, in a more instrumental understanding of the war, be maintained as useful devils).

The emergence of evangelical churches in national politics over the course of the last decades has replaced the socio-political imaginaries at the core of Brazilian national identity (such as the myth of racial democracy) with new “cosmo-political imaginaries of divine ontological power flowing through the nation, which is ‘released’ and put into motion by the enactment of certain neo-Pentecostal rituals and rhetoric acts” (Shapiro 2021: 1). According to Feltran (2020b: 95),

The mass movement that made Bolsonaro is driven by the redemptive promise of resolving Brazil’s social conflicts and ending its social differences. As he sees things, at the end of the revolutionary war he is currently waging, the Brazilian people will be emancipated from oppression and will live in a community of equals in a Christian fatherland.

As in many other countries that have become infused with (neo-)Pentecostal ideational systems, the sociopolitical order—including styles of policing and the production of enemies—becomes sharper, divisive, and, sometimes, more violent (see, e.g. Rio et al. 2017). It should therefore not be a surprise that Bolsonaro’s moralism, which was also present within the UPPs, draws on a neo-Pentecostal cosmology in his bid to reestablish, in a neocolonial fashion, traditional racial and gendered hierarchies in Brazil. Thus, expanding the logic of operation in the favelas onto wider Brazilian society, Bolsonaro employs the state’s security apparatuses to do so, by sanctioning violence against populations that resist this hierarchical ordering of the world (Salem and Bertelsen 2020). Meanwhile, religious theology re-signifies the suffering of urban life as the everyday realities of war are cast as a battle between upstanding citizens and bandits in a worldview that synthesizes evangelical, militaristic, and Old Testament values (an eye for an eye) (Feltran 2020b: 96). Importantly, the evangelical churches have gained a power base in neighborhood police stations, transforming the police into key players in the battle over Brazil’s future. This is not just an issue of police officers congregating in evangelical churches, but also of these churches penetrating police institutions, offering divine legitimation for police terror in the favelas (see Albernaz 2015; Machado 2018; Oosterbaan and Machado 2019; Esperança 2022).¹²

Feltran argues that “the movement that sustains Bolsonaro seeks a central shift away from modern politics” with its categories of party mediation, law, representation, pluralism, and communicative reason, and towards an increased emphasis on the mass movement, male honor, identity, brotherhood, gospel, and raw violence. Noting the totalitarian thrust of this movement, Feltran writes:

The new nation will not have women taking a stand against men, blacks against whites, or employees against employers. The emphasis will be on unity. One God: the Christian one. One theology: the Pentecostal one, based on an Old Testament, eye for an eye, conception of justice. […] In Brazil, mass movements of this nature, just as in the Germany of the 1930s, have been given progressively broader access to the resources of institutional violence. Factions of the police and the military, ideologically mobilized and acting as autonomous militias, are key protagonists of the popular movements. (Feltran 2020b: 97–98)

Now we are arriving at the tension between transgressive and authorized forms of violence, and the Brazilian state’s reliance on unauthorized forms of violence. The key here is the notion of a multiplicity of violent agents that act autonomously but increasingly gain access to “the resources of institutional violence” (see also Hirata and Grillo 2017; Albernaz 2020). That is, the violent formations that I suggest should be analytically grasped through the Deleuzian notion of war machine dynamics (Deleuze and Guattari 1987).

War Machine and State Dynamics

Deleuze and Guattari’s notion of war machine and state dynamics is particularly well suited to the analysis of state processes and effects, and the resistance and opposition towards these processes. This conceptual pair “raise complex questions of agency and structure, intention and logic, command and leadership” (Hoffman 2011: introduction, 3rd paragraph), and highlight the tensions, contradictions, and often violent power struggles at the heart of most state formations. The framework is explicitly formulated for the analysis of processes of violent state power and its contestations and is increasingly being used “to think about the confluence between militarization, social movements, global capital, and the state” (Hoffman 2011: introduction, 3rd paragraph; see also Kapferer and Bertelsen 2009; Mbembe 2003).

In Deleuzian thought, state dynamics are understood as hierarchical, bounding, or territorializing, and oriented towards the conservation of organs of power. War-machine dynamics, on the other hand, are a-hierarchical, counter-systemic, boundless, or deterritorializing, and oriented to the destruction of the structures and hierarchies created by the state (Kapferer and Bertelsen 2009). War machines, despite what their name suggests, do not have war as their primary objective. War is only a consequence or by-product of their exteriority to the state (Hoffman 2011). This exteriority is not necessarily geographical; it is also an exteriority to its rules and regulations. War machines are assemblages that link in illicit or illegitimate ways, often operating in the shadows of security and policing (see Nordstrom 2004; Durão and Argentin 2023). They engage in predatory and often violent forms of accumulation: the drug economy and arms trade in Rio are clear examples of this, and perhaps less evident, so are the logics of global capital which is continuously resisting and challenging the rules and regulations of national state orders (see Kapferer and Bertelsen 2009).

However, while war machines do not have war as their object, they can be captured by states who seek to harness their destructive potential: “It is precisely after the war machine has been appropriated by the State […] that it tends to take war for its direct and primary object […] and that war becomes subordinated to the aims of the state” (Deleuze and Guattari 1987: 418). The war machine dynamics of Bolsonarismo are glaring in Feltran’s description of the central place occupied by a multiplicity of armed agents within the movement. In an almost literal rendition of how Deleuze and Guattari imagine the fraught relations between war machine and state, Feltran (2020b: 107) notes that Brazilian elites “imagined that Bolsonaro’s violent character could serve as a way to corral the masses in the direction of the conservative project they had always defended. After all, in the past, the big slave-owners had always controlled their jagunços” (armed hands or bodyguards that sometimes took the form of private militias).

According to Deleuze and Guattari, militaries are war machines that have been captured by the state. Understanding military institutions as captured war machines has some important implications. First, it signals the coexistence of war machines and state dynamics within the repressive apparatus of the state. This observation is not trivial. Although war machines and state are conceptualized as antagonistic and, brought to their full potential, mutually annihilating forces, they are not dialectical forces of a Hegelian kind, but coexisting dynamics irreducible to each other (see also Kapferer and Bertelsen 2009).¹³ The concepts of war machine and state allow us to move beyond notions of sovereignty as a centralized force, towards an understanding that emphasizes its decentralized, rhizomatically distributed relations. It also equips us with a powerful conceptual toolkit to analyze and describe violent contestations over power, particularly in contexts where “the agents and organs of state power are effectively at war with the populations over whom they claim control” (Kapferer and Bertelsen 2009: 1).¹⁴

War machine and state dynamics can and do coexist within particular state formations and are best understood as modalities of power and can help us untangle the overlapping tension, symbiosis, and blurring between different agents of power. The concept should not be read as a normative framework that posits the superiority of the liberal state in relation to other expressions of violent power. Rather, it offers a framework to analyze how states organize, direct, channel, and instrumentalize the use of force, including unofficial forms of violence, through the logics of capture and escape (Deleuze and Guattari 1987). These logics strike me as particularly useful in their ability to describe a field of policing that, in the words of Jauregui (2016: 10) is continually shifting and indeterminate, constituting “a social order of interconnected official governance and unofficial power relations that is not just continually in motion, but mutating in both content and form.” The concepts of capture and escape allow us to grapple with an empirical reality where the police are both enforcing law and order and operating outside of it; where police terror and massacres coexist with projects that seek to reduce police lethality; where the state engages in symbiotic relationships to agents and forces that challenge its sovereignty; and where a multiplicity of violent agents are given increasing access to the resources of institutional violence.

Such complex relations of hostility and symbiosis are not unique to Rio, and there are numerous historical and contemporary examples of states attempting to harness the potential of unruly adversaries to their own benefit. In Willis’ (2018: 153) theorization of sovereignty by consensus, he points to the relations that the English crown cultivated with pirates and privateers:

The difference between a ‘privateer’ and a ‘pirate’ was based on whether they were seen to have acted under the mandate of the sovereign or not. But for many reasons, both purposeful and not, the boundary between these two was fuzzy. The English crown could claim moral, economic, and political distance from ‘pirates,’ who ostensibly acted outside of sovereign authority, even though it was generally accepted that the relationship between crown and pirates was mutually beneficial. And even while privateers expanded the influence of the crown in expected or coordinated ways, their private alter egos often destabilized the English imperial project, stepping into and out of the unclear realms of legitimate and illegitimate violence and plunder.

A parallel to the multiple sovereignties that operate in Rio seems in place, for while the paramilitary groups that operate under the umbrella of milicia are largely aligned with a sovereign consensus framed according to state interests, their actions clearly operate in the realm of illegality. Nonetheless, these groups have had the explicit sympathy of Bolsonaro both before and during his Presidency. At the same time, while drug gangs are cast as enemies of the state they exist in a symbiotic and mutually beneficial relation with the police, funneling drug money to the state and the formal economy. Furthermore, violent contestations for power between and within police institutions, frequently observed in Rio, add yet another layer to the complexity of the exercise of sovereignty in the city. Examining the logic of violence through the optics of war machine and state dynamics and the logics of capture and escape allows for a much more fluid, context-sensitive, and nuanced analysis capable of explaining the violence of a multiplicity of agents under a common conceptual framework.

Multiple Sovereignties

In Rio de Janeiro, a multiplicity of different armed groups competes for control over the illicit economies territorially bound to the favelas, including at least three different drug gangs with well-established historical trajectories and several de-centered right-wing paramilitary vigilante groups known under the umbrella of militias which exist in a symbiotic relationship to the state’s security forces.¹⁵ This complexity is by no means exclusive to Rio or Brazil, and Tessa Diphoorn and Erella Grassiani (2018: 2) have suggested that the concept of security blurs might account for the “multiple, overlapping set of actors, roles, motivations, values, materialities and power dynamics” in the inception and performance of security, as well as “the overlap and entanglement of the practices and discourses of state and non-state security providers.” In the Rio context, reading the urban conflict as the result of two clearly demarcated normative regimes, or as the product of a killing consensus established by a single sovereign but exercised by multiple agents, only partially accounts for the complex relationship between different violent agents operating across the urban landscape as well as the tensions that I observed within the police.

Ethnographic studies of policing and security from other parts of the world have highlighted the conditionality of police authority. For example, in Beatrice Jauregui’s (2016: 8) study of Indian police, officers continually negotiated their authority to intervene with coercion in various situations, and were “forcefully doubted, and regularly defied among a variety of actors.” This seems better to reflect the situation in Rio de Janeiro where police officers continually negotiate their authority with different agents, including citizens (see, e.g. Misse 2006; Larkins 2015; Albernaz 2020). Accordingly, Jauregui (2016: 12) refutes the idea that there exists “a monopoly on rightful control of the means of coercion, by the state or by anyone else, at least for very long.” But more importantly, she argues that the concept of coercion has gained a monopoly on theorizations of police authority that is unwarranted. She suggests that police authority should be understood as “a relation and provision of sociocultural order making that is co-constituted with configurations of moral right and instrumental exchange,” and that “police authority […] including but not limited to coercive authority, is a contextual and conditional social resource variously demanded, drawn upon, and deployed to help realize human needs and desires” (2016: 13).

I am sympathetic to Jauregui’s call for a multifaceted understanding of police authority, and in particular, her emphasis on “moral right and instrumental exchange” (both are central in my analysis of police authority in Rio’s favelas) but I want to hold on to coercion as a key concept while considering its often fraught and contested nature. In this regard, Michel Misse’s (2008: 379) understanding of the Brazilian social formation as characterized by processes of accusation and justification, criminalization and incrimination that are often autonomous in relation to codified law, and often in permanent tension with it, is instructive of the dynamics that I am discussing. When thinking of urban violence in Brazil, Misse urges us to also consider how the social construction of crime is co-produced by legal and extra-legal orders, signaling the existence of what some scholars have called “legal pluralism” but what Bjørn Enge Bertelsen (2009; 2016) in his research on popular justice in Mozambique suggests is better conceptualized as multiple sovereignties.

The idea that there can be more than one sovereign competing for power within the same national territory is particularly relevant to the analysis of situations where the logic of war is invoked. According to Bertelsen (2016: 255), “multiple sovereignties […] are characterized as shifting, incomplete, and without necessarily corresponding to distinct social groups. Instead, they may be linked to social and cosmological ontologies of justice, rights, and evil or the many letters of state law – colonial, postcolonial, international.” That is, the exercise of sovereignty can emerge from multiple ideational sources and is not constrained within state orders. It is in this sense I wish to approach the analysis of Rio de Janeiro’s police forces and their operation across conceptual divides (peace/war; legality/illegality; security/democracy; state/crime; civil/military; private/public).

Structure of the Book

In this book, I approach the cosmological order and moral universe that emerges from the police’s war in the favelas from the point of view of the officers that were stationed at the UPPs. I understand the pacification project as part of a larger negotiation between different ideas about the future of Brazil and as a transformation of the city’s geography of violence through the logic of war. While this strategy for urban reordering was centered around the 2014 World Cup, 2016 Olympics, and other mega-events, I suggest that it should first and foremost be understood as a process that forcefully asserts police authority across the urban landscape. By zooming in on the militarized practices of policing in Rio de Janeiro’s favelas, I focus on social dynamics whereby Brazilian authoritarianism emerges as a political war machine.

Throughout the book, I raise a series of questions. In Chapter 2: Favela/Asfalto, I ask how the relation between multiple armed groups vying for territorial control in Rio de Janeiro and the state has been shaped by the history of colonialism and explore how stereotype representations of the favelas transform them and the people that live there into objects of interventions. In Chapter 3: Policing in Rio de Janeiro, I explore the racialized and militarized dynamics of policing in Brazil through the oral history of a former police officer and political advisor at the State Legislative Assembly in Rio de Janeiro. Chapter 4: The Postcard and the Frontline focuses on the two extremes of the pacification project: the UPP at Santa Marta, which at the time of research was considered a success story, and the UPP at Alemão, where the police engaged in daily battles with gang members. By looking at these places, I show how police authority is exercised through a logic of elimination on the one hand and transformation on the other. Next, in Chapter 5: Police Masculinities, I engage with analyses that note how traditional gender norms, and particularly an ethos of militant masculinity, shape the dynamics of violence in Rio. Chapter 6: Violent Becomings looks at the impacts of a militarized institutional hierarchy on police subjectivity and morals, while Chapter 7: Modernizing Warriors traces institutional reform attempts as these are implemented and resisted through training. In Chapter 8: A World of Warfare I analyze the predatory economies that the war on drugs in the favelas renders possible, while Chapter 9: The War-Machine analyzes Bolsonaro’s rise to power as the expression of war machine dynamics at the highest level of the state. Throughout the chapters, I am interested in describing what kind of world emerges when social conflicts and policing are conceived of as a war between good and evil forces. It is my hope that these questions I pose can contribute to a broader understanding of the emergent logic of twenty-first-century authoritarianism in Brazil and beyond.

A Note on Ethics and Positionality

The men and women who participated in this research live their lives in contexts shaped by violence. In addition to the armed violence that police officers engage in on the streets, they are subject to an institutional hierarchy that punishes critique as well as institutional dynamics shaped by the blue code of silence. Ensuring the anonymity of the officers has therefore been a top priority. To this end, I have relied on different methods, splitting apart, and piecing together ethnographic observations and conversations to make sure that the persons that I describe cannot be identified even by people who know them well, since they only exist as reassembled fragments that nonetheless are true to the realities that I observed. This is done at the cost of following the same officers throughout the chapters, and the reader will note that most of them appear and disappear across scenes, and that some of them are empirically flat—i.e. they only figure as conversation partners that we know little about, while others characters have more depth. This also reflects the reality of my fieldwork, where I talked with hundreds of police officers, but followed a handful.

In the text, I am by and large letting the ethnographic material speak for itself. This is a stylistic and political decision that is meant to make the text available outside narrow academic circles, and an attempt to engage in what some scholars call a “public” anthropology (see, e.g. Pinheiro-Machado and Scalco 2021). The theoretical and analytical discussions are mostly reserved for the introduction and chapter conclusions, while the chapters are written in a literary style that follows my journey from a perhaps little naïve outsider navigating the field of Brazilian policing to an insider. Like pieces in a puzzle, each chapter adds ethnographic nuance and complexity to a reality that is riddled with contradictions and challenges particular to the local context, history, and politics that make it difficult or even impossible to understand policing in Rio through European notions of what policing is or should be. In some places, I have added footnotes in the text with references to the vast body of literature on policing, state violence, and the far-right in Brazil and beyond. However, this is not an exhaustive analysis, and I hope that the ethnography I present can be interpreted in many ways by the readers.

There are some important limits to these stylistic choices and my methodology. First, since I did most of my fieldwork with the rank and file of the Military Police and my focus is mainly on the everyday practices of policing at some of Rio de Janeiro’s most conflagrated UPPs, I realize that I run the risk of reproducing stereotypes and clichés about patrol officers while letting the upper echelons too easily off the hook. While I have tried to show how my observations fit within larger institutional, political, and economic structures, I acknowledge that my text can be somewhat biased in this sense. Second, while I do not assume a moral relativist position, I have tried to bring nuance and complexity to descriptions of people who are sometimes represented as unidimensional and “savage others.” It has been a challenge to navigate dilemmas between my own political position and a nuanced anthropological analysis; between cognitive and emotional empathy as an analytical tool and the risk of being “seduced” by my research participants (c.f. Kapferer and Gold 2018; Robben and Hinton 2023). I realize that to some, humanizing police officers who have, in the extreme, “killed in the dozens” (as some officers have) is not a legitimate academic endeavor. Others might think that my moral judgements get in the way of an incisive analysis. This tension is present throughout the text and comes from a desire to address police violence and those responsible for it, while desisting from positioning the people I describe within a simplistic good and evil dichotomy, acknowledging that many officers navigate complex moral dilemmas in their profession.

I have tried to explain the processes that lead to the acceptance of and demand for the large-scale deployment of violence—of warfare—as a reasonable solution to social conflicts. My intention has been to show how this militarization also produces a polarization of positions and forces us to “pick a side” in ways that are often anti-ethical to the ethnographic endeavor of understanding the other's point of view. While condemning police violence is important, we must avoid reproducing narratives that assume a strict division between good and evil: These can produce ethnographic accounts that can be analytically lazy and reifying. I believe that humanization—the portrayal of the men and women working in Rio’s Military Police as complex human beings, who can change their outlook and understandings of reality and are concerned with issues of justice, of right and wrong, of good and bad, and of existential meaning—is necessary if we want to develop efficient political strategies that can counter polarization and strengthen our democracies (see Pinheiro-Machado and Scalco 2021).

“The Only Politician That Cares About Us!”

I have only been in Rio for a week. A group of a few hundred people have gathered at the end of Avenida Atlántica, just by the Military Museum and the old fort that lies between the beaches of Copacabana and Arpoador. Many of them are wearing T-shirts displaying a black hand and the word basta (enough). Other shirts have pictures of loved ones who have been killed, the date of their death, and the text saudades eternas (eternally missed). On the beach, someone has erected a mock cemetery with row upon row of crosses. At the front, near the sidewalk, there’s a large cross spattered with red paint that is supposed to resemble blood. The crowd is composed of friends, colleagues, and family members of police officers who have been killed in Rio. Some die in the line of duty but most are killed in confrontations off duty, in personal vendettas carried out by criminals, or as victims of assault. The crowd wants to bring attention to the increasing number of police deaths and demand the state to act. They are supported by the worker union of the Civil Police. Although most of the officers who have been killed are from the Military Police, military police officers are not allowed to unionize and may be sanctioned if they protest in public (Fig. 1.2).

Fig. 1.2

Family members of killed police officers protesting in Copacabana, December 2014

The protesters demand harsher punishments for assaults on police officers. Many want the government to introduce death sentences and lower the age of criminal responsibility from 18 to 14 years. Some are carrying large posters that they hold up so that media and passers-by can read their message. The texts are the patchwork of a worldview I do not yet understand. It espouses that the police officers are victims and scapegoats of a society marked by corruption, moral decay, and violence: “Human rights without the right to live—Basta!”; “The enemy is not the police but the government”; and “Victims of a hypocritical society and a government that does not care.” A woman speaks up: “There are too many thieves in this country! Everybody steals, everyone is corrupt. The crooks, the politicians, the people! But everyone hates the police! My daughter was a good person. She was on duty when she was shot in the back by a common coward—because that is what he was, a coward!” She has been given a microphone and shouts out to the masses, making an ambiguous appeal to the police, who are also known to be corrupt: “You have to make yourselves respected!” I do not understand if she is encouraging them to do their job in a professional manner, or if she wants the police to assert respect through force, but I’m inclined towards the latter. The group has grown since I arrived but is still just a drop in the ocean among the hordes of Sunday beachgoers enjoying the summer heat. Soon they flock around a man who has recently joined the protest. Many try to get a selfie: “It's Bolsonaro!” one of them says. “The only politician who cares about us, the police!”