The “logic” of Renamo civil waR violence: TRans-boRdeR communiTies and Renamo incuRsions in easTeRn

This study investigates the dynamics of civil war violence by Renamo forces among trans-border communities in Honde Valley from the early 1980s to 1992 when the Mozambican Civil War ended. In its venture to understand reasons given by ordinary people in Honde Valley for forging relationships with Renamo, this article does not seek to dismiss but to make sense of the violence. The article analyses the targets of violence to determine whether there was a switch from indiscriminate to selective violence or vice versa. Through adding another stratum of analysis to this discussion of wartime relationships, this study discusses violence in the context of Renamo’s relations with foreign civilians and Mozambican refugees living in a country whose official international policy was anti-Renamo. It concludes that the relationship between the Honde Valley people and Renamo forces was complicated and sometimes qualified as “being with” Renamo, at times “being with” the Zimbabwe security forces and at yet other times as non-aligned. By examining this dimension of the relationship, this article does not seek to sanitise the acts of banditry, looting and killing by the Renamo units, but to expose currently silenced micro-narratives of the incursions in the Honde Valley

Moçambique (Frelimo) and anti-communist forces of the Resistência Nacional Moçambicana (Renamo), stands as one of the bloodiest civil wars in human history. It resulted in about a million deaths, with an estimated 200 000 civilians killed by the rebels or in military rebel activity. 1 Some 1.7 million people fled to neighbouring countries, and another 4.3 million people became so-called deslocados (displaced persons). 2 With these alarming statistics of civil war violence, it is easy to simplify the relationships that developed during this war as merely that of victim and perpetrator. Such a huge number of casualties should be understood in the context of how insurgency and counter-insurgency warfare is conducted. Stathis Kalyvas succinctly describes the conduct as a, "fight [that] must be conducted through the people -like a man who has to hit an opponent through the body of the referee". 3 This, notwithstanding, research on rebel operational zones has demonstrated that violent conflicts often create symbiotic associations between armed groups and populations living in areas that they control but very little is known about the relationships that develop in areas where armed groups may not seek to control or have no realistic prospects of imposing their authority. A variety of reasons, including the civilian search for protection and the way armed groups use different levels of reluctant or voluntary participation to advance their strategic objectives, have been analysed. 4 Discussions on Renamo invariably lead to conversations on their strategic and tactical use of violence but very little has been done to examine households and individuals who remained or even thrived in their operational zones despite the threat of or actual violence. Thus, there is room, as Abbey Steele demonstrates, to question the conventional wisdom that civil war violence necessarily breeds displacement and to interrogate how individuals who remain in affected areas interact with the army and with rebel movements. 5 As has already been intimated, most influential theories on civil war violence cannot fully account for the mass violence deployed by Renamo. Efforts towards filling these lacunae have begun with Lisa Hultman who has developed a theoretical explanation for Renamo's seemingly indiscriminate violence during the civil war. 6 Hultman concludes that Renamo's use of violence was mainly dominant in the south of Mozambique where their political and military position was precarious and in zones where they enjoyed widespread popularity. They tried to uplift locals with very little violence being perpetrated. 7 By adding another layer of analysis to this discussion of wartime relationships, this study discusses violence in the context of Renamo's relations with foreign civilians and Mozambican refugees living in a country whose official international policy was anti-Renamo. In examining this aspect of the relationship, this article does not seek to sanitise the acts of banditry, looting and killing by the Renamo units, but to expose currently silenced micronarratives of their incursions into Honde Valley along Zimbabwe's eastern border region.
Hultman's efforts notwithstanding, Kalyvas's contention that little is known about the dynamics of civil war violence per se, still rings true when considering Renamo's use of violence. 8 Thus, my research intersects with Kalyvas' and Hultman's in that it seeks to explain the dynamics of Renamo's use of violence among trans-border communities in Honde Valley. In doing this, my study further utilises Lee Ann Fujii's examination of popular participation in the Rwandan Genocide for understanding reasons deployed by ordinary people in Honde Valley in order to forge relationships with Renamo fighters. 9 Since civilian support (or collaboration) matters for the outcome of any conflict, it is crucial to explore, through the lens of civilians, the nature of relationships that developed out of the desire by Renamo to procure supplies along the Zimbabwe-Mozambique border. As this article will show, in Honde Valley, Renamo was content with resupplying rather than confronting state security agents in Zimbabwe or imposing its version of social order. Subsequently my intention is not to dismiss but to make sense of the violence that took place, I go further to analyse the targets of this violence in an attempt to determine whether there was a switch from indiscriminate to selective violence or vice versa as is the case in many civil wars? What is clear from the research is that, unlike elsewhere, there was no switch from indiscriminate to selective violence in Honde Valley henceforth this calls for a close analysis of the trajectories of violence in civil wars. Respondents emphasised the selective aspect of Renamo's use of violence right from the start.
Renamo's logistical challenges are well known to merit an in-depth examination here. 10 Suffice to say that it is improbable, as has been intimated in historiography that they sought to deal with logistical challenges through sheer violence. Renamo's commanders would have known that indiscriminate terror is generally inefficient in civil wars because it is likely to be counterproductive through pushing security-minded civilians into the opposite camp. 11 In light of this observation, one can question the simple explanation of violence by Renamo along the border as mere retaliation of involvement by Zimbabwe forces in Mozambique. 12 Patterns of terror, as will be shown, are better explained by their commitment to punishing defaulters in a trading relationship that depended on trust to survive. As observed in other parts of the world, this study will show that Renamo's use of violence was "generally not a haphazard process, but a regulated one, sequentially taking place". 13 Thus, like in the Liberian Civil War in the 1990s, Renamo fighters developed what Steven Ellis has termed "logique de guerre". 14 With regards to the experiences of Honde Valley people, their relationship with Renamo was complex and certainly sometimes qualified to "being with" Renamo and at times "being with" the Zimbabwe security forces and at times being nonaligned. 15

accessing The field
The data presented in this paper is derived from fieldwork carried out at six research sites (Chisuko, Chavhanga, Sagambe, Katiyo, Chinaka, Mandeya) in Honde Valley spread over three months in 2019. During these months, I also made research trips to three adjacent Mozambican villages, Pandagoma, Makore and Nanhanga (Manica Province). Data were gathered through the collection of life histories and semi-structured interviews with over 50 families that experienced violence from Renamo forces in both Mozambique and Zimbabwe. Additional data were collected during interviews with victims' surviving family members, neighbours and traditional authorities. During this time, I was a resident in the communities to grasp local everyday activities and how these were affected by violence. The international border between Mozambique and Zimbabwe in Honde Valley is open hence adjacent border communities share a lot of resources including water, land, schools, dip tanks, clinics, transport and grocery shops. Moreover, these communities share a common language and ethnic identity with minimal reference to national identities. All interviews in Zimbabwe were conducted in Manyika, while those in Mozambique were conducted in Barwe -a socially intelligible Shona dialect spoken by Mozambicans close to the border. I had to cut off my research in October 2019 when the environment in Mozambique became so polarised between Renamo and Frelimo supporters as they headed towards controversial national elections that were won by Frelimo. Media reports and academic papers gave powerful commentary to developments within Mozambique during that period. I chose to do my interviews in Chavhanga, Katiyo, Sagambe, Mandeya, Chisuko and Chinaka areas of Honde Valley partly because these experienced several Renamo incursions. Most of which resulted in fatalities and also partly because these communities harboured Mozambican refugees who were resettled by traditional leaders. I conducted two group interviews at one research site in Pandagoma while observatory visits were also made to Nanhanga and Makore villages (Mozambique). These visits allowed me to obtain as many different perspectives as possible through listening and watching rather than the active asking of questions. Having been introduced by the traditional leader, I took the opportunity to speak to two groups of Mozambican men and women, respectively. Everyone whom I spoke to had lived in either or both Mozambican and Zimbabwean border communities during the civil war.
Getting entry and trust in post-conflict societies is a big challenge, especially for an outsider. 17 Accordingly, I found a research assistant within the Honde Valley community who was known to many of the interviewees and, from then on, was referred to local gatekeepers. In Chavhanga, Sagambe and Pandagoma, Mr Luke Nyachega, a resident and a relative, used his knowledge and experience of the civil war to introduce me to people who ordinarily would not have shared their experiences. In Katiyo, Chisuko and Chinaka, my interlocutor was Mr Charles Gurure, who had lived in that community from the mid-1980s. As a former plantation supervisor and now in semi-retirement but working as a part-time orderly at Chisuko Clinic that serves trans-border communities, Mr Gurure's presence and introduction to interviewees put many locals at ease. Most of the interviews were conducted in people's homes where privacy was guaranteed. Each interview began with self-introduction and a detailed explanation of the entire project. These introductions allowed us to demonstrate that we had nothing to hide and served to allay fear, since in conflict and post-conflict societies suspicions of others, particularly outsiders, run very high. 18 As Fujii observed for Rwanda, the time of my fieldwork occurred 28 years after the end of the Mozambican Civil War. Hence memories undoubtedly changed while some people had forgotten some details and could rearrange chronologies, confuse sequences, and give greater weight to some moments than to others. 19

The moZambican civil waR
Although this paper is not about the Mozambican Civil War, it is paramount to provide the historiographical context in which this study is located. A growing body of literature published during and in the immediate aftermath of the Mozambican Civil War invariably concludes that violence by Renamo insurgents was central to its tactics to gain "support" from both Zimbabweans and Mozambicans living along the border. 20 Not surprisingly, the relationship involving Renamo and people living on the Zimbabwean side of the Mozambique-Zimbabwe border has been largely conceived in simplistic victim-perpetrator terms. This argument, however, cannot go unchallenged. Such a view lacks historical and cultural contextualisation, as the widely diverging experiences of relations between rebels and civilians suggest that the relationship was more complicated than previously assumed. In contrast, the national picture of how this was a war involving Zimbabwe as it aided the Frelimo government to stem the tide of invasion has already been told enough to merit any further close attention. 21 This study adds another layer to this narrative by examining how civilians took advantage of trading networks with the Renamo amidst this chaos. As the study will show, thriving networks of trade involving illegal drugs, biltong, cigarettes, denim jeans, ivory and other commodities were at the centre of this cross-border relationship.
As Harri Englund has observed, the fundamental dividing line in historiography is the question of whether this was a civil war or a national tragedy created by external aggression. 22 Renamo fighters are usually portrayed as "pseudo-terrorists", "pseudo-guerrillas", "puppets" and "warlords". 23 Others view them positively as "freedom fighters against communism and authoritarianism" and Renamo as "a popular movement". 24 It has been widely accepted that Renamo was formed in 1977 with the assistance of Rhodesians who wanted to destabilise guerrilla operations in Rhodesia and Mozambique. 25 Thus, the key proponents of this school view them as an anti-African nationalist proxy movement. 26 However, the problem with this is that it ignores the fact that with or without external influence in Mozambican politics, the failure of Frelimo to institute sound and popular reforms was in itself a substantial cause in the creation of an opposition or rebel movement. 27 Margret Hall notes that sweeping generalisations blur our understanding of Renamo as a movement. 28 In the same vein, Mark Chingono laments that such generalisations unfairly reduced Mozambicans to "mere passive victims 21 See, S Funada-Classen, of manipulations and machinations by powerful external forces". 29 Given this trend of casting insurgent movements as proxy-movements, there has been little appreciation of local causes and dynamics to African civil wars. Alex Vines has offered nuances of the military organisation of Renamo in his 2017 study of Renamo operations in Maringue. Vines concludes that there was a strong focus on discipline, control and that local commanders determined how combatants treated civilians. 30 This is important in explaining networks that were developed at a local level with villagers and also given the rebel group's quandary with regards to the provision of logistical and materiel support to their fighters. 31 These challenges are not to be underestimated as they were real. Narratives of starving suspected Renamo fighters are replete in Honde Valley with some of them allegedly eating unripe avocados and sugar cane as well as begging for food from the tea workers at Aberfoyle Tea Estate and raiding funerals for food. 32 Although Renamo used force to recruit and procure resources within Mozambique, this argument, as will be shown, has to be made with caution when studying Zimbabwean nationals who patronised Renamo controlled markets and zones in Mozambique. Their status as foreigners meant they had a somewhat different relationship with the organisation as they had a higher chance of defaulting. In turn, this also meant the movement had to deploy extra-judicial measures to protect its resources. Vines' work is part of a long tradition of attempts to understand how Renamo operated and used violence. In their studies, Colin Darch, and Ken Wilson articulated the position of religion in Renamo's operations in Zambezia Province and also how this informed their violent streak. 33 These analyses considered ritualised violence and terror as means both to initiate civilians into the movement and to demonstrate the movement's superior command of spirits and medical substances. As will be shown in the case of Charles Muyambo, these cultic and ritual elements are also crucial in understanding 29 Chingono, The State, Violence and Development, pp. 3-35. how some Renamo trading partners in Honde Valley or their families received punishment for defaulting.
In her study of healing in Honde Valley, Schmidt argues that Honde Valley is a borderland characterised by fluidity of identities, mobility, danger and opportunity. 34 This study does not focus on these aspects but rather on how the armed men from Mozambique interacted with society and how these interactions have continued/discontinued in the post-conflict period. It is not about healing per se but about how the experiences have shaped local interactions for both perpetrators and victims as well as how they influence attitudes towards Renamo. Narratives of danger, though accurate, have clouded other existential possibilities on the frontier. Except for Nicholas Nyachega and Enocent Msindo's works on the concept of the "everyday" in Honde Valley, opportunities which emerged out of this war have largely been ignored. 35 Much of the works on Renamo examine their violence in Zambezia Province in the north and Gaza Province in the south. 36 Jocelyn Alexander has examined post-war Manica Province (central Mozambique) and revealed that much of the Mozambican communities in the province were Renamo spheres of influence. 37 Though this may explain why cross-border expeditions along the border were numerous, Renamo cross-border operations into Honde Valley are yet to be examined trenchantly. In the study of Indian borderlands, India's determination to quell an insurgency converted the territory of Jammu and Kashmir into a battleground borderland where nationalist ambitions and opposing constructions of nationhood came into a confrontation. 38  governments to defeat the insurgency militarily turned the area into a battleground. 39 The experiences of violence, and the creation of socio-economic relationships in Mozambican Civil War which this study focuses on connect with the global literature on civil war economies. As already mentioned earlier, works by Hultman, Fujii and Kalyvas have widened knowledge on how violence has been deployed in various civil wars across the world. 40 Whereas Kalvvas, through various examples drawn from across the globe, has managed to explain violence in wars, Fujii utilises an African context (the Rwandan Genocide) to explain popular participation. 41 More important to this study and using the Liberian Civil War as his case study, Mats Utas, articulates female wartime experiences through concepts such as "social navigation" and "victimcy" concluding that the "relationship between "victimhood" and "agency" [is] far from a linear path proceeding from the point of being a victim to that of being a survivor". 42 Drawing on Utas' study, my study of Honde Valley demonstrates, as he does, that the war zone is not "merely a wasteland… but at times may also be a field ripe with possibility for social and economic mobility, even as it may also contain unforeseen pitfalls". 43

Renamo waR PRacTices in honde valley
Renamo activities in Honde Valley generated attention in Zimbabwean state print media in the 1980s and early 1990s. Several robberies and acts of terrorism by Renamo fighters were reported and helped to profile Renamo as a terrorist organisation in the Zimbabwean national psyche. The ruling Zimbabwean National African Union Patriotic Front (ZANU PF) government also took the opportunity to accuse local opposition parties such as Ndabaningi Sithole's Zimbabwe African National Union Ndonga (ZANU Ndonga) of colluding with the Mozambican rebels. 44 A tense political atmosphere amidst insurgent activities and anti-insurgent operations turned the eastern borderland into contested war zones. The following cases reported in the local media at the time will help to contextualise how 39 Nyachega, "Beyond War, Violence, and Suffering", p.137. 40 Hultman  51 In 1990, Renamo was to hog the limelight again following an attack at Mukambachaza homestead which left five people dead.
At the centre of the strategies deployed by the state to stop these Renamo incursions was the curtailment of cross-border movements. Despite widespread state propaganda, stopping cross-border activities proved to be very difficult for security specialists. Increased Renamo activities compelled security officers and politicians to repeatedly appeal to villagers along Zimbabwe's eastern border with Mozambique not to visit their relatives across the border and indulge in illegal cross-border trading. 52 The then commander of the Zimbabwe National Army (ZNA), Lieutenant-General Solomon Tapfumaneyi Mujuru, also intervened in the matter, 45 The Surprisingly, a decorated military official imagined the cessation of cross-border movements when the reality is that in wars, civilians are known to move in their search for relatively safe environments. 54 It betrays the state's misplaced confidence that borderland communities took Renamo as nothing but a menace. This call demonstrates the blind spots in state policy since livelihoods along the border that emerged out of a long history of manoeuvring the border could only adapt to the war rather than collapse entirely. Unlike Mujuru's simplistic reduction of crossing the border as choosing between life and death, people in Honde Valley understood that even in the ebb and flow of the war, life went on. Understanding the movements and engaging a holistic approach would have assisted security officials to understand the historical relationships that Renamo fighters were tapping into. While Mujuru touched on marriage as an important part of relations across the border, that he did not go further to tackle cross-border trade as one of the major factors fuelling these movements and that he presented Renamo attacks as entirely indiscriminate demonstrates how misguided policymakers were. Of course, this is not to dismiss his finding that spates of insurgent activities such as murders, abductions, looting, armed robberies and cattle rustling were on the increase. 55 Indeed, to bolster the ZNA's military efforts along the border, the Zimbabwe Republic Police even urged bus operators and the people travelling along the eastern border to reach their destination before dark to avoid Renamo insurgents' attacks. 56 5. noT jusT abouT waR: social and economic TRading wiTh Renamo  59 Similarly, dagga trade was also rampant across the border throughout the colonial period. 60 In the 1980s, Renamo tapped into these networks to establish crossborder supplies of contraband. Although, the link between armed conflict and the production and trafficking of illicit drugs has been noted in the literature, 61 for Renamo the specific dynamics of this linkage remain poorly understood. pillaging on a wide scale. 62 With the end of the Cold War, Renamo found itself like many other belligerents, more dependent on mobilising private sources of support to sustain its military and political activities. 63 These changes impacted on Renamo's access to materiel resulting in their field commanders having significant power in influencing relationships with civilians and also how the war effort was financed. Furthermore, within Mozambique, Frelimo's counter-insurgency operations such as forced villagisation and deliberate attacks on civilians affected Renamo's support base and forced it to turn to the border communities for food, money and clothes. 64 Following the signing of a General Peace Agreement (GPA) in 1992 between Frelimo and Renamo, acts of banditry along the Mozambique-Zimbabwe border began to decline. 65 Peace amplified trading ties between villagers living on both sides of the border. 66 Besides these trading ties, Renamo fighters also took advantage of other longestablished forms of social transactions between populations on both sides of the border including rain cults, ancestral cults and healing. 67 Throughout the 1980s and early 1990s, Renamo fighters established informal markets close to the Zimbabwean border. 68 In such markets, various illicit commodities were offered for sale on easy terms. In the areas they controlled, Renamo fighters, among other things, grew cannabis for sale in Zimbabwe. 69 Honde Valley residents were aware of such markets at places like Nhanga on the Mozambican side where ngondonga (wildebeest) biltong was one of the main commodities. 70 It was easy for Renamo soldiers to kill game for meat, ivory or horns because Mozambican game parks were no longer protected. 71  Zimbabweans. Maize meal, clothes and soap were exchanged for dried meat. As will be shown, the more enterprising amongst Zimbabwean customers would go for more commercial rather than subsistence commodities such as ivory and rhinoceros horns. 72 Eager to push volumes and also to increase the popularity of their markets, Renamo operatives usually provided their commodities on easy terms, including offering rhino horns on credit and accepting verbal arrangements to pay later. Mike Makuyana explained that the environment in Renamo markets was conducive because "they didn't have guns; it was just like a market place where one trades freely. You would just buy your dried meat or do your barter trade with Renamo fighters and left without being harmed". 73 Defaulters were harshly treated and usually defaulting attracted cross-border punitive expeditions by Renamo that often entailed the massacre of the culprits and, if found absent, their relatives. 74 To save their own lives, some defaulters, as observed by Mozambicans, "reported that the Matsanga soldiers were killing local people. Those who had defaulted to Matsanga came to report to ZNA, and they knew Renamo bases, so they directed the soldiers". 75 In civil wars, selective violence is difficult to achieve because it requires information, typically private information. 76 This is important for understanding Renamo operations inside Zimbabwe as they depended on local intelligence. Understanding the nature of the Renamo informal markets helps to explain how defaulters were tracked down across the border. In their attacks, Renamo fighters tended to move in the opposite direction from that enunciated by Kalyvas', that is, they selected their targets given their forged relationships only to turn indiscriminate once their hard targets were missed. This would lead them to punish family members and neighbours. Renamo had its auxiliaries/police/intelligence agents (mijibhas) who patronised their trading places and kept clients' "debt records". 77 When trading partners defaulted, these operatives sometimes accompanied Renamo fighters across the border to identify their homesteads for punishment. 78 Charles Muyambo explained the nature of the punitive expeditions as follows, Renamo fighters never came here on their own. They came here because locals from this area incited them. They went to borrow different items such as cannabis, ivory and other prohibited goods...when they realised, they are now in trouble, they then stay this side and never go back there…The moment Renamo realised that the payments are now due, they would immediately come after such people. Some of the people who borrowed from Renamo would also give them their names. Renamo would then use those details to track them down, and in most cases, they would come to the specific homestead of the person who would have failed to honour the payments. In areas close to the border, they would move around asking for directions to specific homes. They would even ask to be accompanied to such households. The moment they arrived at that particular home, they would randomly kill people, and they would not care to find out who exactly is involved and who is not. 79 This testimony is typical of the narratives that defy the official description of Renamo fighters. Contrary to popular belief, Renamo fighters were selective in Honde Valley, choosing some homesteads and leaving others, in addition to even asking for directions from locals. Their aim was not to fight local people, but they felt like such people were sabotaging them, and in that, they had justification for killing them. Renamo fighters saw their expeditions as legitimate because they were following up "dishonest" community members who had "stolen" from them. 80 Therefore, for every massacre that occurred, there is an underlying story of the victim having been their "client" or victims having a blood relationship with someone who was connected to them.
Besides using mijibhas to extract information which they would use to track down defaulters, Renamo also tried to establish debtors' physical addresses directly. People in desperate need of Renamo goods also volunteered information, including their physical addresses to convince them. In some cases, the debtor would go with a Renamo representative and physically show him his homestead. As will be shown later in the case of Diason Chimbo, Renamo fighters sometimes delivered commodities to their business associates in Zimbabwe. However, since the trading of rhino or elephant tasks was strictly prohibited, some of these debtors sometimes fell into police custody for possession of contraband. In such situations, the convict's family fell victim to Renamo's punishment because reasons for defaulting would not have been communicated to the creditors. 81
To corroborate this, Shepherd Bello, former Renamo fighter-turned refugee, explained the nature of Renamo's expeditions. He revealed that Renamo punitive expeditions would go inland looking for their debtors leaving homesteads around the border. 82 Revealingly, Bello explained that Renamo forces also had mijibhas (Zimbabweans) in Honde Valley communities who were also crucial in providing intelligence on targets, The mijibhas would escort Matsanga to the homestead they would be looking for, but would not enter the homestead, and they would remain anonymous to the family which would be attacked. Mainly Matsanga attacked at night, so the attacked family would not know the sell-out. 83 Selling out, however, was not always successful as Renamo fighters sometimes investigated claims before delivering punishment. The case of Collins Chisuko is particularly revealing on this aspect. A team of Renamo fighters assigned to carry out punitive justice on a defaulter arrived at Chisuko's homestead looking for him by name, but they immediately noticed that though the name was correct, he was not the debtor. Someone had stolen from Renamo through false pretence. He remembered, I approached them, and we exchanged good words, and they opened up to me that they were looking for a man with trembling hands because I was not the one who owed them although my name was on the list. From the moment they said so, I knew they were looking for Chirambakusakara, and I directed them to his homestead. They didn't find him because he was aware that Matsanga was on his way to his home; thus, he fled away. 84 In the markets themselves, Renamo was also on the prowl in search of other defaulters. After being mistaken as a defaulter, Charles Muyambo's testimony regarding his near-death experience in a Renamo Camp is revealing about the shortcomings of Renamo's verbal arrangements with Zimbabwean cross-border traders, At first, when I went there, I brought a sack full of dried meat in a 90kg container. When I went there for the second time with my son-in-law...he almost left me behind after one Renamo operative had accused me of having taken dried meat on credit from them. I told him it was a case of mistaken identity since it was my first time there. The one who had accused me took me to his superiors and told them that I was one of the people who had failed to honour the payments. So, they resolved to kill me. We went out of the camp down the road...the camp stretched for quite a distance...
We exited the gate in the company of one Renamo fighter. Their leader followed us, walked behind us with his gun. The Renamo officer who had taken custody of me called out to his superior "Hey sir, we have caught that culprit". He further told his superiors that I was the one who had failed to pay for the cannabis, rhino horns, and biltong. The superior asked with surprise if I was the accused person? He looked at me from the bottom to the top, and I remained silent. After carefully looking at me from head to toe, he concluded that I was not the person they were looking for… The leader reprimanded him for almost killing an innocent person. I thanked them and went on to buy my meat and came back home. 85 Besides demonstrating the loopholes in Renamo's loaning arrange ments, Muyambo's testimony also reveals an important aspect of the movement's operations, that is, the place of religion and what Wilson has termed "ritualised" destruction. 86 Muyambo revealed that the officer who saved him from death was a seer whose opinion vetoed that of the arresting officer. Renamo had a specific elite corps, the Grupo Limpa, "a unit of specifically magically imbued and trained soldiers" for maintaining Renamo as a ritually invincible organisation. 87 The story of how a local businessman, Diason Chimbo, was nabbed by ZNA officials in Chavhanga in October 1989 offers an interesting lens to Renamo networks in Honde Valley. As a sole trader, Chimbo's commodities were partly supplied by Renamo, especially ivory, marijuana and reed mattresses. 88 In return, Chimbo provided money, soap, chocolates and other forms of preserved food. 89 Circumstances leading to Diason's outing as a local conduit for Renamo goods are worth narrating. Diason had run out of supplies and, thus, wrote a letter informing Renamo that some of the goods had been bought and others had remained unsold. These interactions came to light after ZNA soldiers intercepted this letter. One of the eyewitnesses to this incident, Dzinduwa, remembered, The army officers said to Diason, "we are here to take the goods from you". Diason asked, "Which goods are you talking about?" He was then handed the letter which he had written to the Renamo forces earlier. After reading through the letter, he realised that it was now a severe issue… They ordered him to go into the house and collect the money before they proceeded to confiscate the remaining items. He retrieved the money and surrendered it to the officers… He was ordered to go and collect all the remaining items. The officers saw buckets of cannabis and quantities of elephant tusks and were surprised. The soldiers asked Diason, "How do you transport all these items? Do you transport them using a car?" Diason replied saying "they bring the 85 Interview: Author with C Muyambo. 86 Wilson, "Cults of violence and counter-violence in Mozambique", p. 534. 87 Wilson, "Cults of violence and counter-violence in Mozambique", p.545. 88 Interview: Author with C Dzinduwa. 89 Interview: Author with C Dzinduwa.
items for me here". The officers also asked if Diason would sometimes help them in bringing the items to his house. He said that he helped them. 90 Chimbo was not the only Honde Valley Renamo middleman to be busted during the civil war. The tragic loss of the life of Steven Muruju, an infamous popular Renamo cannabis middleman, is particularly telling in this regard. Unlike Chimbo who quickly surrendered to the security forces, Muruju was shot dead in an exchange of fire with Zimbabwean security forces. 91 Gurure remembered Muruju as "a good person before his involvement with Renamo. Things changed when he became involved in cannabis, and ivory deals with Renamo. He was now hostile and was now armed. He was caught at his home and exchanged fire with the army. He was then shot and killed". 92 Renamo protected its middlemen from harassment within Honde Valley. Renamo insurgents would kill those who tried to block them from interacting with their trading partners. In Chavhanga, for example, Jipe Nheredzo, a village head, was murdered for warning his subjects to stop interacting with Matsanga. Following the killing of Jipe Nheredzo, Renamo visited Teddy Muomba at his home in Machena area. Muomba, whose son cheated Matsanga killed a Renamo insurgent that same night of April 1990. 93 Dzinduwa explained, Teddy was not aware of the goings-on…His son was given an elephant tusks, but he did not tell his father. One day, he was surprised to see Renamo forces coming to his house saying "Old man we want our money". He did not tell them that he knew nothing about the issue… He pretended to agree to give them the money. He rose from the seat and went with them. He took his stabbing arrow with him. It was very dark on this particular day. A Renamo fighter who was in the front was then stabbed. When he cried for help, his colleagues from Renamo ran away for refuge. They went across the river. He [stabbed fighter] tried to follow his colleagues, but he could not make it. He died. After his death, Renamo was then angered. Teddy Muomba was honoured by the government of Zimbabwe for his bravery and relocated to Rusape for security reasons. 94 That Teddy Muomba was able to defend himself using a home-made weapon was because the government of Zimbabwe had allowed its border civilians to have various assortments of weapons to beef up self-defence against Renamo.

Renamo's cRoss-boRdeR PuniTive exPediTions
While the previous section established the contours of cross-border trade in general, this section explores cross-border punitive expeditions by Renamo. Punishing defaulters in an operational zone dominated by a betterresourced ZNA indeed demanded skill on Renamo's part especially access to insider information. In their "punitive" expeditions, Renamo took advantage of intra-group dynamics including purely private (recurring family feuds) and larger cleavages. 95 This is typified by the Mujangu Family in Mukambachaza Village where Renamo insurgents killed five people: Esitere Changwena (50), Beauty Mujangu (12), Dadirai Mujangu (9), Kudakwashe Mujangu (6) and Doca Chaoneka (5). 96 Four other villagers were injured during the same attack. The Mujangu believe that they were targeted on account of their elevated social status and that those who reported them to Renamo were jealous. These experiences have historical precedence. According to Peter Hart, the typical informer in the Irish Civil War (1922-3), was not someone with a cause but "rather someone with a grudge, a grievance, or with people or property to protect. Others saw the opportunity for gain or to settle old scores". 97 All these are serious examples of what happens in a civil war situation. According to an eyewitness, Amai Noria Marombedza, Doca Chaoneka's mother, the insurgents demanded money and food from her. When she did not comply, they killed her granddaughter, tied her son with a rope and threatened to rape her. 98 She also alleged that Village Head Mukambachaza's son (name withheld) who worked with Renamo from Mozambique had directed the insurgents to her home. 99 Lucia Mujangu, one of the survivors from the attack, indicated that Square Mushamainza, Muchenga and Bothy accompanied the attackers who demanded groceries from her mother. 100 She recollected, "one of the Matsanga took my nephew and banged his head on the floor and he died on the spot. My little sister was killed while sleeping. They took me as they wanted me to escort them carrying the groceries. On the way, I fled, but up to now, I don't even know how I managed to". 101 In an apparent reference to the Mujangu Family feud, Dzinduwa argued, 95 Kalyvas, The logic of violence in civil war. 96 The Attempts to interview the accused or members of his family were futile as they avoided any talk connected to Renamo. However, not all testimonies absolve the Mujangus. An alternative perspective suggests that the family became targets after one of their daughters broke up with a Renamo fighter and double-crossed him with a ZNA soldier. 103 Similarly, the killing of Shingai Muchika in Mandeya Village also appears indiscriminate at face value but further probing revealed that it was connected to "defaulting" payments to Renamo. Muchika had no known connections with Renamo fighters but was caught in the crossfire of their attempts to recover stolen commodities from a fellow villager. His family alleges that the Chief's daughter, Rose Mandeya, took items from Matsanga but failed to clear the debt. When pressured to pay, she then claimed that their ivory had been given to the Muchika Family. 104 In the same Mandeya area, Fanta Zhwati and Mr Mudhomboti who were actively involved with Renamo marijuana, gold and ivory trade were also abducted and were never to be seen again. 105 Whereas those killed in Chavhanga and Sagambe were either macheted or axed, in Mandeya, they were summarily gunned down. 106 As Alexander has observed about Renamo war strategies in Manica Province, these tactical and operational variations suggest that different Renamo groups were operating in different parts of Honde Valley. 107

moZambican Refugees in honde valley
The preceding discussion has established the nature of relations between Renamo and Honde Valley residents. However, this has not included Mozambican refugees, their spatial location in this cross-border "arrangement" and how local communities came to view them in the context of murderous excursions by their compatriots. In this section, there will be no discussion on how Renamo dealt with Frelimo sympathisers and fencesitters in Mozambique as this has already received significant attention in historiography. 108 Suffice to say that Mozambican civilians found themselves squeezed in between two ruthless fighting forces that had little regard for international laws of military engagement. Faced with this situation, a significant number relocated to Honde Valley while some passed through it in their search for safety. Indeed, during the 1980s and early 1990s, thousands of displaced Mozambicans who did not enter refugee camps flooded Zimbabwean rural villages, the border towns of Nyanga, Mutare, Chimanimani and Chipinge and farming and mining areas in the eastern province of Manicaland where most of them were employed in various informal capacities. 109 As Englund has argued for Mozambican refugees on the Mozambican-Malawian borderlands, civil war refugees in Honde Valley came into "exile as persons enmeshed in relationships, and the processes by which they experienced their displacement were indeterminate because of the myriad relationships which they carried with them". 110 Although Mozambicans fleeing the civil war exploited trans-border relationships to settle in Honde Valley, communities in which they were settled in were ill-equipped to distinguish genuine refugees from the infamous Renamo punitive expeditions. Unsurprisingly, local villagers also began to label, name and discriminate against others because of the frequent Renamo attacks. Despite these local perceptions in Honde Valley, traditional leaders loomed large in receiving Mozambican refugees and allocating them resources to meet their basic daily needs. The decision to accommodate Mozambique refugees in Honde Valley is partly explained by familial relationships that cut across borders. Furthermore, the Zimbabwean government actively encouraged them to receive unarmed refugees. 111 In making decisions whether to flee to Honde Valley, some refugees were guided by either the presence or absence of family across the border. There is nothing as revealing about the experiences of these refugees as the following testimony by Simbarashe Kumadzi, My family lived along the border since our homestead was about three kilometres to the border but on the Mozambican side. When the war was at its peak, these borderland homesteads were victims of both Matsanga and Frelimo soldiers as they were constantly attacked. Because of this, my father decided to move to stay at Villa Catandica Camp in Mozambique where there was protection from the soldiers... During this civil war, our fathers wanted to resist the war as they thought that the war was going to be light but that was not the case at the end. They tried to resist also because we had large livestock, so they didn't want to lose these livestock and other resources. When everyone was relocating to safer areas, our fathers remained at their original homes along the border. This prompted Zimbabwean soldiers to view "our fathers" who lived along the border as Renamo supporters since they were not seeking refugees as other Frelimo supporters were doing. Because of these perceptions, it was difficult for them to seek refuge in Zimbabwe. They then decided to go to Villa Catandica still in their country than to go to a foreign land where they were viewed as enemies. 112 So from the preceding discussion one can notice that there were several considerations before Mozambicans sought refuge. Those who went to live in Mozambican camps like Villa Catandica were considered to be Frelimo supporters, while those who decided to live in their original homes to protect their livestock were considered as Renamo supporters hence they too were targeted for punishment by Frelimo soldiers. 113 In Honde Valley, Mozambican refugees would first arrive at the headman's place to inform him that (s)he is a refugee and had come to seek permission to stay. Mostly, they stayed with their relatives on the Zimbabwean side. 114 It was also common for Zimbabwean communities to have welcoming ceremonies for Mozambican refugees with the understanding that they would go back to their homes when the war ended. 115 Kumadzi argued, "We preferred to come to Zimbabwe where there were people who would protect us…our relatives in Zimbabwe offered us a haven. My daughter was married to Mr Mahachi, who said we should come to stay with them". 116 During the liberation struggle for Zimbabwe, Zimbabwean refugees in Mozambique also created strong relationships with Mozambicans which were later reciprocated during the Mozambican Civil War. Many Honde Valley inhabitants had relatives across the border who could choose to stay permanently in Zimbabwe or seek refuge. 117 Yet experiences were quite different for Mozambican refugees who had no close relatives in Zimbabwe. Bello's testimony below puts this into perspective, I was born at Catandica, in Mozambique. I grew up in Mozambique together with my parents and we later separated when l was 17 years old. My parents were separated in the sense that my mother was taken away by Frelimo which was/is the ruling party in Mozambique whilst my father and I were taken by Renamo soldiers… Since I was a young teenager, Renamo wanted to recruit more soldiers on their side, and young teenagers were wanted the most. My colleagues and I spent 3 months in the bush because we didn't want to be recruited as soldiers. In 1982, a family friend who lived in Zimbabwe came to see my father, and I used this opportunity to go with him to the Zimbabwean side… In Zimbabwe, I lived a hard life and it was tough because I had no relatives since it was a foreign land. I lived at some homesteads in Zimbabwe and worked as a houseboy doing various types of chores, be it working in the fields, cooking and laundry. The main reason was not to work, but I only wanted a place to stay and be given food to eat since I had no one to look after. 118 As indicated, Bello and his father were forced into Renamo's structure by the abduction of his mother by Frelimo soldiers. He argues that a senior Frelimo officer took her by force as his wife leaving them with no choice but to join the other side. However, he was to reunite with his mother in Zimbabwe in 1985 after she fled from a Frelimo Camp. His mother was employed as a tea picker at the Eastern Highlands Tea Estate while his father remained a Renamo soldier until his death after the war ended. 119 From his experiences under Renamo, Bello narrated the malnourishment of combatants such that he had to sneak out with his father "to our home to get food". 120 Once he left the life of scrounging for food in Renamo camps, Bello realised that migrants were being received as labourers to work in the fields or cattle herders in Honde Valley. 121 George Manyanga, a Mozambican village head in Pandagoma, corroborated these testimonies by pointing that some Mozambicans directly sought refuge with their relatives from across the border and that this was a driving force to mix or interact with Matsanga. For him, Renamo fighters in their cross-border punitive expeditions did not discriminate Zimbabwean or Mozambican defaulters. 122 A case in point is that of the mass killing of seven members of the Kembo Family after Kembo (a Mozambican refugee in Honde Valley) defaulted in remitting payments for marijuana and elephant tasks that were being smuggled to Harare. 123 With absolute certainty, Mike Makuyana stated, "it's not a secret that the Kembo homestead was 118 Interview: Author with S Bello. 119 Interview: Author with S Bello. 120 Interview: Author with S Bello. 121 Interview: Author with S Bello. 122 Interview: Author with G Manyanga, Mbare Musika, Harare, 5 August 2019. 123 Interview: Author with M Makuyana. The same point was reiterated by G Manyanga destroyed because he took elephant tusks on credits and did not pay back the money". 124 The fact that the Kembo family members were targeted and murdered in a foreign land reveals another dimension to the impact of the Mozambican Civil War on Honde Valley, especially with regards to the settlement of war refugees. An example of collective resettlement of Mozambique refugees in Zimbabwe would be that of Pachije Village which, however, suffer from many Renamo cross-border expeditions. The local security forces believed that Renamo patronised this village because they were following their relatives who had sought refuge there. 125 Resultantly, it was agreed that some refugees were to be sent further away from the border. Thirty-six Mozambican families in Pachije Village were identified for resettlement. In the end, those that had stayed there for at least five years were spared, while those with less than a year in Zimbabwe were relocated further into the Zimbabwean interior. Under this programme, ten families out of the 36 families were sent to the Mazowe area. 126 How these 36 families came to have plots in the same village merits explanation. Village Head Manyanga explained thus, The stands were not bought. For example, when you sought refuge, you would first arrive at your friend's/relative's place, and it was your friend/relative who would take you to the headman and introduce you. For example, the headman would ask you if you have any piece of land you want to give to this refugee, and as your friend/ relative, he would give you a place where you would start your things. As a token of appreciation, you were supposed to pay a little amount to the headman, for example, a hen. It wasn't necessarily about purchasing stands as like nowadays. For example, there are a lot of Mozambican families who now permanently live on the Zimbabwean side in Pachije Village, for instance, Fibiona, Makorora, Chingungu and Mandikisi. 127 As Manyanga has shown, most refugees from Mozambique were not troubled by "the artificial borders" that divided Mozambique and Zimbabwe as they, like those on the Mozambique-Malawi border, were able to mix easily with Zimbabweans of their ethic group. 128 Furthermore, what this reveals is that these relationships also outlived the Mozambican Civil War as some Mozambicans, depending on their circumstances, chose either to stay in Honde Valley or return to their borderland communities from where they continued to interact with Honde Valley residents.

conclusion
This paper has examined how micro-level relationships that developed in Honde Valley influenced Renamo's use of violence during the Mozambican Civil War. It has demonstrated that this rebel movement had a complex relationship with Honde Valley residents stretching beyond the simplistic victim-perpetrator and indiscriminate violence perspectives. In demonstrating these nuances at a local level, this analysis reveals that though officially and historically viewed as indiscriminate in the usage of violence, Renamo's activities were primarily influenced by the complicated nature of its crossborder relationship with the Honde Valley border communities. Renamo's ability to immerse itself in the border communities was partly a result of the limits of the official state in monitoring and curbing informal interactions.
By showing how Renamo became part of these communities, this paper has revealed that cross-border relationships during the Mozambican Civil War emerged out of mutual existential desperation. Although it has not ventured into the post-conflict era, this article sets the platform for further research into how borderlands will continue to serve as zones of opportunities for some borderland communities. Thus violence in these spaces was not as haphazard and indiscriminate as previously thought.

acknowledgemenTs
The research was made possible by a generous Individual Research Grant awarded in May 2019 by the African Peacebuilding Network (APN). I would like to thank Prof. Alois Mlambo, Nicholas Nyachega, Eric Makombe, Joseph Mujere, Sibanengi Ncube and Nathaniel Chimhete, for their valuable comments on earlier drafts.