‘Outsiders’ and ‘Insiders’: Post-Conflict Political Violence and Reconciliation in Malanje, Angola

This article focuses on some episodes from the prolonged Angolan conflict that can potentially broaden our understanding of the past. The analysis is centred on and describes the ways in which this traumatic experience has been represented in the official domestic narrative, relegating local or regional dynamics – such as in Malanje – to the margins. In my opinion, these dynamics are nonetheless of major importance in a discussion of the post-conflict period in Angolan society. The argument is based on the assumption that, although there are always contradictions between local micro-narratives which affect the processes of negotiation,1 the micro-narratives presented in this article assist in the interpretation of the official political-military rituals that are carried out by the Angolan state as an act of appeasement towards the past – with varied and sometimes contradictory implications for the process of national reconciliation. Introduction This text analyses the context of political violence in Malanje, one of Angola’s eighteen provinces, an area traumatised by an armed conflict that laid waste to the entire country. The analysis is a contribution to the theoretical perspective developed by Norbert Elias and John Scotson on the tensions and spaces of belonging. We rely on Elias and Scotson’s approach in an attempt to understand the categories of ‘outsiders’ and ‘insiders’ as they manifested themselves during the armed conflict in Malanje, and we start from their local meanings as an expression of difference within a space of belonging. Elias and Scotson argue that the notion of ‘the established’ is defined by the relationship that a particular individual establishes with another individual, or that exists between groups of individuals who exercise authority within their own environment, and are able to influence others fundamentally through more or less heterogeneous 1 V. Igreja, ‘Negotiating Relationships in Transition: War, Famine, and Embodied Accountability in Mozambique’, Comparative Studies in Society and History, 61, 4, 2019, 774–804.


Introduction
This text analyses the context of political violence in Malanje, one of Angola's eighteen provinces, an area traumatised by an armed conflict that laid waste to the entire country. The analysis is a contribution to the theoretical perspective developed by Norbert Elias and John Scotson on the tensions and spaces of belonging. We rely on Elias and Scotson's approach in an attempt to understand the categories of 'outsiders' and 'insiders' as they manifested themselves during the armed conflict in Malanje, and we start from their local meanings as an expression of difference within a space of belonging.
Elias and Scotson argue that the notion of 'the established' is defined by the relationship that a particular individual establishes with another individual, or that exists between groups of individuals who exercise authority within their own environment, and are able to influence others fundamentally through more or less heterogeneous values or principles. Outsiders, however, are the individuals or groups considered foreign to the environment, possessing values or relational links that are diffuse or less intense than the others. 2 The terms 'established' and 'outsiders' are understood here as denoting those groups of persons who are in relations of conflict based on a refusal to recognise their ties of belonging, or their links to a particular 'social scenario' . Both these groups see themselves as foreign to the social environment in which they move, which in this case creates tensions around identity, tensions that are marked by the social divisions of the colonial past and aggravated by the 'self-justifying narratives' of the armed conflict.
This is to say that the social divisions that already existed in Malanje re-emerged and were exacerbated during the conflict from 1975 onwards -between rural and urban people; between speakers of normative Portuguese and speakers of demotic Portuguese with Kimbundu influences; among supporters of the ruling MPLA party and its military wing FAPLA; between Methodists and Catholics; between people born in Malanje or who settled there before 1975 and those who arrived in 1992; between those who campaigned for and those who campaigned against the ruling party; and among the friends and relatives of party members. These defined categories are considered by 'insiders' to be markers of social hierarchy, inasmuch as their opposites -being born outside the city; being or having been a supporter of the FNLA or UNITA; having settled in Malanje after 1975 or after 1992; having campaigned for UNITA, the FNLA or another party; having friends or family members related in some way to UNITA (above all); speaking Portuguese with Umbundu influences; or even being a natural of Malanje but a supporter or member of UNITA -all these are considered to be the defining markers of the 'outsider' .
It is important to acknowledge that the denial or recognition of 'belonging' and the labelling of insiders and outsiders according to the above mentioned criteria, is in fact a more complex phenomenon than we have space to recognise here.
In this text we adopt a qualitative approach with an interdisciplinary perspective and a special emphasis on micro-sociology. The article is divided into four sections. The first of these briefly frames the analysis of the violent conflict. The second section contextualises -without going into details -the way in which the city of Malanje became a terrain for political-military conflict during Angola's socialist period, and describes the first confrontations in the city centre between the so-called insiders and outsiders. Third, the article analyses the city as space, and how confrontation crystallised its divisions. The fourth and last part points to the effects of war and reconciliation as initiatives of appeasement without, however, assessing what this meant in the lives of the people.
In the context of a war that took place in urban areas and their peripheries, as was the case in Malanje, the concepts of combatant and civilian become muddled. Indeed, their relevance is tenuous and there is only a fragile border separating them.
In this case, the notion of a distinctive identity is conceptualised from the work of Elias and Scotson, which opposes some groups against others using broad markers of constructed identity, or definitions of the 'national' following the ideas of Benedict Anderson. These include 'national consciousness' , the demarcation of territory, language as spoken formally or locally, and memory and forgetfulness. 3 This helps us to broaden our understanding of the local implications discovered in the experience of Malanje, and of the adversity caused by the war, which placed entire families in situations of confrontation, and turned them into combatants and members of militias. 4

On micro-politics and violence in African armed conflicts
The discussion of violence dates from the nineteenth century and, in its turn, the theory of violent conflict dates from the end of the Cold War, which opposed the two dominant superpowers in international politics. In subsequent decades, in the 1970s and 1980s, the focus of such studies has been on conflicts that erupted in Latin America and the African continent, due to defeats in wars in Vietnam and Afghanistan. With particular emphasis on the African continent, the paradigmatic cases that have influenced studies of violent conflict have been, among others, the Rwandan genocide, 5 the wars in Sierra Leone, 6 Liberia 7 and the DRC, 8 and the protracted conflicts in Angola 9 and Mozambique, 10 to mention only a few.
In a broader framework, studies of armed groups in civil wars -such as those carried out by Stathis Kalyvas and Jeremy Weinstein -aim to analyse 'the particular logic of violence in civil wars' , arguing that 'patterns of violence must be understood in the context of the armed groups' relation to the so-called civilian population' . 11 Kalyvas views civil wars as triangular dynamics between incumbents, insurgents and civilians, defining them as 'armed combat within the boundaries of a recognized sovereign entity between parties subject to a common authority at the outset of the hostilities ' . 12 Certainly, looking at methods and paradigms, Kalyvas 'insists that the dynamics of civil wars cannot simply be inferred from politically or ethnically informed 'master cleavages' . Such conflicts violently confront people who share a history of non-violent interaction, who have been neighbours, friends or members of the same family. ' 13 For Kalyvas consideration of this question must include an analysis of local context and inter-community dynamics. We are dealing with divisions between the political and the private, the collective and the individual, and this therefore becomes a central feature of the logic of violence in civil wars. Kalyvas sees the war for the independence of Angola between 1961 and 1975 as similar to the war in Algeria. The conflict that broke out in Angola in 1961 took on the characteristics of irregular warfare, and this defined the response of the colonial authorities, who mobilised an army of 80,000 men. 14 In Malanje district alone 5,000 people were massacred. 15 In this context, it is interesting to note -by way of example and by no means as an exhaustive list -some episodes, some micro-narratives that illustrate the local implications of the Angolan conflict. It should be noted from the beginning that this article is only a brief attempt to broaden discussion of the war in Angola.
Teresa Beck researched the experiences of the war in the central highlands and focussed specifically on UNITA's political project and on the lives of the organisation's ex-combatants before and after demobilisation in Huambo province. She argues that the civil war gave meaning to the lives of these individuals and, when it ended, they found themselves in a situation that was close to abandonment. Given the similarity of these studies to the specific context of Malanje, and because their focus is on the post-war period, my article will follow along the lines defined by these two authors and proceed in dialogue with them.
Beck states that 'empirical accounts of contemporary conflicts suggest that the expansion of war into the sphere of everyday life is one of the most salient characteristics of violent conflict' . 16 Indicators for the expansion of conflicts include the blurring of boundaries between groups of actors as well as the disappearance of the battlefield in the classic sense of the term, processes that in many cases are characterised by a condition of being neither ongoing, nor concluded. 17

Malanje during the socialist period
After independence in 1975 the form of state, which prevailed in Angola, tried to articulate a political discourse that derived its base of legitimacy from urban and rural areas, of which the latter had been the most useful segment in the national liberation struggle against the Portuguese colonial system. Securing government control over 12  agricultural production all over the country was crucial for maintaining the war effort. It was in pursuit of this policy objective that the then Prime Minister of Angola, Lopo do Nascimento, launched a mobilisation campaign in Malanje during his visit to attend the first local party seminar in the province, in 1975. As a senior party and state official, Nascimento was the bearer of an important political message: the peasantry and the provincial structures were to bring together agricultural production on the one hand, with the protection of the local territory from attempts to establish control by rival political groups on the other. The top-level party and state structures were concerned that a lack of mobilisation of peasants in rural areas, as well as in the cities, would facilitate an infiltration by activists or supporters of the two nationalist movements that opposed the MPLA. Alongside this concern, the leadership was worried about how party structures at local level were functioning: according to Nascimento, they 'were unable to put into practice the political line defined by the party' . 18 Malanje became a space for political experimentation at a time when it was deemed necessary to re-mobilise the predominantly peasant civilian population. Nonetheless, it was well-known that the global guidelines outlined by the party and the central organs could not be applied either because of incapacity or incompatibility. This showed that although these functioned as an ideology at the top level, the most obvious contradictions of this socialist model were at the bottom. The socialist model relied on coercion and proved to be inconsistent with the local aspirations of the peasantry in Malanje. In addition, it threw into the mix effective control and the imposition of a way of life and livelihood that created forms of dependence on state structures -quite apart from bungled implementation and other distortions, and the silencing of criticism that occurred on the orders of the state. 19 The maladjustment to which we refer was noted by Brazilian researcher Valdir Sarapu when he analysed the state policy for what he termed agricultural cooperatives on the outskirts of Luanda and Malanje, seeking to organise the peasants so as to increase agricultural production in the green zones. 20 The city of Malanje, like the central highlands, suffered through the different periods of the armed conflict that affected the country rather more than it did the capital. The population had first-hand experience of devastation. The first period occurred at a difficult conjuncture of the socio-cultural contradictions that were characteristic of Angolan society in general. In Malanje communities were separated by the distinction between the 'concrete city' in the centre and the surrounding musseques or townships.
This first period -the so-called 'twelve-day war' in 1975 -affected the province but particularly the city of Malanje. 21 It opposed the military forces of the FNLA supported by the Zairian army, against the MPLA supported by Cuban fighters, and provoked an exodus of the white population and an occupation of the city by the local black population. According to Manuel Gabriel, … the city and almost the entire district had been the scene of frequent and calamitous confrontations between the two military and political movements -the MPLA and the FNLA -which were struggling for power. The MPLA eventually supplanted the FNLA, whose forces withdrew, mainly to Kwanza-Norte and Uige. The European population was generally oppressed and threatened -in Kalandula several Portuguese were murdered -and many eventually fled to Luanda and Huambo where they awaited transport back to Portugal. Many Angolans accompanied them. 22 Socio-economic activity came to a complete halt, since there was no recognised authority capable of maintaining it. Meanwhile, fighting continued, as the following testimony explains: In 1975, in the city of Malanje, there was the armed conflict between the forces of the MPLA and the FNLA. The battle lasted for 12 days, after Agostinho Neto requested reinforcements from Cuba. The FNLA was then pushed out of the city of Malanje. During this conflict, the FNLA used slingshots and the MPLA used AKM's. There was a big man, an MPLA major called Canhangulo, who launched the offensive and succeeded in ending the FNLA in Malanje. But they did not lay mines, like they did in 1992. At that time the sky in the city turned red from all the shooting and bombing … to save the families, Major Canhangulo ordered a withdrawal for about 25 kilometres. Then the population began to loot whatever the whites had left behind, such as shops, machinery, houses, furniture and even food, etc. Because at that time the settlers had also left the city because of the fighting between the FNLA and the MPLA … Later, many dead bodies appeared on the city streets. MPLA troops had to mobilize the population to bury the bodies scattered around the city after the looting of the settlers in Malanje. In this way, within 12 days, the city of Malanje was liberated from the forces of the FNLA. 23 While the fighting continued, a considerable proportion of the population that had occupied the city fled to the rural areas or to nearby towns; others walked to Luanda or to Zaire (now the Democratic Republic of Congo), seeking safety and protection. From very early on Malanje province proved to be a political testing ground because of its economic and strategic importance, located as it was at the intersection of road and railway routes from the far north and the coast, as well as from the provinces of the central highlands and the east. On the other hand, from an economic point of view, the city and province of Malanje had considerable agricultural potential and had been important in terms of the balance of payments and imports.
After independence, the rural character of Malanje led the MPLA-PT to create bases of support among the peasants, on the assumption that the war against its internal opponents would be fought in the countryside, since it was much more difficult for them to strike at the cities where the MPLA dominated. Ever since the uprising in 1961, the small groups of peasants scattered across the provincial fazendas had provided an opportunity for the MPLA to use them and their localities as a political laboratory, so that the struggle could continue after independence. In this way the slogan 'produce and resist' functioned to mobilise the population and it was on that basis that the so-called people's vigilance brigades were created.
Yes, Malanje was a political laboratory in terms of mobilization around agricultural policy in the context of the period after the military skirmishes between the MPLA and the FNLA, and seven years later between the government and UNITA. It was from Malanje that the first agricultural experiment immediately following independence was launched. So much so that it was subsequently attempted to generalize this experience, which had only been possible because of conditions in Malanje. What followed later in the other regions of the country was an absolute failure. 24 The local peasantry's experience of political mobilisation in the 1980s was part of the reason that fierce conflicts erupted in the province in the latter part of the decade, changing in significant ways in the period from the 1990s to 2002. As an agricultural region, Malanje offered -first of all to the MPLA and the FNLA, and later to the MPLA and UNITA -a possibility of penetrating the rural reality of north-eastern Angola, to hijack the legacy of demands of these communities. Indeed, the regions of Malanje, Kwanza-Norte and Kwanza-Sul, and Bengo are made up of local microidentities that have played an important role in functioning as areas that blocked possible foreign military penetration from the region to the capital, Luanda. The political and military forces that were mobilised for this purpose and were stationed in the region were fully aware of the advantages of this geographical reality and of the distinctive local identities.
It is necessary to mention the role played by a key structure that served during the mobilisation and still serves as an intermediary between the policies of the state and the party, and the organisational structures of the peasants in Malanje region. The role of a channel of communication was often played by the sobas 25 -the so-called traditional authorities -acting as the chairs of local committees 26 and integrating in the same way the affiliated structures of the MPLA women's organisation (OMA) and social structures revolving around the Popular Defence Organisation (ODP), which at that time was composed of 17 members. The committee had delegates in small settlements near the villages working to mobilise the peasants for forms of collectivisation. These rural facilitators at the level of the committees constituted a new kind of power structure that was often in conflict with the old (or feudal) local power structures established by the sobas. The soba, usually one of the older peasants, was the head of the so-called African extended family, the repository of the knowledge and power of his ancestors, and continued to dominate in a variety of contexts. The problem of the power struggle that took place in Malanje was aggravated by the attempted coup d' état launched by Nito Alves and his followers. The imposition of methods of mobilising and controlling 'popular power' advocated by the Nitistas, and the Netista 27 reaction to it had alarming consequences for the rural population of Malanje, and was at its most destructive in the centre of the city. 28 The tragedy of 27 May 1977 and its aftermath, just mentioned, had a devastating impact in Malanje. Among the victims was the first Provincial Commissioner after independence, João Manuel da Silva, appointed by the Revolutionary Council (decree no.4/76, 11 February 1976). He was accused of having participated in the coup attempt because the Nitistas had held several meetings in the province and had many supporters and sympathisers there. There were dramatic cases of the deaths of whole families, which is what happened to the family of the Provincial Commissioner.
Because of the worsening military situation from the 1980s onwards, the country was divided into military regions, and the provinces of Kwanza-Norte and Malanje were constituted as the ninth region. These structures were established in July 1983 as part of a strategy of defending both the national territory and the productive sector, but were soon replaced by Regional Military Councils which organised territorial battalions, including workers and peasant militia units. 29 The fourth Provincial Commissioner of Malanje, and the local party committee coordinator, was Ludy Kissassunda, who recalled later, alluding to the military situation in the region, that … the province began to feel the enemy's military presence at the end of 1982, along the border with Bié province. In 1983, we started to introduce the first columns in the areas of Pungo-Andongo, Cacusso and Kangandala with the task of mobilising the population, and in 1984 a group of commandos was sent to open a northern front in the region of Massango. front was established, we [infiltrated] four semi-regular battalions and seven or eight regional columns to carry out actions in the provinces of Malanje, Kwanza-Norte, Uige and Bengo. 30 Actually, from 1985 onwards, UNITA's forces occupied the northern part of the province, specifically the mineral-rich municipality of Massango. This allowed UNITA to build pockets of support in the rural municipalities in the north, and to gain a political advantage in the elections that followed the Peace Accords. In 1989 armed clashes between government forces and the UNITA rebellion resumed as a result of the occupation of northern and southern municipalities in the province, which left only the city centre to endure heavy shelling from UNITA in the municipality of Kangandala. The engagement led to the occupation and control of eight municipalities by UNITA, the destruction of the airport in the provincial capital and the flattening of many residential areas. Three years after the earlier occupation of the province, the social and political landscape changed yet again after armed clashes that followed the 1992 elections, from 31 October to 2 November. Only the provincial capital, where the government structures were located, remained unoccupied. Even during the elections, political speeches delivered in Malanje were the source of a great deal of tension: At a later stage, when it was judged that the conflict had diminished in intensity and consisted of the stockpiling of weapons, the provincial capital again suffered shelling by armed forces and lived in threatening conditions between January and August 1999, in what is considered to be the last stage of the military conflict in Angola. 32 According to the sociologist Christine Messiant … after the elections the war was waged without mercy by both sides against civilians who had become enemies for having 'voted badly' , taking on new contours of cruelty and involving, on the side of the government, a process of militarization of 'their people' in militias that caused massive devastation over a wide area, atomising and destroying society. The United Nations did not offer UNITA, which lost the election, a proposal that it found acceptable, and so the rebellion followed the logic of violence and tried to reverse its electoral defeat by armed force. This led the international community to take the government's side, not only by voting for sanctions and providing diplomatic support, but also through various forms of assistance that were neither political nor transparent. 33 The course of the prolonged war that devastated Malanje city -with people fleeing famine and struggling to survive under military rule, sometimes under the MPLA and sometimes under UNITA, with shells falling on one side and then on the other -vividly illustrates how inadequate the expression 'state of emergency' is to describe the context under consideration.
Agamben 34 sees government confronted by a 'state of exception' , which is not necessarily associated with the notion of a 'state of emergency' , on the basis of which any government confronting a crisis exercises its prerogative to suspend the rule of law and declare a state of war.
Malanje was considered to be a reliably pro-MPLA province, but in truth during the 1992 elections this idea was severely tested, because UNITA, during the period when its forces occupied and controlled the province, had established a sustainable base of support especially in rural areas but not in the provincial capital. This support made UNITA confident, even suggesting a possible victory in the elections.
Margaret Anstee, a United Nations representative in Angola, wrote that UNITA considered that UNAVEM had done 'a very good job' and that the election process had generally gone well. There had been some problems, particularly in Luanda and Malanje. UNITA was putting together a complete picture and would present its position in due course. was closed, since UNITA forces were extremely close and had occupied the outer districts. 37 In the aftermath of the armed clashes in the province and especially in the capital, militia groups began to emerge made up of supporters and political sympathisers of UNITA and the MPLA parties. These groups were armed to fight each other but were located in the tenuous zone that separated combatants from civilians. That is to say, during the day they were civilians and citizens, but at the end of the afternoon they picked up their guns to defend the city from attacks by their political and military rivals. On the MPLA or government side were the so-called civil or civil defence militias, which had gained political and military status since the late 1980s, in the face of the onslaughts of their armed rivals, who were also inside the city. 38 Civilian militias like the ones in Malanje were scattered throughout the country and were sub-divided at the provincial and municipal levels as well as in the rural areas. They had been decommissioned during the run-up to the first general elections, which culminated in the demobilisation of both FAPLA (the Forças Armadas Populares de Libertação de Angola), the MPLA's armed wing, and FALA (the Forças Armadas de Libertação de Angola), UNITA's military force.
Some of them disappeared into the FAPLA veterans group, and from there into another famous group, Fubu, and some individuals from Carreira [the Carreira de Tiro neighbourhood] were completely organised to defend that area. 39 The population was armed by the government itself. I remember at that time in Canâmbua some of our kotas came with guns and distributed them to people's homes. So from that moment youngsters who were a bit more adventurous, some of them even without military training, took up arms for the fight; others already had military training. 40 This is contrary to what Alex Vines claims about the political and military situation: '"Fubo" was composed of units of government special forces, transferred to Malanje to prevent UNITA from infiltrating the city, and according to the same source they were under orders to encourage anti-UN sentiment in an attempt to depict the UN as pro-UNITA. ' 41 In fact, it was a militia group that had emerged in Malanje as a result of a popular initiative by former members of the extinct FAPLA, with individuals who were recruited from the outlying areas of the city. Indeed, in the course of the conflict in Malanje following the siege of the city by UNITA, it was necessary for all the government defence forces to come together in a coordinated military structure, as we shall see later.
At the time when UNITA controlled the city 42 … Malanje was one of the areas hardest hit by the war. This was already in the difficult years, the 1980s. There was a lot of hunger and a lot of people died. Then we were encircled, we were in the middle. UNITA was unable to enter Malanje. It's a small town but they never got in. They attacked from outside launching mortar shells. Caritas did an extraordinary job of supporting the population. When the first refugees arrived, it was the World Food Programme that helped them. [...] They trusted us with all the food they intended to distribute. We had many kitchens. In the city alone we had 75 kitchens for children and 14 for adults. 43 Later the WFP flights restored some normality: 44 We were surrounded, and we couldn't go further than eight kilometres. The city was surrounded, many people died -many families who went to their fields outside the city. Many left in groups and didn't return; or they were killed there. Or they were kidnapped and taken into the bush with the guerrillas. Children were locked up indoors and didn't know what was going on. The children were deformed, they couldn't walk because their legs were swollen … They used to bring them to the kitchens in wheelbarrows, and then come to fetch them in the afternoon. Many people went back to the bush when they thought the war was over. 45 In fact, supporters of both the government and of UNITA did not resist the temptation to provoke each other. In such an atmosphere, the conflict between militias and troops on both sides intensified in a particular neighbourhood of Malanje, and by early July it seemed that there was no solution to the spiralling violence. After a while, early on 13 July, an emergency meeting was convened under the protection of the CPPM to attempt to bring the constant provocations in the area under control. As government and UNITA armed 'civil defence' units in the villages, to supplement and support their regular forces. 46 In the case of the city of Malanje, the scenario was similar to what happened to some extent in all the provincial capitals, as one of our interviewees pointed out: 'The bombardments were launched from Kangandala. There was some degree of harmonious coexistence with the Methodist Church. ' 47 In fact, according to another interviewee 'the war was going on at night. During the day we lived a normal working life, looking for food and medical treatment for the wounded or dead from the bombardment. ' 48 Tony Hodges writes that there were occasions when the Malanje government was negotiating with UNITA about concessions for mineral resources. For example, between 1998 and 1999 governor Flávio Fernandes ordered that lorries from Sonangol were to distribute fuel to UNITA in a municipality located in the east of the province, an operation that lasted several months, until UNITA took control of almost the whole province. This is to say that there were clearly times when provincial government officials were secretly negotiating with the armed rebels: on the one side the government sold resources and on the other side UNITA purchased fuel for the maintenance of the war machine, in exchange for diamonds. 49

Hotel Kijima: UNITA Headquarters in Malanje
Under the terms of the Bicesse Agreement between the government and UNITA, there was to be a ceasefire and general elections were to be held; in addition, the main political and military actors returning from the bush were to be housed in the few city hotels that offered decent accommodation. Such a return of the maquis and its fighters to the cities under UNITA control resulted in what UNITA itself termed the 'return to the cities' , which they had left two decades earlier in a journey to the remote forests of eastern Angola. 50 In Malanje, the hotel chosen for UNITA's military and political officers was the Kijima, located in the heart of the city, on Henrique de Carvalho Street, next to the Praça 4 de Fevereiro. The hotel had opened officially in October 1975 in the heat of the war for national independence, and had been damaged by mortar bombs during the clashes between the MPLA and the FNLA. On the one hand, the location of the hotel in the city centre provided the UNITA provincial delegation the advantage of public visibility, as well as the ease of movement needed for political mobilisation during the elections. It also signalled the status that the organisation had won under the peace agreements. On the other hand, during clashes within the city, the delegation was compelled to take evasive action to escape military pressure from provincial pro-government forces; in fact 'the building had been flattened on three occasions during the armed conflict in the country ' . 51 There were the planes. There was a month, a month and a half when the airport was attacked. That was between 1991 and 1992. And it was a terrible thing. So the services that were here, World Vision, Doctors without Borders, they withdrew. The WFP left a small delegation with local people, and Doctors without Borders left their entire warehouse and gave us the key. We took everything that was needed. And WFP had given us a lot too, we had a full warehouse. We provided the services, and later there was something that was done. It was not made public, but we knew what was going on. There were staff members in Caritas in Luanda who were also UNITA militants. So, when they closed the airport it was not possible to land any aircraft … A secret agreement was then made between the responsible UNITA militants in Caritas, and UNITA. So this was the agreement: Caritas would bring food to Malanje and would also bring food to Jamba, where they had their headquarters and also had many prisoners. A lot of people in the cantonments. They authorised the following, WFP contributed the food and Caritas paid for the transportation. And in August we had 28 flights at the same time bringing food to Jamba. During this period nobody attacked anyone. Only it was agreed secretly to allow the planes to land in Malanje. 52 This is confirmed by the following testimony: In March 1998, UNITA tanks entered the city as far as Vila Matilde, altogether five battle tanks. In Carreira de Tiro neighbourhood, there was a lot of shooting by the tanks. People took their things and fled to Luanda … Malanje has 14 municipalities, and 12 were under UNITA control; only two were not, this was during the 1980s. When the government chased UNITA there was a war. Normally, when it was only UNITA, it was peaceful. But when they were both present, it was a total disaster. By this time, Cacusso was a government-controlled municipality and UNITA was receiving its people in Kaculama. At that time no one from Malanje could go to the river where the bridge is. The river was a real border.

The geography of armed conflict in the city
Relying on Norbert Elias's categories of the 'established' and the 'outsiders' 54 it is possible to understand that, for the armed militias in Malanje, especially in the last phase of the conflict, UNITA militants and supporters were an element alien to the social milieu in which they lived. The context aggravated tensions that had existed ever since the beginning of the war for national independence. Not even the fact that many UNITA supporters were native to the area succeeded in moderating the reaction of the local pro-government armed groups to provocations, intimidation and the revenge killings of known persons. My argument, based on Elias's theory, is that especially in the city, UNITA military and supporters, both families and individual civilians, were considered not to belong to the area; they were strangers to the malanjino social environment. This idea is supported by the fact that many of these local supporters identified with an organisation whose political message differed from the one that was propagated in the city, under the effective control of the government.
In turn, UNITA, for its part, deployed a discourse and held a sense of identity that was constructed from the very beginning of the guerrilha, based on its experience in the bush. After 1992, when it agreed to negotiate and to return to the cities, this maquisarde identity collided with the urban identity of the towns. This explains why the post-election conflict that erupted in Luanda and even in Malanje was so difficult for UNITA to manage. To meet the challenge UNITA's forces had to withdraw abruptly from the urban centres and reorganise themselves strategically again to launch military attacks. This was because, despite their fighting abilities, they were unable to control the cities and remain in them. To use Elias's terminology once again, the maquisarde and urban identities collided like foreign bodies. For some in the groups that opposed each other in the cycle of military confrontation in Malanje, the city was an unknown environment to be tamed, while for others, the presence and penetration of foreign military forces brought a language and a warlike force that was alien to the city milieu. Despite the sympathy that some people felt, UNITA was seen as invasive and provoked disgust in the militias and even among demobilised pro-government forces.
But in the case of Malanje, the confrontation was about more than identity; it was about the spatial control of territory 55 as Justin Pearce points out in his study of the armed conflict in the central highlands. 56 Markers of the independent state also exerted pressure and played out politically as indicators of differentiated political identities: individuals living in MPLA-government-controlled localities could move freely and have access without problems. The question of the city, the question of delimited space and the idea of a civil or 'civilised' identity entered into a collision with a supposed identity of 'the outsiders' built in the heat of war in the 'bush' . The 'bush' was constructed as a marker of political-guerrilla and civil difference, seen as threatening 54  by civilians and urban dwellers alike. Of course, messages of political propaganda played a major role and greatly influenced the various social strata among the population of Malanje. It is no coincidence that even in 1992, at a time when expectations were at their height and there was great sensitivity, UNITA militants were seen as ordinary people and not as 'wild animals' whose natural habitat was the jungle. In the terminology of the period, the 'bush' was synonymous with uncivilised. People believed that Jonas Savimbi had two heads and a tail -cartoons circulated in leaflets distributed in Malanje during the post-election conflict showing Savimbi with horns and a tail. 57 The geography of the armed conflict in Malanje was partly defined by the strategic positioning of the government's military forces, the civilian militias and the UNITA forces, as well as the laying of thousands of land mines throughout the long years of the war. Even today, in fact, the actual consequences of the laying of these explosive devices can be seen among the malanjina population, since, over ten years after the conflict officially ended, mined areas are still being discovered in the provincial capital. The mines can be understood as spatial delimiters of conflict, either for self-defence or as an indication of possible external threats to the political territory. 58 The administration that came into force immediately after the Lusaka Protocol had serious implications for the management of the armed conflict in the municipalities in the interior of the country, and Malanje was no exception. Under the Protocol, signed by both the government and UNITA, administrative representatives were appointed at the municipal and commune level. For example, one can cite the appointment by UNITA, under the Protocol, of an administrator in the municipality of Quela, two deputy administrators in the municipalities of Malanje and Kangandala, and four administrators for the municipalities of Caxinga, Ngola-Luije, Kateco-Kangola and Bângalas. 59 These localities coincided with the areas where UNITA had widespread support after 1994.
The armed conflict in Malanje, and the forms it assumed, illustrate the complex and nuanced impact of the political violence that was to take place in other provincial capitals. The conflict manifested itself sociologically: the food emergency and the establishment of kitchens by the WFP and Caritas de Angola in response to the famine that hit Malanje's population, was something that was common to more or less the whole country. But the conflict in Malanje had some peculiar characteristics that have to be taken into account. These show up especially in the form of confrontations between social configurations historically considered to be reliable. They show up in the destructive and disruptive dimension of urban political violence by local militias and government army groups in reaction to the 'foreign presence' of 'outsiders' in the shape of the UNITA military. They are also illustrated by how these groups of individuals struggled to respond to the food emergency by negotiating a temporary truce with the other belligerents so as to allow the humanitarian agencies freedom of movement. Both sides desperately needed food and fuel to keep the war machine functioning.
The armed conflict in Malanje also revealed how, soon after the Lusaka Protocol was signed in 1994, the real impact of the false peace between the government and UNITA was felt locally, since the city was still under siege by UNITA. One of the 'frontiers' that separated the two forces was the bridge over the Luchimo river, between the municipalities of Malanje and Kaculama. 60 On this bridge it was possible for civilians and for soldiers from both sides to meet and trade vital necessities, since crossing the bridge was the only route. It was also on this Luchimo river bridge that there were signs of goodwill, as the two armies negotiated the opening of an equally necessary transitional road, at a time when the locality of east Kaculama was still the military headquarters of UNITA forces in Malanje province, It had its own administrative structures, with an influence on how the sides co-existed in their daily lives' . 61 The move of UNITA's military coordination centre from the provincial capital to Kaculama was a direct result of the conflict that began in the city between October and November 1992, as we have seen. In the territory controlled by UNITA the movement of people and goods was allowed only with passes or safe conducts. Because of this restriction imposed by UNITA in the areas they controlled, people were unlikely to be allowed to leave and return. The mechanism often used by the socalled 'civil defence' , which were in fact local militia groups, was to encourage people under UNITA's control to flee, while those who were allowed to leave were subjected to intense interrogation when they returned. 62 By contrast with Kaculama, in other diamond mining areas, notably in the municipalities of Quela or Caombo, restrictions were much tighter. 63 The ways in which normality was negotiated thus involved risks that people took every day, driven by the need to survive, against the arbitrary measures enforced by the two belligerent parties.

The 'outsiders' and the 'insiders' after the conflict
According to the official narrative, the notion of reconciliation in Angola first appears in public discourse at the time of the Bicesse Accords in 1991. However, the government had spoken earlier of a policy of clemency and social harmonisation. It was only after the churches took a position advocating reconciliation -charged with the full Christian significance of the term -that the Angolan government began to make use of the word, without really evaluating the consequences in terms of relationships. As a policy of clemency one can understand the idea in terms of verticality -a concession by one party (the government) to the other (UNITA), which would not satisfy UNITA because it was not recognised or treated as an equal partner. Until just before the Bicesse Accords were signed, the Angolan government considered UNITA to be nothing more than an 'armed gang' with a capacity to oppose state institutions with a military arsenal accumulated over many years and which created enormous difficulties for the government's activities.
That said, it should be noted that the post-conflict period, albeit with some differences relative to the earlier phase, and which did allow for the policy of national reconciliation, played out similarly within a framework of a social logic created by the culture of war or an atmosphere of violence as defined by Frantz Fanon. 64 In other words, national reconciliation in the way in which it actually happened did not necessarily interrupt the logic of war attempted in 2002. Official policy sought to emphasise the opposite.
In terms of official policy, peace was considered an indispensable precondition for national development. 65 The policy of peace and national reconciliation was strongly associated with the development of infrastructure that had been destroyed during the various cycles of war, especially in the last phase, in 1998. Strictly speaking, reconciliation is to be understood in several dimensions: a politico-administrative and legal dimension; a spiritual dimension that has to do with traumatic experience and refers exactly to forgetting and forgiveness, especially in a Christian perspective and specifically a Catholic one [Comissão Episcopal de Angola e São Tomé -CEAST] and a Protestant one [Conselho das Igrejas Cristãs de Angola -CICA and also the COEIPA]; 66 and a material dimension, related to national reconstruction and development.
To some extent, in terms of government policy, reconciliation happens alongside reconstruction, since the primary objective is to erase the marks of war in the main cities, 67 just to mention the ones that were hit hardest by the war's destructive impact. From this perspective, it is common knowledge that the positioning of the development project differs in the cities of the Angolan coast, 68 which were mostly places of refuge for people from the south-central region, fleeing armed clashes in search of shelter and security. Even today a large proportion of the population is still concentrated in the cities and urban centres, despite calls for a return to their areas of origin.
That said, the argument can be made that reconciliation in Angola created a series of commemorative rituals through appeals to public forgetting and forgiveness as official policy. 69 More specifically an official policy of silence was linked to the ghost of war, and continues to be used as a way of maintaining a supposed national stability. 70 The first marker of reconciliation was the signing of a ceasefire agreement that permitted a return to the negotiating table, but on terms dictated by the political agenda of the Presidential Palace in the Cidade Alta of Luanda. UNITA, in a position of political and military weakness, was not able to contradict the positions taken by the government. It was in this way that the political-administrative and legal dimensions of national reconciliation were determined. 71 This step placed control of the negotiation process exclusively in the hands of the belligerents, making them the key agents who would determine the end of the long cycle of fratricidal warfare. This was to be done in accordance with an amnesty law covering all crimes committed during the conflict, passed unanimously by the National Assembly that had been elected in 1992. One of the requirements of the democratic process established after 1990 was that the members of the main parties that stood in the first elections were elected to the National Assembly; later on, deputies from UNITA-Renovada were involved.
The second moment of reconciliation can be taken to be the publicity exercise called Ponto de Reencontro (Meeting Point), organising the reunification of families that had been separated during the various cycles of fighting. 72 The moment also helped to create an environment of awareness about forgiveness, which was being pushed for by the religious authorities and which resonated with the official policy of reconciliation. Before the government took it over, this initiative was planned by a group of Angolan and Brazilian journalists, 73 originally with a view to reuniting the family of one of its members who had lost contact in the fierce fighting of late 1997. The Angolan government saw strategic advantage in this initiative by the group of journalists, a timely way of winning political dividends by taking over control and management of the non-governmental programme. From then on the journalists' initiative was disseminated extraordinarily widely and achieved a national impact and dimension, and there were so many adherents competing that other meeting points were created in various provinces all over the country to cope with the hordes of people who flocked to them every day.
A third moment of reconciliation was characterised by political and religious messages about the need for forgetting and public forgiveness of the other. The understanding of the other in effect consisted of recognition of the former enemy and the presently existing brother, once divided by the parties to the conflict. 74 This is more specifically the spiritual dimension, which signifies a high level of complexity and ambiguity.
Reconciliation is often perceived as the pacifying of a traumatic memory of the horrors of war. It is triggered whenever there is a situation that references, in immediate terms, a remembrance of the past, especially if there was loss of life and circumstances that did not allow the mourning of the dead. Sometimes it seems to be a gradual and time-consuming process, in which those who are involved require 71 I am describing here the totality of the legal and political procedures that were adopted, as well as the legislation that conferred legitimacy on the negotiation process. 72 Lázaro, ' Angola: Discursos e Práticas' . 73 The Brazilian journalist Sérgio Guerra coordinated the work and set up a production company for the purpose called Maianga Produções, which, after the end of the project to reunite separated families, continued to provide other services to the Angolan government. 74 G. Lázaro, 'Violência, Trauma e Reconciliação em Angola' , paper presented to the conference 'Constructions of the Nation after Large-Scale Violence' , River Crossing Lodge, Windhoek, 14-19 October 2012. moments to calm their fears, as well as an idea of tolerance so as to coexist in everyday life. This does not necessarily imply forgetting the traumas of the past. The contours taken on by the configuration of politics 75 and their extension to other fields such as the religious, with alliances woven in the period after 2002, all reinforce memories of the traumas of the past.
Reconciliation is gradually taking place. Reconciliation is a very difficult thing. It goes on as the generations renew themselves. A son who saw his father killed, saw his mother killed and disrespected, cannot be friends with the other! It's necessary to understand. It's a great burden of suffering. And today he who killed your father or your mother is capable of killing you. It is more tolerance than reconciliation. Now, as time goes on, the grandchildren and great-grandchildren, for them it no longer has the same weight. There will be a settling down … Nobody likes UNITA. Some have UNITA as their party, but UNITA has done so much harm, just as the MPLA has done so much harm in other UNITA zones. Because in wartime, nobody has clean hands. 76 The influence of the churches there in Malanje, especially the Methodists, is very evident. It is evident not only in urban areas but also in the countryside. 77

Conclusion
The independent Angolan state emerged from the context of a military confrontation that lasted many years and grew to create a bipolar politico-military situation. The experience of Malanje province shows that the armed conflict in Angola was atomised, and varied according to the region under study. Local perceptions indicate an understanding of the war that does not avoid the partisan stereotypes and political clichés created in the field of politics. Perceptions of the war -or of the cycle of armed conflict that unfolded in the province in particular, and in the country in generalare held hostage by an historical interpretation of party-political bipolarity, with two belligerent actors, drawing a distinction between the 'local or native' on one hand, and 'outsiders' on the other -in other words between the 'insiders' , the 'sons of the soil' and 'the ones from elsewhere' as Norbert Elias and John Scotson have described them. 78 Despite the devastating impact that the war had on north-western Angola, there was space for communities in general, with resilience, to rebuild their daily lives in ways that nevertheless have not yet permanently erased the scars left by the fighting.
This contributed to a kind of peace-making between groups of people within the social relations that were established there. Despite its destructive power the war permitted the reinforcement of ties of 'sociation' 79 between people and local bodies dedicated to social action. The local communities in Malanje have deployed numerous social strategies, such as seeking shelter in the homes of the provincial Catholic Church, the strengthening of solidarity among Christian missions, supporting groups for people injured in the bombardments that devastated the city, or the negotiations between Caritas de Angola and the UNITA rebels for a 'non-aggression pact' in exchange for food. In the context of reconciliation, the perceptions of the people in Malanje are quite different to those recounted in the official narrative. There is a limit to what is acceptable that varies in degrees of intensity according to a person's political sympathies. The scars of war are repressed, with people refusing to share their past experience of suffering and trauma; living together in such circumstances is hard.
Reading between the lines of the narrative of the civil war, it can be understood that political violence as a means of winning and keeping power, or even as an accommodation of antagonisms, worked to the extent that the main protagonists have monopolised the discourse about and the practice of national reconciliation, and are the beneficiaries of the division of resources resulting from successive politicalmilitary agreements.
The emergence of UNITA's forces in the centre under the peace accords was seen as a strange phenomenon, with the 'outsiders' who did not share the same political identity disrupting the social order established in the single-party period. The people of Malanje were thus compelled by the peace accords to coexist with UNITA soldiers -the former outsiders -when for a long time the message conveyed by the MPLA propaganda had represented them as enemies of peace and national unity.