Post-war reform of the police in Mozambique in accordance with the rule of law and human rights has had apparently paradoxical results. This is partly because efforts to (re)constitute state authority have relied on both embracing and taming ‘tradition’ as an alternative domain of authority and law. Ethnographic fieldwork at police stations in a rural former war zone shows that the state police handle an increasing number of cases outside their legal mandate. Notably, these include witchcraft and spiritual problems. They also resolve crimes using procedures that mimic those of traditional authorities. This is paradoxical, because police officers strongly advocate a strict boundary between ‘state’ and ‘traditional’ jurisdictions: only the state police should resolve crimes, while chiefs and other community authorities, including healers, should alone handle ‘traditional’ matters. This separation is supported by the 2004 constitutional recognition of legal pluralism. Yet it is difficult to attain in practice.
In this article I argue that the state police's handling of spiritual and social problems does not only reflect a tension between local/customary and state/legal notions of justice. Equally significant is the existence of partial sovereignties. In short, there are different articulations of the authority to make final decisions on life, death, punishments and rewards (Hansen and Stepputat Reference Hansen and Stepputat2005). In particular, the importance of invisible evil forces contradicts what the state/legal system can provide and challenges state sovereignty. A spiritual idiom of power and evildoing constitutes an alternative formation of sovereignty. It is associated with the capacity of invisible forces to give and take life, and it comprises the backbone of local conceptions of transgressions. This is an idiom mastered by chiefs and healers. Police officers engage with invisible forces to ‘get the job done’ in a way that both enhances their popular legitimacy and allows them to manifest state power, even as they act outside their legal mandate. Yet they never really manage to fully tame these forces. Consequently, state police authority remains uncertain, and must be manifested continually – i.e. by enacting hierarchies, by marking out the boundaries of jurisdictions, and by using force.
These aspects are explored in this article in relation to public state ceremonies and everyday work at local police stations. The context is Sussundenga District, a rural area bordering Zimbabwe in the central part of Mozambique. This area was an intense combat zone during the civil war that ravaged the country for sixteen years (1978 to 1992) between Frelimo, the party that has ruled the country since independence in 1975, and Renamo, a rebel movement initially supported by the white minority regimes of Rhodesia (now Zimbabwe) and South Africa. In Sussundenga, the war between Renamo and Frelimo almost immediately succeeded the protracted liberation struggle against colonial Portuguese rule. This meant that Frelimo's new party and state structures, inspired by a Marxist-Leninist societal model, were never fully established. Renamo made its first incursions here from over the border in Rhodesia. To begin with, it was part of Rhodesian and South African efforts to destabilize independent countries in the region that were now supporting the liberation struggles in their own countries. Renamo also received support from Western right-wing movements, as part of their anti-communism campaign in the context of the Cold War. Gradually, Renamo was also able to gain internal support, especially in the central and northern parts of Mozambique, and to take control of a large part of the rural areas. This was partly facilitated by Renamo's collaboration with local chiefs and healers, who had been banned by Frelimo after independence. Dissatisfaction with Frelimo's anti-traditionalist stance gave Renamo rural supporters, and meant that Frelimo had to give up the idea of winning the war with arms. In 1992 a peace agreement was finally signed, which at the same time marked the transition to multiparty democracy and turned Renamo into a political party. At that time, there was a dual administration in place, because Renamo remained in control of many rural areas. In Sussundenga, this history meant that state sovereignty was highly contested. Frelimo/state presence, including through the police, had been very short-lived; where it had existed it had been pushed out by Renamo troops by 1991. Even after the first elections in 1994, when Frelimo won a majority, there were still areas in Sussundenga that were effectively governed by Renamo and by chiefs and healers.
This article is on how the state police tried to (re)encroach upon these contested territories after the civil war and what their ways of doing this implied for state authority.Footnote 1 To this end, there is no substantial comparative literature on state police resolutions of witchcraft or spiritual matters with which to engage.
The role of invisible forces in policing seems to be confined to predominantly anthropological and legal sociological studies of vigilante groups, healers and chiefs, subjects that are usually studied separately from the state police (Pratten Reference Pratten, Pratten and Sen2007; Geschiere Reference Geschiere1996; West Reference West2005; Meneses Reference Meneses2004). This is probably due to the dearth of ethnographic studies of state policing until very recently. Conversely, a cross-disciplinary policing and security literature has begun to use concepts such as plurality, hybridity, nodes and assemblages to describe the complexities of everyday policing and the increase in private and community-based security actors alongside the public police (Wood and Shearing Reference Wood and Shearing2007; Baker Reference Baker2002; Loader Reference Loader2000; Jones and Newborn Reference Jones and Newborn2006; Abrahamsen and Williams Reference Abrahamsen and Williams2009; Lippert and O'Connor Reference Lippert and O'Connor2003). The concepts of assemblages and nodes, in particular, are used to capture how security provision today consists of multiple links between different security actors, rationalities and practices. These links blur and cut across the public–private divide, giving way to ‘hybrid structures’ (Abrahamsen and Williams Reference Abrahamsen and Williams2009: 8). As in legal anthropology since the 1980s (Merry Reference Merry1988), the concept of hybridity helps us to move beyond the dichotomous thinking that underpins most previous understandings of state versus non-state or customary orders.Footnote 2
Hybridity, I suggest, also provides a useful conceptualization of state police practices in Sussundenga. The state police constitute, as Santos (Reference Santos2006) puts it, ‘microstates’ within the state, embodying their own, locally adapted combination of different political and legal cultures. Within this perspective, the concept of assemblages adds a focus on how everyday state police practices develop through the multiple relationships that police officers have with other policing actors, such as chiefs.
Missing in both the legal anthropological and the security studies literature, however, is a discussion of the deeply political foundations of hybridity, which are conditioned, I argue, by partial sovereignties (Bertelsen Reference Bertelsen2009; Stepputat Reference Stepputat, Bouziane, Harders and Hoffmann2013). By focusing too intensely on the intersections and diffusion of power, one can easily lose sight of the simultaneous enacting of divisions, hierarchies and claims to superior authority in which violence often plays a significant role. Based on these considerations, I also draw on recent political anthropological studies of sovereignty in order to understand why the state police in Sussundenga act as they do (Hansen and Stepputat Reference Hansen and Stepputat2005; Das and Poole Reference Das and Poole2004; Comaroff and Comaroff Reference Comaroff and Comaroff2006). I begin with a brief background on post-war police reform and what challenges places such as Sussundenga posed to such reform.
POST-WAR POLICE REFORM: THE DILEMMA OF ‘TRADITION’
Officially, the police has undergone major changes since the sixteen-year war between Frelimo and Renamo ended in 1992. During the war, the police operated side by side with the military and acts of brutality against criminals and enemies of the state were the norm (Baker Reference Baker2002). Police legislation after independence in 1975 also supported a police force that defended the party state rather than the citizen. In contrast, police legislation in 1992 followed the liberal-democratic principles of the 1990 constitution. This marked a shift from a one-party, Marxist-Leninist socialist state to a multiparty democracy, including a strong focus on human rights and the rule of law.Footnote 3
Initially, police reform focused on ‘getting right’ the formal state institutions through professionalization and demilitarization. Another aim was to (re)extend state institutions and law to all the rural areas where they had disappeared as a result of Renamo control (Baker Reference Baker2002). This focus on a state-centric rule of law meant that those popular vigilantes, militias and people's courts that were established by the socialist state to work with the police and the military were abolished. The reform also ignored those traditional chiefs and healers who resolved crimes in the rural areas, despite the fact that they had been banned by the government. Implicitly, it was assumed that all these local authorities would become irrelevant once the state system was in place.Footnote 4 This did not happen.
By 2000 a range of initiatives supported the recognition of local authorities. Decree 15/2000 conceded state recognition of traditional chiefs, who, among other tasks, would assist state officials in conflict resolution and policing.Footnote 5 Healers were allowed to practise traditional medicine, although official law still does not recognize witchcraft (West Reference West2005; Meneses Reference Meneses2004). The revised 2004 constitution consolidated the shift away from state-centrism with its recognition of legal pluralism.Footnote 6 There were both pragmatic and political reasons behind this change in official policy.Footnote 7
The politics behind legal pluralism
Unsuccessful police reform made it increasingly clear that state institutions needed, for pragmatic reasons, to collaborate with local authorities in order to do their job. Everyday state policing was still partisan and paramilitary, and proved incapable of handling the rise in crime in the mid-1990s.Footnote 8 The police faced a severe crisis of legitimacy, especially in rural areas such as Sussundenga, where Renamo had held administrative and military control in the last years of the war. Here, support for Renamo continued after the war, as was evident in the election results. Frelimo was confined to urban administrative centres as a result of frontiers created during the war. Until 2001, some places in Sussundenga were still no-go areas for the police, based on a fear of armed opposition by Renamo supporters. Some of the police officers I knew recalled that they feared those areas because of the ‘strong spirits’ there, since Renamo was known to work with healers. The chief of police in Dombe, Southern Sussundenga, remembered that even in the main town: ‘We were not allowed to be aggressive towards people. We just remained seated and if someone did a crime we would call them in politely to the station and tell them that it was illegal, but we would not arrest them.’ Most rural residents mistrusted the police, who were associated with militarized governance and the forced removal of people into urban Frelimo areas (Alexander Reference Alexander1997).
Chiefs were the preferred option in resolving even serious crimes.Footnote 9 They performed these tasks in collaboration with their madodas (councils of elders), their ma-auxiliares (young police assistants) and the wadzi-nyanga (healers). Among these were also Renamo ex-combatants and civilian police, the mujhibas. These ‘informal sovereigns’ (Hansen and Stepputat Reference Hansen and Stepputat2005) did not represent an integrated system of governance, but they did have a common history of opposition to the Frelimo state. Chiefs and healers were banned after independence partly because chiefs were part of the Portuguese colonial system of indirect rule, and partly because Frelimo saw everything traditional and spiritual as an impediment to modernization and national unity (Kyed Reference Kyed2007a). Renamo capitalized on the resistance this rhetoric created, especially among the rural population, and promised the return of chieftaincy if it won the war (Geffray Reference Geffray1991). Operationally, Renamo soldiers relied on traditional healers, rituals and collaboration with some of those chiefs who had not fled their areas. In Dombe, chiefs were even given back their administrative and dispute resolution roles. In exchange, they had to help Renamo in recruiting and in the provision of food to soldiers. Such kinds of collaboration considerably transformed chiefly authority, as they had done during colonial rule. The chiefs had to enforce highly unpopular functions for their new masters while still trying to nurture popular legitimacy. Hereditary lines were manipulated because Renamo and the colonial regime inserted people who were loyal to them rather than those who were legitimized by the ruling families. During the war, for instance, sons, brothers and uncles of original chiefs were given positions (Alexander Reference Alexander1997).
Despite these manipulations, Renamo members shared with chiefs and healers the use of a spiritual idiom of power in their efforts to govern the rural territories. Within this idiom, power derives from the ancestral spirits, who are believed to be indispensable to the well-being of the land and the people, but who, when dissatisfied, can cause much harm and misfortune, including the taking of life (Kyed Reference Kyed2007a). The dark side of this idiom is that the living can draw on invisible evil forces to boost their power, protect themselves or transgress the order of things. As I address in more detail below, this idiom also informed local perceptions of transgressions.
The police were ill-equipped to engage with the invisible forces, but, in order to gain a footing in the rural areas after the war, they had to relate to them somehow. Unofficial collaboration with chiefs by the police and administrators underscored this need. In fact, in the Frelimo-held villages of Sussundenga, alliances with chiefs for information and advice had been present for some time, even during the war. Also, it was well known that sons and brothers of those chiefs who had been banned took up positions in the new Frelimo structures (Alexander Reference Alexander1997). Secretly, some state officials also supported spiritual ceremonies, because they personally believed that such ceremonies could help in situations of famine and intensified insecurity (Meneses Reference Meneses, Kyed, Coelho, de Souto and Araújo2012: 78). In short, despite it being illegal, some state officials were already engaged with the spiritual domain of social ordering as a way of doing their job. This also supported the move towards recognizing chiefs and healers.
Competition for government power in the new multiparty democracy also sparked the drive towards recognizing legal pluralism, which was surrounded by contestations over traditional authority in the 1990s. While Frelimo was internally split on the issue, Renamo argued for full recognition of chieftaincy. However, after the 1999 elections, when Renamo almost won a majority, a number of influential members of the Frelimo government changed their attitude. This was probably influenced by the fact that they were convinced that Renamo's success in rural elections was due to the party's alliances with chiefs. In Renamo-controlled areas, there were also pockets of resistance by chiefs to the re-establishment of state police and administration (Alexander Reference Alexander1997). These issues pushed the last sceptics within Frelimo to agree on decree 15/2000.
Recognition of legal pluralism was therefore driven by both political and administrative concerns, and the dilemmas relating to popular versus state conceptions of law and order were undermined. In fact, legislation supporting legal pluralism relies on an unproblematic relationship between state/legal and traditional norms, thereby ignoring overlapping jurisdictions and claims to authority (Kyed Reference Kyed2009a). In everyday practice, local state officials have therefore had to deal with the dilemmas on their own, and this in itself has conditioned the way in which state authority is constituted.
EMBRACING TRADITION, CLAIMING ‘STATENESS’
The type of fear that police officers highlighted to me when I first conducted fieldwork in Sussundenga in 2002 gradually changed following the state recognition of chiefs. Alliances with chiefs allowed for the setting up of police posts in the most outlying areas, and enabled the officers to move outside the station. Next, I address the two main ways by which the state police, and the local state more generally, tried to regain terrain: public state rituals and juridical institutional ordering. The dilemmas emerging from embracing ‘tradition’ and, at the same time, separating out the state as a superior entity were common to both.
Ceremonial unification and separation
The recognition ceremonies of chiefs during 2002 allowed for the first visit of state officials after the war to many of the areas outside the main villages (Buur and Kyed Reference Buur and Kyed2005). They highlighted the unification of ‘tradition’ and ‘state’. These visits were theatrically staged with rituals of offerings to the ancestral spirits, communicating to the spirits that the chief was now officially installed. State officials drank doro (locally brewed beer) with the elders from the same calabash as a symbol of reconciliation. Yet this unification was broken by acts that hierarchically distinguished state from tradition.
The state delegation arrived in a Land Rover with flashing headlights, mimicking a miniature version of a presidential visit. Security around the official guests seemed almost comical in what appeared to be a peaceful place: four policemen with AK-47 s spread out across the venue, each taking up a corner to oversee the event, and when the officials moved they followed suit to provide protection. State hierarchies of rank and superiority were displayed in seating and organization. In his speech, the district-level Station Commander, who had been in Sussundenga during the war, spoke about how the police were now in charge of resolving all criminal cases (i.e. as opposed to chiefs and Renamo during the war). The District Administrator similarly talked about obeying government orders and respecting the leaders and laws of the government. The physical dressing-up of the chiefs in state paraphernalia and the inscription of their names into the state register also gave the sense that tradition became ‘fixed’ within a state-defined order.
The state embraced ‘tradition’, but only to reconstitute itself as a superior authority. This also had clear political messages. Renamo was portrayed as the enemy of the unity between state and tradition (Kyed Reference Kyed2007a). This was particularly evident in the District Administrator's speech about respect for authority. He drew an analogy between chiefly authority and Frelimo leadership, tapping into the familiar idiom of power associated with ancestral spirits, yet insinuating the superiority of the Frelimo family. The ceremonies ended with a miniature national celebration. Everyone gathered around the chief's new flagpole to sing the national anthem. This was followed by the consumption of local food prepared for the official visitors, marking the new unity between chiefs and state officials.
The recognition ceremonies opened up more consistent state police encroachment into the hinterlands. They were soon followed by public ‘crime combating meetings’ that were held in Renamo strongholds where there were no police posts. The meetings fulfilled two main functions, the Station Commander told me: to teach people about lawful behaviour and to encourage collaboration between the police and the chiefs in combating crime.
Well aware that there was still much hostility towards the police in these areas, the Station Commander began and ended the meetings by singing ‘Cubatana-Cubatana’ (‘Unity-Unity’). The song is about unity between the police and the people, and between ‘the people’, ‘the chief’, ‘the government’ and ‘Frelimo’. It was accompanied by traditional dancing. Speeches conveyed images of a benevolent police force in the service of the people. These aspects were nonetheless interrupted by serious talks about unlawfulness and threats of police punishments. Cyphers of separation also carried political messages. This became apparent when the Commander physically displayed a number of prisoners from the district-level jail, with the stated aim of law education. This was necessary, the Commander told me, because people in the former war zones ‘don't know how to distinguish what is considered crime according to the law because they have lived with the enemy [Renamo]’. Underlying the display was therefore a politicized delineation of the ‘good citizen’. The prisoners embodied the significant ‘other’ of the political community. This was illustrated by a discourse that combined Frelimo war rhetoric on Renamo as the ‘internal enemy’ with the local idiom of evildoing. The prisoners, like Renamo, were, in the words of the Commander, equated with ‘the bad that has still not been taken away from the war’. He linked this to the crocodiles and leopards that attack secretly at night; in the local idiom of evildoing, these are associated with witches (umroi), the very sources of evil, who make themselves invisible to destroy the community from within. Disguised as animals, the umroi roam around at night and devour close kin so as to gain in strength. The only way to get rid of such evil is to remove the umroi from the community.
Thus, while using state definitions of legality, the Commander also tapped into familiar notions of community belonging and evildoing as elements to outline ‘the good citizen’. He could do this because he had lived in Sussundenga over the past ten years and had learned how Renamo also drew on spirits to wage war. His discourse also reflects a long history of a Frelimo party state where criminals have been equated with internal political enemies. The Commander himself was a strong Frelimo supporter; in fact, he told me, to him it was unthinkable for a police officer of his rank to be a Renamo supporter. Simultaneously, he strongly supported working with the chiefs.
Crime combating meetings were about both gaining police legitimacy and reclaiming territorial sovereignty in enemy zones. This required a careful balancing between enacting unity and marking out the distinctiveness of the state. Merging references to state law with local notions of evildoing did not simply slip into an instance of hybridity, but coexisted with deeply political aspects of boundary marking.
Boundary work: jurisdictions and prohibitions
Institutional (re)ordering of the jurisdictions of the providers of justice and policing constituted a second way in which the Commander and his chiefs of police tried to expand everyday policing operations. This was combined with the outsourcing of policing functions to chiefs, and later to community police.Footnote 10 Chiefs were obliged to localize ‘troublemakers’, arrest them, and bring them to the police station. Yet they were strictly prohibited from actually resolving crimes and from using physical force. If they failed to abide by these rules, they were punished by the police with physical force or with days spent in a cell, as three cases from 2004–05 illustrated.
Collaboration with chiefs thus coexisted with attempts to claim state police monopoly over the use of force, and over defining and prosecuting crimes. Criminal cases included homicide, fights in which blood was spilt, rape, stabbings, larger thefts involving the use of weapons and violence, drugs and arson. This was explained by police officers at meetings with chiefs and elders. They were also told that problems not defined as ‘criminal’ were to be resolved only by those authorities recognized by the state, i.e. the chiefs and community courts. Such cases were called ‘social’ and ‘traditional’ cases. The former included adultery, smaller fights, minor threats and slander, divorce, debt and land disputes between neighbours. The latter covered specific customs (for example, violating sacred places and marriage payments or lobolo) and, intriguingly, also included uroi (witchcraft) and vulí (evil spirit possession), although accusations of such are illegal. Thus the police redefined ‘traditional cases’ by prohibiting chiefs from resolving those transgressions that were defined as a crime, such as ‘the taking of life’. Simultaneously, they recognized illegal parts of ‘tradition’ such as witchcraft accusations, and allowed wadzi-nyanga to handle them. This combination of embracing ‘tradition’ and specifying state sovereignty, as in the ceremonies, recurred as a paradox in everyday policing. The ideal distinction between ‘traditional’ and ‘state’ jurisdictions was constantly faced with obstacles. In fact, chiefs continued to decide criminal matters, while police officers started to receive a growing number of ‘traditional’ and ‘social cases’ at police stations.Footnote 11
EVERYDAY POLICING: JUSTICE OUTSIDE THE COURTS
In ‘social’ and spiritual matters, people went to the police as a sort of appeals institution when a resolution at a chief's banja or community court did not materialize or when one of the parties was dissatisfied with the outcome. On fewer occasions, victims came to the police as a first choice because they did not believe that their chief could make the accused turn up for a hearing.
Rather than sending people back to the chiefs with their social and spiritual problems, as was supposed to be the rule, the commanders at the Dombe police post and the Sussundenga station where I carried out my fieldwork developed quite set routines for dealing with non-criminal cases. Despite some individual differences, more junior officers adhered to the same routines. They assisted people first by ensuring that the accused was ‘brought to trial’ in the sense that they appeared at the station. Secondly, they helped to implement verdicts by facilitating a negotiated settlement between the parties.
The first thing the police did when a person came to them with a case was to issue a notification that obliged the accused to appear for a hearing on a set date. This resembled a common practice of the chiefs’ banjas. It nonetheless differed because the notification was authorized with the police's official stamp, which had a strong enforcing power. In one case, for instance, an elderly man, João, had travelled more than 100 kilometres to reach the Dombe police post with a case concerning his fifteen-year-old daughter who had run away with a man and had become pregnant. The man refused to pay lobolo (bride wealth). Although João knew that this was a case for a chief, he did not believe that his ‘in-laws’ would turn up to pay at the chief's banja. The police officer listened patiently to João and then asked: ‘What is it that you want?’ João just wanted the lobolo. The officer wrote down the name of the accused and the date when he should appear at the police post, adding that ‘if he does not come we will arrest him and educate him’. He moved his hands to show that by ‘education’ he meant a beating. After the hearing, João said to me: ‘He [the accused] sees the notification from the state and then he will be too afraid not to turn up. You know, as he [the officer] said there the police will sjambokear [beat] him if he does not come.’ Thus notifications were seen as an order to which was attached the threat of sanctions enforced by the state police, which notably involved the use of force.Footnote 12
Besides this enforcing power of the police, João had also decided to go to the police because it was expensive to use the chief, he told me: ‘Although you need to pay transport, the police do it for free.’Footnote 13 However, ‘If it turns out that the “in-laws” refuse to pay,’ he explained, ‘then we have to go to the chief so that he can send us to the nyanga, who can see who made my daughter pregnant.’ In other words, João did not turn to the police because he believed they could decide on the veracity of the case; they could simply pressure the accused to turn up for a hearing and make an agreement.
In fact, the officers refused to enforce the verdict. So although the payment of lobolo in João's case was agreed at the station, it had to be overseen by the chief. The notifications were like ‘tickets’ that could be ‘cashed in’ at a banja. The police also ‘returned’ uroi cases that required nyanga consultations to the chiefs, because, as was explained, ‘only a chief can send people to a nyanga’. The notifications provided no guarantee, but they put extra pressure on the parties to implement a verdict.
During the hearings, the officers acted as mediators, supporting resolutions that were proposed by the parties themselves. In doing so the officers adjusted to the parties’ own notions of appropriate justice, rather than basing judgments on written law. However, when trying to convince the parties to abide by a resolution, the officers did refer to state law. They often did so interchangeably with warnings of the potential risks of uroi if the accused did not abide by a verdict. A good example of this was an old divorce case. It ended at the police station because the baby boy of the divorced couple had become ill due to vulí. The father's family had taken the baby, accusing the other grandparents of having ‘sent’ the vulí. The mother's father, Antonio, was furious and wanted the baby back. They had already won the case at the community court, but the others had refused to abide by the court's decision. So they had come to the police to put extra pressure on the father's family, they told me. After each family had spoken, the officer stated: ‘According to the law, a baby of this age has to be with the mother.’ The father then explained to the officer that if this happened the baby would again be possessed. He was clearly concerned. The officer now changed idiom by warning the father of future uroi if they kept the baby: ‘You have to give it back, because if something bad happens to the baby you could be accused of these things of tradition.’ The father then agreed to hand over the baby.
These switches between ‘law’ and ‘tradition’ suggest that, although officers drew on local notions of order and justice, their capacity to facilitate resolutions also relied on enacting the distinctive authority of the police as representatives of the state. This was confirmed in my conversations with people; 85 per cent of my sixty interviewees stated that people turned to the police because they were quicker than the chief. By ‘quick’, they meant that the police, as Antonio had said, ‘can make people turn up and pay, because they beat people and put them in the cells’. In this way, the particular authority of the police was associated with the state police's de jure monopoly on violence. Another common explanation for going to the police was, as João noted, the question of money. Although the police in Mozambique are notorious for corruption, this did not relate to the resolution of social cases and petty crimes in Sussundenga. In fact, in some cases the police could be a way out of a vicious circle of payments to chiefs, healers and contenders.
Elias, for example, came to the police because he was desperate after having spent 1,200 meticais (US$40) in fees and had been imposed a set of fines since his wife, Inês, had been accused of bewitching his nephew. The case had led to various hearings by the chief and several nyanga consultations until the nephew died. Each time they had to pay a fee. An old woman was initially accused of being behind the tragedy, but when she was acquitted by a nyanga, the chief imposed a fine on Elias and Inês for false accusation. The parents also wanted soro u mondu (the price for taking a life). Believing that his wife was innocent, Elias refused to pay the fines and asked the chief to see another nyanga, but his request was denied. After this, Elias and Inês were threatened by the parents of the deceased boy. This is when they went to the police. Elias did not expect the police to resolve the witchcraft case, but hoped that the police would pressure the chief to allow them to see another nyanga so that Inês could go free. The police officer called in the chief and the other contenders for a hearing.
Whatever the individual cause, bringing cases before the police only really made sense because the police adjusted themselves to local perceptions of transgressions and demands for justice. Similar dynamics were involved in the conclusion of criminal cases outside the courts. A core difference was that here the police did enforce the verdicts, which included compensation payments for stolen goods and fines in cases such as rape, physical aggression and arson. This was done in direct response to the wishes of the victims and their families, just as I observed in the banjas. The police commonly gave the option to open a criminal process but then allowed the parties to have the case resolved differently at the police post. It was also common for the police officers to allow the spiritual idiom of evildoing to influence the handling of crimes.Footnote 14 In one case, for instance, the police considered that a man accused of arson had been possessed by an evil spirit, which had been sent by a neighbour who had an unresolved land dispute with his father. The officer on duty judged that the man should help rebuild the house, but referred the case relating to the spirit to the chief, thereby recognizing that the neighbour was also a perpetrator. On some occasions the police also allowed a crime to be reclassified as a traditional case in response to the requests of the victims or their families. This happened in a rape case, where the parents of the victim during the hearing decided that what they really wanted was for the perpetrator to pay a fine for ‘taking the virginity’ of their daughter without prior marriage arrangements. The police officer initially responded by threatening the perpetrator with imprisonment, explaining that rape is a crime punishable by law, but subsequently he gave in to the wishes of the victim's parents.
What made the police different from chiefs in such criminal cases was that they often combined compensational justice with types of punishments that rural residents associated with the police as representatives of the state: physical force, public work for the police and/or detention in the cells. These extra-legal punishments were not seen as an alternative to compensational justice, but as an addition to put extra pressure on the perpetrators to pay up. In short, the police were drawn into the local ideas about justice and order, yet were expected also to act as the state – a state associated strongly with the use of force.Footnote 15 At the same time, people repeated to me that the state was not able to resolve their spiritual problems.
Now, the question is: why did the police officers go through the trouble of all these endless case resolutions outside their official mandate?
Legitimacy and partial sovereignty
Like all the other officers I spoke with, the district-level Station Commander, Senhor Nito, stated again and again to me that the police cannot deal with uroi ‘because it is outside the law’ and there is no evidence. Simultaneously, he was of the opinion that ‘if the practices of uroi diminish, then I also think that crime will diminish’.Footnote 16 When I asked him if he had been bewitched himself, he said no, but he confirmed that police officers could indeed be bewitched, adding:
but I am not afraid, because I chose to work for the people and here in these areas there is a lot of witchcraft … I never went to a healer, I was never bewitched, but when I first came here I was sick for three months and many people thought that it was the witches who came to visit me. I could not walk, my whole body was lame. I prayed to God the whole time.Footnote 17
The police were drawn into spiritual matters because they had to adjust themselves to local conceptions of transgressions in order to be relevant and legitimate. To some extent, police officers such as the Commander shared such conceptions, or at least accepted their profound significance in the area, even if they were from another province, as was Nito.
In Sussundenga, even visible transgressions were associated with the realm of invisible evil forces. Any wrongful act or conflict can potentially have an invisible dimension in the form of a spirit possession, or it can result from or lead to an invisible act of evildoing manifested through illness, misfortune or death. For instance, in the case referred to earlier, the man was believed to have committed arson because he was possessed by vulí due to his father's unresolved conflict. In other cases, uroi was interpreted as revenge over visible crimes or social wrongs that had not been resolved in the right way. Also, any failure to deal with the perpetrator of uroi could lead to counter-uroi with the help of the wadzi-nyanga, or could imply criminal self-redress. Failure to resolve cases could cause a vicious circle of evildoing (kushaisha). To end this vicious circle, justice must be dispensed. In Sussundenga, justice meant compensation to the victim so as to ‘return’ what had been taken or destroyed (material items, virginity, wife, land, etc.). When vulí and uroi were involved, justice also required exorcism by a nyanga to ‘remove’ what had been sent. This also cured the person inflicted. Only in extreme cases, such as the repeated committing of serious crimes or uroi inflictions, was the removal of a perpetrator from the community seen as a desirable form of justice – i.e. imprisonment, eviction or ritual killing. In fact, imprisonment of the perpetrator was seen not only to conflict with the principles of compensatory justice, it also potentially reinforced a vicious circle of evildoing. The source of evildoing remained with the perpetrator or inflicted a family member; for example, the mother of a convicted murderer became possessed by an evil spirit because the family of the deceased had not been compensated.Footnote 18 To end the circle of evildoing, the parties must return to the traditional domain of justice enforcement, because only here can both the visible and invisible dimensions be dealt with.
This explains the difficulty of separating cases into criminal, traditional and social ones. People do not use separate categories, but rather have one word for problem (ndava) and one for transgression or evildoing (kushaisha). When helping to resolve uroi or a social dispute, the police were therefore also engaging in crime prevention. Whether they themselves believed in the spirits or not was a personal matter that was never made fully clear to me. When directly asked, all the officers denied such beliefs, but this is not unusual; as West (Reference West2005) also shows, revealing that one knows about evil spirits makes one potentially complicit in evildoing or at least vulnerable to attacks.
Unlike the Commander, the chief of police of Dombe, Samuel, a slightly younger officer, was born in the area where he worked and was divided on the issue. Having finished training only in 1992, he had fresh memories of the official role of police officers as law enforcers, and was quick to deny that his officers resolved witchcraft. They represented the law, not tradition. But then he went on to explain that when indeed they did receive such cases, he did not send people away or discipline his officers for getting involved with such matters, ‘because it is the police's job to educate people about the law’. What he meant was that the police helped to resolve such cases in order to prevent criminal revenge: ‘It is our tradition. Every area has their working traditions. So many people come here with witchcraft. And if you do not resolve them there will be problems of crime, so we teach them so that they do not go and beat up or kill.’Footnote 19 This familiarization with ‘the people’ and ‘our tradition’ coexisted in the self-representation of Samuel with a distancing of himself as a state representative, and as someone who was more educated and civilized. This was reflected in the constant reference to the law when explaining why he allowed the handling of uroi cases. Questions of police legitimacy in this context merged with a war rhetoric representing the Dombe population, or rather those who had remained during the war, as being less civilized because they had been exposed to Renamo governance. It was the job of police officers like Samuel to come back home and re-educate the people. After he lost his parents in the war, Samuel had fled to live in Frelimo-controlled areas, and he was among the first group of policemen to come back to Dombe in 1996 to set up a police post, which initially was unsuccessful due to attacks by Renamo ex-combatants and chiefs. He explained:
We need to show that the police are there for the people. This is very important in these areas, you know … where some of the people have had this thing of not collaborating with the police because of the war. It is to regain the trust of the people. Some of them have lived in anarchy, with the enemy so many years. We have to now teach the citizen what is crime and what is not. To educate them about the law is our job.Footnote 20
The resolution of crimes outside the courts may seem paradoxical given the strong emphasis on ‘the law’ in statements like this one, and also when one considers that the police got no monetary gains from all those petty crimes they resolved. Here, spiritual beliefs merged with officers’ consideration for people's survival and well-being (Igreja Reference Igreja, Kyed, Coelho, de Souto and Araújo2012). Policemen like Samuel knew that if a person from his area went to prison, the victims would lose compensation for their losses or injuries, and potentially this could lead to new crimes. People are not insured privately; they have no other choice but to negotiate with the perpetrator of the crime. Police officers in posts such as Dombe are not well-off people with private insurance, as salaries are low. So they understand local economic needs, and because of this they also gain popular legitimacy.
Efforts to gain popular legitimacy could not, however, be divorced from the competition over authority with chiefs that the police faced, both as individual officeholders and as part of an institution representing the state. When people took cases to them, this in itself was an act that recognized the specific enforcing power of the police as ‘the state’. As Samuel noted, this made the police more powerful than chiefs:
It is clear that people do this [take uroi and social cases to the police] because the police have instruments that can ensure obedience to law and order. When a person is notified by the police, he becomes very scared because he knows that if he does not obey he will end in prison … or be educated [beaten] … and the chiefs do not have any instrument really to make the undisciplined fear them. We are the ones who have the power to deal with those undisciplined.Footnote 21
In this statement, the paradox resurfaces of how police officers were compelled into operating outside their legal mandate in order to enact state sovereignty – i.e. with the demonstration and threat of force. This is because their authority, I suggest, remained precarious in the face of the chiefs. To show that they had the upper hand they had to tackle crimes and other matters on the spot, in the local arena of authority. That this could develop into personalized power games was apparent in 2005 when I followed a police officer, Tobias, who had been posted in an area near the old Renamo base. His predecessors had not lasted long, but Tobias was determined to win territory. When he caught a sub-chief failing to report an arson case to him, he had him arrested. For a whole week the chief was punished in public: he was made to carry heavy building material for the new police post. ‘What he did was disobedience, because everyone knows what cases go to the police,’ Tobias told me. ‘I had to put him in line.’Footnote 22
The police had the state's instruments of force and this clearly strengthened their position. However, when it came to dealing with invisible forces it was still the chiefs who had the upper hand. The police facilitated the settlement of uroi but could not resolve them. Even when a crime was involved, state police authority was inadequate. It was only by dealing with the invisible forces that order could be restored. State officials were also not immune to evil forces, as the Commander said. In fact, in 2006 the Police Chief, Samuel, died, and this was widely believed to be due to uroi. He had become too involved in local problems, I was told. His successor helped to set up an office for the wadzi-nyanga next to the police post in 2009, which meant that the police did not have to rely on sending people to chiefs for divination and exorcism. Perhaps the new chief of police also thought it could give him better protection against uroi.
In sum, the unofficial practices of the local state police were, I suggest, ultimately founded not only on the normative and procedural inadequacy of official law in dealing with problems of disorder, but also on the coexistence of different articulations of sovereignty. Here uroi, as part of the wider domain of invisible forces that secure societal well-being, yet that can also take life, constitute an alternative articulation of sovereignty that remains untameable by the state (Bertelsen Reference Bertelsen2009: 138). I conclude this article by reflecting on what this implies for the conceptualization of state authority in former war zones such as Sussundenga.
CONCLUSION
In his study of legal pluralism in Mozambique, Santos (Reference Santos2006) introduced the concept of the ‘microstate’ to describe the forms of hybridity that occur at different levels of the state apparatus. Microstates are characterized by having their own combination of different operational logics, because they combine different political and legal cultures, such as the traditional, the former socialist, and the current liberal-democratic cultures. This happens because of the inability of state institutions to guarantee their own efficiency by relying on just the formal procedures and the codified law of the present (Santos Reference Santos2006: 50–4). Such ‘legal hybridization’ challenges ‘conventional dichotomies to the extent that legal practices frequently combine the opposite poles of the dichotomies and contain an infinite number of intermediary situations’ (Santos Reference Santos2006: 46).
The concept of ‘microstate’ and the notion of hybridization capture well how state police officers in Sussundenga combined different norms and procedures that they otherwise saw as ‘opposite poles’. Officers adjusted to local justice demands by enforcing compensational justice and addressing spiritual concerns, yet enacted ‘stateness’ by using official stamps, instruments of force and references to state law. As such, the police can be considered a sort of ‘hybrid’ authority: their legitimacy rests on mixing different logics and registers, beyond those associated with ‘the state’.
Missing from Santos's definition of the microstate, however, is the deeply political foundations of hybrid practices, and in particular the significant role of violence. As Bertelsen argues (Reference Bertelsen2009), Santos assumes a benign form of state and non-state complementarity. While he recognizes the ‘conflict and tension between different legal orders’, his interest is in how interactions – between chiefs and state officials, for example – are open to mutual influences (Santos Reference Santos2006: 45). He does not explore how such interactions may also be used to consolidate particular power positions, such as was the case with police officers in Sussundenga. While the officers drew on assistance from chiefs and spiritual notions of justice, they also set themselves apart as superior state representatives, at times by violently punishing the chiefs who competed with them. Such boundary work also tends to be undermined in the recent policing and security governance literature that emphasizes hybridity, nodes and assemblages (Wood and Shearing Reference Wood and Shearing2007; Loader Reference Loader2000; Jones and Newborn Reference Jones and Newborn2006; Lippert and O'Connor Reference Lippert and O'Connor2003; Abrahamsen and Williams Reference Abrahamsen and Williams2009). The point of departure for these studies is the fragmentation of security governance away from the state. The focus has been on trying to understand how the plurality of auspices and providers – private, public, community-based – interact, compete and collaborate. Whether conceptualized as nodes in a network or as security assemblages, a shared conclusion is that what are developing ‘are not clear-cut hierarchical or vertical relationships, where power and authority runs in only one direction, or from one particular and clearly defined centre’ (Abrahamsen and Williams Reference Abrahamsen and Williams2009: 8). While this certainly also applies to Sussundenga, it is important to take equally seriously the constant efforts of police officers to enact hierarchies and to centralize power. Because of a tendency to assign sovereignty to the state alone, the literature referred to above, I suggest, overlooks the ways in which hybrid practices and assemblages coexist with competing sovereign practices (Stepputat Reference Stepputat, Bouziane, Harders and Hoffmann2013). This is important for understanding the extra-legal violence of the police in Sussundenga.
Microstates, I suggest, emerge in the interstices between police officers’ efforts to consolidate their own power positions within a given local arena and their engagement with the larger project of claiming de jure state sovereignty. As Das and Poole (Reference Das and Poole2004) argue, the extra-legal practices of state officials are made effective because they are also able to enact the ‘supposedly impersonal or neutral authority of the state’. At the same time, the extra-legal practices ‘represent at once the fading of the state's jurisdiction and its continual refounding through its appropriation of private justice’ (Das and Poole Reference Das and Poole2004: 14). The flipside is the prevalence of violence within everyday state operations. This is because state sovereignty remains partial and therefore needs to be constantly re-enacted (Comaroff and Comaroff Reference Comaroff and Comaroff2006). In Sussundenga, state sovereignty was partial, because the police never really managed to master the invisible forces that yield power over life and death and in this sense constitute an informal sovereign (Hansen and Stepputat Reference Hansen and Stepputat2005). The point is that hybridity is conditioned by the existence of partial sovereignties. Hybrid practices coexist with the sort of boundary work that goes into enacting sovereignty – in this case between the state and its ‘other’, i.e. tradition and the invisible evil forces associated with it.
In everyday practice and at public ceremonies, the police officers constantly switched between embracing the ‘other’ and setting it apart from the state. Acts of boundary marking do not convey equality or shared sovereignty, but rather the demarcation of hierarchies between overlapping claims. They are structured around efforts to assert superiority and, to borrow from Jean and John Comaroff's (Reference Comaroff and Comaroff2006: 35) definition of sovereignty, ‘autonomous, exclusive control over the lives, deaths, and conditions of existence of those who fall within a given purview’. This aspect of exclusive control, I suggest, helps us to understand the paradoxes faced by state police officers in legally plural contexts such as Sussundenga: they are caught between alternative articulations of sovereignty embedded in invisible, untameable forces and the quest for state sovereignty. In practice, this makes the local police into hybrid ‘microstates’. However, there is a need to re-politicize these concepts so that an emphasis on links, mergers and assemblages does not overshadow the deeply contested and often very violent ways in which different policing actors attempt to consolidate their power. In short, we should be alert to the boundary work that coexists with different historical layers of hybridity or ‘twilight’ articulations of authority (Lund Reference Lund2006). In doing so, it is important to acknowledge the resilience not only of culturally embedded notions of power and (dis)order, but also of the idea of the state as the centre of sovereignty despite its obvious fragmentation in practice.
ACKNOWLEDGEMENTS
This paper was first presented at a workshop on African policing at Oxford University on 16 and 17 May 2013. I wish to thank the other workshop participants for very useful comments. Special thanks also to the anonymous readers of this journal and to Finn Stepputat, Mark Duffield, Louise Wiuff Moe and Robin May Scott for additional feedback.
