PEOPLE’S WAR: MILITARY SUPPLIES DURING THE MOZAMBICAN CIVIL WAR, 1976-1992

From 1976 to 1992, the government of Mozambique under the leadership of Frente de Libertação de Moçambique (FRELIMO) and Resistência Nacional Moçambicana (RENAMO), the latter sponsored by the right-wing and racist regimes of Rhodesia and South Africa went to war. The independence of Rhodesia (now Zimbabwe) in 1980 and the signing of the Nkomati Non-aggression Pact between the government of Mozambique and South Africa in 1984 led scholars and government officials to claim that the government would win the war because RENAMO had lost its support. These claims proved wrong as RENAMO resisted for another eight years until the signing of the general peace agreement in 1992. The paper argues that the continuation of military confrontations shows that wars are mainly fought with a complex combination of means that are not necessarily military. Claiming that the survival of RENAMO depended on external support represents a misunderstanding of the logistics and morale of both RENAMO and government troops. It is from this perspective that this paper looks at the logistics and enthusiasm of both RENAMO and government military to demonstrate that both lacked adequate military logistics to wage war. It shows that the belligerents depended on civilians and surrounding natural resources to obtain the bulk of supplies of staple foods and recruits. This state of affairs compels scholars to rethink the nature of civil wars and helps to explain the almost decade long delay in achieving peace in Mozambique. It also shows that the burden of the Mozambican civil war fell on the shoulders of civilians. Thus, what is often described as a hotspot of Cold War in Southern Africa or a war of aggression by the apartheid regime was, in practice, a peoples’ war with devastating, yet varied impacts on peoples’ livelihoods. This section provides an overview of the poor material conditions and low morale of the Mozambican armed forces at national level. It includes the voices of people who dealt with this resource scarcity in the interior of southern Mozambique. In the process, it demystifies the idea that only RENAMO forces depended on the extortion of civilians. Based on evidence from fieldwork, it shows that despite having access to armoured vehicles and helicopters, the idea that government troops were well equipped, and disciplined is misleading. It conceals many problems such corruption, and reliance on obsolete communication networks. 92


INTRODUCTION
The academic debate on Mozambique's civil war has been dominated by the analysis of the causes, consequences, the roles of external and internal factors and the predominance of violence against civilians. 1 In these narratives, the government troops are often presented in populist form as having abundant resources and discipline. At the same time, the Mozambique National Resistance (RENAMO) is still frequently portrayed as a band of terrorists without a social base. 2 This narrative of the war in Mozambique conforms with the stereotypical description of African wars as barbaric and primaeval. 3 It is a simplistic understanding of the civil war refuted by many scholars, including Colin Darch, João Cabrita, Jeremy Weinstein, Stephen Lubkemann, and Stephen Emerson. They all focused on Central Mozambique -long and widely regarded as a RENAMO stronghold. 4 Unlike such previous studies, however, this paper is based on data collected in the districts of Mabote, Funhalouro, Chigubo and Massangena which hosted the first RENAMO military bases in the interior of southern Mozambique. Taking advantage of the end of the 16 year war in 1992, and the author's own (subjective) experience during the war in southern Mozambique, this research seeks to understand the dynamics of the armed groups in Mozambique's civil war. It analyses the dynamics of the belligerents in the context of scarce resources able to sustain a prolonged war. 5 It seeks to answer these questions: What means did the RENAMO and Mozambican armies have to wage war? To what extent did the military logistics, morale and discipline affect the conduct of war? In this context, how did the military negotiate their relationship with civilians?
Despite focusing on Mozambique, the paper aims to contribute to the historiography of the dynamics of civil wars in sub-Saharan Africa. 6 It builds up from the works of Carolyn Nordstrom, Jessica Schafer, Stathis Kalyvas Weinstein, Lubkemann an Emerson, which challenge simplistic explanations about the conduct of civil wars in post-colonial Africa. 7 It highlights the heterogeneity of war experience and the importance of civilians and local resources in the conduct of war. Without ignoring the national and international contexts, it explores oral sources from an area relatively neglected by researchers but strategically important for RENAMO's effort to expand to the southern region which has historically been deemed a government stronghold. 8  Despite its proximity to the epicenter of war in central Mozambique, the first military confrontation between government troops and Renamo guerrillas in Inhambane province used widespread violence against civilians. As Elizabeth Lunstrum observed, "although it committed atrocities countrywide, RENAMO reserved its most brutal acts for the south, where support for the Mozambique Liberation Front (FRELIMO) was strongest". 9 Therefore, without taking for granted the violence against civilians perpetrated by both sides of the conflict, this study takes a historical approach to show how the relations between soldiers and civilians changed over time and geographic location. It keeps in mind recent studies such as the work by Wesley Mwatwara showing that the battleground of the Mozambican civil war expanded to neighbouring Zimbabwe with records of selected violence against civilians. This helps to understand the changing nature of the relationship between the military and civilians. 10 This paper is based on literature review, archival research and oral interviews. The fieldwork took place from July to December 2013. Additional information was gathered sporadically in 2017, 2018 and 2019. The archival research included new material such as government and non-governmental reports, public and private letters deposited in the archives of Inhambane Provincial government and the bulk of the documents including newspapers and news bulletins stored at the Mozambique Historical Archives in Maputo. The interviewees were mainly man and women who were at least ten years old when the war reached southern Mozambique in 1982.
Methodologically, the article builds on the work of Africanist scholars who have produced a wealth of literature on the importance of oral accounts in African historiography. Such scholars include Jan Vansina, on the definition of oral history and oral tradition 11 , Joseph Miller's analysis on the challenges of oral historians in non-literate cultures 12 ; John Edward Philips on the methodology of collecting oral traditions 13 and Barbara Cooper, on the importance of oral history to recover the voices and experiences of the disadvantaged and on the importance of multidisciplinary aproach. 14 It uses Vansina's definition of oral history sources as "reminiscences, hearsays, or eyewitness account about events and situations which are contemporary, that is, which occurred during the lifetime of the informant". 15 The paper keeps in mind that the use of memory as a historical source is disputable. It approaches memory like Luisa Passerini who sees it "as collective and tied to both present and past, silence and speech". 16 In this approach, the paper looks at memory as a contentious field subject to reinterpretation and highly influenced by subjective experiences of the informants. Memory is also used considering the importance of genre when dealing with oral history. As Elizabeth Tonkin points out, "oral genre can be signalled by the occasion or the status of the teller". 17 Thus, the article pays attention to the various genres that people use to communicate their encounters with historical events, including tales, riddles, silence, and dance. It also considers, in Cooper words, the "poetic and performative context of the interview". 18 The identification of the interviewees was based on the snowball sampling technique, whereby one informant leads to another. It used nonstructured questions to create flexibility on the side of the informants. Given the sensitivity of the topic, the interviews began with general questions about the interviewee's personal history. As the conversation rolled on, people were asked about their whereabouts during the war and how the war affected their lives and that of their friends and relatives. The sample size consisted of 47 men and 40 women in the provinces of Gaza and Inhambane. The first criteria to participate in the study was having lived in the war zone during the conflict. Particular attention was given to those who witnessed the fighting, suffered violence and assisted the military in various capacities. Those who served combatants as carriers for both government and RENAMO fighters and those who lived in RENAMO military camps also received special attention.
The interviewees were approached bearing in mind Jan Shettler's observation that oral accounts are influenced by diverse factors including the environmental setting in which they are collected, the political and social background of the informants, the gender of informants and researcher, language and instruments of data collection and genre. 19 Thus, the interviewees were approached as key players able to interpret the events according to their interest and exhibit their agency. Like Lubkemann, oral accounts were approached as "constructed (rather than collected) to emphasise the inevitably dialogic and social process of their production, between the author and the subjects in the social settings in which these interviews occurred". 20 The interviewees were also addressed keeping in mind Alessandro Portelli observation that "memory is not a passive depository of facts, but an active process of creation of meanings". 21 Considering that the area of study is seen as a government stronghold, the testimonies were read against the grain of written records and examples from other regions of Mozambique.
Taking advantage of the author's proficiency in four languages spoken in southern Mozambique, namely Xi-Copi, Xi-Tswa, Xi-Changana and Portuguese, the interviewees received an exhaustive explanation of the purpose of the study. Only those interviewees who gave informed consent were interviewed in the language they felt comfortable to tell their stories. The author also explored his own subjective experience of war in the region. He was born and raised in Zavala District located in southern Inhambane which borders with the district of Mandlakazi, northeast Gaza Province. In 1987, he moved to Maputo province after seeing his school set on fire by RENAMO guerrillas. He regularly travelled from Maputo to Inhambane passing by Gaza Province. Heavily armed government troops escorted these trips. After hearing about the war in the early 1980s and witnessing a wave on internally displaced people, the author had his first-hand war experience in 1985 when he was a ten years old schoolboy. During the same year, he was caught between the government and RENAMO guerrillas' skirmishes. For the first time, he witnessed an exchange of gunfire and saw dead civilians and militaries. This experience and subsequent attacks, kidnappings, assassinations and pillage of villagers' properties are vividly present in the author's memory and served as motivation to contribute in the writing of a localised history of the civil war from the perspective of those who were in the frontline of military confrontations. In southern Mozambique, attempts to provide supplies by the sea seem to have produced minor impact thus, in this region, access to military equipment and food became an essential element for RENAMO's military campaigns. 30 As the external support became unreliable, RENAMO looked for local solutions for both food supplies and military equipment. As far as food staples were concerned, the coincidence of severe droughts from the early 1980s to the early 1990s had forced many villagers to relocate to new areas, "constraining RENAMO's ability to subsist off the local population which itself was often on the brink of starvation". 31 Regarding the military supplies, RENAMO obtained them from the battlefield. A significant part of its weapons was captured from government troops or obtained from government army renegades. 32 As one government war veteran stated, "there were some occasions in which we were surprised by finding RENAMO guerrillas wearing new uniforms similar to the ones we had received the same week". 33 The signing of a Non-aggression Pact between the government of Mozambique and the regime of apartheid in 1984 brought hope that RENAMO would be at ropes. Contrary to this expectation, the signing of the Nkomati Non-Aggression Pact with the Apartheid regime did not stop the expansion of RENAMO operations in the countryside. 34 According to David Robinson, a few months before the signing of the agreement, South African Defense Forces delivered massive supplies of weapons to RENAMO. Robinson adds that, even after the agreement, the South African Defense forces continued to supply RENAMO and encouraged diversification of RENAMO's support, including from American conservatives. 35 Owing to this continuous support, the period between 1983 and 1987 witnessed the expansion of RENAMO attacks countrywide. 36 It is noteworthy to remark that the evidence that RENAMO continued receiving supplies after Nkomati Agreement and that it expanded its operations does not mean that it was not affected by the Pact. of overt delivery of supplies affected some of the key logistics, including communications. As Emerson points out, by the late 1980s, RENAMO lost its advantage in the communication network as its communication equipment (like radios) became broken and old, and they lacked access to reliable spare parts. 37 This coincided with the government of FRELIMO's diplomatic endevours which included gradual abandonment of centrally planned economy and adherence to Bretton Woods institutions. 38 This move was signalled by the visits of President Samora Machel to the United States and United Kingdom which marked the beginning of change from radical Marxism-Leninism to a liberal economy thus, laying foundations to cooperation with countries that had been RENAMO traditional sympathisers. 39 In addition to weapons, the guerrillas needed food, which they could not obtain from external sources in the long term. Locally, they could count on requisitions from local communities and harvesting meat or plant foods from the wild. They could farm, but the peripatetic nature of guerrilla warfare and cyclical droughts interfered with such efforts. This situation compelled RENAMO to build networks of popular support in areas under its control. As one former RENAMO guerrilla observed: "in the military bases we used to eat bush meat and the food that we used to request from local people. We used to farm, but it was not viable because we were attacked frequently by FRELIMO, forcing us to abandon our unripe crops". 40 Moreover, the prolonged drought that affected southern Mozambique intermittently from the early 1980s to early 1990s hampered agricultural production and reduced sources of food for guerrillas. Under these circumstances, guerrilla members had to obtain food either from civilians or by attacking government shops and warehouses. 41 According to oral testimonies, when RENAMO guerrillas penetrated the district of Mabote in 1980, they poached wild animals and requested chickens, goats and cows from villagers. 42 As these resources became exhausted, RENAMO guerrillas began to use violence to pressurise people to provide them with food and intelligence. Also, in search of food supplies, RENAMO launched frequent ambushes on civilians' vehicles along the main roads. The majority of these RENAMO attacks were directed at mineworkers, humanitarian aid convoys, and all civilians suspected of transporting valuable 37 See, Emerson, The Battle for Mozambique,p.194. 38 International Monetary  In the case of RENAMO, external support and abundance of local resources in the early stage of war led to less violence. Oral accounts point that the decline of agricultural production due to prolonged drought in the early eighties coincided with the escalation of violence against civilians in the southern region. As some residents of Mabote, Funhalouro, Chigubo and Massangena Districts pointed out, "the 1980s were difficult times. We did not have food because of drought yet, RENAMO guerrillas demanded food from us. If we said we had nothing, they wouldn't understand. They would beat or even kill if you did not give them food". 48 The escalation of violence in Southern Mozambique was not accidental. Some scholars claimed that RENAMO had categorised southern Mozambique as a zone of destruction. 49 Nevertheless, this categorisation does not reflect the reality of the whole southern Mozambique because oral testimonies show that violence was contextual and localised. In areas where RENAMO controlled large areas 43 NOTÍCIAS, 21 October 1987;NOTÍCIAS, 30 November 1987. 44 Emerson of dispersed settlements such as Funhalouro, in Inhambane and Saúte in Chigubo, Gaza Province, people report few episodes of violence. 50 In contrast, in these areas, people reported violence from government troops. According to Deolinda José Macucua "when government troops dislodged RENAMO from Tome, many people were assassinated after accused of collaborating with RENAMO". 51 Another argument which should be analysed with a grain of salt is the idea that RENAMO violence against civilians in southern Mozambique reflected ethnic rivalries between the Shangana (pro-FRELIMO) and Ndau (Pro-RENAMO) dating back pre-colonial time. 52 Recent studies dismiss this justification by showing that unlike other African civil wars, the conflict between FRELIMO and RENAMO revolved around regional economic differences rather than ethnicity. 53 As Fernando Florêncio, remarks, RENAMO proximity with Ndau ethnic group can be explained by its genesis among the Ndau of both Mozambique and Rhodesia. Florêncio explains that RENAMO first recruits and leadership came from vaNdau. This explains why the movement adopted Ndau as a sort of lingua franca. 54 Thus, ethnic differences seem not to have played a significant role in the administration of violence in the region.
It is important to emphasise that it is not the objective of this paper to revive the debate over violence in Mozambican civil war. This paper aims to study the logistics of war with particular emphasis on the role of people and their resources. Thus, violence is analysed as a by-product of a complex relationship between civilians and the military. Unlike Weinstein, who focuses only on rebel groups, this study brings together rebels, government troops and civilians in a single analytical category. It builds on Eric Morier-Genoud's approach of war as a "total" fact. 55 It approaches violence against civilians in southern Mozambique, keeping in mind Kalyvas' and Claire Metelits' argument that rival sides use violence selectively. 56 It also conforms with Lubkemann's claim that Mozambique's civil war was "a fragmented war"(…) in which violence was problematised by the social and cultural formations across which it was staged. 57 Therefore, the experience from southern Mozambique shows that relations between civilians and combatants were complex and could be situated on a spectrum of violence and coercion on one side but collaboration on the other. 58 These relations fit with Lubkermann's assertion of war as a social condition", which does not stop social trajectories and life projects. 59 In Lubkermann's terms, the experience of war was also about "the prosecution of everyday life through more violent means". 60 In fact, oral accounts show that villagers took advantage of the war environment to settle personal differences. 61 As one villager from Mabote Ditrict recalled, "during the war, many people were falsely accused of collaboration with one side of the military contenders. A dispute over a woman would lead to accusation of collaborating with the enemy". 62 Oral accounts provide examples of people who used the war to eliminate competitors and people suspected of being witches. As one villager from Massangena District recalls, "many people enriched themselves during the war by engaging in the business of stolen goods including cattle". 63 Others used the war to force women who they wouldn't marry in a normal setup. According to an interviewee from Chigubo District, "having a gun opened the door to access everything someone wanted including women". 64 Having analysed logistical conditions of RENAMO with particular focus on the interior of southern Mozambique, now the focus is on the motivation of these guerrillas. Some scholars have argued that neo-colonialists and disgruntled, frustrated Mozambicans formed RENAMO. 65 In this approach, the RENAMO leadership is presented as lacking a cohesive and viable political project that could attract people to join the struggle. At the same time, many RENAMO guerrillas were said to be conscripted. 66  guerrilla interviewed in this study were conscripted. In the words of Albino Capitine Chaia: "In 1981, I was kidnapped by Matsanga (RENAMO) when I was returning from South Africa. They found me in Timondzwanine. Someone told us, to not use that way. We insisted because the alternative was too long. We saw boot marks along the way, and we thought they were government troops. When we saw them, we tried to run, but they stopped us with guns. Other young men who were with us were almost killed. One was coming from Maputo. They said, "you are coming from Maputo so, you are a FRELIMO soldier". For some of us, coming from South Africa was not a problem. They took the guy coming from Maputo to a secret place to inspect his legs to see whether he had boot marks. Since he did not have marks, they concluded that he was coming from work, he was not a soldier. They took us to the base, but we were separated there. I don't know where the others went. The two young men received guns, and I was oriented to stay in the base. My job was to cook for wounded soldiers.
In the middle of the war, I became the manager of the food they were looting out there. I was in the Dombe base". 67 Armando Sumbane remembers that he was captured in his home village of Tome in Funhalouro. In his words, they caught me in my farm; they tied me up and marched me to their base, where I spent some time in confinement. After that, I went through a short military training. After some years, I became a commander; I was operating in Panda, Homoine and Morrombene districts. I targeted lorries caring humanitarian aid. I also led my men to areas where cattle were abundant. I ate meat! I met my wife in the base. I had some children over there, after the war I brought them home. 68 Having an army mostly composed of people recruited by force, it is vital to understand how RENAMO survived 16 years of struggle with limited resources to wage war? Many studies have highlighted various reasons impelling the youth to join rebel groups. 69 The accounts of former RENAMO guerrillas confirm that indeed, there are multiple explanations for the motivation and morale of RENAMO guerrillas. As a former RENAMO guerrilla pointed out; to secure the loyalty of its guerrillas, RENAMO used the Janusfaced tactic of seduction and intimidation. In the words of one of RENAMO's ex-combatants, This statement confirms that there was a preoccupation in lifting the morale of guerrillas by promising them a better life after the end of the war. Therefore, while some recruits were coerced, others embraced the guerrilla war heartily hoping to improve their living condition after the end of the war. 71 A former RENAMO guerrilla said that "one of our favourite songs sung in the morning gatherings said that "after the war comes bonanza". 72 It is also important to observe that despite the uncertainty of guerrilla life due to the unpredictability of attacks by government troops, people created families in the military bases. As one informant told Lubkemann, in the war zone, there was also space to have a normal life -to fall in love, have romantic relationships and to experience intimacy. 73 When the war ended in 1992, many guerrillas returned home with spouses and children. A woman who returned from war married to a former guerrilla recalled; "I was kidnapped together with my mother. My mother was assassinated, and I was kept captive for three years. During this period, I had a relationship with one guerrilla. We had two children during the war. When the war ended, we returned home, where I introduced him to my parents. He paid the matrimonial compensation (lobola), and we continue living together. 74 The experience of this couple reveals that for fighters who had family in the war zone, the war was also about the protection of their families.
Another factor that led people to join RENAMO was the easy access to looted goods such as bicycles, cattle, cloth, blankets, radios and sewing machines. 75 For example, when government troops overran RENAMO guerrilla in Manzuile, they found many bicycles, radios sewing machines and other domestic utensils. 76 It is essential to underline that because of the collapse of the post-colonial socialist project, these objects had become luxuries and symbol of high status. 77 As a resident of southern Mozambique and a son of a mineworker, I recall that having a radio cassette, sewing machine and South African heavy cotton blankets was a privilege. Teachers and other public employees used to order these objects from my father and other mineworkers. 78 Moreover, oral accounts point out that very often, RENAMO guerrillas targeted the houses of mineworkers to demand these objects. 79 As far as cattle is concerned, despite its abundance in the region of study, oral accounts highlighted that it is not common for the herders to slaughter cows to eat. The meat was regarded as a luxury food, yet RENAMO guerrillas took it as an integral part of their regular diet. 80 Taking advantage of peoples' "greediness" for meat, RENAMO slaughtered looted cows and goats and distributed meat to civilians. According to a victim of RENAMO kidnapping, in 1987, his friend refused to flee with him from RENAMO's captivity because he could not leave before eating meat. 81 This episode is in line with a statement by a woman interviewed near the former RENAMO base of Manzuile in Funhalouro who confirmed that the meat of cows taken from local communities attracted many villagers to RENAMO. 82 It is also important to remark that RENAMO camps in Mabote, Massangena and Chigubo District were located within Zinave and Banhine National Parks. This ensured them access to game meat, ivory and rhino horns. A recent study by Mwatwara shows that RENAMO used game meat and animal products such as elephant tuskers and rhino horn to establish a complex commercial network with civilians in the borderland of Mozambique and Zimbabwe. 83 In the process, those who honoured their trading commitments enjoyed a good relationship with guerrillas while those who defaulted were selectively punished. 84 Mwatwara's study shows the complicated nature of the "logic" of violence perpetrated by RENAMO and puts into question the idea that RENAMO violence was indiscriminate. Mwatwara, "The "Logic" of Renamo civil war violence".
One of the striking realities about RENAMO in southern Mozambique was its inability to articulate a coherent political ideology of mobilisation. The majority of interviewees for this study did not have a clear idea of what RENAMO was fighting for. Those who were kidnapped and taken to RENAMO bases were physically isolated from the guerrillas for some time before being integrated as guerrillas or forced labour. Because of this isolation, they claim that they did not receive any explanation of the causes of war. The central message that they received consisted of threats of death to those attempting to escape. 85 As one villager kidnapped by RENAMO recalled; "I decided to stay in the base because I was told that if I tried to escape to areas controlled by the government, I would be treated as an enemy and subjected to torture or even execution. They told me that for government troops, everyone who lived in RENAMO camps was an enemy". 86 Many people expressed similar fears to justify their permanence in RENAMO camps. Such testimonies show that fear was the key factor that kept people under RENAMO's control. People feared execution by both government and RENAMO military.
Thus, unlike central and northern Mozambique, in the southern region, with some exceptions, coercion appeared to be the dominant but not the only force behind peoples' loyalty to RENAMO. This might be explained by RENAMO's inability to articulate a strong political message and to develop large administrative territories as attested by Morier-Genoud's research in southern Inhambane. 87 RENAMO guerrillas' relationship with civilians in Inhambane contrasts with Domingos Do Rosário's argument that, in Nampula Province, RENAMO used selective violence and controlled areas that became spaces of socialisation. 88 Zambézia is another province where people adhered voluntarily to RENAMO, and the guerrilla created liberated zones with social services. 89 These findings reinforce this paper's argument that there is no single or mono-causal explanation for the relationship between the military and civilians in a war context. They also complement Morier-Genoud, Michel Cahen and Do Rosário's argument that "the war developed distinctive features in different areas of the country". 90 The lack of open popular support in southern Mozambique as demonstrated by the case study in Ilha Josina Machel in Maputo Province is one such evidence of regional distinctiveness. 91

GOVERNMENT TROOPS RESOURCES DURING THE CIVIL WAR
This section provides an overview of the poor material conditions and low morale of the Mozambican armed forces at national level. It includes the voices of people who dealt with this resource scarcity in the interior of southern Mozambique. In the process, it demystifies the idea that only RENAMO forces depended on the extortion of civilians. Based on evidence from fieldwork, it shows that despite having access to armoured vehicles and helicopters, the idea that government troops were well equipped, and disciplined is misleading. It conceals many problems such as incompetence, corruption, sabotage and reliance on obsolete communication networks. 92 It also neglects the impact of Rhodesian regime attacks on economic and military targets in southern Mozambique. Emerging from ten years of guerrilla warfare against the Portuguese colonial regime, as the civil war began, the government of Mozambique was still in the process of transforming FRELIMO guerrilla forces that had just participated in the liberation struggle against the Portuguese into a regular national army. It is important to remember that before RENAMO's insertion in Mozambique's countryside, Rhodesian troops were bombing Mozambican military and civilian targets in retaliation for Mozambique's implementation of United Nations sanctions against Rhodesia. 93 Moreover, Mozambique's support for Zimbabwe's liberation struggled had also angered the Rhodesian regime. 94 Thus, the civil war waged by RENAMO found the new government of Mozambique unprepared for yet another cycle of a prolonged war. In fact, RENAMO consolidated its offensive while most of the liberation struggle war veterans were expecting to pass into civilian life and enjoy the benefits of their struggle for independence. Faced with the challenge of a fast-expanding RENAMO, the government had to request military assistance from Eastern European countries and Cuba. 95 ). Initially, we were expected to receive a oneyear training but, under Cuban authorities' advice, the training was extended to three years. We returned to Mozambique in 1980 to occupy key positions as commanders and instructors in areas of infantry and anti-aerial defence". 97 As RENAMO actions extended to other regions of Mozambique particularly the corridors of Beira and Nacala which connect countries of the hinterland to Mozambican ports in the coast, FRELIMO requested military assistance from the direct beneficiaries of these corridors. In this context, Zimbabwean troops were requested to protect the Beira corridor, which consists of a road, railway and oil pipeline. As a result, in 1985 there were nearly 15 000 Zimbabwean troops in Mozambique. 98 Further north, Tanzanian troops were deployed to protect the Nacala corridor, which was very important for Zambian and Malawian exports and imports. In 1986, the government signed a joint security agreement with Malawi leading to the deployment of Malawian troops in northern Mozambique. 99 The recourse to external military assistance, together with the increase of recruits impacted the national budget. To create an organised and wellequipped army, the government allocated a disproportionate amount of money to the Ministry of Defense. As is demonstrated in the following chart, in 1986, the funds allocated to Defense reached a record of 42 per cent of the national budget. Considering that in this period Mozambican economy had collapsed due to war and prolonged drought and that many people did not have access to adequate education and health services, allocating near half of the national budget to army says a lot about the government counterinsurgency strategy. It gives the impression that the government failed to understand that to counter the insurgency, one must win the hearts and minds of civilians. 96 See In 1984, President Machel appointed General Armando Panguene and Colonel Sérgio Vieira as deputy ministers of Defense adding the number of Defense Ministers to four. 100 With this concentration of resources on Defense forces, the government expected to defeat RENAMO rapidly and definitively.
To the dismay of the ruling party, the allocation of more resources did not translate into success in the battlefield. As the Chief of Staff of the national armed forces had admitted in 1982, the government was on defensive position while RENAMO took the initiative by attacking government targets. Throughout the war, the Chief of Staff attributed this condition to disorganisation of the armed forces. 101 Although the government received military support from the Soviet Union, on the battlefield, it was proven that weapons alone do not fight a war. Warfare also demanded good planning and management.
During the war period, the Mozambican armed forces encountered severe indiscipline problems, mismanagement and corruption. 102 The mismanagement of funds and resources in the Ministry of Defense had long roots, but it was not addressed openly. Under the one-party regime, the majority of people did not have space to denounce irregularities. Only a few people such as President Machel were in a better position to denounce irregularities committed in the management of public treasure. Machel took his first major action in 1981 when he announced a cleanup operation in the Mozambican Armed forces, people's militias, police and security services. 103 Following setbacks in the battlefield, Machel condemned what he called infiltrated elements in the defence and security forces. He pointed out that "infiltrated elements violated the constitution, the principles of FRELIMO Party and the law of the land through arbitrary imprisonment, beatings, torture, and other forms of abuse of power". 104 In 1989 President Joaquim Chissano, the successor to President Machel, admitted the existence of corruption in the party structures. 105 Chissano gave examples of theft and indiscriminate killing of cattle involving high political officials. 106 The use of expressions such as "infiltration" and "indiscriminate killings" by Machel and Chissano hints the existence of high-ranking military officers not fully committed to the army's objectives. It shows that like RENAMO, some segments of government troops were involved in the acts of banditry and looting of civilians. The fragilities of the armed forces and the weakness of the thesis that they were disciplined were more exposed with the introduction of a multiparty system under the constitution of 1990. Evidence of corruption, abuse of power and shortage of supplies within the armed forces showed that, after all, as it was happening with RENAMO guerrillas, government troops were also operating in precarious conditions. This affected their relationship with civilians. A statement by the Governor of Niassa Province, condemning the commanders of government troops who in 1981 were accused of rape, consuming alcohol while on duty, threatening local people with firearms and stealing large quantities of merchandise destined to supply civilians and military in Marrupa and Mecula districts reveal the collapse of law, order and morals among government troops. 107 The most alarming cases of corruption in the armed forces came from Nampula province. In 1992, the Provincial Governor pointed the failure to pay soldiers their wages as the primary cause of the defeats suffered by the national army in that province. These problems, he said, allowed RENAMO to occupy several towns in the province in 1991-92. 108 Also, in Nampula Province, a military prosecutor brought to light cases of corruption and abuse of power within the provincial military command. He accused the provincial 103 Robinson, Curse on the Land-History, pp.321-233; AHM, AIM, 65, November 1981, p.1. 104 AIM, 65, November 1981 Machel died in 1986 in a plane crash. 106 NOTICIAS, 19 April 1989. 107 AIM, 129, June 1992. 108 AIM, 191, June 1992 commander of suspending him from the post in Nampula Province Military Tribunal because he was determined to fight corruption. The prosecutor also denounced military judges who accepted bribes. He went on to accuse the chief attorney of stealing money from military court, and of being always drunk. 109 While the high commanders were selling military rations for personal benefit, the soldiers on the battlefront were starving. It became common to hear rumours of soldiers left to their fate in remote areas without food supplies and salaries. According to the former commander in chief of government troops in Mabote-Inhambane, the army endured long periods without receiving food supplies. He described an episode in which government troops resorted to eating the leaves of wild-grown trees. Other war veterans provided accounts of soldiers going into combat on the verge of starvation. 110 Arguably, as a result, they extorted food from civilians.
Corruption and mismanagement eroded the morals of government forces. Still, it was not until the opening of the country to liberal democracy and greater freedom of expression from 1990 that it was brought to the attention of the public. A look at the letters to the editors of the main newspapers and news magazines from 1990 reveals the problems faced by servicemen. In an open letter published in TEMPO news magazine, a war veteran, complained, "I had not received my salary for two years. Many soldiers are abandoning their positions to look for food. Is it the ministry of defence policy to train soldiers and then abandon them?" 111 He questioned. Another war veteran said that after joining the army in 1982, he did not receive his salary for four years. He questioned; why do salaries not reach the soldiers while the Ministry of Defence sends them to the Provincial Command? Why are the military commanders eager to send troops to combat missions, but they do not pay them? Why does the Provincial Military Command not act against the irregularities committed against the soldiers? We were recruited before preparing our future. Initially, we were told that there was money. They mobilised us politically, but this does not pay the comrades that suffer day and night. 112 Another veteran expressed his frustration with the poor salaries paid in the armed forces. He underlined that the situation prevented him from working with passion and dedication. He stated that when he joined the armed forces in 1985, he was already married and had two children. He asked how 109 AHM, AIM, 187, February 1992, 21. 110 Interview: Author with DN Chitlango. 111 TEMPO, 26 April 1992. 112 TEMPO, 26 July 1992 someone could survive with low salaries paid in the army, especially given the rising costs of living in the country. 113 Oral accounts from war veterans and civilians confirm that some military units stationed in the interior used to go over six months without receiving supplies. One war veteran recalled that "lacking necessities and supplies, we had to live off the land, but because of the drought, it was difficult to find wild products and water sources". 114 While ordinary soldiers focused their denunciation on their superiors, civilians complained about abuse by soldiers stationed in their community to protect them. Like the military, in the 1990s, civilians used the media to express their dissatisfaction with the military and other state functionaries. In Inhambane Province, civilians wrote to the provincial government denouncing abuses committed by government troops. They accused government troops of acting as judges, extorting civilians' goods and flogging people without reason. 115 In this case, the governor forwarded the letter to Provincial Military Command to verify the allegations and take appropriate measures. 116 The final answer to the letter came eight months later. It blamed the victims, accusing them of taking civil conflicts to the military. It recommended local authorities to instruct civilians to seek justice in local courts while addressing words of comfort to the victim of flogging. 117 This reaction from the provincial military command shows that government troops were loath to recognise their mistakes and overlooked excesses against civilians. In another letter to the editor of TEMPO Magazine published in 1992, a civilian told how he was forced to cancel his trip from Inhambane to Maputo after suffering extortion by government troops in Chongoene, Gaza.
"I identified myself as a war veteran". he wrote, "but despite having documents proving my discharge, they accused me of having bought them. I challenged the soldier who interrogated me to tell me where they sell military service discharge cards. My question infuriated the soldier, and he ordered me to lie down to be flogged. To avoid flogging, I used the money saved for my return from Maputo to bribe the soldiers". 118 This episode, reveals the abuses committed by the government against civilians, including war veterans. Another civilian in Nampula Province complained about theft and rape. He pointed out that civilians were facing two wars, one with government troops and the other with RENAMO guerrillas. He contended that it was not clear what distinguished RENAMO from government forces. He went on to question, where the humble people will live if not in their land? How will the fish survive if not in the water? If the soldier does not respect the people whom are they going to respect? When some courageous people question them, they say that civilians have no right to speak to soldiers. It is not from people that the soldiers do get fat? 119 As many soldiers used to say, he observed; "Who does not get fat in this war will never get fat". 120 These letters to public media and government authorities show that the Mozambique Armed Forces, Foças Armadas de Moçambique (FAM) were rife with disorganisation and indiscipline. They also show that in both RENAMO and government forces, there were people who used the war to pursue everyday life through violence.
Despite the denunciation of irregularities, the condition of the military in the frontline did not improve. Left without adequate supplies, they engaged in actions that put them in conflict with civilians. "By mid-1989 rank-and-file soldiers were virtually on strike due to lack of pay and supplies. Thus, some turned to banditry as a survival strategy". 121 In 1992, government troops went on strike in several parts of the country. The mutineers demanded non-paid wages and the updating of pensions for veterans and the war-disabled. 122 The first mutinies took place in late July among commandos and members of the Maputo City garrison. They spread to Manica province in August. Additional mutinies occurred in Montepuez, Cabo Delgado and Dondo, Sofala Province. 123 As the criticism and protests by both civilians and the military show, the condition of the government forces was precarious. By 1992, lacking adequate supplies and facing the worst drought in 30 years in southern Africa, like RENAMO guerrillas, government forces increased the level of taxation and coercion of the local population. 124 Moreover, disorganisation and corruption detracted potential conscripts. Soldiers lacked morale and 119 TEMPO, 26 April 1992. Among ordinary Mozambicans, being fat is an indication of wellbeing.
In the heyday of socialist propaganda, the petit bourgeoisie was depicted in cartoons with big bellies. 120 TEMPO, 26 April 1992. 121 Robinson, Curse on the Land-History, p.309. 122 AIM, 195, July 1992. 123 AIM, 195, July 1992 Robinson, Curse on the Land-History, p.327. material support to wage war. This forced many soldiers to desert the army.
It also made it difficult to convince the youth to join the army. Consequently, the military services received many requests for the postponement of military enlistment. A letter written by a civil servant shows some of the arguments used to avoid military service. He argued that he was the only one left to take care of his family because his brother was already in the army. He added that he was attending school, and he would like to write exams before being incorporated. 125 Letters like these were often written by young men working in public administration. Managers of state companies wrote similar letters arguing that the government should exempt their employees from military conscription. Such arguments benefited, primarily, the educated and wellconnected people. As one citizen complained, the military service became the burden of those not attending school and the illiterate. 126 While the literate used such tactics, many ordinary people adopted various strategies including nepotism, corruption and migration. 127 Since local party structures coordinated the recruitment in rural areas, it was difficult for the youth to hide. For the ordinary men, the only option available was to flee the countryside to South Africa and Zimbabwe, but this was risky because of the checkpoints installed throughout the main roads. 128 Besides, they needed passports which were not issued to people who had not served in the military. Illegal migration had many risks. Since RENAMO received logistic support from South Africa and it had installed its military bases near the border. Subsequently, many young men were captured by Renamo and integrated into their ranks as guerrillas while attempting to flee government recruitment. 129 Vasco Mbondzo was captured by RENAMO trying to cross the border illegally. He remembers that; everyone was expected to serve the military. My brothers were already incorporated, but that was not enough, they also wanted me. In 1989, I decided to cross the border to South Africa. After walking day and night from Massangena, we reached the border. We met people who offered themselves to show us the way. Surprisingly we found ourselves surrounded by armed men. They told us that we had to join them in the struggle against FRELIMO. We had no choice but to comply. After spending three weeks of training, I managed to escape. I entered the South African territory, where I lived as a refugee until the end of the war in 1992. 130 In urban areas, the youth avoided recruitment by attending school and purchasing forged medical certificates stating that they were suffering from diseases such as asthma, tuberculosis and epilepsy. 131 As an interviewee who worked in a public hospital stated, "selling forged medical certificates was a lucrative business. Many young men approached us, looking for documents that could prevent them from serving the army". 132 Hence, an overview of logistics and morale of both RENAMO guerrillas and government troops shows that contrary to government propaganda about the "parasitic" character of RENAMO, both belligerents lacked reliable military and food supplies. Despite being entitled to benefit from state resources, government troops were often left without ammunition and food, clothing and other indispensable goods. With the near-collapse of government supply chains due to corruption, poor roads and RENAMO attacks on military convoys, it became the duty of civilians to assist government troops by providing food, water, and building barracks and digging trenches. As a resident of Inhambane Province, I used to see my mother and other women carrying water and food to government troops barracks. In 1985, people from my village were responsible for the construction of barracks and digging trenches for government troops in the vicinity of my primary school in Matimbe. 133 This is an example of how government troops, like RENAMO guerrillas, relied on civilians and nature to survive. As it was common among RENAMO guerrillas, the relationship between government troops and civilians combined voluntary and coercive mobilisation. When prolonged drought exacerbated the effects of war by exhausting natural resources and agricultural production, the contenders engaged in systemic violence and extortion of civilians.

125
As the war became prolonged, government troops lost their will to fight. 134 By the late 1980s, government troops passed from an offensive to defensive posture. Information from civilians and war veterans reveal that on various occasions, government troops delayed chasing RENAMO guerrillas while they looted from villagers. 135 In fact, some government troops were more concerned with getting plunder rather than protecting civilians. As one government war veteran recalled how; logistics and morale. Therefore, those who claim that the independence of Zimbabwe, the Nkomati Non-Aggression Pact and the end of apartheid in South Africa left no room for successful military rebellion in Mozambique fail to acknowledge that wars are fought with a complex combination of means that are not necessarily military.
Evidence from the fieldwork confirmed Cahen's argument that the Nkomati Accord "had the opposite effect to what FRELIMO expected: RENAMO ended up settling wholly in Mozambique, and the war consequently became a countrywide civil conflict". 141 Thus, the continuation of war showed that it depended on diverse factors, including the contribution of civilians and natural endowment. This state of affairs compels scholars to rethink the nature of civil wars and helps to explain the almost decade long delay in achieving peace in Mozambique. Shifting from previous studies which focused on the role of external forces and over-relied on data collected in central Mozambique, 142 this paper demonstrated that in southern Mozambique, civilians played an significant role in sustaining the war. The experience of southern Mozambique complements recent studies, particularly by Emerson who also shows that, contrary to pro-government propaganda, the Mozambican state also lacked sufficient resources to wage the war. 143 Like RENAMO, the government forces depended on the contribution of villagers for its logistics. As the war endured for 16 long years, both government and RENAMO military forces lost morale and capacity to fight. Instead, they resorted to violence to extract food staples and other resources. The only difference between government troops and RENAMO guerrillas is that the former had the backup of state media which controlled the narrative of the war experiences, internally and externally. This contributed significantly in the way RENAMO was perceived as lacking any tangible political project and as solely responsible for atrocities against civilians.
Moreover, unlike RENAMO, despite being unreliable, government troops had salaries and could expect to receive regular supplies from the state budget. They were also under the scrutiny of civilians who could report their malpractices to state authorities. This differentiated the relationship between the belligerents and civilians slightly, but both were dependent on civilians to wage war. Thus, what is often described as a hotspot of Cold War in Southern Africa or a war of aggression by the apartheid regime was, in practice, a peoples' war with devastating, yet varied impacts on peoples' livelihoods.